Article

pubs.acs.org/EF

Modified Thermodynamic Equilibrium Model for Biomass

Gasification: A Study of the Influence of Operating Conditions
Maria Puig-Arnavat, Juan Carlos Bruno,* and Alberto Coronas
Department of Mechanical Engineering, Universitat Rovira i Virgili, Avinguda Països Catalans, 26, 43007 Tarragona, Spain

ABSTRACT: This paper presents a mathematical model for biomass gasification processes developed in the equation solver
program Engineering Equation Solver (EES) with an implemented user-friendly interface. It is based on thermodynamic
equilibrium calculations and includes some modifications to be adapted to a real process, in which only a partial approach to
chemical equilibrium is achieved. The model can be used to predict the producer gas composition, yield, and heating value for a
certain biomass with a specific ultimate composition and moisture content. It has been validated with published experimental
data from different authors for downdraft, fluidized-bed gasifiers and different biomasses, showing good agreement between
reported data and modeled values. In addition, it has been used to evaluate the influence of different operating parameters
[equivalence ratio (ER), air preheating, steam injection, and oxygen enrichment] on producer gas. The model predicts the
behavior of different kinds of biomass and becomes a useful tool to simulate the biomass gasification process by allowing its
integration in complete energy supply systems, such as co-generation plants.

1. INTRODUCTION AND OBJECTIVES the sensitivity of the distribution of the product to the
Nowadays, the awareness and concern about the depletion of operating conditions, called for the development of mathemat-
fossil fuels, energy dependency, and global climate change have ical models. The main objectives of these models are to study
called for the development and research on reliable, affordable, the thermochemical processes during the gasification of the
and clean-energy sources. In this context, modern use of biomass and evaluate the influence of the main input variables
biomass is considered a very promising clean-energy option for on the producer gas composition and calorific value.
reducing energy dependency and greenhouse gas emissions; Different kinds of models have been developed for
biomass is considered to be CO2-neutral. In addition, it is the gasification systems, including equilibrium, kinetic, and artificial
only renewable energy source that can directly replace fossil neural networks. A detailed review of recent biomass
fuels because it is widely available and allows for continuous gasification models is available elsewhere,2,3 and only a brief
power generation and synthesis of different fuels and chemicals. description is therefore given here. Equilibrium models predict
Gasification is a highly efficient and clean conversion process the maximum achievable yield of a desired product from a
that converts different feedstocks to a wide variety of products reacting system, while kinetic models predict the progress and
for various applications. In comparison to combustion, product composition at different positions in a reactor.2
gasification has higher efficiencies in power production and a Equilibrium models are less computationally intensive than
more efficient and better controlled heating.1 Gasification can kinetic models, and they are a useful tool for preliminary
be considered in advanced applications in developed countries comparison. However, they cannot give highly accurate results
and also for rural electrification in isolated installations or in in all cases. Equilibrium models usually overestimate the yields
developing countries. of H2 and CO, underestimate those of CO2, and predict an
The gasification conversion process can be defined as a outlet stream free from CH4, tars, and char.3 Equilibrium
partial thermal oxidation, which results in a great proportion of models are considered a good approach when simulating
gaseous products (carbon dioxide, hydrogen, carbon monoxide, entrained-flow gasifiers in chemical process simulators or for
water, and other gaseous hydrocarbons) and little quantities of downdraft fixed-bed gasifiers, as long as high temperature and
char, ash, and several condensable compounds (tars and oils). gas residence time are achieved in the throat. In contrast,
Air, steam, or oxygen can be supplied to the reaction as updraft fixed-bed, dual fluidized-bed, and stand-alone fluidized-
gasifying agents. The quality of gas produced varies according bed gasifiers should be modeled by revised equilibrium models
to the gasifying agent used and operating conditions. The gas or, in some extreme cases, by detailed rate-flow models.4
obtained covers a wide range of calorific values (CVs): low CVs The objective of this paper is to develop a simple but
(4−6 MJ/m3) result from the use of air as the gasifying agent, rigorous gasification model for the design and simulation of a
and medium or high CVs (12−18 or 40 MJ/m3) result when biomass gasification plant. This model, on the basis of
steam or oxygen is used. thermodynamic equilibrium calculations, includes some mod-
The operation of a biomass gasifier depends upon several ifications for adaptation to real processes, in which only a
complex chemical reactions, including several steps, such as
pyrolysis, thermal cracking of vapors to gas and char, Received: November 16, 2011
gasification of char, and partial oxidation of combustible gas, Revised: January 16, 2012
vapors, and char. These complicated processes, coupled with Published: January 18, 2012

© 2012 American Chemical Society 1385 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394
Energy & Fuels Article

Figure 1. Screenshot of the modified equilibrium model, developed in EES software, including feed and product streams entering and leaving the
different units considered. The numerical values of variables in a box are inputs for the model.

