Вы находитесь на странице: 1из 7

Fuel Processing Technology 87 (2006) 633 639

www.elsevier.com/locate/fuproc

Modelling of a wall fired furnace for different operating conditions


using FLUENT
R. Vuthaluru, H.B. Vuthaluru
Department of Chemical Engineering, Division of Engineering and Science, Curtin University of Technology, GPO Box U1987, Perth,
Western Australia 6845, Australia
Received 11 May 2005; received in revised form 9 January 2006; accepted 9 January 2006

Abstract
Studies into the effects of operating conditions in a wall fired furnace using commercial code FLUENT is presented in this paper. The
mathematical model is based on a Eulerian description for the continuum phase and a stochastic Lagrangian description for the coal particles. A
3-D combustor model is used to determine the temperature and heat flux profiles and other thermal characteristics for a typical 500 MW utility boiler
firing medium volatile coal. Temperature profiles for a 100% boiler load as well as reduced boiler loads are predicted using the 3-D model. Particle
trajectories are analyzed to identify the causes of operational problems such as fouling on burners, regions of high-unburnt carbon and temperatures
achieved in various parts of the boiler during the combustion process. The effect of excess air on the furnace temperature pattern is also studied.
Model calculations showed good agreement with experimental measurements in both full and pilot scale systems as well as limited literature data.
Using the experience obtained by these CFD model studies can potentially improve the operation of a boiler, regarding stability and local material
temperature of the walls, which lay a foundation for the boiler operating expert system.
2006 Elsevier B.V. All rights reserved.
Keywords: Wall fired furnace; Grid generation; CFD; Turbulent flow; Multiphase flow; Particle trajectories

1. Introduction
The interest on performance optimisation of large utility
boilers has become very relevant in recent years. All the
optimisation strategies are directed at extending their lifetime,
increasing the thermal efficiency and reducing the pollutant
emissions. Moreover, efficient use of pulverised coal is crucial
to the utility industry. To achieve higher combustion efficiency,
the major influencing factors such as the particle size
distribution, gas and particle temperatures, local heat release,
local oxygen concentration, kinetic parameters for coal devolatilization and char oxidation, char properties should be
understood thoroughly [1,2]. In engineering practice, it is very
difficult to investigate the combustion processes of various
kinds of combustibles directly in the boiler. Rather than
constructing real boilers and trying to check and improve

Corresponding author. Tel.: +61 8 9266 4685; fax: +61 8 9266 2681.
E-mail address: h.vuthaluru@exchange.curtin.edu.au (H.B. Vuthaluru).
0378-3820/$ - see front matter 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.fuproc.2006.01.004

these characteristics, computers are used to experiment with


models of the boilers [37].
One of the techniques to visualise parametric effects in boilers
is to opt for computational fluid dynamics (CFD) modelling,
which potentially can be an accurate and cost effective tool. Over
the last 20 years, the CFD has gained its reputation of being an
effective tool in identifying and solving problems related to
pulverized coal combustion [810]. In particular, it can provide
insights into the combustion characteristics of unfamiliar coals
and for this reason has been extensively applied to evaluate the
combustion performance of coals in bench-, pilot- and full-scale
furnaces [3,814]. However, limited studies have addressed the
effect of operating conditions (such as burner load, excess air
etc.) in the open literature on opposite wall-fired furnaces [1].
Also, earlier, the unstructured code (like FLUENT Version 6)
was not yet established to simulate pulverized coal combustion.
In the present study, an attempt has been made to simulate
and visualise the furnace operation, based on parameters of
interest to the combustion process through CFD modelling.
These parameters include temperatures achieved in various
parts of the boiler, particle residence time, particle velocity, and

634

R. Vuthaluru, H.B. Vuthaluru / Fuel Processing Technology 87 (2006) 633639


Table 2
Analyses of coal samples

Fig. 1. Furnace geometry.

radiation distribution within the furnace. This paper concentrates on the prediction of furnace performance for different
boiler operating conditions. The boiler geometry and operating
conditions along with the mathematical model used are detailed
in the following section. Results simulated are based on real
cases (different operating conditions used in the industry), but
the data sources have necessarily been disguised to the extent
that industry proprietary concerns are not infringed. It must be
mentioned that the verification of the computational simulations
are expensive to obtain, but even if the user has experimental
data available, the comparison between these data and
predictions leaves the user uncertain about both the errors in
the model and experimental data. Nevertheless, using the
experience obtained by a CFD model can significantly improve
the operation of a boiler, regarding stability and local material
temperature of the walls.
2. Furnace geometry and operating conditions
The three-dimensional geometry was created using GAMBIT a FLUENT pre-processor [15]. An isotropic view of the
geometry and grid system is shown in Fig. 1. The meshedgeometry contained 192,950 nodes with hexahedral cells in one
zone and the remaining zones with quadrilateral cells.
In order to make accurate quantitative predictions about
combustion behavior, it is important that an appropriate
geometry is created with suitable meshing techniques. About
Table 1
Specifications of the furnace geometry
Parameter

