Evaluation of different biomass gasification modeling approaches for fluidized bed gasifiers

Biomass and Bioenergy 91 (2016) 69e82

Contents lists available at ScienceDirect

Biomass and Bioenergy
journal homepage: />
Research paper

Evaluation of different biomass gasification modeling approaches for
fluidized bed gasifiers
Guilnaz Mirmoshtaghi*, Hailong Li, Eva Thorin, Erik Dahlquist**
€lardalen University, Box 883, SE-721 23 Va
€sterås, Sweden
School of Business, Society and Engineering, Ma

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 17 November 2015
Received in revised form
28 April 2016
Accepted 1 May 2016

To develop a model for biomass gasification in fluidized bed gasifiers with high accuracy and generality
that could be used under various operating conditions, the equilibrium model (EM) is chosen as a general
and case-independent modeling method. However, EM lacks sufficient accuracy in predicting the content
(volume fraction) of four major components (H2, CO, CO2 and CH4) in product gas. In this paper, three
approachesdMODEL I, which restricts equilibrium to a specific temperature (QET method); MODEL II,
which uses empirical correlations for carbon, CH4, C2H2, C2H4, C2H6 and NH3 conversion; and MODEL III,

which includes kinetic and hydrodynamic equationsdhave been studied and compared to map the
barriers and complexities involved in developing an accurate and generic model for the gasification of
biomass.
This study indicates that existing empirical correlations can be further improved by considering more
experimental data. The updated model features better accuracy in the prediction of product gas
composition in a larger range of operating conditions. Additionally, combining the QET method with a
kinetic and hydrodynamic approach results in a model that features less overall error than the original
model based on a kinetic and hydrodynamic approach.
© 2016 Elsevier Ltd. All rights reserved.

Keywords:
Biomass gasification
Fluidized bed gasifiers
Kinetic
Empirical
Equilibrium model
Generality

1. Introduction
Because of environmental and economic incentives, such as
increasing energy prices and fossil fuel depletion, countries are
changing their energy profiles toward more renewable and sustainable resources.
Thermochemical gasification of carbon-based solid and liquid
materials, which results in product gas consisting of H2, CO, CO2,
CH4 and some light hydrocarbons, has been used and developed for
nearly two hundred years [1]. This technology can convert
renewable resources such as biomass or black liquor to energy
products that substitute for fossil-based fuels.
Among the existing types of gasifiers, the fluidized bed gasifier
has many advantages, such as easy scale-up, flexibility regarding

feedstock type and size, uniform temperature distribution and high
carbon conversion efficiency; therefore, it is suitable for the gasification of biomass. Biomass gasification in fluidized bed gasifiers is

* Corresponding author.
** Corresponding author.
E-mail addresses: (G. Mirmoshtaghi), Erik.
(E. Dahlquist).
/>0961-9534/© 2016 Elsevier Ltd. All rights reserved.

quite a complex process, which means that the operating parameters are influenced by a large number of variables. Therefore,
process modeling and simulation of the gasification process is more
cost effective than performing experiments.
According to the reviews by Puig-Arnavat [2]and Gomez Barea
[3]and the study by Radmanesh [4], there are two major approaches to model gasification in fluidized beds: equilibrium
modeling and dynamic modeling considering the kinetics and hydrodynamics of the bed. Dynamic modeling gives a better interpretation of the real case. However, this approach requires detailed
information on the geometry and design of the reactor, which
makes it dependent on measurements and estimation of these inputs for any further analysis of gasification process [5]. Due to the
complex and quite fast flow regime of different phases in the
gasifier, measuring and calculating residence time is necessary for
developing a correct dynamic model. However defining this
parameter close to reality is an issue which has been studied during
years [3].
In contrast, equilibrium modeling (EM), which is based on
thermodynamic analysis, does not require information on the dimensions, capacity and structure of the gasifier and therefore is
suitable for concept studies, preliminary design and optimization of


70

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

the process [5e7]. EM has been applied to the gasification process
in different waysdfor example, the entire gasification mechanism
is considered to be at equilibrium [6], or only the pyrolysis stage is
assumed to be at equilibrium [7,8].
EM is mostly applicable when the operating temperature is high
and the retention time is longer than the time required for complete gasification. However, the model may not provide accurate
results at low operating temperatures in the range of 750e900  C
[3]. EM also has limitations in predicting the amount of light hydrocarbons and unconverted solid carbon. Several studies have
been performed on how to improve the accuracy of EM in gasification modeling. Some examples are related to the gasification of
coal in a fluidized bed [9] and an entrained flow bed [10], whereas
other examples are of biomass gasification in a downdraft gasifier
[11] and fluidized bed gasifiers [6] [12,13].
In 2001, Kersten [12] reviewed and compared different quasiequilibrium models for biomass gasification in fluidized bed gasifiers. He studied two methods: 1. implementing empirical corre€pfer model [14] and 2. using the quasilations in the Schla
equilibrium temperature (QET) in the Gumz model [15] This
method is explained more in Section 2.1. Kersten concluded that the
Gumz model with QET yields better results. Li and his colleagues in
different studies [6,16] have investigated different methods to
improve the accuracy of EM for biomass air gasification in circulating fluidized beds (CFB). They found that adding empirical correlations for light hydrocarbons (mainly CH4) and carbon
conversion is a successful method for improving EM. Recently, Lim
and Lee [13] also developed a quasi-equilibrium model for fluidized
bed gasifiers. They built their model based on 43 experimental
datasets, which were gathered from different CFB [17,18] and
bubbling fluidized bed (BFB) [19,20] gasifiers. They concluded that
to achieve a higher level of accuracy in quasi-equilibrium models,
the empirical parameters in the correlations for improving EM
models should be adjusted to the experimental data from the same
plant that is modeled. Other attempts have been made to improve
the accuracy of equilibrium models by considering reaction kinetics. In these studies, the pyrolysis step is assumed to be at
equilibrium, whereas char gasification and part of the homogeneous reactions in the gasification are considered to be kinetically

controlled. For example, Bilodeau et al. [8], Nikoo et al. [21] and
Wang et al. [22]included different reaction kinetics and, in some
cases, hydrodynamics of the bed to improve the results of EM.
According to the literature mentioned above and as Gomez and
Leckner described in their review paper [3], the modification of EM
(which is called pseudo-equilibrium in Ref. [3]) for the modeling of
fluidized bed gasifiers can be categorized into three groups: 1.
Modifying the equilibrium temperature by the QET method, 2.
Using quasi-equilibrium by adding empirical correlations for specific components and 3. Introducing the kinetics for specific reactions and adding hydrodynamics of the bed. Gomez and Leckner
[3] evaluated the capability of different modified EMs to predict the
composition of the product gas at different operating conditions to
measure the “generality” of those models. They concluded that
pseudo-equilibrium models give the most accurate results for gas
composition, whereas tar and char content cannot be predicted as
generally as other components.
According to the mentioned studies, although the gasification
system is quite complex and dependent on many interrelated and
independent variables, the “generality” characteristic for a model is
one of the major concerns in the field of gasification modelingemostly, whether it is aimed to be used further in process
design and simulation level. As discussed above, EM is independent
of the gasifier size and type, which makes it suitable as a basis for
developing a general model. However, addressing the limitations of
this model to improve the accuracy of prediction results in some