partial approach to chemical equilibrium is achieved. The constant can be defined as the true equilibrium constant multiplied by
model developed, which has been validated with experimental the degree of approach to equilibrium. In calibrating the model by
published data of other authors, provides the opportunity to Jayah et al.,9 the amount of methane predicted was adjusted to be
evaluate different gasification processes as well as variations in equal to the amount of methane measured in the product gas.
Jarungthammachote and Dutta6 used experimental data from other
fuel and operating conditions. authors5,9,10 to modify their model. They calculated two coefficients
for correcting the equilibrium constant of the water−gas shift reaction
2. MATHEMATICAL MODEL and methane formation reaction that improved the accuracy. The
The model presented in this paper is a modified equilibrium model coefficients were obtained from the average value of the ratio of
based on equilibrium constants, while the process is considered experimental data and calculated data from their model, for CH4 and
stationary. CO. Other authors11 used the equilibrium model to predict the
First, a pure thermodynamic equilibrium model was developed producer-gas compositions, product heating value, and cold gas
following the procedure described elsewhere.5,6 This pure equilibrium efficiency for circulating fluidized-bed (CFB) gasification. To correct
model is based on mass and energy balances together with chemical the deviations that they found between a real gasification process and
equilibrium in the gas phase, using the water−gas shift reaction (eq 3) chemical equilibrium, they developed a phenomenological model to
that results from the combination of Boudouard (eq 1) and water−gas modify the equilibrium to account for important non-equilibrium
(eq 2) reactions5,6 and the methane formation reaction (eq 4). factors. As they knew from a pilot-plant study of the experimental
C + CO2 ↔ 2CO (1) carbon conversion, they applied empirical parameters to modify the
carbon conversion.
C + H2O ↔ CO + H2 (2) In this study, the modifications to the pure equilibrium model
essentially consist of (1) adding a pyrolysis unit that uses correlations
CO + H2O ↔ CO2 + H2 (3) to predict the formation of gas, char, and volatiles in this step of the
gasification process, (2) considering heat losses in pyrolysis and
C + 2H2 ↔ CH 4 (4) gasification units (these heat losses are estimated by the user as a
The chemical formula of feedstock was defined as CHxOyNz, and it can percentage of biomass energy input to the system), (3) adding tar and
be calculated from the ultimate analysis of the biomass and the mass char leaving the gasifier as a percentage of tar and char produced in the
fractions of the carbon, hydrogen, oxygen, and nitrogen. Some pyrolysis unit added, (4) considering particles leaving the gasifier and
assumptions were necessarily made,5,7 and the products leaving the set by the user as mg N−1 m−3 in the producer gas, and (5) setting the
gasifier were considered all gaseous (CO, CO2, H2, CH4, N2, and amount of CH4 produced (for this reason, the equilibrium constant for
H2O). the methane reaction is not taken into account).
The main problems with this pure equilibrium model, as mentioned This modified equilibrium model is built into the equation solver
by other authors,3,6 is the overprediction of H2 and CO yields and program “Engineering Equation Solver (EES)”.12 EES has been found
underestimation of CO2. In addition, it predicts an exit stream free of to be very suitable for modeling this kind of system, because it
CH4, tars, and char. These differences between predicted and contains all of the necessary thermodynamic functions and it is
experimental data can be explained by the fact that a real gasification possible for the model builder to make a user interface, which can
system differs from an ideal reactor at chemical equilibrium. make the model user-friendly.
For this reason and to increase the accuracy of the results of the The gasification model is made up of a series of modules, each
pure thermodynamic equilibrium model, some modifications were containing one process (biomass drying, pyrolysis, gasification, air
introduced. Other authors also previously developed modified or preheating, and steam generation). In this gasification model, the user
pseudo-equilibrium models. Gumz8 stated that a modified equilibrium interface consists of a window, which contains drawings and tables

