Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
Contents lists available at ScienceDirect
Renewable and Sustainable Energy Reviews
journal homepage: www.elsevier.com/locate/rser
Review and analysis of biomass gasification models
Maria Puig-Arnavat, Joan Carles Bruno *, Alberto Coronas
Rovira i Virgili University, Mechanical Engineering Dept., Av. Paı¨sos Catalans, 26, 43007 Tarragona, Spain
A R T I C L E I N F O
A B S T R A C T
Article history:
Received 12 May 2010
Accepted 13 May 2010
The use of biomass as a source of energy has been further enhanced in recent years and special attention
has been paid to biomass gasification. Due to the increasing interest in biomass gasification, several
models have been proposed in order to explain and understand this complex process, and the design,
simulation, optimisation and process analysis of gasifiers have been carried out. This paper presents and
analyses several gasification models based on thermodynamic equilibrium, kinetics and artificial neural
networks. The thermodynamic models are found to be a useful tool for preliminary comparison and for
process studies on the influence of the most important fuel and process parameters. They have the
advantage of being independent of gasifier design, but they cannot give highly accurate results for all
cases. The kinetic-based models are computationally more intensive but give accurate and detailed
results. However, they contain parameters that limit their applicability to different plants.
ß 2010 Elsevier Ltd. All rights reserved.
Keywords:
Biomass
Gasification
Thermodynamic model
Kinetic model
Contents
1.
2.
3.
4.
5.
Introduction and objectives. . . . . . . . . . . . . . . . . .
Biomass gasification principles and technologies.
Performance of biomass gasifiers . . . . . . . . . . . . .
Biomass gasification models . . . . . . . . . . . . . . . . .
4.1.
Kinetic rate models. . . . . . . . . . . . . . . . . . .
4.2.
Thermodynamic equilibrium models . . . . .
4.3.
Aspen Plus gasification models . . . . . . . . .
4.4.
Neural network gasification models . . . . .
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1. Introduction and objectives
As we face the problems of global warming and climate change,
substantial research and development has focused on the use of
biomass as an alternative to fossil fuels. The widespread
availability of biomass has been widely recognised, as has its
potential to supply much larger amounts of useful energy with
fewer environmental impacts than fossil fuels [1].
Biomass can be converted into commercial products via either
biological or thermochemical processes [2–4]. Biological conversion of low-value lignocellulosic biomass still faces challenges
related to low economy and efficiency [2]. Combustion, pyrolysis
and gasification are the three main thermochemical conversion
* Corresponding author.
E-mail address: juancarlos.bruno@urv.cat (J.C. Bruno).
1364-0321/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.rser.2010.07.030
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2841
2842
2842
2844
2844
2846
2847
2849
2849
2850
methods. Biomass is traditionally combusted to supply heat and
power in the process industry. The net efficiency for electricity
generation from biomass combustion is usually very low, ranging
from 20% to 40% [3]. Biomass cofired in existing combustors is
usually limited to 5–10% of the total feedstock due to concerns
about plugging existing coal feed systems [4]. Pyrolysis converts
biomass into bio-oil in the absence of oxygen (O2). The limited uses
and difficulty in downstream processing of bio-oil have restricted
the wide application of biomass pyrolysis technology [5].
Gasification converts biomass through partial oxidation into a
gaseous mixture, small quantities of char and condensable
compounds. It is considered one of the most efficient ways of
converting the energy embedded in biomass, and it is becoming
one of the best alternatives for the reuse of waste solids.
From time to time, attempts are made to explain the complex
nature of gasification. The time has come to review gasification
process modelling in order to highlight the role of gasification
2842
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
models. This review aims to compare and analyse various biomass
gasification models presented by different authors.
2. Biomass gasification principles and technologies
Gasification is partial thermal oxidation, which results in a high
proportion of gaseous products (CO2, water, carbon monoxide,
hydrogen and gaseous hydrocarbons), small quantities of char
(solid product), ash, and condensable compounds (tars and oils).
Steam, air or oxygen is supplied to the reaction as an oxidising
agent. The gas produced can be standardised in its quality and is
easier and more versatile to use than the original biomass (e.g. it
can be used to power gas engines and gas turbines or as a chemical
feedstock for the production of liquid fuels). Gasification adds
value to low- or negative-value feedstock by converting it into
marketable fuels and products.
The chemistry of biomass gasification is quite complex. Broadly
speaking, the gasification process consists of the following stages
[6–9]:
Drying. In this stage, the moisture content of the biomass is
reduced. Typically, the moisture content of biomass ranges from
5% to 35%. Drying occurs at about 100–200 8C with a reduction in
the moisture content of the biomass of <5%.
Devolatilisation (pyrolysis). This is essentially the thermal
decomposition of the biomass in the absence of oxygen or air.
In this process, the volatile matter in the biomass is reduced. This
results in the release of hydrocarbon gases from the biomass, due
to which the biomass is reduced to solid charcoal. The
hydrocarbon gases can condense at a sufficiently low temperature to generate liquid tars.
Oxidation. This is a reaction between solid carbonised biomass
and oxygen in the air, resulting in formation of CO2. Hydrogen
present in the biomass is also oxidised to generate water. A large
amount of heat is released with the oxidation of carbon and
hydrogen. If oxygen is present in substoichiometric quantities,
partial oxidation of carbon may occur, resulting in the generation
of carbon monoxide.
Reduction. In the absence (or substoichiometric presence) of
oxygen, several reduction reactions occur in the 800–1000 8C
temperature range. These reactions are mostly endothermic. The
main reactions in this category are as follows:
Water–gas reaction:
C þ H2 O ! CO þ H2 131:4 kJ=gmol
(1)
By gasifier pressure: Atmospheric or pressurised.
By reactor design:
Fixed-bed (updraft, downdraft, cross-draft and open-core): The
fixed-bed gasifier has a bed of solid fuel particles through
which the gasifying media and gas either move up (updraft),
move down (downdraft) or are introduced from one side of the
reactor and are released from the other side on the same
horizontal level (cross-draft). It is the simplest type of gasifier,
usually consisting of a cylindrical space for fuel and gasifying
media with a fuel-feeding unit, an ash-removal unit and a gas
exit. In the fixed-bed gasifier, the fuel bed moves slowly down
the reactor as the gasification occurs. Fixed-bed gasifiers are
simple to construct and generally operate with high carbon
conversion, long solid residence time, low gas velocity and low
ash carry-over [11,12].
Fluidised-bed (bubbling, circulating and twin-bed): The
gasifying agent is blown through a bed of solid particles at a
sufficient velocity to keep the particles in a state of suspension.
Fuel particles are introduced at the bottom of the reactor, are
very quickly mixed with the bed material, and almost
instantaneously are heated up to the bed temperature. As a
result of this treatment, the fuel is pyrolysed very fast, resulting
in a component mix with a relatively large amount of gaseous
materials. Further gasification and tar-conversion reactions
occur in the gas phase. Twin-bed gasification uses two
fluidised-bed reactors. The biomass enters the first reactor,
where it is gasified with steam, and the remaining char is
transported to the second reactor, where it is burnt with air to
produce heat. The heat is transported to the gasification reactor
by the bed material, normally sand. The flue gas and the
product gas have two separate exits.
Entrained-flow: These gasifiers are commonly used for coal
because they can be slurry-fed in direct gasification mode,
which makes solid fuel feeding at high pressures inexpensive.
These gasifiers are characterised by short residence time, high
temperatures, high pressures and large capacities [13].
Stage gasification with physical separation of pyrolysis,
oxidation and/or reduction zones.
Table 1 compares various types of biomass gasifiers.
A review of gasifier manufacturers in Europe, the United States
and Canada [14] identified 50 manufacturers offering commercial
gasification plants, of which 75% were of the fixed-bed downdraft
type, 20% were fluidised-bed systems, 2.5% were of the updraft
type, and 2.5% were of various other designs.
Bounded reaction:
C þ CO2 $ 2CO 172:6 kJ=gmol
(2)
Shift reaction:
CO2 þ H2 $ CO þ H2 O 42 kJ=gmol
(3)
Methane reaction:
C þ 2H2 $ CH4 þ 75 kJ=gmol
(4)
Gasification reactor designs have been researched for more
than a century, which has resulted in the availability of several
designs at the small and large scales. They can be classified in
several ways [10]:
By gasification agent: Air-blown gasifiers, oxygen gasifiers and
steam gasifiers.
By heat source: Auto-thermal or direct gasifiers (heat is provided
by partial combustion of biomass) and allothermal or indirect
gasifiers (heat is supplied by an external source via a heat
exchanger or an indirect process).
