IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,

Vol.2, No.1, 2012

CFD ANALYSIS OF CIRCULATING FLUIDIZED

BED COMBUSTION
Ravindra Kumar and K.M.Pandey
Department of Mechanical Engineering,
Gurgaon Institute of Technology&Management, Gurgaon, India
National Institute of Technology, Silchar, Assam, India

Abstract- In this paper computational analysis of circulating which is a solid granular particle, is suspended by a flow of
fluidized bed combustion using the fluent software is analysed. air or gas (air). This flow is injected into the combustion
This work is concerned about gas–solid two-phase mixtures chamber from the bottom and from the side. FBC boilers
flowing upwards through the fast beds. The procedures studied can burn fuels other than coal, and the lower temperatures of
are applicable to all types of fluidized-bed boilers. The type of
information resulting from various ways of analyzing the
combustion (800 °C / 1500 °F) have other added benefits as
pressure and temperature in fluidized beds are discussed. well.
This paper presents coal combustion in circulating fluidized In this paper, the gas (air) and solid (coal) is injected at the
bed and the k–є two-phase turbulence model was used to base with different velocities while taking coal particles of
describe the gas–solids flow in a CFB. The analysis of coal different diameters as solid bed. In the 2D CFB combustor
combustion is done by discrete phase model (DPM) and non have the 3 inlet points such as inlet for fluidizing velocity
pre mixed combustion in species model .Predicting the (primary air), inlet for coal particles and inlet for secondary
performance of large scale circulating fluidized bed boilers air. The primary air is used for the fluidized of coal particle
requires reliable and efficient modeling tools. In a CFB and secondary air for proper combustion in the combustor.
furnace, the fuel, air, and other input materials are fed locally
and the mixing of different reactants is limited. The particle
The combustion processes is done by the discrete phase
size has taken 5 mm and fluidizing velocity (4-6) m/s. As a model and use the single injection system for burning of
result of analysis, the variation in mean particle diameter and coal in the combustor.
superficial velocity, does affect the temperature, pressure and
turbulence kinetic energy in different mean fractions in the II. LITERATURE REVIEW
combustion zone. The average temperature is found around
the1370k. A.Williams and M. Pourkashanian [2] worked on the some
of the key issues currently being debated regarding the
Keywords- Coal, fluidized bed combustion, Two-phase flow, combustion of coal and of some biomass materials. It
DPM.
attempts to summarize the present approaches toward the
I. INTRODUCTION quantification of the fundamental processes of solid fuel
combustion for use in computer models. Some aspects of the
‘Getting rid of waste’ was the ultimate goal when the various chemical and physical processes are included, such
fluidized bed combustion (FBC) technology was introduced as heating-up of particles, devolatilization, and subsequent
[1]. This goal evolved over time to ‘clean energy for the char formation. Of particular interest is the prediction of
future’. Since its introduction in the 1970s the technology char properties, such as composition, surface areas, and
has gained acceptance in various industrial applications. morphology, since these impacts on char combustion.
FBC is a combustion technology used in power plants. FBC Konstantin P. Filipov [3] presented on Mathematical model
is known for its ability to burn low-grade fuels with low describing two-phase flow in CFB in framework of external
calorific value, high ash content and high moisture content. model ash was elaborated. It takes into account the all main
Fluidized beds suspend solid fuels on upward-blowing jets processes in high-velocity fluidized bed including processes
of air during the combustion process. The result is a of the coal combustion and the attrition of ash. On base of
turbulent mixing of gas and solids. When a fluidized bed is this model was developed numerical code and carried out
operated above the terminal velocity of the particles, they some calculations of transient and stationary performances
are carried out of the bed. The system of a circulating of CFB. Z.Guangbo [4] focused on a steady state model of a
fluidized bed (CFB) occurs when the particles are separated coal-fired circulating fluidized-bed boiler, based on
from the fluid by the use of cyclones and are recycled to the hydrodynamics, heat transfer and combustion, is presented.
bed. The part of the system where the carryover of solids This model predicts the flue gas temperature, the chemical
transpires is normally referred to as the riser. It is known for gas species (O2, H2O, CO, CO2 and SO2) and char
its ability to burn low-grade fuels with low calorific value, concentration distributions in both the axial and radial
high ash content and high moisture content. The fluidization locations along the furnace including the bottom and upper
process begins when a bed of inert material (usually sand), portion. The model was validated against experimental data