1386 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394

Energy & Fuels Article

with input and output values, diagrams, and hot areas with links to If the pyrolysis stage takes place in a fluidized bed, the correlations
other windows. For example, the pyrolysis unit is a hot area linked obtained by Gomez-Barea et al.19 will then be used. These correlations
with another window where the user can select the pyrolysis were selected because they are the only correlations for non-flash
correlations depending upon the gasifier design (downdraft or pyrolysis in a fluidized bed that could be found in the literature. The
fluidized bed). This way of presenting input and output variables model also gives the opportunity not to use any of these correlations
facilitates the user obtaining an overview of the operating conditions in and to introduce the desired yields manually. It must also be taken into
a certain computation. Figure 1 shows a screenshot of this modified account that only correlations for wood pyrolysis are considered.
equilibrium model developed in EES and including all feed, product However, it is also possible to extend the model by including pyrolysis
streams, and the different units considered. correlations for other kinds of biomass and agricultural residues.
This implemented modified equilibrium model enables work with The correlations used in this pyrolysis unit are as follows:
gasifying agents other than air. It is possible to use air, enriched air, or (1) Wood pyrolysis in a fluidized bed:19
oxygen alone or combined with steam.
Char and particles leaving the gasifier unit are considered to be gas yield (mass %, db)
composed primarily of carbon, and therefore, it was assumed to consist ⎛ Tp ⎞ ⎛ Tp ⎞2
solely of carbon in the model. Char-specific enthalpy is determined by = 311.10 − 351.45⎜ ⎟ + 121.43⎜ ⎟
the regression created from a data set based on the enthalpy of ⎝ Tref ⎠ ⎝ Tref ⎠ (9)
graphite13
char yield (mass %, db)
hchar (kJ/kg) = 0.0004T 2 + 0.8679T − 381.61 (5)
⎛ Tp ⎞ ⎛ Tp ⎞2
where T is the temperature measured in kelvin. = − 15.03 + 50.58⎜ ⎟ − 18.09⎜ ⎟
Tar-specific enthalpy is calculated using a correlation obtained by ⎝ Tref ⎠ ⎝ Tref ⎠ (10)
applying the Joback method, and it is assumed that pyrolysis tar only
consists of seven organic compounds: benzene, toluene, phenol, liquid yield (mass %, db)
guaiacol, methylguaiacol, ethylguaiacol, and isoeugenol14
⎛ Tp ⎞ ⎛ Tp ⎞2
h tar (kJ/kg) = − 4.659 × 10−7(T − 273.15)3 + 0.00193 = − 196.07 + 300.86⎜ ⎟ − 103.34⎜ ⎟
⎝ Tref ⎠ ⎝ Tref ⎠ (11)
2
(T − 273.15) + 0.131(T − 273.15) − 1796.4
⎛ Tp ⎞ ⎛ Tp ⎞2
(6) CO (vol %) = 240.53 − 225.12⎜ ⎟ + 67.50⎜ ⎟
where T is the temperature expressed in kelvin. ⎝ Tref ⎠ ⎝ Tref ⎠ (12)
Because not all carbon contained in biomass is converted into gas
species, it is necessary to define the concept of carbon conversion ⎛ Tp ⎞ ⎛ Tp ⎞2
efficiency (ηc) as CO2 (vol %) = −206.86 + 267.66⎜ ⎟ − 77.50⎜ ⎟
⎝ Tref ⎠ ⎝ Tref ⎠
ηc (%) = (total amount of carbon in the gas outlet stream) (13)
/(total amount of carbon in the biomass inlet stream) ⎛ Tp ⎞ ⎛ Tp ⎞2
× 100 (7) CH 4 (vol %) = −168.64 + 214.47⎜ ⎟ − 62.51⎜ ⎟
⎝ Tref ⎠ ⎝ Tref ⎠
Model parameters, such as pyrolysis temperature, percentage of
(14)
pyrolysis char and tar leaving the gasifier, and heat losses in the gasifier,
can be directly introduced by the user through the user interface, if the ⎛ Tp ⎞ ⎛ Tp ⎞2
information is known, or adjusted automatically. These parameters are H2 (vol %) = 234.97 − 257.01⎜ ⎟ + 72.50⎜ ⎟
automatically adjusted using the experimental and modeled output gas ⎝ Tref ⎠ ⎝ Tref ⎠ (15)
composition for each data set. The least-squares technique is used as
described in the equation below where Tp is the pyrolysis temperature (°C) and Tref = 500 °C.
(2) Correlations for conventional pyrolysis of wood in a fixed-bed
n m reactor obtained from experimental data:17
min ∑ ∑ (pi , j − wi , j)2
i=1 j=1 (8) gas yield (mass %, db)

where p is the yield of the gas species (CO, CO2, H2, and N2) = − 1.09 × 10−6Tp3 + 0.0022Tp2 − 1.392Tp + 288.534
calculated by the model and w is the corresponding experimental (16)
value. n is the number of data points. The minimization is carried out
by the variable metric method available in EES. tar yield (mass %, db)
2.1. Pyrolysis Unit. The main objective of this modeling unit is to
determine the yields of char, tar, and volatiles produced during = 1.33 × 10−6Tp3 − 0.0028Tp2 + 1.797Tp − 339.139 (17)
pyrolysis and to determine the composition of the light gas. For this
reason, experimental data from several authors15−19 have been studied char yield (mass %, db)
to obtain correlations for predicting these parameters as a function of
the pyrolysis temperature. = − 6.6 × 10−7Tp3 + 0.00137Tp2 − 0.93579Tp + 230.5279
After different experimental data for biomass pyrolysis are reviewed
and because different yields on products are obtained, two different (18)
correlations were considered when modeling this unit, depending
water yield (mass %, db)
upon the type of reactor and pyrolysis. Calculated correlations from
the experimental data by Fagbemi et al.17 are used to model the = 2.54 × 10−7Tp3 − 5.24 × 10−4Tp2 + 0, 335Tp − 40.883
pyrolysis stage in a fixed-bed reactor (downdraft, updraft, etc.). These
experimental data were selected because they cover a wide range of (19)
temperatures and are also in good agreement with the results obtained
CO (vol %) = 0.0371Tp + 19.961 (20)
by other authors.15,16

