Waste Management 73 (2018) 476–486

Contents lists available at ScienceDirect

Waste Management
journal homepage: www.elsevier.com/locate/wasman

Fluid dynamics model on fluidized bed gasifier using agro-industrial

biomass as fuel
Tamer M. Ismail a,⇑, M. Abd El-Salam b, Eliseu Monteiro c,d, Abel Rouboa e,f
a
Mechanical Engineering Department, Suez Canal University, Ismailia, Egypt
b
Department of Basic Science, Cairo University, Giza, Egypt
c
CIENER-INEGI/Faculty of Engineering, University of Porto, Porto, Portugal
d
C3i – Interdisciplinary Center for Research and Innovation, Polytechnic Institute of Portalegre, Portalegre, Portugal
e
CIENER-INEGI/Engineering Department of University of Trás-os-Montes e Alto Douro, Vila Real, Portugal
f
MEAM Department, University of Pennsylvania, Philadelphia, PA, USA

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

Article history: The present study shows the experimental and numerical results of thermal gasification of biomass, on
Received 3 February 2017 the energy potential of agro-industrial waste from the Portalegre region. Gasification tests were per-
Revised 4 June 2017 formed in a pilot-scale fluidized bed gasifier, in order to study the behavior of peach stones and miscant-
Accepted 5 June 2017
hus to investigate the effect of gasification temperatures at 750 °C, 800 °C and 850 °C at a constant
Available online 21 June 2017
biomass flow rate of 45 kg/h. In order to optimize the operating conditions of the biomass gasification
process, a numerical model is developed namely COMMENT code. This model is a computer model of
Keywords:
two dimensions describing the biomass gasification processes in a fluidized bed gasifier using peach
CFD
Fluidized bed
stone and miscanthus as fuel. Both phases, solid and gaseous, were described using an Eulerian-
Gasification Eulerian approach exchanging mass, energy, and momentum. The numerical model results are then com-
Peach stone pared with experimental results. The produced results show the impact of the increased temperature in
Miscanthus the calorific value of the syngas. The tests carried out at 750 °C shown an increase in CO2 and N2 and a
Biomass decrease of CO in the range of 5% comparing to the tests carried out at 850 °C. In addition, increased tem-
perature favors a decrease in tar production in thermal gasification process. Numerical results shows to
be in good agreement with the experimental data.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction
At the beginning of the 19th century, gasification technology
was widely applied to convert biomass into synthetic gases used
to produce energy and electricity. With the opportune technology,
energy generation from biomass does not participate in the green-
house effect, because the biomass fuel is carbon neutral to the sur-
roundings (Saidura et al., 2011).
It has been recently observed that there is an outstandingly fast
decrease in fossil fuel sources in coincidence with the increased
demand for fossil fuels, environmental concerns, and energy secu-
rity. Thus, a lot of attention has been focused on technologies of
biomass and municipal solid wastes.
⇑ Corresponding author.
E-mail addresses: temoil@aucegypt.edu, tamer.ismail@eng.suez.edu.eg
(T.M. Ismail), mohamedelsheikh@cu.edu.eg (M. Abd El-Salam), elmmonteiro@
portugalmail.pt (E. Monteiro), rouboa@utad.pt (A. Rouboa).
Among these, combustion and gasification are two of the most
important technologies related to the type of fuel that helps con-
trol greenhouse gas emissions (Kirkels and Verbong, 2011). Cur-
rently, biomass and biofuels are being considered because of
their environmentally-friendly characteristics and their ability to
supply much more energy. One of the most pressing concerns
regarding the use of biomass as a renewable energy source is
increasing its efficiency. In order to improve energy efficiency it
is necessary to optimize the gasification process, by developing
numerical models that optimize the design and operation
conditions.
Therefore, gasification technology offers the possibility of using
clean and efficient energy, from carbon-based feedstock that would
otherwise be expelled away as waste and biomass. Moreover, the
gasification technology will assist developing countries in relying
on renewable energy more than oil in industry, and in generating
more electricity. The resulting gaseous fuel can be employed, as
an alternative of natural gas, to generate electricity and act as

http://dx.doi.org/10.1016/j.wasman.2017.06.018
0956-053X/Ó 2017 Elsevier Ltd. All rights reserved.
 T.M. Ismail et al. / Waste Management 73 (2018) 476–486 477

Nomenclature

A pre-exponent factor, particle surface area [1/s, m2] Sh Sherwood number

CD drag coefficient SU source term
Cmix mixing rate constant Tenv environment temperature [K]
Cp specific heat capacity [J/kg K] Tg gas temperature [K]
Cw,g moisture concentration in the gas phase [kg/m3] Ts solid temperature [K]
Cw,s moisture concentration at the solid phase [kg/m3] t time [min]
Dg mass diffusion coefficient of gas [m2/s] X molar concentration [kmol
kmol
]
DO2 mass diffusion coefficient of oxygen [m2/s] Yv mass fraction of volatile matter
Ds particle mixing coefficient [m2/s] u velocity component in x-direction [m/s]
dp particle diameter [m] v velocity component in y-direction, volume of particle
E activation energy [kJ/mol] [m/s, m3]
Eb blackbody emission of the gas [W]
e coefficient of restitution for particle collisions Greek letters
go radial distribution function a absorption coefficient
Hevp evaporation heat of the solid material [J/kg] b gas-solid interphase drag coefficient
hf enthalpy of formation [J/kg] d Stefan-Boltzmann constant [W/m2 K4]
hrs radiation heat transfer coefficient [m/s] e dissipation rate of turbulent kinetic energy [m2 s3]
hrv effective radiation heat transfer coefficient of the voids
emissivity
[m/s] f random number that obeys the Gauss distribution
hs convective mass transfer coefficient between solid and Hs particle phase pseudo-temperature [m2/s2]
gas [m/s] kg thermal dispersion coefficient
hs0 convection heat transfer coefficient between solid and kmix effective dispersion coefficient
gas [W/m2 K] l dynamic viscosity [kg/m s]
I radiative intensity [W] q density [kg/m3]
K extinction coefficient rp scattering coefficient
k turbulent kinetic energy [m2/s2] ss stress tensor [Pa]
kd diffusion rates [kg/atm m2 s] U dependent variable
kf thermal conductivity of the fluid [W/mK] / void fraction in bed
ks thermal conductivity of the pure [W/mK] v CO/CO2 ratio
keff effective thermal conductivity [W/mK]
keff,0 the thermal conductivity for no fluid flow [W/mK] Subscripts
ls equivalent thickness a layer of solid [m] b bulk
M molecular weight [kg/kmol]
c char burnout
PO2 partial pressure of oxygen [Pa] eff effective
Pw partial pressure [Pa] f fluid
Qcr heat absorbed by the solid [W] g gas
qr radiative flux density [W]
p particle
R gas universal constant [J/kmol K] s solid
Revp moisture evaporation rate [kg/s] sb solid bulk
Rc char consumption rate [kg/s] sg solid to gas
Re Reynolds number
T total
Rv volatile matter in solid rate [kg/s] v volatile matter
Sc Schmidt number