“non-generality” factors. Therefore, a systematic study to evaluate
the advantages and disadvantages of different modification
methods and mapping the barriers and complexities that result in
this “non-generality” would be essential for any further development of any possible generic model. This is one of the major novel
contribution of this study to the field of biomass gasification
modeling. The investigation of further possibilities in improving

the modeling of biomass gasification is another part of this study.
All three modeling approaches presented above are included and
are based on the results of this investigation along with new
models suggested in this paper.
2. Methodology
In this study, three equilibrium-based models from the literature, one model for each modeling approach described in Section 1,
have been selected for evaluation. Two of the models have also
been further modified. The same set of experimental data have
been used for the evaluation of all models.
The simulation tool ASPEN PLUS has been used for the evaluation. As a steady state simulation tool, ASPEN Plus has been widely
used to implement EM to model biomass gasification in fluidized
bed gasifiers [8,23] owing to its powerful database of thermodynamic and chemical properties [24]. According to Puig-Arnavat [2],
ASPEN PLUS is chosen for modeling of gasifiers and further gasification processes to avoid complexity when principal gasification
reactions and some fundamental physical characteristics are
included.
The models evaluated in this study are called MODEL I, MODEL II
and MODEL III, corresponding to the three different modeling approaches described in Section 1.
MODEL I, MODEL II and MODEL III are first replicated in ASPEN
PLUS and verified by comparison with the model results in the
original presented studies. Experimental data from different BFB
and CFB gasifiers have been collected from the literature and used
to evaluate the model performance, with the aim to test whether
the models are also valid for experimental conditions other than
those for which they were originally validated. The input data used
for the simulations are biomass ultimate/proximate analysis, temperature, pressure, and biomass, air and steam flowrates (see Section 2.4). The detailed information on ultimate and proximate
analysis of different biomasses used in this study can be found in
the referred papers for each case, respectively.
To choose the suitable experimental data as the input for evaluating the models, 5 major input parameters have been compared.
The parameters are gasifier type (CFB or BFB), equivalence ratio
(ER) (is a dimensionless index for the ratio of the mass of air input

to the stoichiometric amount of air needed for full combustion
[25]), temperature, load (as an index for the size and residence time
of the gasifier) and the mass ratio of steam to the moisture and ash
free mass of biomass (S/B). These parameters are combinations of
major operating parameters (ER, temperature, S/B) and variables
that can limit the “generality” aspect of the model (gasifier type,
load). The cases with input parameters in different ranges than
those of the original validated experiments have been chosen for
the evaluation step.
In the first part of the paper, the overall accuracy of the models
in predicting the four major components in the product gas (H2, CO,
CO2 and CH4) is the focus. The main results in this part of the paper
are from studying the sensitivity of the models to the major input
parameters, which were mentioned earlier, and analyzing the
variation of accuracy in different cases.
In the second part of the paper, based on the discovered limitations and “bottlenecks” in the existing modified EMs found and
discussed in the first part, MODEL II and MODEL III are further


G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

modified and evaluated. The new models, as extensions of MODEL
II and MODEL III, are called MOD-MODEL II and MOD-MODEL III.
The major evaluation criteria are prediction accuracy and generality of the models. Generality is evaluated for a model in terms of
its ability to be applied to both BFB and FB cases in a larger range of
operating parameters.
2.1. MODEL I-quasi equilibrium temperature (QET)
To improve EM accuracy in predicting product gas composition,
Gumz [15] proposed using QET instead of the actual operating
temperature of the reactor. QET is the temperature, different from

the operating temperature, at which the specific chemical reaction
is assumed to reach equilibrium [2,12]. One way to find QET is to
determine a new temperature at which the difference between the
measured content of product gas components and the values
calculated by the equilibrium model would be at a minimum level.
Since this is an empirical method, the new temperature would be
the result of curve-fitting. It is important to know that the difference between the operating temperature and the best-fit temperature determined by this method is called “degree of approach to
equilibrium” and this is actually the value which is considered in
“restricted equilibrium” sheet under “temperature approach” part
[15,16,26].
In ASPEN PLUS, the Gibbs reactor (RGIBBS), which is based on
the minimization of Gibbs energy, has an option for “restricting the
equilibrium”. The “degree of approach to equilibrium” can be used in
this option either for the whole unit or for a specific reaction.
Using QET for CO-shift, methane reforming and ammonia formation reactions, Doherty et al. [27] developed a model (MODEL-I)
to simulate biomass gasification in a CFB gasifier in ASPEN PLUS.
This model was used to test the effect of air preheating on the gas
yield and composition. The detailed information on the model
development is given in Ref. [27]. More details are given in the
supplementary document.
2.2. MODEL II-empirical correlation
Hannula and Kurkela proposed another model (MODEL II) based
on using empirical correlations to increase the accuracy of EM in
predicting the composition of light hydrocarbons in product gas
[28]. Hydrocarbons, ammonia and total carbon conversions were
correlated to ER, which is an indicator of the amount of oxygen
entering the gasifier for partial oxidation reactions. These correlations were derived by fitting correlations to the empirical results
from measurements on a pressurized CFB gasifier.
The correlations are implemented by the CALCULATOR block in
ASPEN PLUS. The results from this block are used to adjust hydrocarbon formation in a stoichiometric reactor (RSTOIC). In Ref. [28],

all correlations and blocks used for this model are listed. More
details are given in the supplementary document.
2.3. MODEL III-kinetic and hydrodynamic
CH4 conversion, char gasification and tar formation reactions are
kinetically controlled. Sotudeh et al. [29] proposed a model by
combining EM with kinetic models for coal gasification. Similarly,
Nikoo and Mahinpey modified the EM to simulate the biomass
gasification in a BFB [21]. (MODEL-III). This model includes the kinetics of combustion and steam gasification reactions together
with fluidized bed hydrodynamics derived from the simple twophase theory [21]. After the decomposition of the solid biomass
in the model, volatiles and fixed carbon are handled in separate
steps in which volatile combustion and gasification are considered
at equilibrium. However, the char gasification step is modeled

71

including the kinetics of the reaction. As described in detail in
Ref. [21], RGIBBS and RCSTR reactor modules were used for the
respective gasification stages. Char gasification in bed and freeboard were modeled externally as a “user model” and were linked
to the flow sheet in ASPEN PLUS. Another essential point to be
mentioned is that the CH4 content in the product gas for this model
is tuned by a nonlinear regression with specific temperature points:
700, 750, 800, 850 and 900  C.
In fluidized bed gasifiers, the gas velocity is a key parameter
when considering the hydrodynamics of the bed. Gas velocity is
derived from the flowrate of the oxidizing agent and the crosssectional area of the bed. MODEL III describes a specific gasifier
with an internal diameter (ID) of 40 mm. The applicable velocity of
the oxidizing agent and consequently of the biomass is therefore
limited to a specific range. This means that the diameter of the bed
must be provided to use MODEL III correctly. The main limitation of
this model is the requirement of the bed diameter or any type of