1387 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394

Energy & Fuels Article

CO2 (vol %) = 0.000143Tp2 − 0.27808Tp + 139.948 (21)

CH 4 (vol %) = −9 × 10−5Tp2 + 0.1221Tp − 25.206 (22)

H2 (vol %) = 0.04694Tp − 16.96286 (23)

where Tp is the pyrolysis temperature (°C).
In addition to these correlations, the energy, mass, and molar
balances for each element (C, H, O, and N) are set and used to
calculate pyrolysis products. The energy balance was formulated to
include an overall heat loss of the pyrolysis unit. This estimation of the
heat losses can be fixed by the user as a percentage of the product of
dry biomass mass flow entering the system (kg/h) and its lower
heating value (LHV) (kJ/kg).

3. RESULTS AND DISCUSSION

3.1. Validation of the Model with Experimental Data.
A previous literature review2 compared the theoretical results of
equilibrium models to experimental data, mainly for air-blown
downdraft gasifiers. However, in this study, we selected the
published experimental data for downdraft and fluidized-bed Figure 2. Comparison of predicted results from the modified
gasifiers operating with different biomass types and different equilibrium model to experimental data from Jayah et al.9 for a
gasifying agents. downdraft gasifier.
The error in the comparisons between predicted and
experimental values is estimated by the root-mean-square ature is adjusted at 500 °C, and no tar and char production is
(rms) value for each set of data, as given below considered. Heat losses of 5% of the energy input have been
taken into account. The comparison between the predicted
∑in (experimenti − model i)2 values and experimental data is presented in Figure 3. Figure 3
rms =
n (24)
where n is the number of data points.
3.1.1. Experimental Data from Downdraft Gasifiers.
Published experimental data from three different authors and
gasifiers has been selected.9,20,21 The data from Jayah et al.9 has
already been used by other authors6,21−23 to validate their
models. The experimental test rig used to collect data was an 80
kWth downdraft test gasifier with an inner reactor diameter of
0.92 m and a length of 1.15 m. Rubber wood was used as the
feed material for the study. To predict the results and because
of the lack of some information in the papers of authors, the
model parameters were adjusted by minimizing the sum of the
differences between the experimental and predicted results for
producer gas composition. In this case, the model accounts for
no biomass drying, air preheating, and heat losses. The
pyrolysis unit temperature has been adjusted to 440.5 °C,
and a percentage of char leaving the pyrolysis unit has been set
to match the amount of char leaving the gasifier and measured
by the authors (38%). The CH4 percentage leaving the gasifier
is given as an input to the model. The comparison between Figure 3. Comparison of predicted results from the modified
experimental and predicted data is shown in Figure 2. From this equilibrium model to experimental data from Erlich and Fransson20
figure, it can be concluded that the modified equilibrium model for a downdraft gasifier.
predicts with good accuracy the producer gas composition for
all gas species involved, although the differences for H2 are shows that the predicted results generally agree with the
slightly higher than for the others. The rms values obtained are experimental data. The rms values obtained are 1.1, 1.0, 1.4,
1.26, 0.99, 2.71, and 2.91 for CO, CO2, H2, and N2, and 0.2 for CO, CO2, H2, and N2, respectively. These values are
respectively. lower than those obtained in the previous case (Figure 2).
This modified equilibrium model has also been validated
Finally, the experimental results presented by Gautam21 for a
with the experimental data from Erlich and Fransson20 for
wood, sugarcane bagasse from sugar/alcohol production, and mobile 25 kWe downdraft gasifier for sawdust, commercial
empty fruit bunch (EFB) from palm-oil production in a simple wood, and woodchips were also used to validate the model.
constructed pellet-fired downdraft gasifier of about 20 kWth. As Experimental data and predicted data for a reported gasification
with the data of the previous authors, the model parameters temperature of 800 °C, pyrolysis temperature of 416 °C, and
have been adjusted automatically and no biomass drying and air 7% of heat losses are given in Figure 4. As observed, the
preheating have been considered. The pyrolysis unit temper- developed model is able to predict the producer gas
1388 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394
Energy & Fuels Article

