INTRODUCTION
It is widely known that combustion of fossil fuels contributes to the buildup
of Carbon Dioxide (CO_{2}) in the atmosphere which contributes to the
global warming issue all around the world. In 2005, it is reported that the
concentration of CO_{2} was 379 ppm, approximately 180300 ppm more
than the equilibrium concentration for the last 650 000 years (Florin
and Harris, 2008). Hence, the search towards cleaner and renewable alternative
energy is attracting more attention.
Hydrogen is one of the potential alternative energy that could be used to replace
the existing fossil fuels. Hydrogen is expected to become a prominent energy
carrier for stationary and mobile power generation applications such as in transport,
industrial, commercial and residential applications (Florin
and Harris, 2008; Mahishi et al., 2008; Wu
et al., 2006). Clean and renewable hydrogen can be produced through
the biomass gasification process. Biomass gasification is the process of converting
solid biomass into gaseous fuel mainly hydrogen (H_{2}), carbon dioxide
(CO_{2}), carbon monoxide (CO) and methane (CH_{4}) (Foscolo
et al., 2007; Pfeifer and Hofbauer, 2008;
Rapagna et al., 1997) by heating it in a gasification
medium such as air, oxygen or steam (McKendry, 2002).
Biomass gasification process usually took place in various types of fluidized
bed reactor due to its excellent solid mixing, heat and mass transfer properties
(Mahishi et al., 2008; Wu
et al., 2006; Rapagna et al., 1997;
MarquardMollenstedt et al., 2004). For maximum
performance of solid mixing, fluidized bed reactor for biomass gasification
is best to be operated in bubbling flow regime (Foscolo
et al., 2007). The presence of bubble in the fluidized bed reactors
will ensure that the particles are circulated throughout the bed so that the
properties and process condition could be considered uniform. Ideally, good
quality of fluidization must have high population of bubbles, bed should be
large but bubbles should be small in size, homogeneously occupy the bed and
have low rise velocities (Busciglio et al., 2009;
Lim et al., 1995, 2009).
Deeper knowledge of the fluidized bed hydrodynamics would provide the base for
development of a fully predictive model.
Fluid flow, heat and mass transfer, chemical reaction, solid and fluid interaction
and other related phenomena that might involve in the fluidized bed can be model
and simulate using Computational Fluid Dynamic (CFD). Computational fluid dynamic
is a design and analysis tool that uses computers to simulate momentum, mass
and energy transfer equation into virtual fluidized bed reactor model. CFD modeling
technique is becoming widespread in the biomass thermochemical conversion equipment
such as fluidized bed, fixed bed, combustion furnace and many others (Wang
and Yan, 2008).
The recent development of mathematical modeling of particulate solids behavior
together with increasing computation power enables researcher to simulate the
behavior of fluidized powdered biomass and to link fundamental particle properties
directly to the powder behavior and predict the interaction between particles
and gaseous or liquid fluids. In this case, CFD modeling provides a fundamental
tool to support engineering design and research in multiphase system (Busciglio
et al., 2009).
This study highlights the model development for hydrodynamics study on the
effect of steam inlet velocity and solid initial bed height on the solid fluidization
in the fluidized bed gasifier using EulerianEulerian multiphase model coupled
with kinetic theory of granular flow approach (Bokkers
et al., 2004; Papadikis et al., 2009).
This model was validated using experimental and simulation data from Busciglio
et al. (2009) and Tamidi et al. (2009).
The validated model was used to obtain the most optimum condition for fluidized
bed gasifier for biomass gasification.
APPROACH AND METHOD
CFD is the science of predicting fluid flow, heat transfer, chemical reaction
and other related phenomena by solving numerical set of NavierStokes equations
(Wang and Yan, 2008). The results of CFD analysis are
relevant for conceptual studies of new design, detail product development, troubleshooting
and redesign. Besides, CFD modeling is also cost saving, timely, safe and easy
to scaleup (Wang and Yan, 2008). CFD analysis complements
testing and experimentation because CFD can reduce the total effort required
in the experiment design and data acquisition (Lim et
al., 1995). CFD codes turn computers into a virtual laboratory and perform
equivalent numerical experiment conveniently providing insight, foresight and
return on investment.
Various numerical techniques have been employed in the solution of the CFD
model equation and the most widely use numerical technique is discretization
method including finite difference, finite element and finite volumes method.
Finite volume is now the most commonly approached used in CFD code for its ease
in the understanding, programming and versatility. The most routinely used commercial
codes include ANSYS Fluent, ANSYS CFX, CFD2000 and many others (Busciglio
et al., 2009; Wang and Yan, 2008). For this
project, ANSYS Fluent v.6.3 was used for simulation of biomass gasification
in fluidized bed reactor.