3. Performance of biomass gasifiers
The performance of biomass gasifiers could be characterised by
several parameters. Here, we will review two such parameters:
producer-gas composition, which directly influences the heating
value of the gas, and gasification efficiency.
The composition of the gas obtained from a gasifier depends on
a number of parameters, such as fuel composition, gasifying
medium, operating pressure, temperature, moisture content of the
fuels, mode of bringing the reactants into contact inside the gasifier
(gasifier design), etc., and this is why it is very difficult to predict
the exact composition of the gas from a gasifier [15]. By way of
example, Table 2 shows typical gas composition data as obtained
from commercial wood downdraft gasifiers operated on low- to
medium-moisture-content fuels [16] and Table 3 shows typical
producer-gas composition and operating conditions for atmospheric bubbling fluidised-bed gasifiers [17].
The oxidant or gasifying agents can be air, pure O2, steam, CO2
or mixtures thereof. Air, while a cheap and widely used gasifying
agent, contains a large amount of nitrogen, which lowers the
2843
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
Table 1
Salient features and comparative evaluation of different designs of biomass gasifiers [9,106,107].
Downdraft
Simple and proven technology.
Fuel specificity in terms of both type and size.
Suitable for biomasses with low moisture.
Producer gas with moderate calorific value and low tar and ash
(or particulate) content.
High exit gas temperature.
Suitable for capacities of 20–200 kW.
High residence time of solids.
High overall carbon conversion.
Limited scale-up potential with maximum capacity of 250 kW.
Bubbling fluidised bed
High fuel flexibility in terms of both size and type.
Flexibility of operation at loads lower than design load.
Ease of operation.
Low feedstock inventory.
Good temperature control and high reaction rates.
Good gas–solid contact and mixing.
In-bed catalytic processing possible.
Producer gas with moderate HHV but low tar levels and high particulates.
Carbon loss with ash.
High conversion efficiency.
Suitable for large-scale capacities (up to 1 MW or even higher).
Good scale-up potential.
Entrained-flow bed
Relatively complex construction and operation.
Fuel specificity in terms of particle size (costly feed preparation).
Low feedstock inventory.
High temperature gives good gas quality.
Problems with construction materials at high temperature.
Good gas–solid contact and mixing.
Producer gas with moderate HHV and low tar content.
High conversion efficiency.
Suitable for high capacities (>1 MW).
Very good scale-up potential.
Table 2
Typical producer-gas composition from commercial wood for downdraft gasifiers
operated on low- to medium-moisture-content fuels [16].
Component
[%]
H2
CO2
CH4
CO
N2
Heating value
12–20
9–15
2–3
17–22
50–54
5–5.9 MJ/m3
Table 3
Typical producer-gas composition and operating conditions for atmospheric
bubbling fluidised-bed gasifiers [17].
Operating conditions
ER
S/B (kg/kg daf)
T (8C)
Gas composition
H2 (vol%, dry basis)
CO (vol%, dry basis)
CO2 (vol%, dry basis)
CH4 (vol%, dry basis)
C2Hn (vol%, dry basis)
N2 (vol%, dry basis)
Steam (vol%, wet basis)
Yields
Tars (g/kg daf)
Char (g/kg daf)
Gas (Nm3/kg daf)
LHV (MJ/Nm3)
Air [108]
Steam (pure)
[109]
Steam–O2
mixtures [110]
0.18–0.45
0.08–0.66
780–830
0
0.53–1.10
750–780
0.24–0.51
0.48–1.11
785–830
5.0–16.3
9.9–22.4
9.0–19.4
2.2–6.2
0.2–3.3
41.6–61.6
11–34
38–56
17–32
13–17
7–12
2.1–2.3
0
52–60
13.8–31.7
42.5–52.0
14.4–36.3
6.0–7.5
2.5–3.6
0
38–61
3.7–61.9
na
1.25–2.45
3.7–8.4
60–95
95–110
1.3–1.6
12.2–13.8
2.2–46
5–20
0.86–1.14
10.3–13.5
na: not available; daf: dry ash-free basis; ER: equivalence ratio; S/B: steam-tobiomass ratio (H2O (kg/h)/biomass (kg daf/h)).
Updraft
Simple and proven technology.
Low exit gas temperature.
High thermal efficiency.
Producer gas with moderate calorific value but high tar and ash
(or particulate) content.
High residence time of solids.
High overall carbon conversion.
Extensive gas cleanup required before it can be used in engines.
Suitable for capacities up to 250 kW.
Limited scale-up potential.
Circulating fluidised bed
High fuel flexibility in terms of both size and type.
Flexibility of operation at loads lower than design load.
Ease of operation.
Low feedstock inventory.
Good temperature control and high reaction rates.
In-bed catalytic processing possible.
Producer gas with moderate tar levels but high particulates.
High carbon conversion.
Good gas–solid contact and mixing.
Suitable for large-scale capacities (up to 1 MW or even higher).
High conversion efficiency.
Very good scale-up potential.
Twin fluidised bed
Relatively complex construction and operation.
Producer gas with moderate HHV and moderate tar levels.
Cleaning of gas required before it can be fired into engines.
In-bed catalytic conversions possible.
Good gas–solid contact and mixing.
Relatively low efficiency.
Suitable for high specific capacities (>1 MW).
Good scale-up potential but relatively complex design.
heating value of the syngas produced. If pure O2 is used as the
gasifying agent, the heating value of syngas will increase but the
operating costs will also increase due to the O2 production costs.
Partial combustion of biomass with air or O2 can provide heat for
drying the biomass, raising the biomass temperature and driving
the endothermic gasification reactions, and generate water and
CO2 for further reduction reactions [15]. The heating value and H2
content of the syngas can be increased if steam is used as the
gasifying agent, in which case the heating value of the product gas
is about 10–15 MJ/Nm3 [17,18], compared with 3–6 MJ/Nm3 for air
gasification of biomass [19,20]. The use of CO2 as the gasifying
agent is promising because of its presence in the syngas. CO2 with a
catalyst such as Ni/Al can transform char, tar and CH4 into H2 and/
or CO, thus increasing H2 and CO contents [21–23]. Pure steam or
CO2 requires an indirect or external heat supply for endothermic
gasification reactions [24–27]. Alternatively, a mixture of steam or
CO2 and air or O2 can be used as the gasifying agent, and the partial
combustion of biomass with air/O2 provides the heat required for
gasification [17,28–30].
Some authors, like Prins et al. [31,32], Ptasinski et al. [33] and
Ptasinski [34], have focused their studies on the efficiency of
biomass gasification. Efficiency is either based on energy (lower
heating value, LHV) (Eq. (5)) or exergy (chemical, and chemical and
physical, exergy) (Eq. (6)). All efficiencies are defined as the ratio
between the exergy (energy, respectively) of the syngas to the
exergy (energy, respectively) of the biomass:
Energy efficiency ð%Þðper kg of biomassÞ : h ¼
ngas LHVgas
LHVbiomass
(5)
Exergy efficiencyð%Þðfor an adiabatic gasifier using air as the
gasifying mediumÞ : c ¼
ngas ðech;gas þ eph;gas Þ
ech;biomass þ nair eair
(6)
2844
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
where ngas is the molar amount of product gas (kmol); nair is the
molar amount of air (kmol); ech,gas is the chemical exergy of
product gas (kJ/kmol); eph,gas is the physical exergy of product gas
(kJ/kmol); ech,biomass is the chemical exergy of biomass (kJ/kmol);
eair is the specific molar exergy of air (kJ/kmol).
Ptasinski [34] analysed the efficiency of biomass gasification
using the triangular C–H–O diagram, considering a biomass fuel
that can be represented by a general formula of CH1.4O0.59N0.0017.
At the equivalence ratio of 0.26, the chemical and total exergy of
the gas reach the maximum at the carbon boundary. The carbon
boundary point (CBP) is the optimum point for operating an airblown gasifier and it is obtained when exactly enough gasifying
medium is added to avoid carbon formation and achieve complete
gasification. Desrosiers [35] and Double and Bridgwater [36]
proved that the CBP is the optimum point for gasification with
respect to energy-based efficiency, and Prins et al. [37] proved that
it is the optimum point with respect to exergy-based efficiency, as
cited by Ptasinski et al. [33]. At this point, according to Ptasinski
[34], the exergetic efficiency of biomass gasification is the highest,
equal to 80.5%. For gasification with steam, operation at the CBP is
also optimal from the point of view of exergetic efficiency. The
optimal gasification occurs at the steam-to-biomass ratio of
1.30 kg/kg and results in an overall exergetic efficiency of 87.6%.