gasification. The rms values are very similar in both parts of

Table 1, which means that predictions have the same precision
in both cases. It can be seen that predictions for H2 yields are
better for downdraft gasifiers than for fluidized-bed gasifiers,
with a higher rms value. However, the model predicts
reasonably well the values for CO2 and CO yields. As a result,
it can be concluded that the modified equilibrium model also
predicts reasonably good values for fluidized-bed gasifiers,
although its predictions are better for downdraft gasifiers.
3.2. Effect of Operating Parameters on Producer Gas
Composition and LHV. After the developed model was
validated with data reported by various researchers, it was used
to predict the effect of different operating parameters
[equivalence ratio (ER), air preheating, steam injection, and
oxygen enrichment] on producer gas composition and LHV.
These predictions are compared to those reported by other
authors.25−31 Different biomass and model parameter values are
used in the different cases studied for this reason.
3.2.1. Effect of ER. The ER is defined as the ratio of the
Figure 4. Comparison of predicted results from the modified moles of oxygen supplied to the gasifier to those required for
equilibrium model to experimental data from Gautam21 for a stoichiometric combustion. Figure 5 shows the variation of
downdraft gasifier. producer gas composition as a function of the ER in an
adiabatic gasifier of woodchips with a moisture content of 10%.
composition and also the higher heating value (HHV) (not In an autothermal gasifier such as this one, the gasification
plotted) with good accuracy for different types of biomass temperature depends upon the amount of air fed to the gasifier
feedstock. (Figure 6). In consequence, varying the ER or gasification
From the comparison of the predicted and experimental data temperature has the same effect on producer gas composition,
for these three different cases and different biomasses, it can be heating value, and gasification efficiency. For this reason, only
concluded that the model predicts the behavior of downdraft ER is plotted against producer gas composition and LHV.
biomass gasifiers and different biomasses reasonably well. These results were compared to those published by Plis and
3.1.2. Experimental Data from a Fluidized-Bed Gasifier. Wilk,25 Mathieu and Dubuisson,26 and Baratieri et al.27 The
Campoy24 conducted an experimental research study in an three models and the model herein show the same qualitative
atmospheric pressure bubbling fluidized-bed biomass gas- and quantitative tendencies. The percentage of CH4 remains
ification pilot plant of 100 kWth. Experimental results for very low and decreases when the ER increases. The H2
wood pellets and using different gasifying agents are used for percentage decreases from 22.5 to 6.5% when the ER increases
validating this model with fluidized-bed gasifiers. The model is from 0.3 to 0.6; the same behavior was observed by Plis and
adjusted on the basis of a reported air preheating temperature Wilk,25 with a decrease for the same range of ER values from 20
of 400 °C, pyrolysis temperature of 600 °C, heat losses of 7% of to 6.5%. While H2 decreases, CO2 increases slightly from 8.6 to
energy input, and 42 and 21% of pyrolysis char and tar, 11.6% (from 8.5 to 11.3% for Plis and Wilk25) and the CO
respectively, leaving the gasifier. The experimental and percentage decreases from 26.2 to 16.5% in this model and
predicted values are presented in Table 1. The table is divided from 25.9 to 16.6% for Plis and Wilk.25
into two parts, with the first accounting for air biomass If 3% of the total heat input is considered to be heat losses
gasification and the second accounting for air−steam biomass and 2% of carbon is also lost in the ash, as Doherty et al.28

Table 1. Comparison of Predicted Results from the Modified Equilibrium Model to Experimental Data from Campoy24 for a
Bubbling Fluidized-Bed Gasifier
gas heating valve
gas composition (vol %, dry basis) (MJ N−1 m−3, dry basis)
model experimental24 model experimental24 model experimental24 model experimental24
3 −1
dry biomass (kg/h) air (N m h ) steam (kg/h) CO CO2 H2 LHV
20.5 17 0 20.3 18.2 14.6 14.2 18.4 13.2 6.7 5.9
15 17 0 18.7 17.6 13.9 14.9 11.3 12.6 5.4 5.4
11.5 17 0 13.6 15.8 15.9 15.1 5.0 8.7 4.1 4.8
rms 1.9 0.8 3.8 0.6
17.5 17 3 16.6 13.8 16.3 16.9 17.6 14.6 5.8 5.2
19.1 15.5 5 14.3 11.5 19.0 18.6 22.6 16.2 6.3 5.3
15 17 3.2 15.4 15.0 16.2 16.2 14.5 14.0 5.2 5.1
15 17 6 12.4 11.9 18.4 18.6 14.9 16.2 5.1 5.1
12.2 17 2.5 12.9 15.4 16.7 15.9 8.4 11.9 4.3 4.9
12.2 17 5.1 11.1 13.8 18.1 17.0 10.3 13.3 4.2 4.8
rms 2.2 0.6 3.5 0.6

1389 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394

Energy & Fuels Article

Figure 5. Effect of the ER on the composition and LHV of producer gas for gasification in adiabatic conditions of woodchips with a 10% moisture
content.

found when studying hemlock woodchip gasification, the

results presented in Figure 7 would be obtained. The
tendencies observed are in good agreement with those
observed by the authors.28 CO and H2 reach a maximum,
and their content decreases steadily after these peaks. CO2
decreases rapidly to an ER of 0.35 and then increases slowly.
CH4 decreases and eventually reaches 0 between an ER of 0.4
and 0.45. As Doherty et al.28 stated, these trends may be
explained as follows: (1) The Boudouard reaction (eq 1) is
endothermic. As a result, the production of CO rises when the
temperature rises. The amount of available char for the
Boudouard reaction is enough for ERs up to 0.35. However, for
ERs higher than 0.35, the available char is insuficient; therefore,
CO decreases, while CO2 increases. (2) The water−gas
reaction (eq 2) is also endothermic. For this reason, increasing
the ER and temperature results in a higher consumption of char
and H2O to produce more CO and H2. (3) The methane
reaction (eq 4) is exothermic. The production of CH4
Figure 6. Effect of the temperature on the ER in adiabatic conditions decreases with the increase of the ER and gasification
when increasing the temperature for woodchip gasification with a 10% temperature. This leaves more H2 in the gas. (4) CO2 is
moisture content. produced through the reaction of CO and the available O2. (5)
The CO shift reaction is exothermic. This reaction produces