raw material to make chemicals and liquid fuels. This process
requires a gasifying agent such as air, air/steam or pure oxygen
under certain temperatures and pressure to completely convert
carbonaceous materials into a producer gas (Jangsawang et al.,
2006).
Computational fluid dynamics (CFD) modeling is a very elegant
and economical tool to study gasification and combustion pro-
cesses. So when a reliable CFD model is developed, it supports
the setting and optimization of such systems. CFD draws tempera-
ture profiles of solid and gas phases in bed, predicting the temper-
ature profile of furnaces, the concentration of species and heat flux.
A reliable CFD model is considered as one of the most important
and the best approaches to predict the critical results and the
important requirements to control and guarantee an efficient
result.
There are two different models used to key out the gas-solid
flow: discrete element method (DEM) and the continuum descrip-
tion model. Discrete element method is the most popular model
used; merely due to the increasing in the particle number the
model becomes more expensive computationally, because it
requires high memory resources to overcome these calculations
(Zhong et al., 2006). On the other hand, the continuum model,
based on the kinetic theory of granular flow (KTGF), became the
method generally employed in identifying the two phase model,
particularly because it is less demanding in terms of central pro-
cessing unit and memory resources (Ding and Gidaspow, 1990;
Gidaspow, 1994; Schmidt and Renz, 2000; Subramanian et al.,
2011; Couto et al., 2015).
Fluidization is a phenomenon of the movement of solid parti-
cles. Each particle is subjected to the force of gravity, to the impul-
sion of Archimedes and to the drag force. The force balance makes
it possible to determine the minimum fluidization velocity, which
is the minimum velocity at which the gas must circulate in order to
move the particles. This velocity can be determined from empirical
correlations. Many researchers have been studying the modeling of
biomass gasification based on many parameters, such as the type
of gasifier reactor, gasification agent, and type of fuel used, as well
as the chemical process reactions (Benyahia et al., 2000;
478 T.M. Ismail et al. / Waste Management 73 (2018) 476–486

Goldschmidt et al., 2002; Wang et al., 2008; Gómez-Barea and available in Portugal at relatively low cost, which is why it was
Leckner, 2010). These simulations predict the behavior of gas and chosen to figure out in the present study.
solid flow within the gasifier.
In order to optimize the operating conditions of the biomass
gasification process, a numerical model was developed, namely 2. Materials and experimental set-up
the COMMENT code. It was previously adapted and validated to
combustion and gasification of waste and biomass in different This study was carried out using a pilot plant of thermal gasifi-
reactors (Ismail and Abd El-Salam, 2015; Ismail et al., 2016; cation of biomass, in a fluidized bed, (as shown in Fig. 1) and bio-
Ismail and Abd El-Salam, 2017). This two-dimensional model mass characterization equipment of the chemistry laboratory. The
describes the biomass gasification processes in updraft and flu- process of thermal gasification by bubbling fluidized bed occurs
idized bed gasifiers using different types of fuel such as peach through the addition of air and a catalyst (dolomite and/or olivine),
stone and miscanthus. Both phases, solid and gaseous, were burning some of the produced char and tar. This step occurs at
described using an Eulerian-Eulerian approach exchanging mass, temperatures approximately 750–850 °C, and the admission of
energy and momentum. atmospheric oxygen, in amounts below the stoichiometric reac-
Given the growing concerns about climate change, namely seri- tion, which are favorable for the combustion of the remaining char
ous dependence on fossil fuels and rising energy costs, many coun- and tar.
tries, including Portugal, have been promoting renewable energy From the heterogeneous reactions between the solid phase and
sources. The use of biomass comes forward as a contributing solu- the gas phase, the resulting gases are hydrogen, carbon monoxide
tion, since it is carbon neutral. In this study two kinds of biomass and carbon dioxide, methane, water vapor and nitrogen. The com-
were chosen, an agro-industrial residue (peach stone) and energy position of the gases depends, essentially on the rate of heating, the
crop (miscanthus), because they have a high potential in the final temperature and the amount of biomass and air introduced
country. into the reactor.
The recent Alqueva irrigation perimeters created in the Alentejo The catalytic mechanisms of the gasification reactions are
region of Portugal a high potential for the implantation of energy described below:
crops. The soil and ecological requirements of eight cultures (Pop-
ulus spp., Miscanthus  giganteus, Salix spp., Cynara cardunculus L., C þ 0:5O2 ! CO ðR1Þ
Panicum virgatum L., Arundo donax L., Hibiscus cannabinus L., and
Jatropha curcas L.) were previously studied. Although results C þ CO2 $ 2CO ðR2Þ
showed that all crops analyzed are able to grow in the Alqueva
perimeter, Miscanthus  giganteus is the most interesting from an C þ H2 O $ CO þ H2 ðR3Þ
ecological perspective. Moreover, the considerable potential for
energy production makes this biomass attractive for gasification C þ 2H2 $ CH4 ðR4Þ
and selected in the present study.
There are various studies on agro-industrial biomass gasifica- CO þ 0:5O2 ! CO2 ðR5Þ
tion including various types of fruit kernels. Among these, peach
stones depict a prominent biomass source, easily available in the
H2 þ 0:5O2 ! H2 O ðR6Þ
Mediterranean region, particularly in Portugal. The major advan-
tages accomplished by the thermochemical conversion of peach
CO þ H2 O $ CO2 þ H2 ðR7Þ
stone are the emission of less noxious flue gases and higher effi-
ciencies and higher heating values achieved when compared to
CH4 þ H2 O $ CO þ 3H2 ðR8Þ
coal or other biomass species. Moreover, peach stone is largely