Fig. 1: 
Domain dimension for biomass gasifier 
Table 1: 
Simulation parameters 

Working domain: Figure 1 shows the dimension of the working domain in the fluent. It is the representation of a 2D fluidized bed gasifier or reactor. Reactor inlet was specified as velocity inlet of steam and reactor outlet was specified as pressure outlet, which mean only compressible fluid can escape the reactor. The wall of the reactor was specified as stationary wall.
Simulation parameters: Table 1 shows the simulation parameters that are incorporated in the simulation during solver setting. Water vapor or steam is the primary phase in the system. Steam enters the reactor through the reactor inlet at a specified velocity. The secondary phase in the system is the carbon solid particle that is chosen to represent biomass in the reactor with modified parameter to suit the properties of biomass.
The properties of the solid phase are likely to have the sand like behavior of Geldart B Group. There is also gravitational force of 9.81 m sec^{2} is acting in downward direction of the gasifier.
Table 2: 
Value of variable 

Variables: The main objective of this simulation is to study the solid fluidization in the fluidized bed gasifier with respect to the change in inlet velocity of steam and solid initial bed height. The variables value of each parameter is shown in Table 2. The control variables are the initial condition of the system.
Mathematical model: The EulerianEulerian Multiphase Flow Model coupled
with granular kinetic theory has been used for the simulation of fluidized bed
gasifier. For the present case of twophase flow, the model has to solve several
equations related to scalar continuity balance equation, mass and momentum balance
equations.
Volume fraction: For EulerianEulerian model, the volume fractions are
assumed to be continuous functions of space and time and their sum is equal
to one since the volume of one phase can never be occupied by the other phase
as shown below (http://www.scribd.com/doc/20480734/FluentTraining):
Continuity equation: The continuity equation for gas and solid phases
are given by (http://www.scribd.com/doc/
20480734/FluentTraining):
Momentum equation: The momentum balance equations for each phase are
derived based on the assumption of there are no mass transfer between the two
phases and no lift force, external body force and virtual mass force acting
on the secondary phase of the system. The momentum balance equations for each
phase are as follow (http://www.scribd.com/doc/20480734/FluentTraining):
The interphase momentum change is further defined by the granular kinetic theory in order to estimate the rheological properties for the solid phase. Standard drag models are also employed to estimate the momentum exchange between phases at the particle boundaries.
The fluidsolid momentum exchange coefficient in the interphase momentum change
equation for dense fluidized bed can be further described by Gidaspow
et al. (1992) drag function as follow:
For ε_{g}>0.8:
For ε_{g}≤0.8:
Where:
For granular flows in the compressible regime where the solid volume fraction
is less than its maximum allowed value, a solid pressure is calculated independently
and used for the pressure gradient term in the granularphase momentum equation.
Because a Maxwellian velocity distribution is used for the particles, a granular
temperature is introduced into the model and appears in the expression for the
solid pressure and viscosities (Lun et al., 1984).
The solid pressure is composed of a kinetic term and particle collisions term:
Radial distribution function is a correction factor that modifies the probability
of collisions between grains when solid granular phase becomes dense. For one
solid phase, the equation for radial distribution function is as follow (http://www.scribd.com/doc/20480734/FluentTraining):
Solid stress tensor contains shear and bulk viscosities arising from particle
momentum exchange due to translation and collision. In this simulation, frictional
component of viscosity is assumed negligible. The solid shear stress equation
with collisional and (Gidaspow et al., 1992)
kinetic viscosity is as follow:
Solid bulk viscosity accounts for the resistance of the granular particles
to compression and expansion. In Fluent, the solid bulk viscosity has the following
form from Lun et al. (1984).
Kinetic Theory of Granular Flow (KTGF):
The transport equation derived from kinetic theory takes the form as follow
(http://www.scribd.com/doc/20480734/FluentTraining):
The diffusion coefficient is further described by Gidaspow
et al. (1992) as follow:
The collisional dissipation energy represent the rate of energy dissipate within
the solid phase due to collisions between particles. This term is represented
by the equation derived by Lun et al. (1984):
RESULTS AND DISCUSSION
Effect of steam inlet velocity to solid fluidization in the gasifier: Figure 2 shows the instantaneous expended bed height at different steam inlet velocity. From this graph, it is observed that at lowest inlet velocity which is at 3 U_{mf}, the expended bed height is the lowest and the expended bed height increase as the steam inlet velocity increase. However, at high steam inlet velocity (5 U_{mf}), the expended bed height is not stable and keeps changing with respect to time. This is because at higher inlet velocity, bigger bubbles are formed and the eruption of this bigger bubble affects the stability of expended bed height. The most stable expended bed height is observed at steam inlet velocity of 3.5 U_{mf}.
The fluidization of solid particles by steam at inlet velocity of 3.5 U_{mf} is shown in Fig. 3a. From this figure, it revealed that the solid fluidization is good at the bottom of the bed as the solid volume fraction is about 0.4380.613 (green and yellow color). However, at higher bed height, the solid phase become denser and solid volume fraction become higher. This is probably because at higher bed height the steam had not enough energy to support and fluidized the solid particles that have higher potential energy due to elevation and at this point, the gravitational effect is much dominant compared to the lifting energy by the steam.
Figure 3b shows the flow regimes inside the fluidized bed reactor for steam inlet velocity of 5 U_{mf}. As the steam inlet velocity increase, the steam has more energy to fluidize the solid particle and the flow regime become more vigorous. However, the bubbles formed in the bed is too big and the void region increase in the bed as the volume fraction of solid is low as 0.350.438 (green color). This condition is not favorable especially for gasification reaction to occur inside the reactor because the solid particles and the steam do not mix very well.