Steam gasification is thus more efficient than air gasification, but
this advantage is reduced when the exergy losses during steam
production are taken into account. The exergetic efficiency of
gasification also depends on the chemical composition of the
biofuel used as feedstock. According to Ptasinski [34], the energetic
efficiencies of vegetable oil, straw, treated wood, untreated wood
and grass are comparable with that of coal, whereas the efficiencies
of sludge and manure are considerably lower (Fig. 1). This can be
expected, because gasification at the optimum operating point is
not possible for these streams. In sludge and manure gasification,
oxygen is added mainly to generate heat and to evaporate moisture
present in the fuel. If the exothermic oxidation reactions could
drive endothermic gasification reactions, rather than the endothermic evaporation of water, the gasifier would work much more
efficiently. Efficiency based on chemical exergy is higher for coal
and vegetable oil than for the other biomass streams. The same
trends can be observed for gasification efficiencies based on
chemical and physical exergy. Because coal and vegetable oil are
gasified at higher temperatures, their gasification efficiencies are
relatively much improved by the inclusion of the physical exergy.
Drier biomass, such as treated wood or straw, may be slightly
preferred over fresh biomass, such as untreated wood and grass.
These results assume that gasification reaction rates are fast
enough and residence time is long enough for the equilibrium state
to be reached.
Fig. 1. Comparison of gasification efficiency of various fuels [34].
4. Biomass gasification models
The efficient operation of a biomass gasifier depends on a
number of complex chemical reactions, including fast pyrolysis,
partial oxidation of pyrolysis products, gasification of the resulting
char, conversion of tar and lower hydrocarbons, and the water–gas
shift reaction. These complicated processes, coupled with the
sensitivity of the product distribution to the rate of heating and
residence time in the reactor, required the development of
mathematical models. The main goals of these models are to
study the thermochemical processes during the gasification of the
biomass and to evaluate the influence of the main input variables,
such as moisture content, air/fuel ratio, producer-gas composition
and the calorific value of the producer gas.
Some studies only consider the final composition of chemical
equilibrium, while others take into account the different processes
along the gasifier, distinguishing at least two zones. The models
can be divided into kinetic rate models, thermodynamic equilibrium models and neural network models. Some models use the
process simulator Aspen Plus [101] combining thermodynamic
and kinetic rate models.
4.1. Kinetic rate models
Kinetic models provide essential information on kinetic
mechanisms to describe the conversion during biomass gasification, which is crucial in designing, evaluating and improving
gasifiers. These rate models are accurate and detailed but are
computationally intensive [38]. Nevertheless, numerous researchers have focused extensively on kinetic models of biomass
gasification: Wang and Kinoshita [39], Di Blassi [102], Fiaschi
and Michelini [40], Giltrap et al. [41], Yang et al. [42], Roshmi et al.
[43], Dennis et al. [44], Babu and Sheth [54], Radmanesh et al. [45],
Gobel et al. [46], Sharma [38], Fermoso et al. [47], Zhong et al. [48]
and Roy et al. [49].
Kinetic models describe the char reduction process using
kinetic rate expressions obtained from experiments and permit
better simulation of the experimental data where the residence
time of gas and biomass is relatively short.
The kinetic model proposed by Wang and Kinoshita [39] is
based on a mechanism of surface reactions in the reduction zone
assuming a given residence time and reaction temperature. Giltrap
et al. [41] developed a model of the reduction zone of a downdraft
biomass gasifier to predict the composition of the producer gas
under steady-state operation, adopting the kinetic rate expressions
of Wang and Kinoshita [39]. The accuracy of the model is limited by
the availability of data on the initial conditions at the top of the
reduction zone; pyrolysis and cracking reactions are not considered because the number of possible pyrolysis products, along with
all the possible reactions and intermediate products, would make
the model very complex. It assumes that all the oxygen from the air
inlet is combusted to CO2 and that the pyrolysis products are
completely cracked. Solid carbon, in the form of char, is considered
to be present throughout the reduction region. It is assumed that
the char reactivity factor (CRF), which represents the reactivity of
the char and is a key variable in the simulation, is taken as constant
throughout the reduction zone. These authors tested the model
with experimental data for two different downdraft gasifiers (Chee
[50,105]). Fig. 2 compares the composition of the dry producer gas
predicted by this model with those found experimentally. The
model produced reasonable agreement with the experimental
results for all components except CH4. According to Giltrap et al.
[41], this over-prediction was the result of the assumption that O2
in the air reacts only with char. The pyrolysis products are cracked
in a region of high temperature and in the presence of O2, so it is
probable that some of the CH4 produced will undergo combustion
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
Fig. 2. Predicted results from the model of Giltrap et al. [41] compared with
experimental data from Chee [105] and Senelwa [50].
Fig. 3. Predicted results from the model of Giltrap et al. [41] compared with
experimental data from Chee [105] and Senelwa [50] when the initial conditions of
the model assumed that the CH4 produced by the cracking of pyrolysis products
reacts with O2 entering the gasifier through the air inlet.
with O2. Fig. 3 shows the results when the initial gas concentration
is altered under the assumption that all of the O2 in the air reacts
with the CH4 from the cracking of the pyrolysis gas. This
assumption reduces the amount of CH4 predicted, but the
prediction is still higher than the concentrations found experimentally.
Jayah et al. [51] developed a model based in the early work of
Chen [52] but with a few modifications. Chen’s model was
intended to estimate the length of the gasification zone and the
diameter of the reactor, and to investigate the dependence of the
reactor’s performance on operating parameters such as feedstock
moisture content, chip size, reactor insulation, input air temperature and gasifier load. Chen’s model consists of three parts. The first
part determines the amount of oxygen needed for a fixed input of
fuel at a specific operating condition. The fuel-to-air ratio
estimated from this first part of the model is then used as an
input in the second part, where the drying, pyrolysis and
combustion zones are all lumped together and considered as a
single zone. The outputs from this ‘‘lumped’’ zone are the product
concentrations and the temperatures of the gaseous and solid
phases leaving the zone. These calculated concentrations and
temperatures are then used as inputs in the third part of the model,
which predicts the temperature profile along the axis of the
gasification zone, the gas composition, the conversion efficiency
and the length of the gasification zone at any given time interval.
2845
The main weakness of Chen’s model is the over-prediction of the
gas exit temperature from the ‘‘lumped’’ zone due to an
unrealistically low estimate of heat loss and the omission of CO
and H2 in the pyrolysis gas. Jayah et al. [51] therefore introduced
modifications to overcome these deficiencies and also to suit a
reactor with a variable rather than constant gasification zone
diameter. For this reason, the authors incorporated Milligan’s [53]
flaming pyrolysis sub-model instead of the algorithms used by
Chen [52]. The aim of Milligan’s Daming pyrolysis zone model is to
calculate the composition of the product gas entering the
gasification zone in terms of CO, H2, CO2, H2O, CH4 and N2. As a
result, the model used in the study by Jayah et al. [51] consists of
two sub-models, namely of the Daming pyrolysis and the
gasification zones. The Daming pyrolysis zone sub-model is used
to determine the maximum temperature and the product
concentration of the gas leaving that zone. The gasification zone
sub-model assumes that a single char particle moves vertically
downwards along the vertical axis of the gasifier. This sub-model
includes a description of the physical and chemical processes, the
flow equations, the transport phenomena and the conservation
principles. The model is limited to considering the effect of packed
char particles in the reduction zone.
Babu and Sheth [54] modified Giltrap’s model suggesting an
exponentially varying CRF in order to predict better simulation of
the temperature profile in the reduction reaction zone. The CRF
value was increased both linearly and exponentially along the
length of the reduction bed in the model. The model was simulated
with a finite difference method to predict the temperature and
composition profiles in the reduction zone. The model predictions
were compared with the experimental data reported by Jayah et al.
[51] (Fig. 4). Simulations were performed for varying CRFs ranging
from 1 to 10,000, and linearly and exponentially as well.
Simulations were also performed for different values of CRF (1,
10, 100 and 1000) held constant throughout the reduction zone.
The authors of the study concluded that the CRF must be varied
along the reduction zone of the downdraft gasifier and that their
simulated results were in very good agreement with the
experimental data of Jayah et al. [51]—better, in fact, than the
mathematical model of Jayah et al. [51], which considered an
exponentially varying CRF value.
Recently, Sharma [38] presented a model for a downdraft
gasifier in which the reduction zone was modelled using a finite
rate of reaction following the chemical kinetics. The pyrooxidation zone, prior to the reduction zone, was also modelled
considering thermodynamic equilibrium. However, the author did
not take into account any char combustion in the pyro-oxidation
zone and also neglected the formation of methane there. The
Fig. 4. Comparison of various producer-gas compositions that have varying CRF
values (Babu and Seth [54]) with experimental data from Jayah et al. [51].