Figure 7. Effect of the ER on the composition of producer gas for hemlock woodchip gasification with a moisture content of 11.7%, assuming 3%
heat losses and 2% carbon losses.

1390 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394

Energy & Fuels Article

less CO2 and H2 at higher temperatures. This means that less an ER = 0.29, the air could be heated to 850 °C because the
CO and H2O are used. (6) The steam−methane reforming gasification temperature stays below 967 °C.
reaction reduces CH4. This reaction is endothermic; therefore, Figure 9 shows the predicted product gas composition for
an increase of the temperature benefits the forward reaction. the same operating conditions as Figure 8. Increasing the
3.2.2. Effect of Air Preheating. Air preheating is useful to temperature favors the products of endothermic reactions and
achieve a higher conversion efficiency of the gasification simultaneously the reactants of exothermic reactions. For low
process. The gasification temperature is increased by the ERs, the air temperature has a higher influence on the product
sensible heat in the air. This increase in the temperature gas composition. For example, over the air temperature range,
influences the product gas composition, LHV, and gasifier cold CO and H2 contents increased by 5.5 and 5.4 percentage
gas efficiency by increasing the production of combustible points, respectively, for an ER of 0.29. However, for an ER of
gases, CO, and H2. For this reason, air preheating can be 0.35 (not plotted), CO and H2 contents increased by only 4
considered as an alternative and more economical approach and 0.1 percentage points, respectively, for the same temper-
compared to oxygen-blown systems. However, to increase the ature range. The air temperature also has a significant influence
overall efficiency of the process, the heat required for air on composition but only to a certain level, after which
preheating should be recovered from the gas cooling section of additional preheating has little effect. For both of the ERs
the plant. The size of the plant is reduced when high- mentioned above, this level is reached at about 700 °C, which
temperature air is used as an oxidant because a smaller volume agrees with Lucas et al.,30 who reported an increase of the H2
of air is needed to bring the gasifier to the required operating content with an increasing air preheating temperature but no
temperature.29 In addition, the size of the reactor and gas rise between 700 and 830 °C. Yang et al.31 also refer to a critical
cleanup system needed are also reduced. air temperature at which air preheating is no longer efficient if
Figure 8 shows the influence of air preheating on the the aim is to maximize the yield of gaseous products. As
gasification temperature for hemlock woodchips with moisture expected,26,28 the LHV and cold gas efficiency were also
observed to increase when the air inlet temperature increased.
3.2.3. Effect of Steam Injection. The influence of steam
injection on the gasifier performance was studied (Figure 10)
for an ER of 0.34 and compared to the results presented by
Doherty et al.28 The steam injection rate was varied from 0 to
10.5 kg/h, as performed by Doherty et al.28 The producer gas
LHV decreased slightly from 5.25 to 5.14 MJ/kg. Because
steam injection rises the H2O content, a lower LHV is obtained.
CO and CH4 are shifted and reformed, respectively, with the
contents of the additional H2O, decreasing and producing more
CO2. The main effect of steam injection is the rise in the H2
content, which, in this case, increases by 1% over the range of
steam injection. Doherty et al.28 observed a slight increase in
cold gas efficiency, from 66.1 to 66.5%, while a slight decrease
of 1% for the same whole range is observed here. Increasing
steam injection decreases the gasifier temperature because of
highly endothermic reforming and water−gas reactions, unless
heat is supplied from an external source. As stated by Doherty
et al.,28 a decrease in the temperature is undesirable because
this would degrade the performance of the gasifier and could
lead to a high tar yield. For this reason, air preheating should be
Figure 8. Effect of air preheating on the gasification temperature for taken into account when using high moisture fuels and/or
hemlock woodchip gasification with a moisture content of 11.7% and steam injection.
ER = 0.29. 3.2.4. Effect of Oxygen Enrichment. The effect of oxygen
enrichment in the air on producer gas composition and LHV
content of 11.7%, heat losses of 3%, and carbon loss of 2%. The was also studied, and the results are presented (Figure 11).
results showed that the gasification temperature increased Figure 11 shows how the composition of producer gas changes
almost linearly with the air temperature for all ERs.22,28 As with the oxygen fraction in the air for woodchip gasification,
with an initial moisture content of 10%, ER = 0.3, and no air
Doherty et al.28 stated, there is a limit on the level of air preheating. The N2 yield decreases as the oxygen fraction
preheating for each ER. This limit is imposed by the increases, as expected. The methane content is very low, at a
effectiveness of the heat-exchange equipment and the operating percentage of less than 1%. The percentage of hydrogen in the
temperature constraints of the reactor. For fluidized-bed producer gas increases continuously with the oxygen fraction,
gasifiers, the operating temperature should not be above 1000 from about 25 to 32%, for an increase in the oxygen fraction
°C, to avoid reaching the ash melting temperature. This would from 25 to 50%. A similar tendency is also observed for CO,
bring agglomeration and defluidization problems. Air preheat- but in this case, the increase is from 30 to 42%. CO2 remains
more or less constant at around 10%. In addition, the reaction
ing at high ERs is limited to a low level. According to this temperature increases from 1100 to 1200 K when the oxygen
model, for a CFB at an ER = 0.37, an air temperature of no fraction increases from 25 to 50%. For the same increase in the
more than 170 °C would be recommended because the oxygen fraction, the LHV of the producer gas increases from 6
corresponding gasification temperature is 978 °C. In the case of to 7.8 MJ N−1 m−3. The increase of LHV is due to the increase
1391 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394
Energy & Fuels Article