Value (m)

Distance between top burner level and furnace outlet


Clearance between burners:
Horizontal
Vertical
Burner to wall
Clearance between bottom burner level and furnace hopper

30
2.29 & 1.98
4.57
2.48
2.14

Proximate analysis

(% db)

Ultimate analysis

(% daf)

Moisture
Ash content
Volatile matter
Fixed carbon
Total sulfur

17
4
37
43
0.83

Carbon
Hydrogen
Nitrogen
Sulfur
Oxygen

74.42
4.9
1.5
1.0
18.13

60% of the project time is involved in creating the model


geometry. In particular, the geometry of typical combustor
exhibits two features which can impede its solution: the
discrepancy of scale between the burners and of the much
larger volume of the combustion chamber, and the discrepancy
between the typically three dimensional cylindrical geometry of
the boiler and the largely Cartesian geometry formed by the
confining walls of the combustion chamber. These difficulties
were surmounted by using flexible unstructured grid wherever
necessary and subdividing the solution domain into fine-grid
regions near the burners and patching them to more coarsely
girded regions elsewhere. Primary and secondary inlet details
are taken according to the actual burner dimensions. The swirl
angle is not considered for the present study.
The furnace geometry of the simulated boiler can be seen in
Fig. 1. The opposite wall fired coal furnace is 58 m high, 19.5 m
wide and 18 m deep and with an installed capacity of 500 MW.
Thirty burners are arranged in an array of six burners deposited
in different levels on two opposite walls of the furnace. The
specifications of the furnace geometry and original drawing are
given in Table 1 and Fig. 1. The main boiler operating conditions
considered in this paper include base case (corresponding to full
load with all burners in service, 3 rows of burners situated on the
front wall of the furnace and two rows on the opposite wall), 22
firing (two rows of the furnace situated on the front wall and two
rows on the opposite wall of the furnace), 21 firing (two rows
on the front wall of the furnace and one row on the opposite wall
in operation), respectively. Besides the above main operating
conditions, effect of 20% excess air on the boiler performance is
also considered in the present work.
3. Mathematical model
The mathematical model used here is based on the
commercial CFD code, FLUENT [15], where the gas flow is
described by the time averaged equations of global mass,
Table 3
Operating conditions of the boiler
Case description

Base case
32 firing

22
firing

21
firing

32 firing
with excess
air

Total primary air flow rate


(m3/s) & temperature (K)
Total secondary air flow rate
(m3/s) & temperature (K)
VM in coal (%)
Total coal feed rate (kg/s)

19.95 &
343
89.57 &
573 K
37
68.1

15.96 &
343
71.66 &
573 K
37
54.4

11.97 &
343
53.74 &
573 K
37
40.8

28.48 & 343


99.38 &
573 K
37
68.1

R. Vuthaluru, H.B. Vuthaluru / Fuel Processing Technology 87 (2006) 633639

635

Fig. 2. Scaled residuals for initial non-reacting flow solution.

momentum, enthalpy and species mass fraction. The particle


phase equations formulated in Lagrangian form and the
coupling between phases is introduced through particle sources
of Eularian gas-phase equations. The standard k- turbulence
model, single mixture fraction probability density function
(PDF) and the P1 radiation models are used in the present
simulations.
For the bulk of engineering combustion systems the mixing
process proceeds much more slowly than the chemistry and as a
result the mixing rate almost always determines the rate of
combustion [11]. Therefore, for the predominant industry case of
diffusion or non-premixed combustion for example, it normally
suffices to solve a conservation equation for a mixing variable
called the mixture fraction in order to determine the temperature
and concentrations of major species. In this study, a simplified coal
combustion furnace is modeled using the non-premixed combustion model for the reaction chemistry. The following steps describe
the method followed, after creating the required geometry.

volume form of the discretization equations are solved with


SIMPLE-based approaches. It is also assumed that the flow
field is at a steady-state and the solution procedure is simplified
by solving a steady-state form of the equation of motion. For
many combustion processes, radiation is not only the dominant
energy transport mechanism but also one of the most complex
problems. The accuracy of the radiation calculation depends on
a combination of the accuracy of calculation method and the
accuracy to which the properties of the radiating media and
surrounding walls are known [9]. The P1 radiation model is
used to account for the exchange of radiation between gas and
particulates, with cell based WSGGM model (weighted-sum-ofgray-gases model), which specifies a composition dependent
absorption coefficient. All thermodynamic data including
density, specific heat, and formation of enthalpies are extracted
from the pre-PDF chemical database.