2846
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
water–gas shift equilibrium was considered at the outlet of the
pyro-oxidation zone. In the reduction zone, a linear variation of
CRF was adopted. Following this previous work, Roy et al. [49]
developed a model for a downdraft gasifier based on chemical
equilibrium in the pyro-oxidation zone and finite-rate kineticcontrolled chemical reactions in the reduction zone. The CRF was
optimised by comparing the model’s predictions against the
experimental results from the literature.
4.2. Thermodynamic equilibrium models
Kinetic rate models always contain parameters that limit their
applicability to different plants. Thus, thermodynamic equilibrium
calculations, which are independent of gasifier design, may be more
suitable for process studies on the influence of the most important
fuel process parameters. At chemical equilibrium, a reacting system
is at its most stable composition, a condition achieved when the
entropy of the system is maximised while its Gibbs free energy is
minimised. However, thermodynamic equilibrium may not be
achieved, mainly due to the relatively low operation temperatures
(product gas outlet temperatures range from 750 to 1000 8C)
(Bridgwater [106]). Nevertheless, models based on thermodynamic
equilibrium have been used widely. Some recent efforts include the
work done by Bacon et al. [55], Double et al. [56], Ruggiero and
Manfrida [57], Zainal et al. [58], Schuster et al. [59], Altafini et al. [60],
Li et al. [61,62], Melgar et al. [63], Jarungthammachote and Dutta
[71,73], Yoshida et al. [64], Karamarkovic and Karamarkovic [65],
Huang and Ramaswamy [66] and Haryanto et al. [67] to predict the
performance of commercial gasifiers. These authors have shown
reasonable agreement between equilibrium predictions and experimental data.
Equilibrium models have two general approaches: stoichiometric and non-stoichiometric. The stoichiometric approach
requires a clearly defined reaction mechanism that incorporates
all chemical reactions and species involved. In the non-stoichiometric approach, no particular reaction mechanisms or species are
involved in the numerical simulation. The only input needed to
specify the feed is its elemental composition, which can be readily
obtained from ultimate analysis data [62]. The non-stoichiometric
equilibrium model [68] is based on minimising Gibbs free energy
in the system without specifying the possible reactions taking
place. The stoichiometric chemical equilibrium model is based on
selecting those species that are present in the largest amounts, i.e.
those which have the lowest value of free energy of formation. As
noted by Prins et al. [37], Desrosiers [35] showed that under
gasification conditions (with temperatures between 600 and
1500 K) the only species present at concentrations higher than
104 mol% are CO, CO2, CH4, H2, N2, H2O and solid carbon
(graphite). For this system of species, there are three independent
chemical reactions (Eqs. (1), (2) and (4)), according to Duhem’s
theory [69]. For the homogeneous system that consists of CO, CO2,
CH4, H2, N2 and H2O, there are two independent chemical
reactions, resulting from the combination of Eqs. (1) and (2) and
also Eqs. (1) and (4).
As shown by various authors [70,71], the two approaches
(stoichiometric and non-stoichiometric) are essentially equivalent.
A stoichiometric model may also use free energy data to determine
the equilibrium constants of a proposed set of reactions.
Equilibrium models are based on some general assumptions
that are in better agreement with some specific types of reactors
for which equilibrium models have better predictive capabilities.
Prins et al. [31] presented these assumptions:
The reactor is implicitly considered to be zero-dimensional.
The gasifier is often regarded as a perfectly insulated apparatus,
i.e. heat losses are neglected. In practice, gasifiers have heat
losses to the environment, but this term can be incorporated in
the enthalpy balance of the equilibrium model.
Perfect mixing and uniform temperature are assumed for the
gasifier although different hydrodynamics are observed in
practice, depending on the design of the gasifier.
The model assumes that gasification reaction rates are fast
enough and residence time is long enough to reach the
equilibrium state.
No information about reaction pathways and the formation of
intermediates is given in the model.
Tars are not modelled.
Due to these assumptions, equilibrium models yield great
disagreements under some circumstances. Typical pitfalls at
relatively low gasification temperatures are the overestimation
of H2 and CO yields and the underestimation of CO2, methane, tars
and char (in fact, null values for these last three components above
800 8C) [72]. For this reason, and as detailed below, several authors
have modified and corrected the equilibrium model or used the
quasi-equilibrium temperature (QET) approach.
Zainal et al. [58] modelled the biomass gasification process on
the basis of stoichiometric thermodynamic equilibrium. They
predicted the composition of the producer gas for different
biomass materials. Jarungthammachote and Dutta [71] developed
the thermodynamic equilibrium model based on the equilibrium
constant for predicting the composition of a producer gas in a
downdraft gasifier. They used coefficients for correcting the
equilibrium constant of the water–gas shift reaction and the
methane reaction in order to improve the model. Those coefficients
were obtained from the comparison between the model and the
results of other researchers’ experiments. The predicted results
from the modified model satisfactorily agree with experimental
results reported by Jayah et al. [51] (Table 4).
Jarungthammachote and Dutta [73] applied the non-stoichiometric equilibrium model to three types of gasifiers: a central jet
spouted bed, a circular split spouted bed and a spout-fluid bed. The
simulation results from the model showed a significant deviation
from the experimental data, especially for CO and CO2. One
important factor was carbon conversion. Thus, the model was
modified to consider the effect of carbon conversion. The results
improved and were closer to the experimental data (Table 5).
However, this model could not give results with high accuracy for
the spouted-bed gasification process. The heating value was also
an important parameter because it is usually used to estimate the
energy that could be gained from using that producer gas. The
modified model predicted heating values that were generally
higher than those from experiments because of the over-prediction
of the CO content in the producer gas.
An equilibrium model for studying biomass gasification with
steam in a fluidised-bed gasifier was presented by Schuster et al.
[59]. The results of the equilibrium model for the gasifier
(LHV gas yield) were in the range of the measured results,
though the CH4 content in the product gas was overestimated. It
Table 4
Comparison of the results from the modified model of Jarunthammachote and Dutta
[71] with the experimental data of Jayah et al. [51].
Gas composition
% mol dry basis
H2
CO
CH4
CO2
N2
m
Model data [71]
Experimental data [51]
MC (16%)
MC (14%)
MC (16%)
MC (14%)
16.81
17.86
1.05
12.10
52.18
0.4472
16.80
18.52
1.06
11.68
51.94
0.4415
17.00
18.40
1.30
10.60
52.70
0.3361
12.50
18.90
1.20
8.50
59.10
0.3927
2847
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
Table 5
Comparison between experimental results, original model and modified model of Jarunthammachote and Dutta [73].
H2 (vol%)
Central jet spouted bed at 1323.3 K
Experiment
12.56
Original model
11.08
Modified model
13.55
Circular split spouted bed at 1388.3 K
Experiment
10.98
Original model
10.26
Modified model
12.45
Spout-fluid bed ER = 0.35 at 1148.7 K
Experiment
8.43
Original model
14.99
Modified model
16.07
Spout-fluid bed ER = 0.30 at 1127.65 K
Experiment
11.86
Original model
15.45
Modified model
16.72
CO2 (vol%)
CO (vol%)
CH4 (vol%)
N2 (vol%)
O2 (vol%)
HHV (MJ/Nm3)
RMS error
14.56
2.6
8.73
14.97
30.36
19.18
0.7
0
0
54.96
55.96
58.53
2.27
–
–
3.906
5.44
4.302
8.020
3.319
13.7
3.17
9.16
16.41
29.23
18.15
0.88
0
0
57.47
57.34
60.22
0.55
–
–
3.961
5.183
4.022
6.807
2.385
14.95
10.42
14.42
11.61
20.68
13.71
2.52
0
0
61.55
53.9
55.8
–
–
–
3.891
4.688
3.917
5.935
4.133
14.48
10.43
14.5
13.03
21.08
13.76
2.95
0
0
56.87
53.3
55.02
–
–
–
4.01
4.801
4.01
4.400
2.459
was shown that the discrepancies in the prediction of the gas
composition did not significantly influence the overall efficiency.
Li et al. [61] used a non-stoichiometric equilibrium model
(minimisation of Gibbs free energy) to predict the producer-gas
composition from a circulating fluidised-bed coal gasifier. Li et al.
[62] employed the equilibrium model to predict the producer-gas
compositions, product heating value and cold gas efficiency for
circulating fluidised-bed gasification. They observed that real
gasification processes deviate from chemical equilibrium. Therefore, in order to correct the deviations, they developed a
phenomenological model to modify the equilibrium-based framework to account for key non-equilibrium factors. As they knew
from a pilot-plant study the experimental carbon conversion and
methane yield, it was possible to correct non-equilibrium effects
by withdrawing the corresponding carbon and hydrogen from the
equilibrium system. Fig. 5 compares the experimental data and the
predicted values from the modified model. This method was also
applied successfully to coal gasification [61] and to steam–
methane reforming [74], where hydrogen was preferentially
removed through perm-selective membranes.