Fig. 1. Sketch of the biomass gasification pilot plant. (a) Feed system; (b) Bubbling fluidized bed gasifier; (c) Heat exchangers; (d) Bag filter; (e) Condenser; (f) Flare; (g)
Condensate storage tank; (h) Vacuum pump; (i) Air compressor; (j) Air fan; (k) Air fan.
 T.M. Ismail et al. / Waste Management 73 (2018) 476–486 479

Table 1 gasifier is 4.15 m high and 0.5 m wide, with an admission capacity
Biomass properties. of 100 kg/h of biomass and 50 kg of dolomite in the bed.
Biomass properties Peach stone Miscanthus The synthesis gas is cooled by two heat exchangers; one of them
Proximate analysis (wt.% wet basis) also responsible for the preheating of the air used as gasifying
Volatile 63 64.4 agent. The ashes and the unconverted carbon particles are removed
Fixed carbon 29 22.1 in a bag filter. Another heat exchanger serves to remove tars from
Ash 1 2.1 the system through condensation, which are then forwarded to a
Moisture 7 11.4
tank. The last organ of the system is a vacuum pump, which
Ultimate analysis (wt.% dry and ash free basis) ensured total control of the system, and created a negative pres-
N 4.9 5.3
C 41.0 44.5
sure gradient. Once the predetermined temperature and the bio-
H 5.7 5.2 mass inlet flow were assured, the system stabilized for about
O 48.4 45.0 two hours and then the test was started. The tests lasted for two
S 0.0 0.0 hours each, with duplicate collection of the synthesis gas and
3
Density (kg/m ) 480 600
accounting for temperatures, flows, ashes and tars. An elaborative
Higher heating value 18.8 18.6
(MJ/kg biomass) description of the gasification pilot plant can be found in (Silva
Empirical formula CH1.668O0.885N0.102 CH1.402O0.758N0.102 et al., 2014).
Biomass properties analysis (Table 1) is integrated into the pre-
sent model. Table 2 shows the analysis of the experimental condi-
 Reaction (1) (R1) is endothermic, which means that the temper- tions and the syngas. Data will be tested and verified with the
ature increase leads to the decrease of CO and the increase of experimental data.
CO2.
 Reaction (2) (R2) is endothermic. Therefore, as the temperature 3. Computational model approach
increases inside the gasifier due to higher reactivity between
CO2 and the char present, this rises CO. However, when the tem- The two-dimensional mathematical model developed by Ismail
perature reaches 850 °C in the tests performed, there is still et al. (2016) was applied in this study. The gasification was mod-
enough char present to give continuity to the reaction, although eled using Fluent data base for a two-dimensional model to solve
not enough to maintain the high temperature inside the gasifier, the governing equations of mass, momentum, energy and species.
and as a result the amount of CO2 increases, while CO decreases. A multi-fluid Eulerian model, incorporates the kinetic theory of
 Reaction (3) (R3) is endothermic, which means that the increase granular flow, taking into consideration the process rates and
in temperature increases the production of CO and H2, increas- including tar and light hydrocarbons. Numerical procedure can
ing the consumption of char and water. also be seen in Ismail et al. (2016), for more details and validations
 Reaction (4) (R4) is exothermic. With increasing temperature of the current model.
there is a decrease in CH4 concentration, leaving more H2 pre-
sent in the synthesis gas.
 CO reacts with the available O2 (R5) producing CO2. 3.1. Chemical equations
 H2 reacts with O2 (R6), resulting H2O.
 Reaction (7) (R7) is exothermic; produces less CO2 and H2 at 3.1.1. Drying
high temperatures, which means that CO and H2O are less The rate of moisture release from solids can be expressed as
consumed. (Bruch et al., 2003):
 CH4 is reduced during reaction 8 (R8). Is an endothermic reac-
Rev p ¼ As hs ðC w;s  C w;g Þ T s < 100  C ð1Þ
tion and is favored by the increase of temperature, so CH4
decreases while H2 and CO increases the concentration.
Q cr
Or Rev p ¼ T s ¼ 100  C ð2Þ
H ev p
Biomass gasification apparatus used in the experiment is
depicted in Fig. 1. The system consists of two biomass silos, con-
Q cr ¼ Aðhs ðT g  T s Þ þ
dðT 41  T 4s ÞÞ
0
nected to a worm screw, which constitutes the feed system. The ð3Þ

Table 2
Experimental operating conditions and syngas analysis.