1) A PDF table for a pulverized coal fuel using the pre-PDF


preprocessor has been prepared for medium volatile coal.
2) FLUENT inputs for non-premixed combustion chemistry
modeling were estimated.
3) A discrete second phase of coal particles was defined.
4) Simulations were carried out first without involving
reactions and discrete phase coal particles and then once
the convergence is achieved, simulations were continued
involving reacting discrete phase coal particles.

The computational model has been applied to the furnace of


500 MW boiler fired with high-ash, medium volatile coal. The
input data for simulations (including boiler operational conditions) is selected in accordance with data related to the
experimental tests of the considered boiler. The properties and
the lower heating value of fuel mass as received basis at a
local power station are assumed in calculations. The yield of
volatile matter was taken according to the data provided (daf).
Although, the full-scale boiler is equipped with 30 conventional
burners of the swirling type arranged on the front and rear

The composition in terms of atom fraction of H, C, N, O, along


with the lower heating value and heat capacity of the fuel are
defined using the data provided by a local power station
personnel. The fuel composition inputs were determined using
proximate and ultimate analysis data provided. Table 2
summarizes the proximate and ultimate analysis data (db: dry
basis; daf: dry-ash-free basis), which was used to derive the
elemental composition of the volatile stream. The volatile release
model is based on a single kinetic rate model [15,16]. This model
states that the rate of production of volatile gases is given by a first
order reaction and the rate constant is expressed in an Arrhenius
form, which correlates rates of weight loss with the temperature.
As in most practical fossil-fuel combustion simulations, the
incompressible form of the equations of motion using a finite-

4. Results and discussion

Fig. 3. Scaled residuals for total combustion process including the discrete phase.

636

R. Vuthaluru, H.B. Vuthaluru / Fuel Processing Technology 87 (2006) 633639

1.80e+03
1.73e+03
1.66e+03
1.59e+03
1.52e+03
1.45e+03
1.38e+03
1.31e+03
1.24e+03
1.17e+03
1.10e+03
1.03e+03
9.60e+02
8.90e+02
8.20e+02
7.50e+02
6.80e+02
6.10e+02
Y
5.40e+02
4.70e+02
4.00e+02 Z

Full load 3-2 firing

Case 2-2 firing

Case 2-1 firing

Fig. 4. Temperature fields in a coal fired furnace at three different operating conditions. Plane z = 10.65 m crossing first burner in each row.

furnace walls in three tiers, only five of the burner sets are in
operation at any particular time and one set is used as a standby.
Hence, only five sets are considered as full load (hereafter
referred to as base case) for modelling the operation.
The calculated primary and secondary air velocities are 20
and 30 m/s, respectively, which have been determined from the
mass flow of combustion air and temperature of preheated air.
Another boundary condition is the outflow that is expressed as
outlet with the external radiation temperature. This temperature
is equal to the wall temperature and approximately corresponds
with vaporization of water in evaporator. In this case, the
temperature of the walls is equal to 600 K with an emissivity of
1. The summary of boundary conditions (inlets and outlets
conditions) is given in Table 3.
4.1. Solution approach to base model
It is generally recommended to start and obtain a nonreacting flow solution for coal combustion [57]. Despite the
2.10e+03
2.02e+03
1.93e+03
1.85e+03
1.76e+03
1.68e+03
1.59e+03
1.51e+03
1.42e+03
1.34e+03
1.25e+03
1.17e+03
1.08e+03
9.95e+02
9.10e+02
8.25e+02
7.40e+02
6.55e+02
5.70e+02
4.85e+02
4.00e+02