Another approach is the use QET, whereby the equilibria of the
reactions defined in the model are evaluated at a temperature that
is lower than the actual process temperature. This approach was
introduced by Gumz [75]. For fluidised-bed gasifiers, the average
bed temperature can be used as the process temperature, whereas
for downdraft gasifiers, the outlet temperature at the throat exit
should be used. Li et al. [61] found that the kinetic carbon
conversion for pressurised gasification of sub-bituminous coal in
the temperature range 747–877 8C is seen to be comparable to
equilibrium predictions for a temperature about 250 8C lower.
Bacon et al. [76] defined QETs for each independent chemical
reaction. Based on 75 operational data points measured in
circulating fluidised-bed (CFB) gasifiers operated on biomass,
Kersten et al. [77] showed that, for operating temperatures in the
range 740–910 8C, the reaction equilibrium of Eqs. (1), (2) and (4)
should be evaluated at much lower temperatures (respectively,
531 25 8C, 583 25 8C and 457 29 8C). These QETs appear to be
independent of process temperature in this range.
This literature review has shown that equilibrium models are
useful tools for preliminary comparison, but that they cannot give
highly accurate results for all cases. As mentioned above,
thermodynamic equilibrium models do not require any knowledge
of the mechanisms of transformation. Moreover, they are
independent of the reactor and not limited to a specified range
of operating conditions. They are valuable because they predict the
thermodynamic limits of the gasification reaction system. Thus, in
order to describe the behaviour of gasifiers more accurately,
modifications have been made to equilibrium models. According to
Villanueva et al. [72], chemical equilibrium is a good approach
when simulating entrained-flow gasifiers in chemical process
simulators. Similar conclusions were made, by the same authors,
for downdraft fixed-bed gasifiers as long as high temperature and
gas residence time are achieved in the throat. In contrast, updraft
fixed-bed, dual fluidised-bed and stand-alone fluidised-bed
gasifiers should be modelled by revised equilibrium models or,
in some extreme cases, by detailed rate-flow models. In these
gasifiers, knowledge of biomass devolatilisation is crucial for a
successful prediction of performance. In dual fluidised-bed
gasifiers, char is converted in the combustor, while in fluidisedbed gasifiers, char conversion is rather limited and its prediction
seems to be a key parameter for proper prediction.
4.3. Aspen Plus gasification models
Fig. 5. Comparison between the experimental gas composition and the gas
composition predicted with the modified model. Data from the study of Li et al. [62].
Some authors, trying to avoid complex processes and develop
the simplest possible model that incorporates the principal
gasification reactions and the gross physical characteristics of
the reactor, have developed models using the process simulator
Aspen Plus [101]. Aspen Plus is a problem-oriented input program
that is used to facilitate the calculation of physical, chemical and
biological processes. It can be used to describe processes involving
solids in addition to vapour and liquid streams. Aspen Plus makes
model creation and updating easier, since small sections of
2848
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
complex and integrated systems can be created and tested as
separate modules before they are integrated. This process
simulator is equipped with a large property data bank containing
the various stream properties required to model the material
streams in a gasification plant, with an allowance for the addition
of in-house property data. Where more sophisticated block
abilities are required, they can be developed as FORTRAN
subroutines.
Aspen Plus has been used to simulate coal conversion;
examples include methanol synthesis [78,79], indirect coal
liquefaction processes [80], integrated coal gasification combined
cycle (IGCC) power plants [81], atmospheric fluidised-bed
combustor processes [82], compartment fluidised-bed coal gasifiers [83], coal hydrogasification processes [84] and coal gasification simulation [85]. It has also been used to model and simulate a
tyre pyrolysis unit within a gasification-based plant (Gómez et al.
[103]).
However, the work that has been done on biomass gasification
is less extensive. Mansaray et al. [86–88] used Aspen Plus to
simulate a dual-distributor-type fluidised-bed rice husk gasifier.
Two thermodynamic models were developed: a one-compartment
model, where the hydrodynamic complexity of the fluidised-bed
gasifier was neglected and an overall equilibrium approach was
used; and a two-compartment model, where the complex
hydrodynamic conditions presented within the gasification
chamber were taken into account. The models were capable of
predicting the reactor temperature, gas composition, gas higher
heating value, and overall carbon conversion under various
operating conditions, including bed height, fluidisation velocity,
equivalence ratio, oxygen concentration in the fluidising gas, and
rice husk moisture content. Because of the large amount of volatile
material in biomass and the complexity of biomass reaction rate
kinetics in fluidised-beds, the authors ignored char gasification and
simulated the gasification process by assuming that biomass
gasification follows the Gibbs equilibrium. The reactions considered in the development of the model were pyrolysis, partial
combustion and gasification. Predictions of the core, annulus and
exit temperatures, as well as the mole fractions of the combustible
gas components and product-gas higher heating value, agreed
reasonably well with experimental data. Correlations of the overall
carbon conversion were not very good. The discrepancies between
experimental and predicted overall carbon conversions were
attributed to uncertainties in the sampling procedure.
Mathieu and Dubuisson [68] modelled wood gasification in a
fluidised bed using Aspen Plus. The model was based on the
minimisation of the Gibbs free energy and the process was
uncoupled in pyrolysis, combustion, Boudouard reaction and
gasification. The authors performed a sensitivity analysis and
concluded that there is a critical air temperature above which
preheating is no longer efficient, that there is an optimum oxygen
factor, that the oxygen enrichment of air plays an efficient role
under a certain value, and that the operating pressure has only a
slight positive effect on process efficiency.
Mitta et al. [89] modelled a fluidised-bed tyre gasification plant
with air and steam using Aspen Plus and validated their results
with the gasification pilot plant located at the Chemical
Engineering Department of the Technical University of Catalonia
(UPC). Their gasification model was divided into three different
stages: drying, devolatilisation-pyrolysis and gasification-combustion. Fig. 6 shows the Aspen Plus flowsheet of the model. When
the raw material is fed, the first step is the heating and drying of the
particles. A ‘‘RSTOIC’’ module was used to model this instantaneous
drying. Due to the high content of volatiles in the tyre, the authors
considered the devolatilisation step of its conversion. This
devolatilisation process, namely the fast pyrolysis mechanism,
produces volatile gases, tars and char. The ‘‘RYIELD’’ block was used
to model the pyrolysis/devolatilisation part of the model. It was
assumed that the total yield of volatiles equals the volatile content
of the parent fuel determined by the proximate analysis. The
‘‘RGIBBS’’ reactor module was used to model the gasification and
combustion reaction. The stream from the ‘‘RYIELD’’ block and the
preheated oxygen and steam were directed into the ‘‘RGIBBS’’
module, which can predict the equilibrium composition of the
produced gas from ‘‘RYIELD’’ at a specified temperature and
pressure. The ash from the gasification process was removed from
the ‘‘RGIBBS’’ module. In the model, an overall equilibrium
approach was employed by neglecting the hydrodynamic complexity of the gasifier. Although higher hydrocarbons, tars and oils
were produced in the gasifier, they were considered nonequilibrium products in order to decrease the complexity of the
model. Therefore, CH4 was the only hydrocarbon taken into
consideration in the calculation. All of the results of the model
were normalised to make them free from tars. The sulphur in the
tyre was assumed to be converted mainly into H2S. Steady-state
conditions were assumed. The model was able to predict the
composition of the produced gas under various working conditions, including the flow rate, composition and temperature of the
feed materials, as well as the operating pressure and temperature.
Nikoo and Mahinpey [90] developed a model capable of
predicting the steady-state performance of an atmospheric
fluidised-bed gasifier by considering the hydrodynamic and
reaction kinetics simultaneously. They used four Aspen Plus
Fig. 6. Simulation diagram in Aspen Plus for a fluidised-bed tyre gasification process [89].
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
2849
presented an approximate gasifier model that can be used for
dynamic analysis using Aspen Dynamics. They used a highmolecular-weight hydrocarbon that is present in the Aspen library
as a pseudofuel and the proposed approximate model captured the
essential macroscale thermal, flow, composition and pressure
dynamics. Doherty et al. [94,95] developed a model for a
circulating fluidised bed and studied the effect of varying the
equivalence ratio, temperature, level of air preheating, biomass
moisture and steam injection on the product gas composition, the
gas heating value and the cold gas efficiency. Van der Meijden et al.