Experimental conditions Miscanthus Peach Stone

Run 1 Run 2 Run 3 Run 4 Run 5 Run 6
Temperature of gasification (°C) 750 800 850 750 800 850
Admission biomass (kg/h) 45 45 45 45 45 45
Air flow rate (Nm3/h) 35 50 90 55 57 60
Equivalence ratio 0.18 0.26 0.47 0.29 0.30 0.32
Syngas fraction (%vol. dry basis)
H2 9.8 11.0 9.7 8.5 7.4 7.7
CO 17.4 13.3 12.0 15.7 14.9 12.9
CH4 6.0 4.0 3.5 4.1 4.2 3.7
CO2 15.2 15.8 16.8 15.1 15.2 16.5
N2 47.6 49.7 52.5 51.1 52.2 53.5
C2H2 0.1 0.1 0.1 0.1 0.1 0.2
C2H4 1.5 0.6 0.3 1.1 1.1 0.9
C2H6 0.1 0.0 0.0 0.2 0.1 0.0
Syngas LHV (MJ/Nm3) 6.4 4.7 4.1 5.2 5.0 4.4
480 T.M. Ismail et al. / Waste Management 73 (2018) 476–486

3.1.2. Pyrolysis C þ 0:5O2 ! CO

From a chemical point of view, pyrolysis involves an indefin-
able number of intermediate reactions between the degradation
of wood and the formation of pyrolysis products. The modeling
of all these reactions is not reasonable or particularly useful for
the development of a reactor model. In order to describe the
conversion of the particle into pyrolysis products more easily,
simplified reaction schemes combining compounds formed in
generic entities such as non-condensable gases, condensable
gases (tars) and biomass have been developed. These models
may have parallel, consecutive, or even competitive reactions
(Boroson et al., 1989).
Cx Hy Oz ! char þ volatile gases ðCO2 þ H2 O þ CH4 þ H2 þ COÞ þ tar
C þ CO2 $ 2CO
C þ H2 O $ CO þ H2
Regarding the kinetic reactions for biomass in the present
model, data was obtained from (Gungor, 2008; Prando et al.,
2014; Loha et al., 2014).
The heterogeneous reactions are influenced by many factors,
namely: reactant diffusion, breaking up of char and interaction of
reactions and turbulence flow. In order to include both diffusion
and kinetic effects, the Kinetic/Diffusion Surface Reaction Model
(Zhang et al., 2012) was applied.

 P O2
Rc ¼
dY v Ev 1
þ K1
Rv ¼ qsb ¼ qsb Y v Av exp   ð4Þ Kr d
dt RT

0:75
5:06  107 Ts þ Tg
Ev Kd ¼ 
Av ¼ 3:63  104 1=s; ¼ 9340 K ð5Þ dp 2
R
 
In the present model, the pyrolysis product of biomass to gas, Ec
tar, and char; and secondary pyrolysis reactions of tar in the gas
K r ¼ Ac T s exp
RT s
and char is modeled in the form of a one-step global model (five-
reaction model). The complete set of reaction parameters for pyrol- where Rc is the char consumption rate, Kd is the diffusion rate and Kr
ysis is given in Table 3. These data were taken from (Chan et al., is the kinetic rate. Ac is the pre-exponential factor rate equal to
1985; Liden et al., 1988 and Gungor, 2008). 3 kgm2s1Pa1K1 and Ec/R is equal to 10300 K.
The competitive reaction scheme allows the products of pyrol-
ysis to vary depending on the conditions under which pyrolysis 3.2. Governing equations for gas and solid phases
occurs. The kinetic rate constant is expressed in terms of Arrhenius
equation: 3.2.1. Continuity equations


Ei
ki ¼ Ai exp   Gas phase:
RT s @ð/qg Þ
þ rð/qg ug Þ ¼ Ssg ð6Þ
@t
3.1.3. Gasification
Products of the gasification (reduction) zone are primarily dom-  Solid phase:
inated by heterogeneous reactions and homogeneous reactions in a @ðð1  /Þqs Þ
þ rðð1  /Þqs us Þ ¼ Ssg ð7Þ
complete absence of oxidants. In the reduction zone, equilibrium @t
may not be established because the reduction zone is intrinsically / is the void fraction in the bed, and can expressed as follows:
slower than the oxidation reactions by several orders of magnitude
(Di Blasi, 2004). _
/ ¼ / ð8Þ
Much char is formed in the combustion and pyrolysis zones and _
is expected to exist in the gasification, and due to the assumption
of Giltrap et al. (2003) the oxygen has been overwhelmed with the _
¼ 1  a1 ðRdry  Rdry Þ  a2 ðRv  Rv Þ  a3 ðRc  Rc Þ ð9Þ
reaction of C + O2 = CO2, and the products of pyrolysis have been _
totally cracked into components of low molecular weight (Loha
et al., 2014; Zhang et al., 2014). The five main chemical reactions, _ is the initial volume, _ is the volume of the particle, a1 ; a2 ; a3
together with the reaction rate expressions, are listed in Table 4 are coefficients equal to 1 or 0 according to the appearance of
(Siau, 1984; Bryden and Hagge, 2003): moisture, devolatilization, and char burnout respectively, which
appears in the form of source term Ssg .
3.1.4. Char combustion Ssg ¼ Rev p þ Rv þ Rc ð10Þ
The most used overall simplified heterogeneous reactions are
(Loha et al., 2014):
3.2.2. Momentum equation

Table 3  Gas phase:

Reaction parameters for pyrolysis. @ð/qg ug Þ
þ rð/qg ug ug Þ ¼ /rPg þ /qg g  bðug  us Þ þ r/sg
Reaction A (S1) E (kJ mol1) Heat of reaction (MJ kg1) @t
k1 1.44  104 88.6 -0.42
ð11Þ
biomass ! gas
k2
biomass ! tar
4.13  106 112.7 0.42 The gas–solid inter-phase drag coefficient, b, is calculated as fol-
k3 7.38  105 106.5 0.42 lows (De Souza-Santos, 1989; Di Blasi, 2004):
biomass ! char
k4
tar ! gas
4.28  106 107.5 0.04 ð1  /Þ2 lT qg ð1  /ÞjU g  U s j
1  105 107.5 0.04
b ¼ 150 2
þ 1:75 ð12Þ
k5
tar ! char /dp dp
 T.M. Ismail et al. / Waste Management 73 (2018) 476–486 481

Table 4
Main five chemical reactions.