similar factors that is equivalent to the diameter. Therefore, to use
MODEL III for gasifiers that operate at other flowrates, the equivalent flowrate is calculated assuming the equality of load between
that gasifier and the one in the original model. Table 1 summarizes
the characteristics of EM-based models studied in this work.
Additional information are given in the supplementary document.
2.4. Experimental data used for model evaluation
The experimental data by which each model has been validated
are listed in Table 2. Basically each model has been validated with
exactly the same data used by the authors that developed that
model. However due to further use of the input data presented in
this table for evaluation of models, it is important to report all of the
experimental data used for validation of the original models.
Further, the experimental data that have been used for the
evaluation of all models are listed in Table 3. Each “set point” is a
combination of one set of ER value, S/B ratio, temperature, pressure
and load. Since in some cases (gasifiers) the combination of these
operating parameters are changing at the same time (not one
parameter at the time), it is not scientifically correct to report the
results versus variation of only one of the operating parameters.
Therefore in such kind of cases, set point is used to show which set
of operating parameters are used as an input to the model.
2.5. Model evaluation
To evaluate the original and modified models, the results for the
volume percentage of H2, CO, CO2 and CH4 in the gasification
product gas have been compared with experimental data, and the
errors are calculated based on following equations.
The relative error for each component at each data point is
calculated by

yie À yip



relE i ¼


yie


(1)

where yie and yip are the experimental and predicted results of the
volume percentages, respectively. To evaluate the total performance of the model in predicting the product gas composition, the
overall error (OEi) at each set point is also calculated and considered as the basis for comparison.

OEi ¼

n
X
ðrelE i  yie Þ

(2)

i¼1

To evaluate the performance of different models in a range of set
points, two other indexes are required. The first is the average


72

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

Table 1
Modified EMs simulated in ASPEN PLUS for biomass gasification.
Models name

MODEL I

MODEL II

MODEL III

Reference to
models
Oxidizing
agent(s)
Temp. range
( C)
ER range
S/B range
Flowrate (kg/h)
Modification
methods
from EMs

[27]

[28]

[21]

Air

Air/Steam

Air/Steam

730e815

856e955

700e900

0.22e0.53
e
15.5e48.5
Assigning temperature lower than equilibrium
temperature in RGIBBS block for specific components
and reactions

0.28e0.39
0.08e0.28
18e58
Using the empirical correlation for nonequilibrium components in product gas

by calculator block

0.19e0.27
0e4.04
0.445e0.512
Including reaction kinetics and hydrodynamics of the
bed as an external FORTRAN subroutine.
Including an empirical nonlinear regression with
specific temperatures (700, 750, 800, 850 and 900  C)
for CH4 concentration.

and “n” represents the number of set points in the analyzed dataset.

overall error (OE), and the other is the variation width (VW). These
are calculated as follows:

OE ¼

n
X

!,
OE i

n

VW ¼ OE imax À OE imin

(3)

(4)

i¼1

OEimin and OEimax are the minimum and maximum values in the
analyzed dataset, respectively.

OEi is the overall error at each data point taken from Equation (4),

Table 2
Experimental data used for validation in original paper and partly for model evaluation in this study.
Test rigs

Set point no.

ERa

S/Bb

Temp ( C)

Pressure (kPaa)

Load (Mg.me2$he1)

Ref

0.53
0.45
0.40

0.52
0.37
0.43
0.34
0.35
0.4
0.38
0.22
0.26
0.3
0.46

0
0
0
0
0.004
0.03
0
0
0
0
0
0
0
0

740
718
766

815
772
787
718
730
752
789
701
728
739
805

165
119
119
119
119
119
119
119
119
119
119
119
119
119

3.051
2.918
3.427

3.330
3.684
3.227
4.283
4.507
3.952
4.051
6.174
5.819
5.275
1.981

[17]

CFB-EXP I

1
2
3
4
5
6
7
8
9
10
11
12
13
14

0.28
0.39
0.43
0.32
0.3
0.34
0.31
0.31
0.39
0.39

0.074
0.26
0.21
0.16
0.15
0.16
0.156
0.13
0.17
0.16

822.5
900
885
840
855
890
860

885
900
930

400
400
400
400
400
400
400
500
500
500

1.181
0.712
0.807
0.917
0.983
1.013
0.939
1.005
0.785
0.858

[30]

CFB-EXP II

1
2
3
4
5
6
7
8
9
10
1
2
3
4
5

0.19
0.21
0.23
0.25
0.27

1.56
1.56
1.56
1.56
1.56

800
800

800
800
800

100
100
100
100
100

0.408
0.408
0.408
0.408
0.408

[31]

6
7
8
9
10

0.22
0.22
0.22
0.22
0.22

2.7
2.7
2.7
2.7
2.7

700
750
800
850
900

100
100
100
100
100

0.354
0.354
0.354
0.354
0.354

[31]

11
12
13
14

15

0.22
0.22
0.22
0.22
0.22

0
1.35
2.02
2.7
4.04

800
800
800
800
800

100
100
100
100
100

0.354
0.354
0.354
0.354

0.354

[31]

BFB-EXP III

a
b

Equivalence Ratio ¼ Air (kg)/air stoichiometry (kg).
Steam (kg)/biomass (kg dry ash free).


G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

73

Table 3
Experimental data used for model evaluation.
ERa

S/Bb

Temp ( C)

Pressure (kPaa)

0.17
0.19
0.2

0.22
0.26

0
0
0
0
0

686
690
720
919
974

100
100
100
100
100

2.070
1.951
1.951
1.919
1.672

[32]

CFB I

1
2
3
4
5

0.32
0.34
0.34
0.37
0.43

0
0
0
0
0

861
855
822
850
805

100
100
100
100
100

1.790
1.783
1.904
2.796
2.548

[33]

CFB II

1
2
3
4
5
1
2
3
4

0.22
0.25
0.28
0.30

0
0
0
0

688
726
784
779

100
100
100
100

0.438
0.438
0.632
0.438

[34]

CFB III

1
2
3
4

0.38
0.38
0.38
0.38

1.13
1.45
0.93
1.22

775
770
780
780

100
100
100
100

1.942
11.942
2.219
2.219

[18]

CFB IV

0.26
0.28
0.33
0.34
0.44

0
0
0
0
0

800
800
800
800
800

100
100
100
100
100

0.245
0.227
0.193
0.187
0.145

[35]

BFB I

1
2

3
4
5
1
2
3
4
5
6
7
8
9
10

0
0.09
0.18
0.28
0.37
0.18
0.18
0.18
0.18
0.18

1.4
1.4
1.4
1.4
1.4

1.7
1.7
1.7
1.7
1.7

850
850
850
850
850
750
800
850
900
950

100
100
100
100
100
100
100
100
100
100

0.241
0.241

0.241
0.241
0.241
0.273
0.273
0.273
0.273
0.273

[36]

11
12
13
14
15

0
0
0
0
0

1.1
1.4
1.8
2.7
4.7

800

800
800
800
800

100
100
100
100
100

0.370
0.289
0.225
0.161
0.088

[36]

0.19
0.24
0.27
0.32

0
0
0
0

800

800
800
800

100
100
100
100

0.271
0.438
0.438
0.199

[20]

Test rigs

BFB II

BFB III

a
b

Set point no.

1
2
3

4

Load (Mg.me2$he1)

Ref

[36]

Equivalence Ratio ¼ Air (kg)/air stoichiometry (kg).
Steam (kg)/biomass (kg dry ash free).