[96] also used Aspen Plus as a modelling tool to quantify the
differences in overall process efficiency for producing synthetic
natural gas in three different gasifiers (entrained-flow, allothermal
and circulating fluidised-bed gasifier).
4.4. Neural network gasification models
Fig. 7. Simulation diagram in Aspen Plus for an atmospheric fluidised-bed
gasification [90].
reactor models and external FORTRAN subroutines for hydrodynamics and kinetics nested to simulate the gasification process
(Fig. 7). The Aspen Plus yield reactor, ‘‘RYIELD’’, was used to
simulate the decomposition of the feed. In this step, biomass was
converted into its constituting components, including carbon,
hydrogen, oxygen, sulphur, nitrogen and ash, by specifying the
yield distribution according to the biomass ultimate analysis. A
separation column model was used to separate the volatile
materials and solids in order to perform the volatile reactions. The
Aspen Plus Gibbs reactor, ‘‘RGIBBS’’, was used for volatile
combustion, in conformity with the assumption that volatile
reactions follow the Gibbs equilibrium. The Aspen Plus CSTR
reactor, ‘‘RCSTR’’, performed char gasification using reaction
kinetics, written as an external FORTRAN code. The hydrodynamic
parameters divided the reactor into two regions, bed and
freeboard, and each region was simulated by one ‘‘RCSTR’’. The
authors validated their model using different sets of operating
conditions for a lab-scale pine gasifier with air and steam. They
found good qualitative agreement between the model’s prediction
and the experimental data, but they considered further improvements to the model, such as implementing tar production by
defining non-equilibrium products in the ‘‘RGIBBS’’ reactor, as well
as parameters considering mass transfer inside solid particles and
heat transfer inside the particles, between phases, and between the
material and the wall. However, the authors observed that the
production of hydrogen increases with temperature, thereby
enhancing carbon conversion efficiency. The equivalence ratio is
directly proportional to CO2 production and carbon conversion
efficiency. Increasing the steam-to-biomass ratio increases hydrogen and carbon monoxide production and decreases CO2 and
carbon conversion efficiency. The average particle size, ranging
from 0.25 to 0.75 mm, does not seem to contribute significantly to
the composition of the product gases.
Other authors have worked with Aspen Plus to model the
gasification process for coal and biomass. Yan and Rudolph [83]
developed a model for a compartmented fluidised-bed coal gasifier
process, Sudiro et al. [91] modelled the gasification process to
obtain synthetic natural gas from petcoke. Paviet et al. [92]
describe a very simple two-step model of chemical equilibrium in
the wood biomass gasification process. Robinson and Luyben [93]
Non-mechanistic, non-equilibrium modelling using neural
networks for biomass gasification has also been reported
[97,98]. Artificial neural networks (ANN) have been extensively
used in the fields of pattern recognition, signal processing, function
approximation and process simulation. Sometimes a hybrid neural
network (HNN) model is synthesised for process modelling [99].
This modelling approach usually combines a partial first-principles
model, which describes certain characteristics of the process being
simulated and involves a multilayer feedforward neural network
(MFNN) that serves as an estimator of unmeasured process
parameters that are difficult to model from first principles. MFNN
is a universal function approximator, which has the ability to
approximate any continuous function to an arbitrary precision
even without a priori knowledge of the structure of the function to
be approximated [100]. In Guo et al. [97], a HNN model was
developed to simulate biomass gasification in a steam fluidisedbed gasifier. A series of gasification runs were conducted on a
bench-scale facility, with four types of biomass as feedstock. The
data obtained from these experiments were used to train the HNN
model.
Brown et al. [98] developed an equilibrium reaction modelling
approach for the efficient design of biomass gasifiers. Fuels and
chars were defined as pseudospecies with properties derived from
their ultimate analyses; tars were defined as a subset of known
molecular species and their distribution was determined by
equilibrium calculations. Non-equilibrium behaviour for gas, tar,
and char formation was explained by reaction temperature
differences for a complete set of stoichiometric equations. A
nonlinear regression, with an artificial neural network, related
changes in temperature differences to fuel composition and
operational variables. This first-principles approach, illustrated
with fluidised-bed reactor data, improves the accuracy of
equilibrium calculations and reduces the amount of data required
by preventing the NN from learning atomic and heat balances.
5. Conclusions
Models of several different types have been developed for
gasification systems—kinetic, equilibrium and artificial neural
networks. Unlike kinetic models that predict the progress and
product composition at different positions along a reactor, an
equilibrium model predicts the maximum achievable yield of a
desired product from a reacting system. It also provides a useful
design aid in evaluating the possible limiting behaviour of a
complex reacting system that is difficult or unsafe to reproduce
experimentally or in commercial operation. Equilibrium models
are less computationally intensive than kinetic models and they
are a useful tool for preliminary comparison. However, they cannot
give highly accurate results for all cases.
2850
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
Some authors, trying to avoid complex processes and aiming to
develop the simplest possible model that incorporates the
principal gasification reactions and the gross physical characteristics of the reactor, have developed models using the process
simulator Aspen Plus.
References
[1] Maniatis K, Guiu G, Riesgo J. The European commission perspective in
biomass and waste thermochemical conversion. In: Bridgwater AV, editor.
Pyrolysis and gasification of biomass and waste, Proceedings of an Expert
Meeting. 2002. p. 1–18.
[2] Lin Y, Tanaka S. Ethanol fermentation from biomass resources: current state
and prospects. Appl Microbiol Biotechnol 2006;69:627–42.
[3] Caputo AC, Palumbo M, Pelagagge PM, Scacchia F. Economics of biomass
energy utilization in combustion and gasification plants: effects of logistic
variables. Biomass Bioenergy 2005;28:35–51.
[4] Yoshioka T, Hirata S, Matsumura Y, Sakanishi KW. Biomass resources and
conversion in Japan: the current situation and projections to 2010 and 2050.
Biomass Bioenergy 2005;29:336–46.
[5] Faaij APC. Bio-energy in Europe: changing technology choices. Energy Policy
2006;34:322–42.
[6] McKendry P. Energy production from biomass (part 1): overview of biomass.
Bioresour Technol 2002;83:37–46.
[7] McKendry P. Energy production from biomass (part 3): gasification technologies. Bioresour Technol 2002;83:55–63.
[8] Li X. Biomass gasification in circulating fluidized bed. PhD dissertation.
Vancouver, Canada: University of British Columbia;2002.
[9] Kishore VVN, editor. Renewable energy engineering & technology: a knowledge compendium. New Delhi: TERI Press; 2008.
[10] Rauch R. Biomass gasification to produce synthesis gas for fuels and chemicals, report made for IEA Bioenergy Agreement, Task 33: Thermal Gasification
of Biomass;2003.
[11] Carlos L. High temperature air/steam gasification of biomass in an updraft
fixed bed type gasifier. PhD thesis. Stockholm, Sweden: Royal Institute of
Technology, Energy Furnace and Technology;2005.
[12] Reed TB, Das A. Handbook of biomass downdraft gasifier engine systems.
Colorado: Solar Energy Research Institute; 1988.
[13] Knoef HAM. Handbook biomass gasification. Meppel, The Nederlands: BTG
Biomass Technology Group B.V.; 2005.
[14] Knoef HAM. Inventory of biomass gasifier manufacturers and installations.
Final Report to European. Commission, Contract DIS/1734/98-NL, Biomass
Technology Group BV, University of Twente, Enschede; 2000.
[15] Basu P. Combustion and gasification in fluidized beds. London: Taylor &
Francis Group/CRC Press; 2006.
[16] Stassen HEM, Knoef HAM. Small scale gasification systems. The Netherlands:
Biomass Technology Group, University of Twente; 1993.
[17] Gil J, Corella J, Aznar MP, Caballero MA. Biomass gasification in atmospheric
and bubbling fluidized bed: effect of the type of gasifying agent on the
product distribution. Biomass Bioenergy 1999;17:389–403.
[18] Rapagna S, Jand N, Kiennemann A, Foscolo PU. Steam gasification of biomass
in a fluidised-bed of olivine particles. Biomass Bioenergy 2000;19:187–97.
[19] Gabra M, Pettersson E, Backman R, Kjellstrom B. Evaluation of cyclone gasifier
performance for gasification of sugar cane residue—part 1: gasification of
bagasse. Biomass Bioenergy 2001;21:351–69.
[20] Zainal ZA, Rifau A, Quadir GA, Seetharamu KN. Experimental investigation of
a downdraft biomass gasifier. Biomass Bioenergy 2002;23:283–9.
[21] Devi L, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar
elimination in biomass gasification processes. Biomass Bioenergy
2003;24:125–40.