Reaction Equation Reaction rate (mol m3 s1)

 
Boudouard reaction (R1) C þ CO2 ! 2CO r B ¼ 1:272ms T exp 22;645 CCO
T
 
Water-gas reaction (R2) C þ H2 O ! CO þ H2 r wg ¼ 1:272ms T exp 22;645
CH2 O
T

Methanation reaction (R3) C þ 2H2 ! CH4 r M ¼ 1:368  103 ms T exp 8078  7:087 CH2
  T
Shift reaction (R4) CO2 þ H2 ! CO þ H2 O r S ¼ 6:4  109 T exp 39;260 C0:5
T H2 CO2
 
Steam reforming reaction (R5) CH4 þ H2 O ! CO þ 3H2 5
r sr ¼ 3  10 T exp 15;042
C CH2 O
T CH 4

The gas phase stress tensor as follows: Where Hs granular temperature; e is the coefficient of restitution
h i 2
for particle collisions; g o is the radial distribution function.
sg ¼ lg rug þ ruTg  lT ðrug Þ ð13Þ For the radial distribution function of solid phase, g o , is
3
expressed as (Arastoopour, 2001):
lT ¼ lg þ lt ð14Þ "
1 #1
3 ð1  /Þ 3
go ¼ 1 ð23Þ
C l is constant, set as 0.09. 5 ð1  /Þmax
Fully turbulent flows develop within the fluidized bed. This is
why the turbulence model k-e is used. This model is based on The granular temperature Hs is a pseudo-temperature, which
the writing of velocities in the Navier-Stokes equations as the can be defined as:
sum of a mean flow and an instantaneous flow. The governing 3 1
transport equations for k and e respectively are: Hs ¼ u0s u0s ð24Þ
2 2


@ l The u0 s is the fluctuating velocity of the particles and can be
ð/qg kÞ þ rð/qg ug kÞ ¼ r / t rk þ /Gk  /qg
ð15Þ
@t rk determined by turbulence kinetic energy as follows:

 u0s ¼ fð2k=3Þ
0:5
@ l
ð/qg eÞ þ rð/qg ug eÞ ¼ r / t re þ /ðC e1 Gk  C e2 qg
Þ ð16Þ
@t re where f is a random number that obeys the Gauss distribution,
In the above equations Gk represents the generation of turbu- 0  f  1.
lence kinetic energy due to the mean velocity gradients and is
expressed as follows; 3.2.3. Energy equation

2  Gas phase:
Gk ¼ lt rug :½rug þ ruTg   rug ðlt rug þ qg kÞ ð17Þ
3 @ð/qg cpg T g Þ 0
þ rð/qg ug cpg T g Þ ¼ rðkg :rT g Þ þ As hs ðT g  T s Þ þ ST g
C e1 = 1.44 and C e2 = 1.92, the turbulent Prandtl numbers for k @t
and e are rk = 1 and re = 1.3, respectively (Launder and Spalding, ð25Þ
1974).

 Solid phase:
 Solid phase:
@ðð1  /Þqs cps T s Þ
@ðð1  /Þqs us Þ þ rðð1  /Þqs us cps T s Þ
þ rðð1  /Þqs us us Þ @t
@t ;
¼ rðkeff :rT s Þ þ ðrqr Þ  As hs ðT g  T s Þ þ ST s ð26Þ
¼ ð1  /ÞrPs þ ð1  /Þqs g  bðug  us Þ þ rð1  /Þss ð18Þ
The radiative flux density is given by Rosseland (1936) as
Here the stress tensor of the solid phase is expressed as follows:
follows:


2
ss ¼ lb  ls rus þ ls ðrus þ uTs Þ ð19Þ 16rT 2 16rT 3 2
3 rqr ¼  ðrTÞ2 þ ðr TÞ ð27Þ
K 3K
The equation of the solid shear viscosity, ls, is derived from as
The thermal dispersion coefficient kg can be expressed as
follows:
(Gidaspow, 1994; Huilin et al., 2003):
4
lb ¼ ð1  /Þqs dp g o ð20Þ kg ¼ keff ;0 þ 0:5  dp  U g  qg  C pg ð28Þ
3
In the above equation represents the bulk viscosity, which may ð1  /ÞDl
keff ;0 ¼ /ðkf þ hrv DlÞ þ   ð29Þ
be obtained as follows: kf
1= lv þ hrs þ ls =ks
rffiffiffiffiffiffi
4 Hs
ls ¼ ð1  /Þqs dp g o ð1 þ eÞ where ls = 2dp/3, ks is the thermal conductivity of the pure solid, lv,
5 p
pffiffiffiffiffiffiffiffiffiffi  2 hrv, hrs, Dl are written as follows (Zhou et al., 2005):
10qs dp pHs 4

þ 1 þ g o ð1  /Þð1 þ eÞ ð21Þ
96ð1 þ eÞ
g o 5 lv ¼ 0:151912Dl
kf
ð30Þ
kair
The solid pressure Ps is as following:

1
n
Ps ¼ ð1  /Þqs Hs þ 2ð1 þ eÞð1  /Þ2 g o qs Hs ð22Þ /ð1 
Þ Ts
hrv ¼ 0:1952 1 þ ð31Þ
2ð1  /Þ
100
482 T.M. Ismail et al. / Waste Management 73 (2018) 476–486


n

Ts adapted in each time step to the height of the gasifier. The stag-
hrs ¼ 0:1952  dp ð32Þ
2
100 gered grid was utilized, which set vectors at the boundaries of cells
and scalars at the center. The partial differential equations were
Dl ¼ 0:96795dp ð1  /Þ1=3 ð33Þ discretized using the Upwind Difference Scheme. A convergence
study, not shown here, shows that the solution is not sensitive to
kair is the air thermal conductivity, the mesh. This number of cells is sufficient because the solution
no longer evolves for denser meshes.
kair ðT g Þ ¼ 5:66  105 T g þ 1:1  102 ð34Þ
Mesh sensitivity studies were performed as part of the model
(n) is an empirical parameter related to the fuel packing condi- development work and this mesh density proved to be a reason-
tions. In this model, able compromise between the competing needs for accuracy and