3. Results
3.1. Model verification
In Table 4, the results of replicated models and originally published results from the literature are shown for the major components. The results show a good agreement between the original
models and the replicated ones. The small differences between the
results of replicated and original model are expected since not all
the detailed information for the model set up in ASPEN plus (for
example, selection of property model) is given in the published
material of each original model. This can therefore cause slightly
different results that can still be accepted.
In the following subsections, the results from simulations of the
replicated models are compared with experimental data.
3.2. Evaluation of the original EM-based models
To form a common basis for comparison of the different models,
the same set of experimental data has been used for the evaluation

of all models. The overall error has been calculated for all tested
cases, and a descriptive statistical analysis has been made on the set
of OEis. The results of these tests are illustrated and analyzed
further in Section 3.2.1. The experimental setups chosen for evaluation of the models are listed in Table 3. Moreover, set points 1 to 6

of CFB-EXP II and 1 to 15 of BFB-EXP III from Table 2 are also used
for model evaluation. These points are highlighted with italic outlined numbers in Table 2.
3.2.1. Evaluation of original models in different operating conditions
Using MODELS I, II and III for different ranges of input data results in variation of the overall error. This variation illustrates the
impact of changing the operating parameters on the accuracy and
performance of each model in predicting product gas
concentration.
 Gasifier type (CFB and BFB)
In Fig. 1, the average overall error and the variation of this value
in the dataset for different gasifier types (CFB and BFB) are shown.


74

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

Table 4
Verification of original models replica.
Input parameters

Results

NO.a

Temp ( C)

ERb

S/Bc

Pressure (kPaa)

Model name

1
2
3
4
5
6
7
8
9

718
930
905
890
980
700
750
800
850

0.34
0.39
0.35
0.36
0.4
0.22

0.22
0.22
0.22

0
0.2
0.09
0.14
0.12
2.7
2.7
2.7
2.7

120
500
500
500
500
100
100
100
100

MODEL
MODEL
MODEL
MODEL
MODEL
MODEL

MODEL
MODEL
MODEL

a
b
c

I
II
II
II
II
III
III
III
III

Original model volume fraction (%)

Replicated model volume fraction
(%)

H2

CH4

CO

CO2

H2

CH4

CO

CO2

5.5
8
10
10.5
10.2
39
39
40
42

7.7
4
6.5
5
3.6
10.5
10
9.3
9

16.8

12
16
15.5
17
36
36
36
36

14.6
13.5
14
13
11
14
10
14
14.5

5.5
7.7
9.7
10.1
9.6
38.1
38.7
39.4
40.5

6.9

4.3
6.3
5.3
4
10.3
9.8
9.3
8.8

16.1
11
15.4
14
17
36.6
36.6
36.3
35.5

14.6
14.1
13.1
13
11
15
10
15
15.1

NO. 1 refers to [17], NO. 2 to 5 refer to [28], NO. 6 to 9 refer to [21].

Equivalence Ratio ¼ Air (kg)/air stoichiometry (kg).
Steam (kg)/biomass (kg dry ash free).

Of the 63 tested set points, 24 points represent CFB gasifiers and
39 points represent BFB gasifiers. Comparing the average values of
overall error shows that MODEL I has almost the same level of inaccuracy in predicting the volume fraction of different components
in the product gas for both BFB and CFB gasifiers; therefore, it can
be concluded that the gasifier type does not affect the level of
prediction accuracy for MODEL I.
Conversely, MODEL II and MODEL III show lower levels of OE for
CFB and BFB, respectively. This is because both of these models
were originally built for these gasifier types. Therefore, without any
further analysis, it can be expected that MODEL III, which is based
on the kinetics and hydrodynamics of the bed/freeboard of a specific BFB, is a more suitable approach for BFB gasifiers, whereas
MODEL II, with the implemented empirical correlations taken from
one specific CFB, has a better prediction capability for CFBs.
The other input parameter that is essential for air gasification is
the ER value. To investigate the impact of this parameter on the
performance of each model, three groups of set points with fixed ER
values, as shown in Table 5, have been analyzed based on average
overall error (OE) and variation width (VW). OE identifies the level
of accuracy, whereas VW shows the extent to which the changes of
other parameters at that fixed ER can affect the accuracy of the
model. The smaller the width, the lower the level of impact on the
model performance by the variation of other parameters.

In Tables 5e8, the minimum OE and VW are in boldface, and the
maximum values for these factors are shown by outlined numbers.
According to the results in Table 5, MODEL III, which modifies
EM by including the reaction kinetics and bed hydrodynamics,

has the lowest level of OE and the smallest VW compared to
the other models for ER values 0.18 and 0.22. These ER
values actually represent set points 6e10 from BFB II and 6e15
from BFB-EXP III, respectively. Because these points are taken
from BFB gasifiers, this can be the main reason for the better performance of MODEL III compared to the other models. However,
in the case of ER ¼ 0.38, the most accurate model is MODEL I,
whereas the smallest VW is from MODEL II. This means that
MODEL II, which is modified using correlations with ER for
non-equilibrium components, is less affected by the variation of
the other parameters when ER is greater than 0.3. Actually,
based on the dataset used in this paper, ER values equal to or
greater than 0.3 are more typical for CFB gasifiers; therefore, less
variation of results in this ER range for MODEL II is expected.
However, better performance by MODEL I, which modifies EM by
the QET method, can be connected to a temperature level less than
or equal to 800  C, which is connected to set points analyzed for
ER ¼ 0.38.
 Temperature

CFB gasifiers

60

Overall Error (%)

Overall Error (%)

60

50

40

30

50

40

30

20

20

10

10

0

0

MODEL I

MODEL II

MODEL III

Fig. 1. Average overall error level and variation for prediction of gas composition from BFB and CFB gasifiers.

 Equivalence ratio (ER) value

BFB gasifiers

70

70

MODEL I

MODEL II

MODEL III


G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82
Table 5
Average overall error (OE) and variation width of overall errors of original models at
fixed ER (À).
OE

MODEL I
MODEL II
MODEL III

75

Table 7
Average overall error (OE) and variation width of overall errors of original models at
fixed S/B (À).

VW

OE

0.18

0.22

0.38

0.18

0.22

0.38

37.76
42.26
32.37

40.54
63.69
18.37

23.74
35.83
56.98

25.78

28.79
11.96

42.86
50.52
20.08

15.65
4.63
6.67

MODEL I
MODEL II
MODEL III

VW

1.4

1.56

1.7

2.7

1.4

1.56

1.7

2.7

37.76
42.29
32.37

33.016
63.09
17.18

45.65
53.81
36.34

44.23
66.2
17.81

5.2
7.44
1.72

45.23
69.84
18.7

25.78
18.6
11.96

34.13
31.72
20.08

The minimum OE and VW are boldface, whereas the maximum values for these
factors are shown by outlined numbers.

The minimum OE and VW are boldface, whereas the maximum values for these
factors are shown by outlined numbers.

The other important parameter that affects the performance of a
gasifier to a large extent in reality is temperature. A general model
should be predictive at different temperatures, so to find the nongenerality connected to a specific temperature point or range,
three fixed temperature values are considered in this part of the
study.
Table 6 shows the OE and VW in the errors of the original
models when the temperature is fixed. Based on the results in this
table, MODEL II shows the largest VW for all fixed temperature
cases. This means that MODEL II, which is based on correlating nonequilibrium components to ER value, is more sensitive to the
variation of other parameters when applying fixed temperature
values.
Conversely, MODEL I at temperatures of 780 and 800  C shows
the smallest VW, which is equivalent to the larger impact of these
fixed temperatures than the variation of the other input parameters
on the performance of the model. This can be observed because the
quasi-temperature (as described in 2.1), which is set for nonequilibrium reactions in this model, is chosen by assuming an
equilibrium temperature of 800  C. In the case of 850  C, which is
from a BFB gasifier and is also one of the specific temperatures for
which MODEL III is tuned, this model shows the lowest OE and VW.