163
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

generated in a 35 t/h commercial boiler with low circulation of the gas phase, the temperature change of particles and the
ratio. radiative parameters of a particle group. The results provide
Afsin Gungor [5] worked on a dynamic two dimensional sufficient information to improve the understanding of heat
model is developed considering the hydrodynamic behavior transfer processes near the immersed surface. J.C.S.C.
of CFB. In the modeling, the CFB riser is analyzed in two Bastos focused on [9] Radial solids velocity profiles were
regions: The bottom zone in turbulent fluidization regime is computed on seven axial levels in the riser of a high-flux
modeled in detail as two-phase flow which is subdivided circulating fluidized bed (HFCFB) using a two-phase 3-D
into a solid-free bubble phase and a solid-laden emulsion computational fluid dynamics model. The computed solids
phase. In the upper zone core–annulus solids flow structure velocities were compared with experimental data on a riser
is established. Simulation model takes into account the axial with an internal diameter of 76 mm and a height of 10 m, at
and radial distribution of voidage, velocity and pressure a high solids flux of and a superficial
drop for gas and solid phase, and solids volume fraction and
particle size distribution for solid phase. The model results velocity of . Several hundreds of experimental and
are compared with and validated against atmospheric cold numerical studies on CFBs have been carried out at low
bed CFB units' experimental data given in the fluxes of less than , whereas only a few
literature.Hideya Nakamura [6] presented on modeling of limited useful studies have dealt with high solids flux.
particle fluidization behaviors in a rotating fluidized bed A three-dimensional model of a circulating fluidized bed
(RFB) was conducted. The proposed numerical model was gasifier was developed by [10] I. Petersen and J.Werther,
based on a DEM (Discrete Element Method)-CFD which uses continuous radial profiles of velocities and solids
(Computational Fluid Dynamics) coupling model. Fluid hold-up with regard to the description of fluid mechanics. A
motion was calculated two dimensionally by solving the complex reaction network of sewage sludge gasification is
local averaged basic equations. Particle motion was included in the model. In the simulation calculations the
calculated two-dimensionally by the DEM. Calculation of influence of the axial location and the number of feeding
fluid motion by the CFD and particle motions by the DEM points was examined for gasifiers of different scales. It was
were simultaneously conducted in the present model. found that due to the very fast decomposition of the volatiles
Geldart group B particles (diameter and particle density and the high volatile content in the sewage sludge, lateral
were 0.5 mm and 918 kg/m3, respectively) were used for mixing of the gas around the feeding port is not complete,
both calculation and experiment. Aboozar Hadavand [7] and plumes with high pyrolysis gas concentrations are
presented on a mathematical model of the circulating formed. Vidyasagar Shilapuram [11] worked on
fluidized bed combustion system based on mass and energy Experiments were conducted in a liquid–solid circulating
conservation equations were successfully extracted. Using fluidized bed (LSCFB) to study the flow regimes,
these correlations, a state space dynamical model oriented to operational instability, critical transitional velocity to
bed temperature has been obtained based on subspace circulating fluidized bed (CFB) regime, solids holdup and
method. Bed temperature, which influences boiler overall solids circulation rate by three experimental methods. The
efficiency and the rate of pollutants emission, is one of the results indicate that the operational instabilities such as arch
most significant parameters in the operation of these types formation, liquid–solid separator blockage and solids return
of systems. Having dynamic and parametric uncertainties in pipe blockage were observed in two of these methods at
the model, a robust control algorithm based on linear matrix large primary and auxiliary liquid velocities. The critical
inequalities (LMI) have been applied to control the bed transitional velocity that demarcates the expanded bed from
temperature by input parameters, i.e. coal feed rate and CFB regime was observed to be different by these three
fluidization velocity. Circulating fluidized bed (CFB) methods. The macroscopic flow properties (flow regimes,
combustion systems are increasingly used as superior coal onset of average solids holdup, average solids holdup and
burning systems in power generation due to their higher solids circulation rate in the riser) are different by different
efficiency and lower emissions. methods of operation.Jack T. Cornelissen [12] discussed on
L.X. Kong and P.D. Hodgson [8] worked on To improve a multiplied Eulerian computational fluid dynamics (CFD)
the understanding of the heat transfer mechanism and to find model with granular flow extension is used to simulate a
a reliable and simple heat-transfer model, the gas flow and liquid–solid fluidized bed. The numerical simulations are
heat transfer between fluidized beds and the surfaces of an evaluated qualitatively by experimental data from the
immersed object is numerically simulated based on a double literature and quantitatively by comparison with new
particle-layer and porous medium model. The velocity field experimental data. The effects of mesh size, time step and
and temperature distribution of the gas and particles are convergence criteria are investigated. The Eulerian CFD
analyzed during the heat transfer process. The double simulations based on FLUENT software for water
particle-layer and porous medium model has the ability to fluidization of glass spheres give results which are generally
simulate the gas flow and the heat transfer near the surface in good qualitative and reasonable quantitative agreement
of an immersed object in fluidized beds, and was with experimental results. Courant number between 0.03
successfully used in calculating the dynamic characteristics and 0.3 gave good predictions in this work.