 manageable run times.
ðm:g  0:39Þ
n ¼ 1:93 þ 0:67 exp  ð35Þ Grid resolution is very important in a numerical simulation. A
0:054 finer mesh commonly gives better calculation results, however, it
Source term of the energy equation for both gas and solid is cal- induces higher computational time. Therefore to choose an appro-
culated as follows: priate mesh size is crucial for accurate results. The currently avail-
able Eulerian-Eulerian model is usually closed with constitutive
STg ¼ Rev p  hf ;CO ð36Þ laws that are based on the assumption of homogeneity at the level
  of computational cells. In this particular case, the chosen cell size
M CO Y CO was about 12 times larger than the average particle size which
STs ¼ Rev p   ½hf ;CO2  hf ;CO   1 ð37Þ
M CO2 2 was shown to be able to effectively capture the hydrodynamics
in the fluidized bed gasifier (Gelderbloom et al., 2003; Agrawal
3.2.4. Species equations et al., 2001; Andrews et al., 2005; Guenther et al., 2002; Wang,
2008).
 Gas phase: In general, the continuous increase in mesh density may lead to
@ð/qg Y ig Þ slightly better results that are more grid-dependent. However, the
þ rð/qg ug Y ig Þ ¼ rðDig rð/qg Y ig ÞÞ þ SY g ð38Þ computational power currently available is still a significant
@t
restriction when using a finer mesh. Based on the previous studies,
for two-dimensional and three-dimensional simulations, the rea-
sonable grid size to be used in the current study would be a med-
 Solid phase: ium size grid as to reduce the computational time without
@ðð1  /Þqs Y is Þ compromising in the results accuracy.
þ rðð1  /Þqs us Y is Þ ¼ SY s ð39Þ
@t In such a complex model, it is sometimes difficult to define a
good initial condition. For this reason, the process was first simu-
3.3. Boundary conditions and numerical solution lated considering only flow and non-reacting heat transfer (also
known as ‘‘cold flow”) and after reaching conversion, reactive mul-
The solving of the previously described transport equations is tiphase flow was added.
carried out using the COMMENT software tool. COMMENT is a code
dedicated to coupled transfer applications in the fields of thermal, 4. Results and discussion
fluid mechanics and chemical engineering. The energy and mass
equations are implemented, as presented, in the above mathemat- 4.1. Results of thermal gasification testing
ical model. Uses the finite volume method for solving the equa-
tions (Patankar, 1980; Tao, 2001) by using the SIMPLE algorithm The experimental results are collected from the described pilot-
(Wakao and Kagei, 1982). Transport equations are generalized into scale heat gasification apparatus. Due to their cost, these experi-
a standard form: mental results are seldom collected and used to test and verify a
numerical model, and laboratory or small-scale gasifiers that can
ai;j Ui;j þ ai1;j Ui1;j þ aiþ1;j Uiþ1;j þ ai;j1 Ui;j1 þ ai;jþ1 Ui;jþ1 ¼ Si;j ð40Þ
be used instead. Fig. 2 shows a comparison between the present
The whole geometrical domain of the bed is divided into a num- model and the experimental work carried out in the pilot-scale flu-
ber of small cells and Eq. (40) is discretized over each cell and idized bed apparatus of the Polytechnic Institute of Portalegre,
solved numerically. The total number of mesh cells is 200,000. under different gasification conditions as described in Table 2.
The time step is 104 s and the gasification time of the biomass The relative errors between the simulated and experimental
was resolved by 18,000 time steps. The computational grid was data are detailed in this study. The relative error is defined as the

Fig. 2. Comparison between experimental and numerical syngas composition for miscanthus (left, run#1) and peach stone (right, run #4).
 T.M. Ismail et al. / Waste Management 73 (2018) 476–486 483

Fig. 3. Numerical contours for syngas composition produced within the fluidized bed gasifier for peach stone under operating conditions of run# 2.

Fig. 4. Numerical contours for syngas composition produced within the fluidized bed gasifier for miscanthus under operating conditions of run# 5.

absolute difference between the simulated and experimental val- From these comparisons, we can deduce that the kinetic theory
ues divided by the experimental value; the results obtained by this of granular flow introduced in the present study, with a heat trans-
model show a good agreement with the experimental results as fer model based on the Rosseland model between particle and gas
shown Fig. 2. phases with the bed wall, increases the accuracy of predicting the
484 T.M. Ismail et al. / Waste Management 73 (2018) 476–486

Fig. 5. Comparison between experimental work and numerical model for the effect of temperature on syngas composition for miscanthus under the operating conditions of
run#1, run# 2 and run# 3.

Fig. 6. Comparison between experimental work and numerical model for the effect of temperature on syngas composition for peach stone under the operating conditions of
run#4, run# 5 and run# 6.