large density variations that occur in reacting flows, the velocity


and turbulence fields are similar for reacting and isothermal
combustors. The trajectory of the DPM (Discrete Phase Model)
coal particles is critical, and sensitive to the velocity and
turbulence fields, hence starting from a non-reacting solution is
often necessary for convergence. The non-reacting solution was
obtained by disabling energy, radiation, and DPM for equations
and setting the pressure discretization to PRESTO!. This is
often useful for buoyant flows where velocity vectors near walls
may not align with the wall due to the assumption of uniform
pressure in the boundary layer for these highly buoyant cases.
The residuals in obtaining the initial solutions (for non-reacting
flow) are shown in Fig. 2.
Once the non-reacting solution is converged coal particles
are set to release by enabling the energy, P1 and DPM model
equations. With sudden heat release, the velocity and turbulence
fields may change very quickly and cause divergence. To
control this, momentum and pressure under-relaxation factors
have been reduced for the first order solution. The value of 0.5
for both pressure and momentum is chosen. The Discrete Phase
Sources under-relaxation factor is set to 0.25, which is more
stable than the default of 0.5. Fig. 3 shows the residuals with
energy, radiation and discrete phase models enabled.

Table 4
Comparison of predicted temperatures with measurements in full- and pilotscale furnaces and (literature) reported data

Y
Z

Full load with excess air


Fig. 5. Temperature field in a coal fired furnace full load 32 firing with excess
air (20%). Plane z = 10.65 m crossing first burner in each row.

Case

Base case
32 firing

22
firing

21
firing

Base case 32
firing with
excess air

Current model
Full scale boiler
Pilot scale furnace
Literature datab

1332
12901350
1331a
1287

1304

1260

1168

1067

1495

a
A firing rate of 218 kWth with excess air adjusted to give 2.6% (v/v, dry)
oxygen in flue gas.
b
Reported values correspond to base-case (all 20 burners), 14 rows (16
burners) and partial load 34 rows (8 burners) in service, respectively.

R. Vuthaluru, H.B. Vuthaluru / Fuel Processing Technology 87 (2006) 633639


Table 5
Comparison of radiative heat transfer (kW) to walls with measurements in fullscale furnace and literature reported data
Case

Base case
32 firing

22
firing

21
firing

Base case 32
firing with
excess air

Current model
Full scale
Pilot scale
Literature data a

5.4 105
3.0 105

4.98 105
3.18 105

3.91 105
2.94 105

8.16 105

3.3 105

3.2 105

1.65 105

Reported values correspond to base-case (all 20 burners), 14 rows (16


burners) and partial load 34 rows (8 burners) in service, respectively.

By fitting the size distribution data (1090 m) to the


RosinRammler equation, the required particle distribution is
achieved. This initial pulverized fuel size distribution was
chosen based on the sieving data provided by the utility personnel. The model can be further improved to simulate actual
combustion process by adopting a size breakage model to
control the distribution better.
4.2. Model results
The predicted temperature distribution in a vertical plane
crossing the burners is shown in Figs. 4 and 5 for the configurations considered. Initially some general trends observed
are reported from the analysis of the temperature distribution,
and then the comparison of each case is presented. The
temperature distribution for all the cases are shown at a chosen
vertical plane, z = 10.65 m from the left wall, the plane crossing
first burner of each row considered. A similar scale range was
used for all the cases, except for the excess air case for
comparison.
It can be seen that the temperature approaches a maximum in
the vicinity of the burner zone or at active combustion zone. The
maximum temperature is about 1800 K, the upper limit esFront
Wall
1.18e+05
7.75e+04
3.72e+04
-3.09e+03
-4.34e+04
-8.37e+04
-1.24e+05
-1.64e+05
-2.04e+05
-2.45e+05
-2.85e+05
-3.25e+05
-3.66e+05
-4.06e+05
-4.46e+05
-4.86e+05
-5.27e+05
-5.67e+05 Y
-6.07e+05
-6.48e+05 Z
-6.88e+05