164
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

Xiao-Bo Qi [13] presented on Friction between co-current cracking (FCC) particles in the riser of a high density
down flow gas–solid flow and column wall was investigated circulating fluidized bed (HDCFB). The implementation of
by measuring apparent and actual solids concentrations in a correct inlet conditions was found to be critical for the
circulating fluidized bed (CFB) downer. A new model to successful simulation of the hydrodynamics. The simulated
predict pressure drops due to friction between the gas–solid profiles of gas and solid velocity and volume fraction were
suspension in the fully developed zone and the downer wall overall in good agreement with experimental data reported
was developed. The results show that the friction between in the literature. However, due to the difficulties in accurate
the gas–solid suspension and the downer wall causes a modeling of the solid segregation toward the wall, the solid
significant deviation of the apparent solids concentrations volume fraction was under predicted near the walls. The
from the actual ones, especially for those operating effect of modeling parameters including different drag
conditions with higher superficial gas velocities and solids models, wall restitution coefficient values, and solid slip
circulation rates. When the superficial gas velocity is greater conditions have been evaluated.
than 8 m/s, the actual solids concentrations in the fully Ernst-Ulrich Hartge [18] presented on a model of the fluid
developed region of the downer can be up to two to three mechanics in the riser of a circulating fluidized bed (CFB)
times of the apparent values. Particle diameters have has been implemented using computational fluid dynamics
different influences on the frictional pressure drops under (CFD). The model developed shall be used in future as the
different superficial gas velocity. basis of 3D-reactor model for the simulation of large scale
Haiyan Zhu [14] focused on detailed local flow structures CFB combustors. The two-fluid model (TFM) approach is
are investigated in bubbling and turbulent fluidized bed with used to represent the fluid mechanics involved in the flow.
FCC particles. The operating conditions ranges from 0.06 to The computational implementation is accomplished by the
1.4 m/s. Extensive experiments are carried out using a commercial software FLUENT. As result of this study, a
newly developed optical fiber probe system, which can formulation is established from the variety of available
measure the solids concentration and velocity at multi- models. As a major finding of this part it has been shown,
points. The results reveal that with increasing Ug, local that the commonly used drag correlations with an increasing
solids concentrations go through three evolution stages, drag with increasing solids concentration fail to predict a
reflecting a gradual regime transition process. Under all dense bottom zone in the CFB riser, as is observed in
operating conditions, up flowing and descending particles reality. Taking the sub-grid heterogeneity of the flow into
co-exist at all measuring locations. Results show that in the account, as is done here with the help of the EMMS model,
turbulent fluidized bed a stable two-phase flow structure allows predicting the dense bottom zone.
breaks down and the solids concentration and velocity Xiao-Bo Qi and Hui Zhang [19] worked on the investigate
distribution is not at all uniform in both radial and axial solids concentration in the fully developed region of co-
directions. Zhongxiang Chen [15] described on the current downward gas–solid flow, actual solids
performance of fluidized bed methane reformers with three concentrations were measured in a circulating fluidized bed
models—a simple equilibrium model and two kinetic (CFB) downer with 9.3 m in height and 0.1 m in diameter
distributed models, based on different assumptions of using a fiber optical probe. The results obtained from this
varying sophistication. Membranes are incorporated to work and in the literature show that the average solids
improve reactor performance. Eighteen cases are simulated concentration in the fully developed region of the CFB
for different flow regimes and membrane configurations. downers is not only a function of the corresponding terminal
Predictions for the fast fluidization and turbulent flow solids concentration, but the operating conditions and
regimes show that the rate-controlling step is permeation particle properties also have influences on the average solids
through the membranes. Bubbling regime simulations concentration in the fully developed region of the CFB
predict somewhat less hydrogen production than for downers. Jinsen Gao and Jian Chang [20] presented on the
turbulent and fast fluidization, due to the effects of experimental and computational studies on the flow
interphase crossflow and mass transfer. behavior of a gas–solid fluidized bed with disparately sized
Stephen J. Goidich [16] focused on Circulating fluidized binary particle mixtures. The mixing/segregation behavior
bed technology has moved into the utility power generation and segregation efficiency of the small and large particles
market with units as large as 250 MWe in operation, two (2) are investigated experimentally. Particle composition and
300 MWe units in the construction phase, and designs operating conditions that influence the fluidization behavior
developed for units as large as 600 MWe in size. of mixing/segregation are examined. Based on the granular
Technology improvements are continually being developed kinetics theory, a multi-fluid CFD model has been
and incorporated in the designs to enhance performance, developed and verified against the experimental results. The
increase operational flexibility, and improve reliability in a simulation results are in reasonable agreement with
cost-effective way. Most notable is the Compact CFB boiler experimental data. The results showed that the smaller
with INTREXTM heat exchanger which Foster Wheeler particles are found near the bed surface while the larger
introduced in the early 1990’s.Adnan Almuttahar [17] particles tend to settle down to the bed bottom in turbulent
performed on CFD modeling of air and fluid catalytic fluidized bed. Ernst-Ulrich Hartge [21] invented on a model