Fig. 2: 
Expended bed height with respect to different inlet velocity 

Fig. 3: 
Contours of solid volume fraction (a) u_{g} = 3.5
U_{mf}, (b) u_{g} = 5 U_{mf}. s: sec 
Overall, as the steam inlet velocity increase, the steam will have more energy
to fluidize the solid particles. However, increasing the steam inlet velocity
will increase the void region in the bed, which is in agreement with literature
(Lim et al., 2009) and at the same time increases
the steam operating cost. Thus, operating steam at lower inlet velocity is much
favorable. Therefore, the inlet velocity of 3.5 U_{mf} is the optimum
steam inlet velocity as it gives the best solid fluidization in the reactor
with average solid volume fraction of 0.4380.613 throughout the bed which is
comparable with literature that gives the optimum inlet velocity of 4 U_{mf}
(Lim et al., 2009). This steam velocity is sufficient
in order to fluidize the solid particles in the fluidized bed reactor at bed
height of 5 cm.
Effect of solid initial bed height to solid fluidization in the gasifier:
Figure 4 shows the instantaneous expended bed height for solid
particles that has been gasified by steam at different initial bed height. It
is observed that, the higher the initial bed height, the higher the instantaneous
expended bed height in the fluidized bed reactor.

Fig. 4: 
Expended bed height with respect to different solid initial
bed height 
All expended bed heights are almost stabilized except for 12.5 cm initial
bed height. The expended bed for 12.5 cm initial bed height keeps fluctuating
with time as the result of bigger bubble eruption at the surface of the bed.
From the graph, it is observed that 5 cm initial bed height gives the most stabilized
extended bed height which is in good agreement with literature (Busciglio
et al., 2009; Lim et al., 2009; Papadikis
et al., 2009).
Figure 5a shows the flow regime of gasified solid particles at initial bed height of 5 cm. For the first 0.8 sec, the flow developed from jetting flow regime (0.2 sec) to bubbling flow regime (0.4 sec) and to slugging flow regime (0.6 sec). As the slug reaches the surface of the extended bed height, it erupted and the solid particles flow downward due to gravitational effect (0.8 sec). Stabilized bubble flow is observed after 2 sec with small and homogeneously dispersed bubbles form at the bottom of the bed and the bubbles is getting bigger as it reaches the surface of the bed.
Figure 5b shows the contours of solid volume fraction for
solid particles at initial bed height of 12.5 cm. The flow development from
0.2 sec until 0.8 sec is the same as previous ones which is from jetting flow,
to bubbling flow and to slugging flow. However at higher solid bed height, bigger
bubbles tend to form through the bed. These big bubbles turn into slug as it
moves towards the surface of the bed. Formation of big bubbles and slug shows
not a very good fluidization condition as the bubbles are not homogeneously
dispersed and the condition could not be considered as uniform. This observation
is found to be similar with other literature (Busciglio
et al., 2009; Lim et al., 1995).