Figure 9. Effect of the air inlet temperature on the composition of producer gas of hemlock woodchip gasification with a moisture content of 11.7%
and ER = 0.29.

Figure 10. Effect of steam injection on the composition of producer gas of hemlock woodchip gasification with a moisture content of 11.7% and ER
= 0.34.

Figure 11. Effect of oxygen enrichment on the composition of producer gas for woodchip gasification with a moisture content of 10% and ER = 0.3.

in the amount of CO and H2. The results are in good 4. CONCLUSION AND FINAL REMARKS
A model for biomass gasification has been developed in this
22
agreement with those obtained by Babu and Sheth. study. It is based on thermodynamic equilibrium calculations
1392 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394
Energy & Fuels Article

and includes some modifications to be adapted to a real pj = yield of the gas species j (CO, CO2, H2, and N2)
process, in which only a partial approach to chemical calculated by the model
equilibrium is achieved. It is a simple but rigorous model rms = root mean square
implemented in the equation solver program EES, with a user T = temperature (K)
interface that makes the model user-friendly and facilitates the Tp = pyrolysis temperature (°C)
user obtaining an overview of the operating conditions in a x, y, and z = normalized coefficient of atomic hydrogen,
certain computation. oxygen, and nitrogen for the biomass molecule
The model can be used to predict the final producer gas wj = experimental yield of the gas species j (CO, CO2, H2,
composition and its main characteristics, such as the heating and N2)
value, for a certain biomass with a defined ultimate composition ηc = carbon conversion efficiency (%)
and moisture. It has been validated with the data reported by
various researchers for downdraft, fluidized-bed gasifiers and
different biomasses and shows good agreement with the
■ REFERENCES
(1) Faaij, A. Potential Contribution of Bioenergy to the World’s Future
experimental data. In addition, it has been used to evaluate the Energy Demand; International Energy Agency (IEA) Bioenergy:
influence of different operating parameters on producer gas, Rotorua, New Zealand, 2007; http://www.idahoforests.org/img/pdf/
presenting the following conclusions: (1) For an adiabatic PotentialContribution.pdf.
process, increasing the ER also entails increasing the gas- (2) Puig-Arnavat, M.; Bruno, J. C.; Coronas, A. Renewable Sustainable
ification temperature and decreasing the LHV of producer gas. Energy Rev. 2010, 14, 2841−2851.
(2) The use of high-temperature preheated air allows for the (3) Gómez-Barea, A.; Leckner, B. Prog. Energy Combust. Sci. 2010, 36,
use of a smaller volume of air to become the same temperature 444−509.
(4) Villanueva, A. L.; Gomez-Barea, A.; Revuelta, E.; Campoy, M.;
in the gasifier bed and, in consequence, achieves downsizing of Ollero, P. Guidelines for selection of gasifiers modelling strategies.
the plant. However, it has been proven that the air temperature Proceedings of the 16th European Biomass Conference and Exhibition;
has a significant influence on composition only up to a certain Valencia, Spain, June 2−6, 2008.
level, and it is limited by the effectiveness of the heat-exchange (5) Zainal, Z. A.; Ali, R.; Lean, C. H.; Seetharamu, K. N. Energy
equipment and the operating temperature constraints of the Convers. Manage. 2001, 42, 1499−1515.
reactor. (3) Steam injection in biomass gasification increases (6) Jarungthammachote, S.; Dutta, A. Energy 2007, 32, 1660−1669.
the H2 content of producer gas. (4) The LHV, CO, and H2 (7) Ramanan, M. V.; Lakshmanan, E.; Sethumadhavan, R.;
yields of producer gas increase when the oxygen fraction of air Renganarayanan, S. Braz. J. Chem. Eng. 2008, 25, 585−601.
increases. (8) Gumz, W. Gas Producers and Blast Furnaces; Wiley: New York,
In conclusion, the model helps to predict the behavior of 1950.
(9) Jayah, T. H.; Aye, L.; Fuller, R. J.; Stewart, D. F. Biomass Bioenergy
different biomass types, can be adapted to different gasifier 2003, 25, 459−469.
designs by changing the values of the model parameters, and is (10) Altafini, C. R.; Wander, P. R.; Barreto, R. M. Energy Convers.
a useful tool for preliminary calculations, design, and operation Manage. 2003, 44, 2763−2777.
of biomass gasifiers. It is also a first step and can be used as an (11) Li, X. T.; Grace, J. R.; Lim, C. J.; Watkinson, A. P.; Chen, H. P.;
input to the combustion model of an internal combustion Kim, J. R. Biomass Bioenergy 2004, 26, 171−193.
engine or another gas to energy engine to model a whole (12) F-Chart Software. Engineering Equation Solver (EES), Academic
biomass co- or tri-generation plant. Professional Version V 8.649; F-Chart Software: Madison, WI, 2010.