temperature profile, as well as the composition of the produced
gas.
With regard to the numerical model results, Figs. 3 and 4 show
the mole fraction distribution obtained inside the reactor of the
gaseous species, from the peach stone and miscanthus gasification
respectively. The first column of each gas species defines the scale
(lower values – blue; higher values – red).
The syngas produced, H2, CH4, CO2, N2, CO, O2 and tar are
released in the bed due to devolatilization reaction, and indicated
by local high values. Higher amounts of CO and CO2 occur in the
freeboard region, due to the char burning with the oxygen remain-
ing in the bed. Tar is further decomposed into light gasses in the
freeboard, with a very small amount of tar left at the reactor exit.
 T.M. Ismail et al. / Waste Management 73 (2018) 476–486
Fig. 7. Comparison between experimental work and numerical model for the effect of temperature and admission biomass on LHV of miscanthus (left) and peach stone
(right).
In the following section, the thermal gasification was studied
with variation in gasification temperature (750 °C, 800 °C, and
850 °C).
4.2. Effect of temperature in the synthesis gas
The temperature of the gasifier is a major contributor for the
composition of the synthesis gas, since there are certain gasifica-
tion reactions that require specific temperatures. The main reac-
tions of the gasification are endothermic and thus enhanced by
increasing temperature.
In an autothermal gasification process the increase of the gasi-
fication temperature is obtained by adding more oxidant in order
to enhance the combustion of gaseous species. Therefore, increas-
ing the temperature also means increasing the equivalence ratio.
Figs. 5 and 6 show the comparison between experimental work
and numerical model of the syngas composition obtained, respec-
tively, from the miscanthus and peach stone gasification at differ-
ent temperatures and equivalence ratios.
The explanation of the results presented in Figs. 5 and 6 can be
formulated based on two directives:
– According to Le Chatelier’s principle, higher temperatures favor
products in endothermic reactions. Therefore, both H2 and CO
will increase and, on the other hand, CH4 and CO2 will decrease.
– An excessively low equivalence ratio value (<0.2) results in
numerous problems, including incomplete gasification and a
low heating value of the syngas. On the other hand, an exces-
sively high equivalence ratio (>0.4) results in excessive forma-
tion of combustion products, such as CO2 and H2O, at the
expense of fuel gases like H2 and CO.
These directives show that the effects of the temperature and
the equivalence ratio on producer gas composition are opposite,
which makes the explanation of the obtained results a hard task.
Nevertheless, it is possible to verify the good agreement between
the experimental and numerical results.
Fig. 7 shows the comparison between experimental work and
numerical model of the syngas composition obtained from the
peach stone gasification at different temperatures and equivalence
ratios.
From Fig. 7 it is possible to verify that the calorific value
decreases with increasing temperature and equivalence ratios for
miscanthus and peach stone gasification. The increase of the gasi-
fication temperature results in a higher conversion of biomass to
carbon dioxide, and the reduction of the concentration of CH4.
Simultaneously, the equivalence ratio is also increased, which
means that the H2 and CO contents are reduced. The combined
effect of the temperature and equivalence ratio leads to a reduction
485
in the gas heating value. These results are in agreement with
Karamarkovic and Karamarkovic, 2010.
5. Summary and conclusion
The present study examined the thermal gasification in the
bubbling fluidized bed at gasification temperatures in the range
of 750–850 °C. The effects of the temperature in the gasifier are
introduced. They play an important role in the final quality of
the synthesis gas. The tests performed at a higher temperature
and equivalence ratios resulted in a higher conversion of biomass
into a poorer gas production.
Less elevated temperatures tend to favor the thermal cracking
reactions and increased carbon monoxide concentration, which
increase the calorific value of the synthesis gas. In addition, this
study aims to demonstrate the feasibility of transforming waste
or crops into valuable gases and other products highly relevant
in terms of calorific value and interest to the chemical industry.
It also showed that the thermal gasification unit had a performance
very acceptable during rehearsals.
Mathematical models of gasification can be used for biomass
practical applications. The developed model was validated against
experimental data. Peach stone and miscanthus are used as bio-
mass substrate in a pilot-scale fluidized-bed gasifier to provide
experimental data for model validation proposes. The gas compo-
sition produced in the simulation was found to be in agreement
with the gas composition produced experimentally. The developed
model is capable of predicting the effects of gasification tempera-
ture and provides a sensitive analysis of the produced gas compo-
sition. The obtained data are essential to describe scenarios
concerning the potential use of these biomasses as a source of
energy in the Alentejo region of Portugal.
References
Agrawal, K., Loezos, P.N., Syamlal, M., Sundaresan, S., 2001. The role of mesoscale
structures in rapid gas-solid flow. J. Fluid. Mech. 445, 151–185.
Andrews, A.T.I., Loezos, P.N., Sundaresan, S., 2005. Coarse-grid simulation of gas-
particle flows in vertical risers. Ind. Eng. Chem. Res. 44, 6022–6037.
Arastoopour, H., 2001. Numerical simulation and experimental analysis of gas/solid
flow systems: 1999 Fluor-Daniel Plenary lecture. Powder Technol. 119 (2), 59–
67.
Benyahia, S., Arastoopour, H., Knowlton, T., Massah, H., 2000. Simulation of particles
and gas flow behavior in the riser section of a circulating fluidized bed using the
kinetic theory approach for the particulate phase. Powder Technol. 112 (1), 24–
33.
Boroson, M.L., Howard, J.B., Longwell, J.P., Peters, W.A., 1989. Product yields and
kinetics from the vapor phase cracking of wood pyrolysis tars. AIChE J. 35 (1),