[22] Garcia L, Salvador ML, Arauzo J, Bilbao R. CO2 as a gasifying agent for gas
production from pine sawdust at low temperature using Ni/Al coprecipitated
catalyst. Fuel Process Technol 2001;69:157–74.
[23] Ollero P, Serrera A, Arjona R, Alcantarilla S. The CO2 gasification kinetics of
olive residue. Biomass Bioenergy 2003;24:151–61.
[24] Hofbauer H, Fleck T, Veronik G, Rauch R, Mackinger H, Fercher E. The FICFBgasification process. In: Bridgwater AV, Boocock DGB, editors. Developments in thermochemical biomass conversion. London: Blackie; 1997 . p.
1016–25.
[25] Pletka R, Brown RC, Smeenk J. Indirectly heated biomass gasification using a
latent heat ballast—part 1: experimental evaluations. Biomass Bioenergy
2001;20:297–305.
[26] Pletka R, Brown RC, Smeenk J. Indirectly heated biomass gasification using a
latent heat ballast—part 2: modeling. Biomass Bioenergy 2001;20:307–15.
[27] Cummera K, Brown RC. Indirectly heated biomass gasification using a latentheat ballast—part 3: refinement of the heat transfer model. Biomass Bioenergy 2005;28:321–30.
[28] Hanaoka T, Inoue S, Uno S, Ogi T, Minowa T. Effect of woody biomass
components on air–steam gasification. Biomass Bioenergy 2005;28:69–76.
[29] Lv PM, Xiong ZH, Chang J, Wu CZ, Chen Y, Zhu JX. An experimental study on
biomass air–steam gasification in a fluidized bed. Bioresour Technol
2004;95:95–101.
[30] Lucas C, Szewczyk D, Blasiak W, Mochida S. High-temperature air and steam
gasification of densified biofuels. Biomass Bioenergy 2004;27:563–75.
[31] Prins MJ, Ptasinski KJ, Janssen FJJG. From coal to biomass gasification:
comparison of thermodynamic efficiency. Energy 2007;32:1248–59.
[32] Prins MJ, Ptasinski KJ, Janssen FJJG. More efficient biomass gasification via
torrefaction. Energy 2006;31:3458–70.
[33] Ptasinski KJ, Prins MJ, Pierik A. Exergetic evaluation of biomass gasification.
Energy 2007;32(4):568–74.
[34] Ptasinski KJ. Thermodynamic efficiency of biomass gasification and biofuels
conversion. Biofuels Bioprod Bioref 2008;2:239–53.
[35] Desrosiers R. Thermodynamics of gas–char reactions. In: Reed TB, editor. A
survey of biomass gasification. Colorado: Solar Energy Research Institute;
1979.
[36] Double JM, Bridgwater AV. Sensitivity of theoretical gasifier performance to
system parameters. In: PalzW, Coombs J, Hall DO, editors. Energy from
biomass, Proceedings of the third EC conference, 1985.
[37] Prins MJ, Ptasinski KJ, Janssen FJJG. Thermodynamics of gas–char reactions:
first and second law analysis. Chem Eng Sci 2003;58:1003–11.
[38] Sharma AK. Equilibrium and kinetic modelling of char reduction reactions in
a downdraft biomass gasifier: a comparison. Solar Energy 2008;52:918–28.
[39] Wang Y, Kinoshita CM. Kinetic model of biomass gasification. Solar Energy
1993;51(1):19–25.
[40] Fiaschi D, Michelini M. A two-phase one-dimensional biomass gasification
kinetic model. Biomass Bioenergy 2001;21:121–32.
[41] Giltrap DL, McKibbin R, Barnes GRG. A steady state model of gas–char
reactions in a downdraft gasifier. Solar Energy 2003;74:85–91.
[42] Yang YB, Yamauchi H, Nasserzadeh V, Swithenbank J. Effect of fuel devolatilization on the combustion of wood chips and incineration of simulated
municipal wastes in packed bed. Fuel 2003;82:2205–21.
[43] Roshmi A, Murthy J, Hajaligol M. Modeling of smoldering process in porous
biomass fuel rod. Fuel 2004;83:1527–36.
[44] Dennis JS, Lambert RS, Milne AJ, Scott SA, Hayhurst AN. The kinetics of
combustion of chars derived from sewage sludge. Fuel 2005;84:117–26.
[45] Radmanesh R, Chaouki J, Guy C. Biomass gasification in a bubbling fluidised
bed reactor: experiments and modeling. AIChe J 2006;52:4258–72.
[46] Gobel B, Henriksen U, Jensen TK, Qvale B, Houbak N. The development of a
computer model for a fixed bed gasifier an its use for optimization an control.
Bioresour Technol 2007;98:2043–52.
[47] Fermoso J, Arias B, Pevida C, Plaza MG, Rubiera F, Pis JJ. Kinetic models
comparison for steam gasification of different nature fuel chars. J Therm Anal
Calorim 2008;91:779–86.
[48] Zhong LD, Mei WH, Hong Z. Kinetic model establishment and verification of
the biomass gasification fluidised bed. In: Proceedings of the 8th International Conference on Machine Learning and Cybernetics; 2009.
[49] Roy PC, Datta A, Chakraborty N. Modelling of a downdraft biomass gasifier with
finite rate kinetics in the reduction zone. Int J Energy Res 2009;33:833–51.
[50] Senelwa K. The air gasification of woody biomass from short rotation forests.
PhD Thesis. New Zealand: Massey University;1997.
[51] Jayah TH, Aye L, Fuller RJ, Stewart DF. Computer simulation of a downdraft
wood gasifier for tea drying. Biomass Bioenergy 2003;25:459–69.
[52] Chen JS. Kinetic engineering modelling of co-current moving bed gasification
reactors for carbonaceous material. PhD thesis. New York: Cornell University;
1987.
[53] Milligan JB. PhD thesis. Birmingham, UK: Aston University;1994.
[54] Babu BV, Sheth PN. Modeling and simulation of reduction zone of downdraft
biomass gasifier: effect of char reactivity factor. Energy Convers Manage
2006;47(15–16):2602–11.
[55] Bacon DW, Downie J, Hsu JC, Peters J. Modeling of fluidized bed gasifiers. In:
Milne TA, Mudge KL, editors. Fundamentals of thermochemical biomass
conversion. London, UK: Elsevier Applied Science; 1985. p. 717–32.
[56] Double JM, Smith EL, Bridgwater AV. Computer modeling of fluidized bed
gasifier. In: Ferrero GL, Maniatis K, Buekens A, editors. Process and gasification. London, UK: Elsevier Applied Science; 1989. p. 651–5.
[57] Ruggiero M, Manfrida G. An equilibrium model for biomass gasification
processes. Renew Energy 1999;16:1106–9.
[58] Zainal ZA, Ali R, Lean CH, Seetharamu KN. Prediction of performance of a
downdraft gasifier using equilibrium modeling for different biomass materials. Energy Convers Manage 2001;42:1499–515.
[59] Schuster G, Löffler G, Weigl K, Hofbauer H. Biomass steam gasification: an
extensive parametric modeling study. Bioresour Technol 2001;77:71–9.
[60] Altafini CR, Wander PR, Barreto RM. Prediction of the working parameters of a
wood waste gasifier through an equilibrium model. Energy Convers Manage
2003;44:2763–77.
[61] Li X, Grace JR, Watkinson AP, Lim CJ, Ergüdenler A. Equilibrium modeling of
gasification: a free energy minimization approach and its application to
circulating fluidized bed coal gasifier. Fuel 2001;80:195–207.
[62] Li XT, Grace JR, Lim CJ, Watkinson AP, Chen HP, Kim JR. Biomass gasification in
a circulating fluidized bed. Biomass Bioenergy 2004;26:171–93.
[63] Melgar A, Pérez JF, Laget H, Hornillo A. Thermochemical equilibrium modelling of gasifying process. Energy Convers Manage 2007;48:59–67.
[64] Yoshida H, Kiyon F, Tajima H, Yamasaki A, Ogasawara K, Masuyama T. Twostage equilibrium model for a coal gasifier to predict the accurate carbon
conversion in hydrogen production. Fuel 2008;87:2186–93.
[65] Karamarkovic R, Karamarkovic V. Exergy and energy analysis of biomass
gasification at different temperatures. Energy 2009. doi: 10.1016/j.energy.
2009.10.022.
[66] Huang H-J, Ramaswamy S. Modeling biomass gasification using thermodynamic equilibrium approach. Appl Biochem Biotechnol 2009;154:193–204.
M. Puig-Arnavat et al. / Renewable and Sustainable Energy Reviews 14 (2010) 2841–2851
[67] Haryanto A, Fernando SD, Pordesimo LO, Adhikarid S. Upgrading of syngas
derived from biomass gasification: a thermodynamic analysis. Biomass
Bioenergy 2009;33:882–9.