■
(13) Gøbel, B.; Bentzen, J. D. Technical Report, ET-PER 95-13;
AUTHOR INFORMATION Laboratoriet for Energiteknik, Technical University of Denmark
(DTU): Lyngby, Denmark, 1995.
Corresponding Author (14) Fock, F.; Thomsen, K. Technical Report, ET-EP 99-10A; Institut
*Telephone: +34-977257891. Fax: +34-977559691. E-mail: for Energiteknik, Technical University of Denmark (DTU): Lyngby,
[email protected]. Denmark, 1999.
Notes (15) Di Blasi, C.; Signorelli, G.; Di Russo, C.; Rea, G. Ind. Eng. Chem.
Res. 1999, 38, 2216−2224.
The authors declare no competing financial interest.

■
(16) Figueiredo, J. L.; Valenzuela, C.; Bernalte, A.; Encinar, J. M. Fuel
1989, 68, 1012−1016.
ACKNOWLEDGMENTS (17) Fagbemi, L.; Khezami, L.; Capart, R. Appl. Energy 2001, 69,
The authors thank the European Commission for the financial 293−306.
support received as part of the European Project Polycity (18) Schröder, E. J. Anal. Appl. Pyrolysis 2004, 71, 669−694.
(Energy Networks in Sustainable Communities, TREN/ (19) Gómez-Barea, A.; Nilsson, S.; Vidal Barrero, F.; Campoy, M.
Fuel Process. Technol. 2012, 91, 1624−1633.
05FP6EN/S07.43964/51381).

■
(20) Erlich, A.; Fransson, T. H. Appl. Energy 2011, 88, 899−908.
(21) Gautam, G. Master’s Thesis, Auburn University, Auburn, AL,
NOMENCLATURE 2010.
C, H, O, and N = carbon, hydrogen, oxygen, and nitrogen (22) Babu, B. V.; Sheth, P. Modeling and simulation of biomass
fractions in biomass (dry basis) gasifier: Effect of oxygen enrichment and steam-to-air ratio. Proceedings
CV = calorific value (kJ/kg) of the International Congress on Renewable Energy (ICORE-2005); Pune,
ER = equivalence ratio India, Jan 20−22, 2005.
(23) Melgar, A.; Pérez, J. F.; Laget, H.; Hornillo, A. Energy Convers.
hchar = specific enthalpy of char (kJ/kg) Manage. 2007, 48, 59−67.
htar = specific enthalpy of tar (kJ/kg) (24) Campoy, M. Ph.D. Thesis, University of Seville, Seville, Spain,
HHV = higher heating value (kJ/kg) 2009.
LHV = lower heating value (kJ/kg) (25) Plis, P.; Wilk, R. K. Energy 2011, 36, 3838−3845.
m = number of gas species in the producer gas (26) Mathieu, P.; Dubuisson, R. Energy Convers. Manage. 2002, 43,
n = number of data points 1291−1299.

1393 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394

Energy & Fuels Article

(27) Baratieri, M.; Baggio, P.; Fiori, L.; Grigiante, M. Bioresour.

Technol. 2008, 99, 7063−7073.
(28) Doherty, W.; Reynolds, A.; Kennedy, D. Biomass Bioenergy
2009, 33, 1158−1167.
(29) Sugiyama, S.; Suzuki, N.; Kato, Y.; Yoshikawa, K.; Omino, A.;
Ishii, T.; Yoshikawa, K.; Kiga, T. Energy 2005, 30, 399−413.
(30) Lucas, C.; Szewczyk, D.; Blasiak, W.; Mochida, S. Biomass
Bioenergy 2004, 27, 563−575.
(31) Yang, W.; Ponzio, A.; Lucas, C.; Blasiak, W. Fuel Process.
Technol. 2006, 87, 235−245.

1394 dx.doi.org/10.1021/ef2019462 | Energy Fuels 2012, 26, 1385−1394