This means that for BFB gasifiers with temperatures higher than
800  C, the variation of other parameters affects the performance of
MODEL III less than the other models.
In steam gasification or the cases in which steam is also used as
one of the oxidizing agents, the steam-to-biomass ratio (S/B) is
important. This parameter is defined as the ratio between steam
input mass flow and the biomass flow in kg/h. A general model,
therefore, is expected to reflect the impact of variation in this
parameter and other operating parameters. In Table 7, the OE and
VW of the three original models in predicting the product gas
composition at a fixed S/B for four different tests at two BFB gasifiers are shown.
Based on the results in Table 7, MODEL II shows the largest OE
compared to the other models. This can be attributed to the fact
that MODEL II was originally developed for air gasification in a CFB
gasifier and that the non-equilibrium components are not correlated to oxidizing agents other than ER value, which can result in
inaccuracy in this model when steam is one or the only oxidizing

agent. Therefore, “not being connected to steam flow” can be
considered as a source of non-generality in this model. Moreover,
the largest VW for MODEL II with S/B ¼ 1.4 and 1.56 shows a larger
impact by variation in other input parameters than the fixed S/B on
the performance of MODEL II.
Conversely, MODEL III shows the smallest OE and VW for all set
points in this analysis compared to other models. This means that
the performance of MODEL III is less affected by the variation of
other parameters than by the fixed S/B. This finding is observed
because in MODEL III, CO-shift is considered as one of the kinetically controlled reactions; thus, steam flowrate and the ratio between steam flow and input biomass (S/B) affects the prediction of
product gas composition in this model.
In the cases of air gasification without any steam flow, as illustrated in Fig. 2, MODEL III has the largest OE and VW whereas
MODEL II shows the smallest OE and VW. The results shown in this

figure are based on 23 set points from 3 CFB and 2 BFB gasifiers. As
discussed earlier, a large VW indicates a higher impact by changes
from other input parameters than by the fixed value (in this case,
zero steam flow), and a small VW indicates a lower impact by other
parameters than by the fixed value. This means that, at least for the
air gasification cases analyzed in this study, the approach of
correlating the non-equilibrium components to the ER value, as in
MODEL II, is more suitable than using kinetics for combustion and
steam gasification considering the hydrodynamics of the bed, as in
MODEL III.

Table 6
Average overall error (OE) and variation width in overall errors of original models at
fixed temperature ( C).
 Steam to biomass ratio (S/B)

In addition to the operating parameters discussed above, the
size of the gasifier, which affects the residence time for gas and char
in the reactor, is also important. The load, which is defined as
biomass mass flowrate per cross-sectional area of the gasifier, not
only is an index for the size of the reactor but also shows the input
rate of the feedstock. The OE and VW presented in Table 8 are for
the cases in which the load is fixed for the number of set points.
According to Table 8, for loads less than 0.430 Mg.me2$he1,
which basically represents BFB gasifiers in the analyzed dataset,

Table 8
Average overall error (OE) and variation width in overall errors of original models at
fixed load (Mg.me2$he1).
OE

780

800

850

780

800

850

MODEL I
MODEL II
MODEL III

16.08
33.94
59.27

32.36
57.51
30.84

45.65
53.82
36.34

0.6

1.34
0.74

37.22
67.01
42.48

45.23
69.84
18.7

MODEL I
MODEL II
MODEL III

OE

MODEL I
MODEL II
MODEL III

 Load

VW

The minimum OE and VW are boldface, whereas the maximum values for these
factors are shown by outlined numbers.

0.241

0.273

0.354

0.407

0.438

1.942

1.951

2.219

45.65
53.81
36.34

37.76
42.3
32.37

40.55
63.7
18.37

33.016
63.09
17.18

38.36
38.78
52.8

17.46
22.33
60.23

31.4
37.73
54.7

16.08
33.94
59.27

45.23
69.84
18.7

25.78
18.6
11.96

53
50.52
20.08

5.2
7.44

1.72

10.69
15.29
22.03

8.02
1.27
13.82

0.05
0.44
3.44

0.6
1.34
0.74

The minimum OE and VW are boldface, whereas the maximum values for these
factors are shown by outlined numbers.


76

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

No Steam/Biomass

70

60

Overall Error (%)

50

40

30

20

10

0

Fig. 3. Regression equation for modifying MODEL II and developing MOD-MODEL II.

MODEL I

MODEL II

MODEL III

Fig. 2. Average overall error level and variation for no steam/biomass.

MODEL III shows the smallest OE and VW whereas MODEL II shows
the largest OE and VW except for load ¼ 0.273 Mg.me2$he1. Small
VW in this range of loads (for BFB gasifiers) leads to variation of
other operating parameters being less effective on the accuracy of

MODEL III than the application of fixed loads. Conversely, MODEL II
is more significantly affected by the changes in other parameters
when the load is fixed for values less than 0.430 Mg.me2$he1.
Because MODEL III considers the hydrodynamics of the bed and
freeboard, the approach is quite sensitive to the size of the reactor
and the biomass and gas flowrates. In a fixed-size gasifier, which is
close to the original case of MODEL III, the model has better performance and predictive capability. However, MODEL II was originally developed for a pilot-scale CFB gasifier with a larger volume.

Therefore, this model is less affected by a fixed load compared to
other parameters.
These results can also be related to the discussed factor of
gasifier type, which significantly affects the performance of
MODELS II and III.
In the case of MODEL I, at fixed loads greater than 0.430
Mg.me2$he1, the VW (and mostly OE) are the smallest compared to
the other models. This can be interpreted in terms of the changes in
other operating parameters than these fixed loads having a lesser
effect on the performance of MODEL I. Because the temperatures
for the set points taken for fixed loads greater than 0.430
Mg.me2$he1 are all less than 800  C and this temperature is
assumed to be the operating temperature in the original model, the
model performance is better than that at temperatures above
800  C.
According to the discussed behavior of three modified equilibrium models (MODEL I, II and III) for a range of 5 major operating
criteria (gasifier type, fixed ER, temperature, S/B and load), the

Table 9
Overview of equilibrium-based models for biomass gasification in fluidized bed gasifiers. The italic and bold text indicates the limitations that have been the focus of the
modified models.
Equilibrium model (EM)

Special criteria

Limitations
Overprediction of H2, CO, Underprediction of CH4, tar and CO2

Minimization of the system Gibbs free energy
MODEL I

MODEL II

MODEL III

Special criteria

Limitations

Special criteria

Limitations

Special criteria

Limitations

Restricting equilibrium to
temperature lower than
operation temperature
“approach equilibrium”
(QET method)

-Large inaccuracy
when steam is one or
the only oxidizing agent
-The prediction is only
correct for a reactor
temperature of
approximately 800  C
-No prediction of light
hydrocarbons and tar
other than CH4
-Trial-and-error
method to find
temperature approach

Using empirical correlations
relating ER value to carbon
conversion, hydrocarbon
conversion and NH3

-The model is valid
only within the
range of the data
used for fitting
-Not applicable for
ER > 0.44
-Not accurate for
ER < 0.3 mostly for
BFB gasifiers
-Not accurate for
different values of

load

Using kinetic equations
together with
hydrodynamics for char
gasification (heterogeneous)
reactions

-Large inaccuracy for CFB
gasifiers
-Inflexible to large loads
because it has been developed
for a BFB gasifier with
ID ¼ 0.04 m
-No prediction of light
hydrocarbons and tars other
than CH4
-Cl is not considered in the
ultimate analysis of the
feedstock
-Functions only for specific
temperatures: 700e750e800
e850e900  C and not the
temperatures in between.