165
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

of the fluid mechanics in the riser of a circulating fluidized underestimation or overestimation of the drag force between
bed (CFB) has been implemented using computational fluid the gas and the solid phase.
dynamics (CFD). The model developed shall be used in
future as the basis of 3D-reactor model for the simulation of III. GOVERNING EQUATIONS OF FLUID DYNAMICS
large scale CFB combustors. The two-fluid model (TFM)
III.I CONTINUITY EQUATION
approach is used to represent the fluid mechanics involved
in the flow. The computational implementation is The conservation of mass equation or continuity equation
accomplished by the commercial software FLUENT. [28] is given by
Different closure formulations are tested on a simplified
geometry. The work presented here consists of two major
parts. The first part involves a formulation study, where (3.1)
several combinations of granular temperature formulation,
turbulence models, approaches to solids phase turbulence, Where is the density, is the velocity vector.
drag correlations and solid–solid restitution coefficients are
tested. Macroscopic (semi-empirical) models for fluid III.II MOMENTUM EQUATION
dynamics of circulating fluidized bed (CFB) units are
presented [22] by Filip Johnsson, with emphasize on Applying the Newton’s second law
applications for conditions relevant to industrial units such the conservation of
as fluidized-bed combustors. In order to make a structured momentum equation is given by
analysis of the models, the CFB unit is divided into 6 fluid
dynamical zones, which have been shown to exhibit (3.2)
different fluid-dynamical behaviour (bottom bed, freeboard,
exit zone, exit duct, cyclone and downcomer and particle
seal). The paper summarizes the main basis and assumptions Where is the density, is the velocity vector, is the
for each model together with major advantages and static pressure, and is the stress tensor (described below),
drawbacks. Dongsheng wen and Yurong He [23] developed
on gas–solid two-phase mixtures flowing upwards through and and are the gravitational body force and external
packed beds. An Eulerian-based two-fluid model coupled body forces (e.g., that arise from interaction with the
with a newly proposed porosity distribution model is used to dispersed phase), respectively. The stress tensor is given
simulate the flow behaviour. The results are compared with by
recently published experimental results in terms of both
hydrodynamics and solids motion. It is found that the use of
the newly proposed porosity model not only gives better (3.3)
agreement with experimental porosity data, but also Where is the molecular viscosity, is the unit tensor,
provides a much better prediction of the pressure drop than and the second term on the right hand side is the effect of
other porosity models could do. The results also show that volume dilation.
the model predicts very well the dynamic hold-up of
suspended particles, and captures the main features of the III.III ENERGY EQUATION
radial distributions of the suspended solids concentration Energy is neither created nor destroyed. It is always
and the axial solids velocity.Peng Li [24] performed on the conserved. The conservation of energy equation is given by
suitability of various drag models for predicting the
hydrodynamics of the turbulent fluidization of FCC particles
on the Fluent V6.2 platform. The drag models included
those of Syamlal–O’Brien, Gidaspow, modified Syamlal–
O’Brien, and McKeen. Comparison between experimental ( 3.4)
data and simulated results showed that the Syamlal–
O’Brien, Gidaspow, and modified Syamlal–O’Brien drag
models highly overestimated gas–solid momentum (3.5)
exchange and could not predict the formation of dense phase where sensible enthalpy is defined for ideal gaseas
in the fluidized bed, while the McKeen drag model could
not capture the dilute characteristics due to underestimation
of drag force. The standard Gidaspow drag model was then
modified by adopting the effective particle cluster diameter (3.6) For incompressible flows as
to account for particle clusters, which was, however, proved
inapplicable for FCC particle turbulent fluidization. The
simulated results showed that they all failed due to either

166
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

(3.7) (4.1)

These equations form a set of coupled, nonlinear partial

differential equations. It is not possible to solve these
equations analytically for most engineering problems.
(4.2)
However, it is possible to obtain approximate computer-
based solutions to the governing equations for a variety of
engineering problems. This is the subject matter of ,
Computational Fluid Dynamics (CFD). In these equations, Gk represents the generation of
turbulence kinetic energy due to the mean velocity
III.IV ENERGY EQUATION FOR NON-PREMIXED
gradients. Gb is the generation of turbulence kinetic energy
COMBUSTION MODEL
due to buoyancy. YM represents the contribution of the
When the non-adiabatic non-premixed combustion model is fluctuating dilatation in compressible turbulence to the
enabled, FLUENT solves the total enthalpy form of the overall dissipation rate. C2 and C1Є are constants. σk and σЄ
energy equation are the turbulent Prandtl numbers for K and Є respectively.
Sk and SЄ are user-defined source terms.

(3.8)
V. CFD MODEL OF FLUIDIZED BED
Under the assumption that the Lewis number (Le) = 1, the
conduction and species diffusion terms combine to give the The furnace model considered for analysis of coal
first term on the right-hand side of the above equation while combustion is 15m height and 3.2 m diameter of the lower
the contribution from viscous dissipation appears in the non- part of the furnace. The purpose of the work was to study
conservative form as the second term. The total enthalpy the char at the lower part of the furnace. The model volume
is defined as was restricted to a 15 m height section of the furnace
[28].Model geometry in two-dimensional and three-
dimensional as shown in fig. 6.2.Circulating bed is fluidized
by primary air through the grid. Secondary air is introduced
at level 2 – 4 m from grid. From the furnace, the gas and
(3.9)
solids enter the separator, which separates the solids from
flue gas. A circulating fluidized bed is one where fuel is
Where is the mass fraction of species and burnt in a fast fluidized bed regime. In CFB boiler furnace
the gas velocity is sufficiently high to blow all the solids out
of the furnace.
The majority of solids leaving the furnace is captured by
(3.10) gas-solid separators.
is the formation enthalpy of species at the VI. COMPUTATION GRID GENERATION
reference temperature .
VI.I MODEL CREATION AND GRID GENERATION IN
IV. STANDARD K - Є MODEL GAMBIT

The turbulence models are the two-equation models. The For the furnace as explained in a two dimensional model is
simplest one is the standard k - є model, which is proposed created in GAMBIT 2.3.16. The 2D view of the furnace
by Launder and Spalding. It is widely used in turbulence after modeling is as shown in Fig.1.
simulations because of its general applicability, robustness VI.II GRIDS SELECTED FOR THE PRESENT WORK
and economy. The two transport equations for the kinetic
energy and dissipation rate are solved to form a The grids are selected for all the meshes for doing CFD
characteristic scale for both turbulent velocity and length. analysis. As the furnace for which analysis is carried out,
These scales represent the turbulent viscosity. quadrate type mesh is selected. This specifies that the mesh
The modeled transport equations for K and Є in the is composed primarily of quadrate mesh elements. The
realizable K - Є model are qualities of the created mesh are checked.