637

tablished for the burner zone outlet, i.e., the non-slagging


furnace operation is ensured. The figures show that combustion
occurs mainly in the near burner region. As the furnace chamber
is 58 m, the temperatures shown in the vertical plane implies
that most of the combustion was occurring about half to three
quarters of distance along the chamber.
Based on the plant observations at power station boiler, it
was found that the exit temperature of flue gas must be in the
range of 12901350 K. It can be seen from Fig. 4, this
requirement has been achieved in our simulations. Establishing
these temperatures are essential in order to avoid slagging of the
water walls at boiler loads close to the 100% rated load [4,17].
Comparison of temperatures suggests that as the load or number
of burners in service increases, the output temperature increase
(see contour plots in Fig. 4). The flow field for all the operating
conditions is typical for an opposite wall fired furnace. Average
furnace exit temperatures predicted from the present simulations
are summarised in Table 4.
Experimental tests are conducted with the same coal in a
pilot scale combustor that simulates the combustion and heat
transfer processes occurring in a full-scale boiler. It consists of a
down-fired refractory lined furnace fitted with a variable swirl
burner. The furnace combustion chamber is 680 mm in diameter
and 2.5 m high, to give a residence time of approximately 3 s.
The furnace has a normal firing rate of 150 kW, equivalent to a
coal feed rate of approximately 25 kg/h depending on coal
specific energy. All combustion tests performed in the furnace
on the coal tested used a nominal pulverised fuel size of 70%
passing 75 m. Gas temperatures in the radiative and transition
sections were measured using a suction pyrometer and are
accurate to within 30 C of the real gas temperature.
Experimental data collected with similar coal in this system
indicated exit temperatures of 1331 and 1456 K at different
firing loads (150 and 218 kW, respectively). The effect of boiler
load is evident in pilot scale runs also with temperatures rising
Left
Wall

Rear
Wall

Fig. 6. Radiation heat flux (w/m2) to walls for the case of excess air.

Right
Wall

638

R. Vuthaluru, H.B. Vuthaluru / Fuel Processing Technology 87 (2006) 633639

Fig. 7. Particle traces coloured by particle residence time for selected burner
ports. a) Base case; b) Base case with excess air.

with load. These observations are in close agreement with our


model predictions (see Table 4).
Measured temperatures for a front wall fired furnace reported
in the literature [1] are also used for comparison. The reported
values ranged from 1287 K for base-case (with all six of the
burners in operation), 1257 K for 14 burners in service and 2
5 firing with similar measured temperatures. For partial load the
measured temperature was reported to be 1067 K indicating the
same trends obtained from our simulations. As the load decreases, the outlet temperature decreased significantly as expected. These measured observations are in general agreement
with our predicted values with some differences as expected due
to different furnace configuration (front wall fired as opposed to
opposite wall fired furnace).
As expected simulations with excess air showed significantly
higher temperatures compared to stoichiometric air ratios used
in all other cases. Typical radiative heat transfer to the walls
(kW) presented in Table 5 substantiates this fact further with
higher total energy to walls in case of excess air simulations.
Fig. 6 shows the contour plots of radiative heat fluxes (w/m2) to
walls for the case of excess air. Comparison of our simulation
results with literature reported values [1] indicated close
agreement (Table 5).

The next parameter chosen is the residence time, which is


paramount importance for combustion process. Fig. 7 shows
particle traces (colored by particle residence time) for lower and
upper rows of the front and rear walls for base case with and
without excess air. As it can be seen, left panel of the picture
shows particles from selected burners generally traverse in
straight lines with relatively shorter residence time. However,
particles injected from one of the lower most burners trace back
into the hopper section with relatively longer residence time
with higher recirculation of particle streams. The differences in
residence times of the particles will lead to different proportions
of unburnt carbon and alter ash deposition propensities. It
should be noted that unburned carbon will not necessarily
depend on residence time as other conditions (such as temperature, oxygen) also equally dictate the extent of combustion
in a boiler.
In general, several factors affect the carbon burnout of a fuel
under a given burner/furnace design and operating conditions
including coal rank, quality and petrographic characteristics,
presence of low melting inorganics in coal ash, residence time
available for combustion in furnace, type and number of
burners, type of milling system and primary air control system,
percentage of coarser particles, primary air to secondary air
ratios, excess air at the burner/furnace and burner tilt [1,3,7,18].
For example, the amount of oxygen will be usually small at the
bottom section of the boiler indicating that there is a possibility
that unburned carbon of the lowest burners is higher than that of
the upper burners. This is expected to increase the fraction of
unburnt carbon in the flue gas stream and thus reduces the
efficiency of a boiler. Additionally, the particle trajectories as
reported in Fig. 7b for simulations with excess air indicated
much shorter residence times, which are not practical for
efficient combustion operations in power stations. It is possible
to simulate with a range of excess air levels as opposed to one
value used in the current studies to optimize the residence times
for smooth operation of boilers. It is also necessary to present
the relation between the residence time distribution and the
calculated unburned carbon in each burner rows to substantiate
our visual observations. This requires further attention in future
studies.
5. Conclusions
Modelling of a wall-fired furnace is carried out using CFD
commercial flow solver FLUENT. Simulation results with full
and partial loads with different burners in operation suggested a
significant impact on the distribution of temperatures and
furnace exit temperatures. This indicates that extreme care must
be taken in terms of maintenance and service of burners so as to
minimise the effects on combustion and continue smooth
operation of boilers. Simulations carried out with excess air
indicated that the temperatures within the boiler will increase.
This could lead to severe ash deposition problems. Predicted
particle trajectories indicate differences in particle residence
times within the boiler. In particular, particles emerging out
lower burner sections tend to reside within the boiler longer than
the particles at other locations. This may lead to poor