MOD-MODEL II

MOD-MODEL III

Special criteria

Special criteria
Using empirical correlations for CH4 content based on regression of experimental Restricting equilibrium to temperature lower than operation temperature
data from different tests for both BFB and CFB gasifiers in addition to MODEL II “approach to equilibrium” for methanation reaction in addition to MODEL III


G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

(a)

77

(b)

Fig. 4. -a. Comparison of MODEL II for BFB III [20] (set points 1 to 4) with MOD-MODEL II after applying improvement ideas. -b. Comparison of MODEL II for CFB III [34] (set points 1
to 4) with MOD-MODEL II after applying improvement ideas.


78

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

Fig. 5. MOD-MODEL III flow sheet in ASPEN PLUS.

following results can be obtained:
1. MODEL I, with a quasi-temperature approach that restricts COshift and steam reforming reactions, shows an OE between 35
and 45% in all cases. Fixing the temperature at approximately
800  C has a greater effect on the performance of this model
than the other operating conditions. Moreover, large loads,
which are equivalent to CFBs at temperatures lower than 800  C
in the experimental data used for validation, also have a positive

impact on the accuracy of this model.
2. MODEL II, with the approach of correlating non-equilibrium
components to ER values, does not show larger sensitivity for
any of the analyzed operating parameters. However, it functions
better for CFB gasifiers with ER equal to or higher than 0.3.
3. MODEL III, with the approach of including the kinetics for char
combustion and steam gasification reactions and hydrodynamics of the bed and freeboard, is accurate only for specific
operating parameters and therefore is quite sensitive to any
changes in the major input data. Therefore, this model is inaccurate for a wide range of input data. MODEL III shows higher
accuracy only if it is applied to BFB gasifiers with steam flow and
a smaller load when ER is less than 0.3 and the temperature is
greater than 800  C.
The results of observations made in this section are summarized
in Table 9. Moreover, some preliminary ideas for the further
improvement of EM using these three approaches are given. The
new models are called MOD-MODEL II and MOD-MODEL III and are
discussed below.
3.3. Evaluation of the modified EM-based models for different ER
values
According to the abovementioned points, MOD-MODEL II is
developed based on the correlations derived from a wider range of
experimental data, and MOD-MODEL III is built as the combination
of two approaches used in MODEL I and MODEL III. The new models
are further described briefly in this section along with the visual
results of testing new models for the cases with a wide variation of
input data. Finally the accuracy of prediction through modified
models is compared with that of original models and discussed.
3.3.1. Description and results of MOD-MODEL II
As discussed earlier in the paper, one of the limitations of
MODEL II is the strong impact of gasifier type on the accuracy of this

model in the prediction of product gas composition. Therefore,
including equations that distinguish between CFB and BFB gasifiers

is one step towards “generality” for this model. However, MODEL II
is more accurate for ER values greater than or equal to 0.3 (which is
more typical for CFB gasifiers). This “non-generality” factor can also
be tackled using empirical regression with a wider range of ER. The
new dataset that is used for the definition of new empirical equations is collected from experimental studies of different BFB
[20,31,35e41] and CFB gasifiers [17,32,33,42e45] with a wider
range of ER, temperature, S/B and load compared to the empirical
equations used in the original model. A new equation, Equation (5),
which relates CH4 content to the ER value, is obtained from Fig. 3
and used to substitute the corresponding equation in the original
model. In this equation CH4 content is calculated as volume fraction
of the product gas. Implementing this equation in the CALCULATOR
block, the “execution sequence” of result should be before RSTOIC to
adjust the “molar extend” of the respective reaction. As given in the
setup explanation of RSTOIC in ASPEN plus, molar extend is defined
as the number of moles generated for any component divided by its
stoichiometric coefficient. All these terms are explained in ASPEN
plus unit operation user guide. Equation (5) is derived from 92 set
points from tests on CFB gasifiers and 44 set points from tests on
BFB gasifiers

CH4 ¼ 1:207 Â ERÀ0:92

(5)

In Fig. 4-a, the results of MOD-MODEL II are compared with
those of the original MODEL II and the experimental data for a BFB

gasifier from Kim et al. [20]. This gasifier is operated isothermally at
800  C while ER is varied from 0.19 to 0.32. The circular points are
experimental results, and the dashed lines represent MODEL II and
MOD-MODEL II. For all 4 components, the accuracy of the modified
model is improved compared to the original model, and the average
overall error for the modified model is 30.9% compared to 44% for
the original model.
Fig. 4-b illustrates the comparison about MOD-MODEL II and the
original MODEL II for the CFB gasifier tested by Miao et al. [34]. This
gasifier is operated in low temperature (688e784  C) while ER
value is varied in a small range of 0.22e0.3. As mentioned before,
circular points are experimental points and dashed lines of
different size represent MODEL II and MOD-MODEL II. As shown in
Fig. 4-b, there is not remarkable difference between the two
models.
3.3.2. Description and results of MOD-MODEL III
According to Table 9 and the limitations highlighted for MODEL I
and MODEL III, combining MODEL I and MODEL III is suggested as
another modification strategy to improve the accuracy of prediction for wider range of input data and consequently improve the
generality of the original models by overcoming the limitations.


G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

(a)

79

(b)

Fig. 6. -a. Comparison of MODELS I and III for CFB I [32] (set points 1 to 5) with MOD-MODEL-III after applying modifications. -b. Comparison of MODELS I and III for BFB II [36] (set
points 1 to 5, ER variation) with MOD-MODEL-III after applying modifications.


80

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

Fig. 7. Average overall error level of modified models compared to original models for CFB gasifier, BFB gasifiers and all gasifiers.