167
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

phase inertia, hydrodynamic drag, and the force of

gravity, for both steady and unsteady flows
2) Combusting particles, including volatile evolution
and char combustion to simulate coal combustion
3) These modeling capabilities allow FLUENT to
simulate a wide range of discrete phase problems
including particle separation and classification,
spray drying, aerosol dispersion, and bubble
stirring of liquids, liquid fuel combustion and coal
combustion.

Fig.1. View of the model in 2D after modeling in GAMBIT VIII. PROBLEM DESCRIPTION
2.3.16
GAMBIT 2.3.16 was used for making 2D furnace geometry The problem consists of a two phase fluidized bed in which
with width of 3.2m from the lower part and height 15m. gas and solid (coal) enters at the separate portion of the
Coarse mesh size of 0.01m was taken in order to have 9365 domain. For a 2D CFB furnace, estimate the primary air is
cells (18952 faces) and 9588 nodes for the whole geometry. flowing at 503 k and 4m/s which is flowing through the
It was used in order to have better accuracy. But using mesh nozzles .The secondary air is injected at the level of 2.3 m
results in 9365 cells (18952 faces), which requires smaller from the level of the bottom at 473 k and 2m/s and the fuel
time steps, more number of iterations per time step and 4 feed rate 0.5 kg/s is injected at the level of 1.5m from the
times more calculation per iteration for the solution to level of the bottom, corresponding to very fuel-lean
converge. The model of combustion chamber with boundary conditions in the flow. The bed cross-section is 3.2 m x15 m
conditions as below. below and 6.7 m x15 m above this level. The Reynolds
number, based on inlet conditions and the flow is turbulent.
The combustion is modelled using the mixture-fraction
approach. The bed consists of solid material (coal particles)
of uniform diameter and the size of solid particles is 5000
µm.
Data for coal analysis is given below.
Coal – Lignite (composition on dry) basis
C – 67%, H – 6%, O– 25%, N – 1%, S – 1%, HHV –
25MJ/Kg

IX. ANALYSIS PARAMETERS

Table I:

Fig.2. Combustion chamber in 3D Parameter Value

VII. DISCRETE PHASE MODEL (DPM) Primary air velocity 4-6m/s

Secondary air velocity 2m/s

In addition to solving transport equations for the continuous
phase, FLUENT allows you to simulate a discrete second Coal particle size 5000 µm
phase in a Lagrangian frame of reference. This second phase
consists of spherical particles (which may be taken to Primary air temperature 503k
represent droplets or bubbles) dispersed in the continuous
phase. FLUENT computes the trajectories of these discrete Secondary air temperature 473k
phase entities, as well as heat and mass transfer to/from
them. The coupling between the phases and its impact on Mass flow rate 0.5kg/s
both the discrete phase trajectories and the continuous phase
flow can be included. Oxygen concentration at inlet 21%
FLUENT provides the following discrete phase modeling
options: Back flow temperature 1800k
1) Calculation of the discrete phase trajectory using a
Lagrangian formulation that includes the discrete Initial pressure

168
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

Coal heat capacity 1000J/kg-k

Coal density 1350kg/m3

Mean mixture fraction 0.09

Oxide in PDF table 600k

X. SELECTION OF MODELS FOR ANALYSIS

Fig.3. Plot of residuals for k-epsilon solver method as the
FLUENT 6.2.16 was used for analysis. 2D segregated 1st iterations proceeds.
order implicit steady solver is used. (The segregated solver XII. BOUNDARY CONDITIONS
must be used for multiphase calculations). Standard k-ε
model with standard wall functions were used. The boundary conditions are as equally important as the
The model constants are tabulated as: selection of the proper mathematical model. Initially, solid
Table II:Model constants used for analysis as given below particle velocity was set at in minimum fluidization and gas
Cmu 0.09 velocity was assumed to have the same value everywhere in
the bed. The temperature of the primary and secondary air
C1-Epsilon 1.44
was also set to 503k and 473k respectively. At the inlet, all
C2-Epsilon 1.92 velocities of all phases were specified. At the outlet, the
C3-Epsilon 1.3 pressure was assumed to be Atmospheric pressure. The gas
TKE Prandtl Number 1 tangential normal velocities on the wall were set to zero (no
TDR Prandtl Number 1.3 slip condition). The following boundary equation was
Energy Prandtl Number 0.85 applied for the tangential velocity of Particle on the wall.