Overall, the increase of solid initial bed height does not significantly improve
the solid fluidization in the gasifier at constant steam inlet velocity. At
high solid initial bed height, which is 12.5 cm, more fluctuations in the expended
bed and formation of slug are observed.

Fig. 5: 
Contours of solid volume fraction for solid initial bed height
of (a) 5 cm, (b) 12.5 cm. s: sec 
This is because solid at higher initial bed height, is much more difficult
to be fluidized by the steam. The bubbles form in the bed also tends to be bigger
due to coalescing with other bubbles in the bed, which will increase the voidage
between the solid particles. The result obtained from simulation shows that
initial bed height of 10 cm or lower is much suitable for the gasifier.
CONCLUSIONS
As a conclusion, the difference in steam inlet velocity and solid initial bed height shows significant changes to the hydrodynamics of the fluidized bed reactor. As the steam inlet velocity increases, the steam will have more energy to fluidize the solid particles. However, increasing the steam inlet velocity will increase the voidage in the bed and at the same time increase the steam operating cost. Therefore, the inlet velocity of 3.5U_{mf} is the most optimum steam inlet velocity as it gives the best solid fluidization in the gasifier with average solid volume fraction of 0.4770.513 throughout the bed. Solid initial bed height also may affect the hydrodynamic inside a fluidized bed reactor. With the range used in this work, the increase in solid initial bed height does not give significant improvement on the solid fluidization in the bed. However, there is a trend where the expended bed height tends to fluctuate more at very high initial bed height, which is 12.5 cm. From the simulation, it is observed that solid initial bed height of 10 cm or lower gives the best solid fluidization in the gasifier.
The hydrodynamics model developed is capable to predict the solid fluidization behavior in the fluidized bed. Thus, this provides the insight to the researcher on what is actually happening inside the gasifier during fluidization that is difficult to be observed from experimental setup. The results obtained from this model also can be utilized in various applications such as in design of gasifier, optimization of reaction condition and many others.
NOMENCLATURE

: 
Interphase momentum change 

: 
Stressstrain tensor 
C_{d} 
: 
Drag coefficient 
d_{p} 
: 
Particle diameter (m) 
d 
: 
Column diameter (mm) 

: 
Coefficient of restitution for particle collision 

: 
Gravitational force (m sec^{2}) 
g_{0} 
: 
Radial distribution function 
ke_{i} 
: 
Diffusion coefficient 

: 
Velocity (m s^{1}) 
F 
: 
Interphase momentum exchange coefficient 
P 
: 
Pressure (Pa) 
Re 
: 
Reynolds number 
U_{mf} 
: 
Minimum fluidization velocity (m sec^{1}) 
u 
: 
Superficial velocity (m sec^{1}) 
Greek symbols
γ_{θi} 
: 
Collisional dissipation energy 

: 
Maximum packing limit of species i 

: 
Granular temperature

μ_{g} 
: 
Shear viscosity 
φ_{gs} 
: 
Energy exchange between gas and solid phase 
ε 
: 
Volume fraction 
ρ 
: 
Density (kg m^{3}) 
Subscripts
g 
: 
Gas phase 
s 
: 
Solid phase 