The new modified model is called MOD-MODEL III. As shown in
Fig. 5, the devolatilization and char gasification parts are taken
directly from MODEL III, whereas the volatile gasification mainly
controlled by steam and restricting the equilibrium for methane
formation reaction by the QET method is added based on the
approach in MODEL I. Methane formation reaction ”temperature
approach” is set to À350  C.
Applying this modification has been evaluated by experimental
data from CFB gasifier by Wu and et al. [32] and BFB gasifier by Turn
and et al. [36]. The results are shown in Fig. 6-a and -b. The overall
accuracy and the generality of MODEL III has been improved in both
cases. In the case of CFB I, the overall error is reduced from 50% to
40% while in the case of BFB II, it is reduced from 36% to 8%.
Moreover, the predictive capability of MODEL III has been
improved, especially for CH4 and H2 at high temperatures and high
ER conditions. This can be seen in Fig. 6-a at set points 4 and 5, also
in Fig. 6-b at ER values around 0.2 and more. On the other hand, in
the case of CFB I, the accuracy in predicting CO and CO2 content is
decreased in MOD-MODEL III, but the trend of changes in value of
contents by variation of ER and temperature, follows the experimental results. The link between CO, H2 and CO2 content is CO-shift

reaction. Additionally, since H2 and CH4 are related by the methanation reaction, which is assigned as restricted equilibrium by the
QET method, the amount of these two components are accurately
predicted.
However, even though H2 content of product gas, as one of the
effective components in the CO-shift reaction, is well predicted, CO
is still overestimated and CO2 is underestimated. The lack of accuracy for these two components compared with the performance
of each original model (MODEL I and III) can be connected to the
kinetic and hydrodynamic equations used for MODEL III. However,
the reason why MOD-MODEL III shows worse results than MODEL

III for these components can be explained by referring to the COshift reaction. Based on this reaction, when H2 amount is reduced
(MOD-MODEL III predicts lower content of H2 than MODEL III), the
reaction is shifted to the left side, which results in more CO and less
CO2, which can be clearly observed from the MOD-MODEL III results compared to MODEL III.
In the case of BFB II, as shown in Fig. 6-b, the accuracy of the
modified model is also improved.
Fig. 7 summarizes the average overall error of each model in
predicting the content of major components in the product gas. It
clearly shows that the accuracy of MODEL III has been improved in
the case of CFB gasifiers, BFB gasifiers and all the gasifiers by MODMODEL III. However in case of CFB gasifiers the most accurate
model is MODEL II which has not been improved substantially in
MOD-MODEL II. This shows that changing the conversion equation
to the Equation (5) does not improve the results for CFB gasifiers.
Based on the results for BFB gasifiers, the accuracy of MODMODEL II is slightly better than the original model. Generally a
small improvement in accuracy of MODEL II has been obtained by
developing MOD-MODEL II, as shown in the case of all gasifiers.
Generally it can be concluded that applying some minor
modification and changes in the presented original models can
improve the accuracy of models especially in the case of MODMODEL III which shows the most accurate prediction compared
to all other models. This further adds up to the generality of the

model to give more accurate prediction in different operating
condition and gasifier design.
4. Conclusions
Using the equilibrium approach as the basis for modeling gasification enables the model to be independent of the detailed design


G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82

information of the gasifier. However, improving the accuracy of EM
by different approaches can introduce limitations and sources of
non-generality to EM.
In this study, the performance of three EM modification approaches have been studied. To investigate the effect of different
operating parameters on the predictive potential of the models,
different experimental cases have been selected for validation. The
quasi-temperature modeling approach (MODEL I) shows the least
sensitivity to gasifier type compared to other models. The modeling
approach based on including empirical correlations with ER values
(MODEL II) shows the highest inaccuracy and higher sensitivity to
all tested parameters, and the approach of including kinetics and
hydrodynamics (MODEL III) shows the best accuracy but only for
specific and limited input data.
By including more experimental data for the determination of
the empirical correlations, as performed in MOD-MODEL II, better
accuracy in the prediction of product gas composition at a larger
range of operating conditions is reached. This is mainly the case for
BFB gasifiers. Additionally, combining the QET method used in
MODEL I with the kinetic and hydrodynamic approach in MODEL III
results in a modified model (MOD-MODEL III) that shows lower
overall error than not only MODEL III but also the other models
presented in this study. However, CO and CO2 content of product

gas are not well predicted in all of the conditions.

[6]

[7]

[8]

[9]
[10]

[11]
[12]

[13]

[14]

[15]
[16]

Acknowledgments

[17]

This work is carried out in the Swedish Gasification Centre
consortium. The Swedish Energy Agency and the academic and
industrial partners are gratefully acknowledged.

[18]

Appendix A. Supplementary data
[19]

Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.biombioe.2016.05.002.
[20]

Nomenclature
Number of samples À
Relative error À
The experimental results Volume fraction (%)
The predicted results Volume fraction (%)
The overall error À
Average overall error À
Variation width
Equivalence ratio (kg) air/(kg) stoichiometric air
Steam to biomass ratio (kg) steam/(kg) dry and ash free
biomass
CFB
Circulating fluidized bed
BFB
Bubbling fluidized bed
QET
Quasi equilibrium temperature
RGIBBS Gibbs reactor in ASPEN plus
RSTOIC Stoichiometric reactor in ASPEN plus
RCSTR Continuous steering tank reactor in ASPEN plus
n
relE i

yie
yip
OEi
OE
VW
ER
S/B

[21]
[22]
[23]

[24]
[25]
[26]

[27]

[28]

[29]
[30]

References
[31]
[1] G.J. Stiegel, R.C. Maxwell, Gasification technologies: the path to clean,
affordable energy in the 21st century, Fuel Process Technol. 71 (2001) 79e97.
[2] M. Puig-arnavat, J.C. Bruno, A. Coronas, Review and analysis of biomass
gasification models, Renew. Sustain. Energy Rev. 14 (9) (2010) 2841e2851,
/>[3] A. Gomez Barea, B. Leckner, Modeling of biomass gasification in fluidized bed,

Prog. Energy Combust. Sci. 36 (2010) 444e509.
[4] R. Radmanesh, J. Chaouki, C. Guy, Q. Hc, Biomass gasification in a bubbling
fluidized bed reactor: experiments and modeling, AICHE J. 52 (12) (2006).
[5] B. Buragohain, P. Mahanta, V.S. Moholkar, Thermodynamic optimization of

[32]

[33]
[34]
[35]

81

biomass gasi fi cation for decentralized power generation and Fischer
-Tropsch synthesis, Energy 35 (6) (2010) 2557e2579, />10.1016/j.energy.2010.03.003.
X. Li, J. Grace, C. Lim, A. Watkinson, A. Ergüdenler, Equilibrium modeling of
gasification: a free energy minimization approach and its application to a
circulating fluidized bed coal gasifier, Fuel 80 (2001) 195e207.
S.S. Sadaka, A.E. Ghaly, M.A. Sabbah, Two-phase biomass air-steam gasiÿcation
model for uidized bed reactors: part III d model validation, Biomass Bioenergy 22 (2002) 479e487.
J.F. Bilodeau, N. Therien, P. Proulx, S. Czernik, D.D.G. Chimique, A mathematical
model of fluidized bed biomass gasification, Can. J. Chem. Eng. 71 (1993)
549e557.
V. Gururajan, PA, JA, Mathematical modeling of fluidized bed coal gasifiers,
Trans.IChemE 70a (1992) 211e238.
H. Yoshida, F. Kiyono, H. Tajima, A. Yamasaki, Two-stage equilibrium model
for a coal gasifier to predict the accurate carbon conversion in hydrogen
production, Fuel 87 (2008) 2186e2193.
Z.A. Zainal, A. Rifau, G.A. Quadir, K.N. Seetharamu, Experimental investigation
of a downdraft biomass gasifier, Biomass Bioenergy 23 (2002) 283e289.