The non premixed combustion model is used for combustion Boundary Condition
and in DPM (discrete phase model) single type injection Inlet1 Primary(Fluidized) Velocity inlet
system is used for the solid fuel combustion. Inlet2 Secondary Velocity inlet
XI. SOLUTION Particle flow Mass flow rate
Outlet Pressure-outlet
The finite volume method was applied to discretize the
governing equations on computational grids. Under
XIII. NUMERICAL METHODOLOGY
relaxation factor for pressure, momentum and mixture
fraction were taken. The discretization scheme for
A numerical method adopted to approximate the governing
momentum, mixture fraction, turbulence kinetic energy and
equations, along with the relevant boundary conditions, by a
turbulence dissipation rate were all first order upwind.
system of linear algebraic equations is known as a
Pressure-velocity coupling scheme was Phase Coupled
discretization method. Thus, a problem involving calculus is
SIMPLE. The solution was initialized from all zones.
transformed into an algebraic problem which can then be
Iterations were carried out for time step size of 0.01-0.001
solved on a computer by using a solution methodology. A
depending on ease of convergence and time required to get
discretization technique and a solution methodology
the result for fluidization. Convergence and accuracy is
constitute the numerical methodology used to solve a heat
important during solution. This can be seen by the residual
transfer and fluid flow problem. There are many
plots in fig.3. If not then we have to change the solution
discretization methods, but the most commonly used are the
parameters and sometimes solution method also. Currently,
Finite difference method (FDM), the Finite volume method
K-epsilon method is used for the combustion study of the
(FVM) and the Finite element method (FEM). During the
fluidized bed.
early days of Computational fluid dynamics (CFD) finite-
difference methods were the most popular. They are
algorithmically simple, efficient, and accurate. However,
they are best used on uniform grids and hence on regular
computational domains. With advances of CFD, and its
application to industrial problems, there is a need for
methods for computing flows in complex geometries. To
adapt the finite difference method to such geometries, we

169
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

can map the complex domain into simple domains, either

globally or locally, and solve the equations there. However,
such transformation makes the governing equations take
quite complicated forms and may lead to a loss of
computational efficiency and accuracy. Alternatively, one
can use schemes based on the finite volume methods
directly on the physical domain (i.e. without
transformation).
Finite volume methods are essentially a generalization of the
finite-difference method, but use the integral form of the
governing equations of flow rather than their differential
form. This gives greater flexibility in handling complex
domains, as the finite volumes need not be regular. The
FLUENT code, which is used to simulate the flow field is Fig.5. x-y plot of total temperature (k)
based on the finite volume discrtization scheme and is one
of the best application software for this purpose.

XIV. RESULT AND DISCUSSION

XIV.I ANALYSIS OF COMBUSTION AT FLUIDIZING

VELOCITY 4 m/s

TOTAL TEMPERATURE

When the fuel and air enter into the combustor, it burns due Fig.6 x-y plot of total temperature (k)
to high velocity and temperature and then temperature
increase rapidly in the combustor. Finally obtained the STATIC PRESSURE
result, the total temperature of the coal combustion is
1370(k). The value of static pressure is found in the combustor is
The figure 4 shows that the temperature profile in (Pascal) after combustion process. The static pressure
circulating fluidized bed combustor. In this figure, the bed is rapidly increased when air fuel velocity enters in the
and temperature are increasing as soon as the coal particles combustion chamber. In the combustor, maximum static
are burning and finally obtained the maximum value of total pressure is at the inlet points of the furnace.
temperature after coal combustion is K.

Fig.7 contours of static pressure (Pascal)

Fig.4 contours of total temperature (k) TURBULENCE KINETIC ENERGY
Fig 5 shows the profile between the total temperature and The maximum turbulence intensity is found to be
the position of the combustor on all conditions such as
default interior, pressure-outlet, primary air, secondary air after the combustion. In case of turbulence of
and wall but fig.6 shows the plot the x-y diagram between kinetic energy, when the velocity of a gas is increased above
total temperature and position of the combustor on pressure the minimum bubbling velocity, the bed starts expanding. A
outlet condition. continued increase in the velocity may eventually show a

170
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

change in the pattern of bed expansion. In the turbulent

kinetic energy regime, the bubble phase loses its identity
due to rapid coalescence and break up the bubbles. This
results in a violently active and highly expanded bed with a
change in the pattern of bed expansion.

Fig.10 Contours of Static Pressure (Pascal)

TURBULENCE KINETIC ENERGY

The maximum turbulence intensity is found to be

after the combustion. In case of turbulence of
Fig.8 contours of turbulence kinetic energy (k) (m2/s2) kinetic energy, when the velocity of a gas is increased above
XIV.II ANALYSIS OF COMBUSTION AT the minimum bubbling velocity, the bed starts expanding. A
FLUIDIZING VELOCITY 5m/s continued increase in the velocity may eventually show a
change in the pattern of bed expansion. In the turbulent
TOTAL TEMPERATURE kinetic energy regime, the bubble phase loses its identity
due to rapid coalescence and break up the bubbles. This
The figure 40 shows that the temperature profile in results in a violently active and highly expanded bed with a
circulating fluidized bed combustor. In this figure, the bed change in the pattern of bed expansion.
and temperature are increasing as soon as the coal particles
are burning and finally obtained the maximum value of total
temperature after coal combustion is k.