[68] Mathieu P, Dubuisson R. Performance analysis of a biomass gasifier. Energy
Convers Manage 2002;43:1291–9.
[69] Daubert TE. Chemical engineering thermodynamics. Singapore: McGraw-Hill
book company; 1987.
[70] Smith WR, Missen RW. Chemical reaction equilibrium analysis: theory and
algorithms. New York, USA: Wiley; 1982.
[71] Jarungthammachote S, Dutta A. Thermodynamic equilibrium model and
second law analysis of a downdraft waste gasifier. Energy 2007;32:1660–9.
[72] Villanueva AL, Gomez-Barea A, Revuelta E, Campoy M, Ollero P. Guidelines for
selection of gasifiers modelling strategies. In: 16th European biomass conference and exhibition; 2008.
[73] Jarungthammachote S, Dutta A. Equilibrium modeling of gasification: Gibbs
free energy minimisation approach and its application to spouted bed and
spout-fluid bed gasifiers. Energy Convers Manage 2008;49:1345–56.
[74] Grace JR, Li XT, Lim CJ. Equilibrium modelling of catalytic steam reforming of
methane in membrane reactors with oxygen addition. Catal Today
2001;64:141–9.
[75] Gumz W. Gas producers and blast furnaces. New York: Wiley; 1950.
[76] Bacon DW, Downie J, Hsu JC, Peters J. Modeling of fluidized bed wood
gasifiers. In: Overend RP, Milne TA, Mudge LK, editors. Fundamentals of
thermochemical biomass conversion. London: Elsevier; 1982. p. 717–32.
[77] Kersten SRA, Prins W, Van der Drift A, Van Swaaij WPM. Interpretation of
biomass gasification by ‘‘quasi’’-equilibrium models. In: Proceedings of the
Twelfth European conference on biomass for energy, industry and climate
protection; 2002.
[78] Kundsen RA, Bailey T, Fabiano LA. Experience with ASPEN while simulating a
new methanol plant. AIChE Symp Ser 1982;78:214.
[79] Schwint KT. Great plains ASPEN model development, methanol synthesis
flowsheet. Final topical report. USA: Scientific Design Co., Inc.;1985.
[80] Barker RE. ASPEN modeling of the tri-state indirect-liquefaction process. Oak
Ridge, USA: Oak Ridge National Laboratory; 1983.
[81] Phillips JN, Erbes MR, Eustis RH. Study of the off-design performance of
integrated coal gasification. In: Combined cycle power plants, computeraided engineering of energy systems, vol. 2—analysis and simulation, Proceedings of the winter annual meeting of the American Society of Mechanical
Engineers; 1986.
[82] Douglas PL, Young BE. Modelling and simulation of an AFBC steam heating
plant using ASPEN/SP. Fuel 1990;70:145–54.
[83] Yan HM, Rudolph V. Modeling a compartmented fluidized bed coal gasifier
process using ASPEN PLUS. Chem Eng Commun 2000;183:1–38.
[84] Backham L, Croiset E, Douglas PL. Simulation of a coal hydrogasification
process with integrated CO2 capture. Combust Canada 2003;3A(4).
[85] Lee HG, Chung KM, Kim C, Han SH, Kim HT. Coal gasification simulation using
ASPEN PLUS. In: US–Korea joint workshop on coal utilization technology.
1992. p. 447–74.
[86] Mansaray KG, Al-Taweel AM, Ghaly AE, Hamdullahpur F, Ugursal VI. Mathematical Modeling of a Fluidized Bed Rice Husk Gasifier: Part I - Model
Development, Energy Sources, Part A: Recovery. Utilization and Environmental Effects 2000;22:83–98.
[87] Mansaray KG, Ghaly AE, Al-Taweel AM, Hamdullahpur F, Ugursal VI. Mathematical Modeling of a Fluidized Bed Rice Husk Gasifier: Part II - Model
Sensitivity, Energy Sources, Part A: Recovery. Utilization and Environmental
Effects 2000;22:167–85.
[88] Mansaray KG, Ghaly AE, Al-Taweel AM, Ugursal VI, Hamdullahpur F. Mathematical Modeling of a Fluidized Bed Rice Husk Gasifier: Part III - Model
Verification, Energy Sources, Part A: Recovery. Utilization and Environmental
Effects 2000;22:281–96.
2851
[89] Mitta NR, Ferrer-Nadal S, Lazovic AM, Perales JF, Velo E, Puigjaner L. Modelling and simulation of a tyre gasification plant for synthesis gas production.
In: Proceedings of 16th European symposium on computed aided process
engineering and 9th international symposium on process systems engineering; 2006. p. 1771–6.
[90] Nikoo MB, Mahinpey N. Simulation of biomass gasification in fluidized bed
reactor using ASPEN PLUS. Biomass Bioenergy 2008. doi: 10.1016/j.biombioe.2008.02.020.
[91] Sudiro M, Zanella C, Bressan L, Fontana M, Bertucco A. Synthetic natural gas
(SNG) from petcoke: model development and simulation. In: Proceedings of
the 9th International Conference on Chemical and Process Engineering
(ICheaP-9); 2009.
[92] Paviet F, Chazarenc F, Tazerout M. Thermochemical equilibrium modelling of
a biomass gasifying process using Aspen Plus. Int J Chem React Eng
2009;7:A40.
[93] Robinson PJ, Luyben WL. Simple dynamic gasifier model that runs in Aspen
dynamics. Ind Eng Chem Res 2008;47:7784–92.
[94] Doherty W, Reynolds A, Kennedy D. Simulation of a circulating fluidised bed
biomass gasifier using ASPEN Plus—a performance analysis. In: Ziebik A,
Kolenda Z, Stanek W, editors. Proceedings of the 21st international conference on efficiency, cost, optimization, simulation and environmental impact
of energy systems, 2008.
[95] Doherty W, Reynolds A, Kennedy D. The effect of air preheating in a biomass CFB
gasifier using Aspen Plus simulation. Biomass Bioenergy 2009;33:1158–67.
[96] Van der Meijden CM, Veringa HJ, Rabou LPLM. The production of synthetic
natural gas (SNG): a comparison of three wood gasification systems for
energy balance and overall efficiency. Biomass Bioenergy 2009. doi:
10.1016/j.biombioe.2009.11.001.
[97] Guo B, Li D, Cheng C, Lu Z, Shen Y. Simulation of biomass gasification with a
hybrid neural network model. Bioresour Technol 2001;76(2):77–83.
[98] Brown D, Fuchino T, Marechal F. Solid fuel decomposition modelling for the
design of biomass gasification systems. In: Marquardt W, Pantelides C,
editors. Proceedings of the 16th European symposium on computer aided
process engineering and 9th international symposium on process systems
engineering. 2006. p. 1661–6.
[99] Psichogios DC, Ungar LH. A hybrid neural network-first principles approach
to process modelling. AIChE J 1992;38:1499–511.
[100] Hornik K. Approximation capabilities of multilayer feedforward networks.
Neural Networks 1991;4:251–7.
[101] Aspen Plus. Cambridge, MA, USA: Aspen Technology, Inc., 2006.
[102] Di Blassi. Dynamic behaviour of stratified downdraft gasifiers. Chem Eng Sci
2000;55:2931–44.
[103] Gomez CJ, Mitta NR, Velo E, Puigjaner L. Modelling and simulation of a biomass
pyrolysis unit within a gasification-based plant. In: Proceedings of the 17th
European symposium on computer aided process engineering; 2007.
[105] Chee CS. The air gasification of wood chips in a gasifier bed. MSc Thesis.
Kansas State University;1987.
[106] Bridgwater AV. The technical and economic feasibility of biomass gasification
for power generation. Fuel 1995;74:631–53.
[107] Beenackers AACM. Biomass gasification in moving beds. A review of European technologies. Renew Energy 1999;16:1180–6.
[108] Narváez I, Orio A, Corella J, Aznar MP. Biomass gasification with air in a
bubbling fluidized bed. Effect of six operational variables on the quality of the
produced raw gas. Ind Eng Chem Res 1996;35:2110–20.
[109] Herguido J, Corella J, González-Saiz J. Steam gasification of lignocellulosic
residues in a fluidized bed at small pilot scale. Effect of the type of feedstock.
Ind Eng Chem Res 1992;31:1274–82.
[110] Gil J, Aznar MP, Caballero MA, Francés E, Corella J. Biomass gasification in
fluidized bed at pilot scale with steam-oxygen mixtures. Product distribution
for very different operating conditions. Energy Fuels 1997;11:1109–18.