R. Vuthaluru, H.B. Vuthaluru / Fuel Processing Technology 87 (2006) 633639

combustion, which is undesirable for efficient operation of


power station boilers.
Overall the model results showed good agreement with
measured and reported data, which provides a solid foundation
for future work. Data presented in the paper demonstrates the
implications and possible options for burner operations in terms
of their maintenance and serviceability. Current model in
combination with ash deposition sub-model can provide useful
insights into the slagging and fouling associated with firing of
these coals. The results obtained with such models would prove
to be of practical value to optimize the performance of boilers.
References
[1] Minghou Xu, J.L.T. Azevedo, M.G. Carvalho, Modeling of a front wall
fired utility boiler for different operating conditions, Comput. Methods
Appl. Mech. Eng. 190 (2001) 35813590.
[2] S. Chapman, T.G. Cowling, The Mathematical Theory of Non-Uniform
Gases, Cambridge University Press, Cambridge, UK, 1990.
[3] Scott C. Hill, L. Douglas Smoot, A Comprehensive three-dimensional
model for simulation of combustion systems: PCGE-3, Energy Fuels 7
(1993) 874883.
[4] V.I. Kouprianov, Modelling of thermal characteristics for a furnace of a
500 MW boiler fired with high-ash coal, Energy 26 (2001) 839853.
[5] M.P. Mathur, D. Gera, M. Freeman, Computational Fluid Dynamics
Modelling Analysis of Combustors, Fluent, Inc., Collins Ferry Road,
Morgantown, West Virginia, 2002.
[6] P.L. Stephenson, Mathematical modelling of semi-anthracite combustion
in a single burner furnace, Fuel 82 (2003) 20692073.

639

[7] T. Sabel, B. Risio, S. Unterberger, U. Schnell, K.R.G. Hein, M. Kab,


U. Priesmeier, H.U. Thierbach, Full scale measurements and mathematical
modelling studies for the investigation of the combustion behaviour of a
modern bituminous coal-fired boiler, IFRF Combust. J. (2001) 120
(Article number 200102).
[8] S. Niksa, Coal Combustion Modelling, IEAPER/31, IEA Coal Research,
London, 1996.
[9] M. Eaton, L.D. Smoot, S.C. Hill, C.N. Eatough, Components, formulations, solutions, evaluation, and application of comprehensive combustion
models, Prog. Energy. Combust. Sci. 25 (1999) 387436.
[10] T. Abbas, P.G. Costen, F.C. Lockwood, 26th Symposium (International) on
Combustion, The Combustion Institute, 1996, p. 3041.
[11] P.G. Costen, D. Dajnak, M. Messina, F.C. Lockwood, T. Abbas, C.
Bertrand, N.H. Kandamby, V. Sakthitharan, I. Siera, S. Yousif, On the
prediction and control of industrial combustors by mathematical
modelling, Paper presented at 2002 Australian Symposium on Combustion
and The Seventh Australian Flame Days, Adelaide, February 2002.
[12] J. Truelove, D. Helcombe, 23rd Symposium (International) on Combustion, The Combustion Institute, 1990, p. 963.
[13] B.S. Brewster, L.D. Smoot, S.H. Barthelson, D.E. Thornock, Energy Fuels
9 (1995) 870.
[14] J.M. Jones, P.M. Patterson, M. Pourkashanian, A. Williams, A. Arenillas,
F. Rubiera, Fuel 78 (1999) 1171.
[15] FLUENT, FLUENT 6.0 Users Guide, , 2002 Lebanon, USA.
[16] Williams, R. Backreedy, R. Habib, J.M. Jones, M. Pourkashanian,
Modelling coal combustion: the current position, Fuel 81 (2002) 605618.
[17] Eddy H. Chui, M.J. Douglas, Y. Tan, Modelling of oxy-fuel combustion
for a western Canadian sub-bituminous coal, Fuel 82 (2003) 12011210.
[18] V.T. Sathyanathan, K.P. Mohammad, Prediction of unburnt carbon in
tangentially fired boiler using Indian coals, Fuel 83 (16) (November 2004)
22172227.

Вам также может понравиться