Fig.11 Contours of turbulence kinetic energy (k) (m2/s2)

XIV.III ANALYSIS OF COMBUSTION AT

Fig.9 Contours of total temperature (k) FLUIDIZING VELOCITY 6m/s

TOTAL TEMPERATURE
STATIC PRESSURE
The figure 12 shows that the temperature profile in
The value of static pressure is found in the combustor is circulating fluidized bed combustor. In this diagram the bed
(Pascal) after combustion process. The static pressure and temperature are increasing as soon as the coal particles
is rapidly increased when air fuel velocity enters in the are burning and finally obtained the maximum value of total
combustion chamber. In the combustor, maximum static temperature after coal combustion is k.
pressure is at the inlet points of the furnace.

171
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

Fig.12 contours of total temperature (k)

Fig.14 Contours of turbulence kinetic energy (k) (m2/s2)
STATIC PRESSURE
XV. BEHAVOIUR OF TOTAL TEMPERATURE AT
The value of static pressure is found in the combustor is
DIFFERENT FLUIDIZING VELOCITIES
(Pascal) after combustion process. The static pressure
is rapidly increased when air fuel velocity enters in the Following are the trends of Total temperature vs. inlet air
combustion chamber. In the combustor, maximum static velocity obtained at different inlet fluidizing velocities,
pressure is at the inlet points of the furnace. which show that temperature increases when air velocity is
increased.

Fig.13 Contours of static pressure (Pascal)

Fig.15 Temperature versus different fluidizing velocity

XVI. BEHAVIOUR OF PRESSURE AT DIFFERENT

TURBULENCE KINETIC ENERGY FLUIDIZING VELOCITIES

Following plots of various pressures vs. inlet air velocity

The maximum turbulence intensity is found to be
obtained at different fluidizing velocities. This plots show
after the combustion. In case of turbulence of kinetic
that pressure increases as air velocity is increased. This is
energy, when the velocity of a gas is increased above the
because with increase in air velocity and mass flow rate of
minimum bubbling velocity, the bed starts expanding. A
fuel in the combustor increases thereby increasing the
continued increase in the velocity may eventually show a
pressure across the combustor. In this fig.16,when the
change in the pattern of bed expansion. In the turbulent
fluidizing velocity is 4m/s at the inlet of combustor then the
kinetic energy regime, the bubble phase loses its identity
static pressure is observed 5.63 Pascal after coal combustion
due to rapid coalescence and break up the bubbles. This
in the combustor of CFB i.e. the static pressure is increases
results in a violently active and highly expanded bed with a
with increasing the inlet velocity. The same conditions are
change in the pattern of bed expansion
for total and dynamic pressure.

172
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

The resolution of the grid has a great quantitative impact

over the results obtained. There exists a level of refining of a
computational domain beyond which there is no significant
quantitative changes in the results achieved. The
computational domain at this level of refinement is said to
enter the regime of grid independence.
The temperature variation along the combustor is studied for
these levels of refinement. The results are compared and the
computational grid structure with the level of refining from
which no change in the temperature variation along the
combustor is seen with further refinement, is chosen for
detailed analysis. This particular grid is said to undergo grid
independence test.

Fig.16 Pressures versus different fluidizing velocity.

XVIII. CONCLUSION

In this work, analysis of coal combustion in circulating

fluidized bed has been performed at three different
fluidizing velocities with fluent software.
Following conclusions are drawn from the computational
analysis in this present work.
The following conclusions are:
1) It is observed that the change in maximum total
temperature for the three fluidizing velocities i.e.
4m/s, 5m/s and 6m/s is insignificant. Fig 17: Plot temperature profile vs. number of cells

2) Maximum static pressure increases from 5.63Pa to To find out the most independent grid for CFD analysis of
10Pa, when fluidizing velocity is increased from model of CFB combustor, grid independency of the solution
4m/s to 6m/s. This increase in the static pressure was established. Based on the different grids, analysis have
inside the combustion chamber is because there is been made and it was observed that after refining the grid
no leakage of air from the combustion chamber. from cells 15435, results are not varying significantly. So,
cells15435 have been used for further analysis.
3) Maximum total pressure increases from 10.4 Pa to
22 Pa, when fluidizing velocity is increased from Notation
4m/s to 6m/s. Maximum total pressure values are
more than maximum static pressure values at the = Density
same fluidizing velocity as it includes dynamic
= Velocity vector
pressure also.
= Static pressure
4) The maximum turbulent intensity of the mixture = Stress tensor
increases from 67.6% to 97.8%, when fluidizing = Gravitational body force
velocity is increased from 4m/s to 6m/s. P= Pressure
µeff= Effective viscosity
5) In fluidized bed combustion, a number of
K= Turbulent kinetic energy
parameters are important such as temperature,
ε= Dissipation rate of turbulent kinetic energy
pressure, as it has been observed that all the
T= Temperature
parameters temperature and pressure are better for
g= Acceleration due to gravity
combustion at fluidizing velocity 6m/s. Therefore,
a fluidizing velocity of 6m/s is suitable for = Effective conductivity
fluidized bed combustion as compared to 4m/s and = Turbulent thermal conductivity
5m/s. = molecular viscosity
= unit tensor
XVII. GRID INDEPENDENCE TEST
Sensible enthalpy