120–128.
Bruch, C., Peters, B., Nussbaumer, T., 2003. Modelling wood combustion under fixed
bed conditions. Fuel 82 (6), 729–738.
Bryden, K.M., Hagge, M.J., 2003. Modeling the combined impact of moisture and
char shrinkage on the pyrolysis of a biomass particle. Fuel 82 (13), 1633–1644.
486 T.M. Ismail et al. / Waste Management 73 (2018) 476–486
Chan, W.C.R., Kelbon, M., Kriege, B.B., 1985. Modelling and experimental verification
of physical and chemical processes during pyrolysis of a large biomass particle.
Fuel 64 (11), 1505–1513.
Couto, N., Silva, V., Monteiro, E., Brito, P., Rouboa, A., 2015. Using an Eulerian-
granular 2-D multiphase CFD model to simulate oxygen air enriched
gasification of agro-industrial residues. Renew. Energy 77, 174–181.
De Souza-Santos, M.L., 1989. Comprehensive modelling and simulation of fluidized
bed boilers and gasifiers. Fuel 68 (12), 1507–1521.
Di Blasi, C., 2004. Modeling wood gasification in a countercurrent fixed-bed reactor.
AIChE J. 50 (9), 2306–2319.
Ding, J., Gidaspow, D., 1990. A bubbling fluidization model using kinetic theory of
granular flow. AIChE J. 36 (4), 523–538.
Gelderbloom, S.J., Gidaspow, D., Lyczkowski, R.W., 2003. CFD simulations of
bubbling/collapsing fluidized beds for three Geldart Groups. AIChE J. 49, 844–
858.
Gidaspow, D., 1994. Multiphase Flow and Fluidization: Continuum and Kinetic
Theory Descriptions. Academic press.
Giltrap, D.L., McKibbin, R., Barnes, G.R.G., 2003. A steady state model of gas-char
reactions in a downdraft biomass gasifier. Sol. Energy 74, 85–91.
Goldschmidt, M.J.V., Beetstra, R., Kuipers, J.A.M., 2002. Hydrodynamic modeling of
dense gas-fluidized beds: comparison of the kinetic theory of granular flow
with 3D hard-sphere discrete particle simulations. Chem. Eng. Sci. 57, 2059–
2075.
Gómez-Barea, A., Leckner, B., 2010. Modeling of biomass gasification in fluidized
bed. Prog. Energy Combust. 36, 444–509.
Guenther, C., Syamlal, M., Shadle, L., Ludlow, C., 2002. A Numerical Investigation of
An Industrial Scale Gas-Solids CFB. Presented at Proceedings of the Seventh
International Conference on Circulating Fluidized Beds, Ontario, Canada, pp.
483–488.
Gungor, A., 2008. Two-dimensional biomass combustion modeling of CFB. Fuel 87
(8), 1453–1468.
Huilin, L., Gidaspow, D., Bouillard, J., Wentie, L., 2003. Hydrodynamic simulation of
gas-solid flow in a riser using kinetic theory of granular flow. Chem. Eng. J. 95
(1), 1–13.
Ismail, T.M., Abd, El.-Salam M., 2015. Numerical and experimental studies on
updraft gasifier HTAG. Renew. Energy 78, 484–497.
Ismail, T.M., Abd El-Salam, M., 2017. Parametric studies on biomass gasification on
updraft gasifier high temperature air gasification. Appl. Therm. Eng. 112, 1460–
1473.
Ismail, T.M., Abd El-Salam, M., Monteiro, E., Rouboa, A., 2016. Eulerian-Eulerian CFD
model on fluidized bed gasifier using coffee husks as fuel. Appl. Therm. Eng.
106, 1391–1402.
Jansawang, W., Klimanek, A., Gupta, A., 2006. Enhanced yield of hydrogen from
wastes using high temperature steam gasification. J. Energ. Resour.-ASME 128
(3), 179–185.
Karamarkovic, R., Karamarkovic, V., 2010. Energy and exergy analysis of biomass
gasification at different temperatures. Energy 35 (2), 537–549.
Kirkels, A.F., Verbong, G.P., 2011. Biomass gasification: still promising? A 30-year
global overview. Renew. Sust. Energy Rev. 15 (1), 471–481.
Launder, B.E., Spalding, D.B., 1974. The numerical computation of turbulent flows.
Comput. Method. Appl. M. 3 (2), 269–289.
Liden, A.G., Berruti, F., Scott, D.S., 1988. A kinetic model for the production of liquids
from the flash pyrolysis of biomass. Chem. Eng. Commun. 65 (1), 207–221.
Loha, C., Chattopadhyay, H., Chatterjee, P.K., 2014. Three dimensional kinetic
modeling of fluidized bed biomass gasification. Chem. Eng. Sci. 109, 53–64.
Patankar, S.V., 1980. Numerical heat transfer and fluid flow. Hemisphere.
Prando, D., Patuzzi, F., Pernigotto, G., Gasparella, A., Baratieri, M., 2014. Biomass
gasification systems for residential application: an integrated simulation
approach. Appl. Therm. Eng. 71 (1), 152–160.
Rosseland, S., 1936. Theoretical Astrophysics: Atomic Theory and the Analysis of
Stellar Atmospheres and Envelopes. Clarendon, Oxford, UK.
Saidura, R., Abdelaziz, E.A., Demirbas, A., Hossain, M.S., Mekhilef, S., 2011. A review
on biomass as a fuel for boilers. Renew. Sustain. Energy Rev. 15 (5), 2262–2289.
Schmidt, A., Renz, U., 2000. Numerical prediction of heat transfer in fluidized beds
by a kinetic theory of granular flows. Int. J. Therm. Sci. 3, 885–9871.
Siau, J.F., 1984. In Transport Processes in Wood. Springer, Berlin Heidelberg.
Silva, V., Monteiro, E., Couto, N., Brito, P., Rouboa, A., 2014. Analysis of syngas
quality from Portuguese biomasses: an experimental and numerical study.
Energy Fuel 28 (9), 5766–5777.
Subramanian, P., Sampathrajan, A., Venkatachalam, P., 2011. Fluidized bed
gasification of select granular biomaterials. Bioresource Technol. 102 (2),
1914–1920.
Tao, W.Q., 2001. Numerical Heat Transfer. Xi’an Jiaotong University, Xi’an.
Wakao, N., Kagei, S., 1982. Heat and Mass Transfer in Packed Beds. Taylor & Francis.
Wang, J., 2008. High-resolution Eulerian simulation of RMS of solid volume fraction
fluctuation and particle clustering characteristics in a CFB riser. Chem. Eng. Sci.
63, 3341–3347.
Wang, L., Weller, C.L., Jones, D.D., Hanna, M.A., 2008. Contemporary issues in
thermal gasification of biomass and its application to electricity and fuel
production. Biomass Bioenerg. 32 (7), 573–581.
Zhang, Q., Dor, L., Fenigshtein, D., Yang, W., Blasiak, W., 2012. Gasification of
municipal solid waste in the Plasma Gasification Melting process. Appl. Energy
90 (1), 106–112.
Zhang, Y., Wang, J., Zhang, X., Sun, G., 2014. Experimental study of petroleum coke
steam gasification catalyzed by black liquor in a fluidized bed. Energy Procedia
61, 472–475.
Zhou, H., Jensen, A.D., Glarborg, P., Jensen, P.A., Kavaliauskas, A., 2005. Numerical
modeling of straw combustion in a fixed bed. Fuel 84 (4), 389–403.
Zhong, W., Zhang, M., Baosheng, J.I.N., Zhulin, Y.U.A.N., 2006. Three-dimensional
simulation of gas/solid flow in spout-fluid beds with kinetic theory of granular
flow. Chinese J. Chem. Eng. 14, 611–617.