S.R.A. Kersten, A. van der Drift, W. Prins, W.P.M. van Swaaij, Interpretation of
biomass gasification by “Quasi”-equilibrium models, in: 12th, European
Biomass Conference; Biomass for Energy, Industry and Climate Protection,
2002. Amsterdam.
Y. Lim, U. Lee, Quasi-equilibrium thermodynamic model with empirical
equations for air e steam biomass gasification in fluidized beds, Fuel Process
Technol. 128 (2014) 199e210, />P. Schl€
apfer, J. Tobler, Theoretische und praktische Untersuchungen über den
Betrieb von Motorfahrzeugen mit Holzgas. schweizerische gesellschaft fur das
stadium der motorbrenstoff, d. Gesellschaft, Bern: Selbstverl, 1937.
W. Gumz, Gas Producers and Blast Furnaces, John Wiley & Sons, Inc., New
York, 1950.
X.T. Li, Biomass Gasification in Circulating Fluidized Bed, University of British
Columbia, 2002.
X.T. Li, J.R. Grace, C.J. Lim, A.P. Watkinson, H.P. Chen, J.R. Kim, Biomass gasification in a circulating fluidized bed [Internet], Biomass Bioenergy 26 (2)
(2004) 171e193. Available from: />S0961953403000849.
X. Meng, W. de Jong, N. Fu, A.H.M. Verkooijen, Biomass gasification in a 100
kWth steam-oxygen blown circulating fluidized bed gasifier: effects of
operational conditions on product gas distribution and tar formation, Biomass
Bioenergy
35
(7)
(2011)
2910e2924,
/>j.biombioe.2011.03.028.
mez-barea, F.B. Vidal, P. Ollero, Air e steam gasification of
M. Campoy, A. Go
biomass in a fluidised bed: process optimisation by enriched air, Fuel Process
Technol.
90

(5)
(2009)
677e685,
/>j.fuproc.2008.12.007.
Y. Doo Kim, C. Won Yang, B. Jong Kim, K. Su Kim, J. Woo Lee, J. Hong Moon, et
al., Air-blown gasification of woody biomass in a bubbling fluidized bed
gasifier, Appl. Energy 112 (2013) 414e420, />j.apenergy.2013.03.072.
M.B. Nikoo, N. Mahinpey, Simulation of biomass gasification in fluidized bed
reactor using ASPEN PLUS, Biomass Bioenergy 32 (12) (2008) 1245e1254.
Y. Wang, C.M. Kinoshita, Kinetic model of biomass gasification, Sol. Energy 51
(1) (1993) 19e25.
J. Corella, A. Sanz, Modeling circulating fluidized bed biomass gasifiers. A
pseudo-rigorous model for stationary state, Fuel Process Technol. 86 (2005)
1021e1053.
Aspen Plus User Guide, 2006. Cambridge,MA.
G. Mirmoshtaghi, H. Li, E. Dahlquist, E. Thorin, Bio-methane production
through different biomass gasifiers, in: ICAE2013, 2013, pp. 1e10.
D. Bacon, J. Downie, J. Hsu, J. Peters, Modeling of fluidized bed gasifiers, in:
R.P. Overend, K. TM, L.K. Mudge (Eds.), Fundamentals of Thermochemical
Biomass Conversion, Elsevier Applied Science, London, 1985, pp. 717e732.
W. Doherty, A. Reynolds, D. Kennedy, The effect of air preheating in a biomass
CFB gasifier using ASPEN Plus simulation, Biomass Bioenergy 33 (9) (2009)
1158e1167, />I. Hannula, E. Kurkela, A semi-empirical model for pressurised air-blown
fluidized-bed gasification of biomass, Bioresour. Technol. 101 (12) (2010)
4608e4615. />R. Sotudeh-Gharebaagh, R. Legros, J. Chaouki, J. Paris, Simulation of circulating
fluidized bed reactors using ASPEN PLUS, Fuel 77 (4) (1998) 327e337.
E. Kurkela, P. Stahlberg, J. Laatikainen, Pressurized Fluidized Bed Gasification
Experiments with Wood, Peat, and Coal at VTT[ Espo, Finland] in 1991e1992.
1: Test Facilities and Gasification Experiments with Sawdust. Espo, 1993.
P.M. Lv, Z.H. Xiong, J. Chang, C.Z. Wu, Y. Chen, J.X. Zhu, An experimental study

on biomass air e steam gasification in a fluidized bed, Bioresour. Technol. 95
(2004) 95e101.
W. Jianzhi, X. Bingyan, L. Zhenfang, Z. Xiguang, Performance analysis of a
biomass circulating fluidized bed gasifier, Biomass Bioenergy 3 (2) (1992)
105e110.
A. Van Der Drift, J. Van Doorn, J.W. Vermeulen, Ten residual biomass fuels for
circulating fluidized-bed gasification, Biomass Bioenergy 20 (1) (2001) 45e56.
Q. Miao, J. Zhu, S. Barghi, C. Wu, X. Yin, Z. Zhou, Model validation of a CFB
biomass gasification model, Renew. Energy 63 (2014) 317e323.
I. Narvaez, A. Orio, M.P. Aznar, J. Corella, Biomass gasification with air in an


82

[36]
[37]

[38]

[39]

[40]

G. Mirmoshtaghi et al. / Biomass and Bioenergy 91 (2016) 69e82
atmospheric bubbling fluidized bed. Effect of six operational variables on the
quality of the produced raw gas, Ind. Eng. Chem. Res. 5885 (95) (1996)
2110e2120.
S. Turn, C. Kinoshita, D. Ishimura, An experimental investigation of Hydrogen
production, Int. J. Hydrogen Energy 23 (8) (1998) 641e648.
W.A. Wan, A. Karim Ghani, R. Alipour Moghadam, M.A. Mohd Salleh, A.B. Alias,

Air gasification of agricultural waste in a fluidized bed gasifier: hydrogen
production performance, Energies 2 (2009) 258e268.
U. Arena, L. Zaccariello, M.L. Mastellone, Gasification of natural and waste
biomass in a pilot scale fluidized bed reactor, Combust. Sci. Technol. 182
(2010) 625e639.
V. Skoulou, G. Koufodimos, Z. Samaras, A. Zabaniotou, Low temperature
gasification of olive kernels in a 5-kW fluidized bed reactor for H2-rich producer gas, Int. J. Hydrogen Energy 33 (22) (2008) 6515e6524, http://
dx.doi.org/10.1016/j.ijhydene.2008.07.074.
R. Radmanesh, J. Chaouki, C. Guy, Q. Hc, Biomass gasification in a bubbling

[41]

[42]

[43]

[44]

[45]

fluidized bed reactor: experiments and modeling, AICHE J. 52 (12) (2006)
4258e4272.
P. Subramanian, A. Sampathrajan, P. Venkatachalam, Fluidized bed gasification of select granular biomaterials, Bioresour. Technol. 102 (2) (2011)
1914e1920, />K.N. Sheeba, J.S. Chandra, S. Jaisankar, Energy for sustainable development air
gasification characteristics of coir pith in a circulating fluidized bed gasifier,
Energy Sustain Dev. 13 (2009) 166e173.
E. Kurkela PS and JL. Pressurized Fluidized Bed Gasification Experiments with
Wood, Peat, and Coal at VTT 1: Test Facilities and Gasification Experiments
with Sawdust,” VTT Espo, Finland;
~ ez, A. Cabanillas, J.M. Sa

nchez, Gasification of leached orujillo
P. García-Iban
(olive oil waste) in a pilot plant circulating fluidised bed reactor. Preliminary
results, Biomass Bioenergy 27 (2) (2004) 183e194.
Lars Waldheim, C. Fredriksson, Evaluation of the Pilot Plant Test on Cane
Trash, 2002.