173
 IRACST – Engineering Science and Technology: An International Journal (ESTIJ), ISSN: 2250-3498,
Vol.2, No.1, 2012

REFERENCES [16] Stephen J. Goidich, Foster Wheeler Compact CFB

Boilers for Utility Scale, Foster Wheeler Energy
[1] Martin Junginger, Development of fluidized bed International, Perryville Corporate Park Clinton,
combustion—An overview of trends, performance New Jersey, USA 08809-4000.
and cost. Progress in Energy and Combustion [17] Adnan Almuttahar, Computational fluid dynamics
Science 2007; 33: 19–55. of high density circulating fluidized bed riser:
[2] A. Williams and M. Pourkashanian, The Study of modeling parameters. Powder Technology
Combustion of Coal and Some Other Solid 2008; 185:11–23.
Fuels.2000; 28: 2141–2162. [18] Ernst-Ulrich Hartge, CFD-simulation of a
[3] Konstantin P. Filipov ,Numerical Modeling Of circulating fluidized bed riser, Particuology 2009;
Coal Combustion Processes In Ecological Clear 7: 283–296.
Circulating Fluidized Bed Boiler Units, [19] Xiao-Bo Qi and Hui Zhang, Solids concentration in
Novosibirsk State Technical University 630092. the fully developed region of circulating fluidized
[4] Z.Guangbo, A coal combustion model for bed downers, Powder Technology 2008;183: 417–
circulating fluidized bed boilers. Fuel 2000; 79: 425.
165–172. [20] Jinsen Gao and Jian Chang, Experimental and
[5] Afsin Gungor, Hydrodynamic modeling of a computational studies on flow behavior of gas–
circulating fluidized bed, Powder Technology solid fluidized bed with disparately sized binary
2007; 172: 1–13, particles, Particuology 2008; 6: 59–71.
[6] Hideya Nakamura, Numerical modeling of particle [21] Ernst-Ulrich Hartge, CFD-simulation of a
fluidization behavior in a rotating fluidized bed, circulating fluidized bed riser, Particuology 2009;
Powder Technology 2007; 171: 106–117. 7: 283–296.
[7] Aboozar Hadavand, An innovative bed [22] Filip Johnsson, Macroscopic modelling of fluid
temperature-oriented modeling and robust control dynamics in large-scale circulating fluidized beds,
of a circulating fluidized bed combustor, Chemical Progress in Energy and Combustion Science 2006;
Engineering Journal 2008; 140: 497–508. 32: 539–569.
[8] L.X. Kong and P.D. Hodgson, Computational [23] Dongsheng Wen and Yurong He, Modelling of the
simulation of gas flow and heat transfer near an behaviour of gas–solid two-phase mixtures flowing
immersed object in fluidized beds, Advances in through packed beds, Chemical Engineering
Engineering Software 2007;38: 826–834. Science 2006; 61:1922 – 1931.
[9] J.C.S.C. Bastos, Modelling and simulation of a [24] Peng Li, Drag models for simulating gas–solid
gas–solids dispersion flow in a high flux flow in the turbulent fluidization of FCC particles,
circulating fluidized bed (HFCFB) riser, Catalysis Particuology 2009; 7: 269–277.
Today 2008; 130: 462–470. [25] Nan Zhang, Virtual experimentation through 3D
[10] I. Petersen and J.Werther, Three-dimensional full-loop simulation of a circulating fluidized bed.
modeling of a circulating fluidized bed gasifier for Particuology 2008; 6: 529–539.
sewage sludge. Chemical Engineering Science [26] Ranjeet P. Utikar, Single jet fluidized beds:
2005; 60: 4469–4484. Experiments and CFD simulations with glass and
[11] Vidyasagar Shilapuram, Comparison of polypropylene particles, Chemical Engineering
macroscopic flow properties obtained by three Science 2007; 62: 167 – 183.
methods of operation in a liquid–solid circulating [27] Scott Cooper, CFD simulations of particle mixing
fluidized bed, Chemical Engineering and in a binary fluidized bed. Powder Technology
Processing 2009;48: 259–267. 2005; 151: 27– 36.
[12] Jack T.Cornelissen, CFD modelling of a liquid– [28] Fluent User Guide, Fluent Inc. January 11, 2005.
solid fluidized bed, Chemical Engineering Science [29] Kari Myöhänen and Vesa Tanskanen, CFD
2007; 62:6334 – 6348. Modelling Of Fluidized Bed Systems,Lappeenranta
[13] Xiao-Bo Qi, Friction between gas–solid flow and University of Technology, FI-53851 Lappeenranta,
circulating fluidized bed downer wall, Chemical Finland.
Engineering Journal 2008;142: 318–326.
[14] Haiyan Zhu, Detailed measurements of flow
structure inside a dense gas–solids fluidized bed,
Powder Technology 2008; 180: 339–349.
[15] Zhongxiang Chen, Comparison of fluidized bed
flow regimes for steam methane reforming in
membrane reactors: A simulation study. Chemical
Engineering Science 2009; 64: 3598 – 3613.

174