Accepted Manuscript

Numerical Investigation on the Lifetime Decline of Burners in a Wall-Fired Dual-Fuel

Utility Boiler
M. Pourramezan, M. Kahrom, M. Passandideh-Fard
PII:

S1359-4311(15)00199-4

DOI:

10.1016/j.applthermaleng.2015.02.067

Reference:

ATE 6419

To appear in:

Applied Thermal Engineering

Received Date: 18 November 2014

Revised Date:

13 February 2015

Accepted Date: 24 February 2015

Please cite this article as: M. Pourramezan, M. Kahrom, M. Passandideh-Fard, Numerical Investigation
on the Lifetime Decline of Burners in a Wall-Fired Dual-Fuel Utility Boiler, Applied Thermal Engineering
(2015), doi: 10.1016/j.applthermaleng.2015.02.067.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Numerical Investigation on the Lifetime Decline of Burners in a Wall-Fired Dual-Fuel

Utility Boiler
M. Pourramezan, M. Kahrom, M. Passandideh-Fard

Department of Mechanical Engineering, Faculty of Engineering, Ferdowsi University of Mashhad,

RI
PT

Mashhad, Iran

Abstract

SC

This paper reports the results of the numerical investigation on the causes of burners lifetime decline in a
dual-fuel, front-wall-fired utility boiler. For this purpose, a three-dimensional modeling was performed

M
AN
U

using Computational Fluid Dynamics (CFD) to simulate a 150 MW boiler. The whole boiler, including the
water-wall, low-NOx swirling burners with both types of nozzles, etc., was modeled in the real dimensions.
Furthermore, the heat exchangers, like superheaters and economizers, were treated as porous media. Major
attention was paid to the furnace zone, and especially region close to the burners where the failures have
occasionally occurred due to the metal overheating. In order to achieve the objective, effects of the fuel

TE
D

type, the burners swirl ratio, droplet diameter of the fuel, and air-fuel ratio were investigated. Comparisons
between numerical results, field measurements and plant observations showed the simulation can predict
the flow and temperature fields in the boiler with acceptable accuracy for both fuels. It was found that, the

EP

most damage occurs while the plant works with natural gas, since the flame establishes closer to the
burners. In this situation, the average heat flux received by the rear wall is about 0.25% higher than the

AC
C

received heat flux due to heavy fuel oil. Additionally, the rise in the swirl ratio of the burners can be
another parameter, which strongly influences the flow and temperature field within the boiler and causes
overheating damages to the burners and water-wall. Although variations in other investigated parameters
have dramatic effects on performance characteristics of the boiler, they cannot lead to the overheating
failure.
Key words: CFD modeling; Wall-fired boiler; Duala-fuel firing; Burner; Temperature distribution

Corresponding author. Tel: +989153219376; Fax: +985132768836.

E-mail address: m.pourramezan@stu.um.ac.ir (M. Pourramezan).

ACCEPTED MANUSCRIPT

1. Introduction

In the past decade, Iran has started the privatization of its power-generation industry and, particularly, its
power plants. As a result, the electricity market environment in this country has become increasingly
competitive due to the deregulation, along with the globalization of the industry. Therefore, all power plants
in Iran, like other countries, have made great efforts to optimize and extend the lifetime of equipment and

RI
PT

thereby minimizing their downtime. This has led to giving greater consideration to the elements of power
plants like boilers, as the worlds largest consumers of fossil fuels. Given Irans vast resources of oil and
gas, all utility boilers in this country consume heavy fuel oil or natural gas, and some of them are also

SC

capable of using both, alternately.

A modern utility boiler must have the maximum lifetime, minimum rupture of the boiler tubes, highest

M
AN
U

thermal and combustion efficiency, lowest pollutant emissions and maximum power generation. The furnace
efficiency mainly depends on the fuel type and its features. In addition, fuel combustion can be affected by
several factors such as burners characteristics, droplet size distribution and rate of devolatilization of these
droplets (for nongaseous fuels), temperature distributions of the flame, air-fuel ratio, etc. [1]
Due to the expensive price of measurements of the combustion characteristics and the limitation by

TE
D

geometry, time, the number of costly instruments and skills required, only a few detailed experimental
works on large-capacity boilers are reported in the literature [2-7]. As a result, the use of Computational
Fluid Dynamics (CFD) full-scale modeling has become a useful tool to predict the performance and

EP

operational conditions of the boilers.

Numerical simulation of boilers requires comprehensive modeling for fuel combustion processes, species

AC
C

reactions, turbulence, radiation, etc. [8, 9]. With the impressive advances and pervasive accessibility of
modern computers, recent studies could tackle more practical power plant issues. Some examples include
optimization of the operating conditions [10-12], reduction of pollutant emissions [13], co-fired and
recovery boilers [14, 15], NOx formation [16-18], boilers tube rupture [19, 20] and uneven wall
temperature distribution [20, 21].
The present paper describes the results of a numerical investigation on the causes of the lifetime decline of
burners and water-wall located within the furnace zone of a dual-fuel, heavy fuel oil and natural gas, frontwall-fired utility boiler. A special attention is exerted to the furnace zone since the frequent failures,
thought to be due to the metal overheating, occur during the power plant operation. CFD simulations and gas
2

MANUSCRIPT
temperature measurements, together ACCEPTED
with the evaluation
of the relevant plant data have been conducted to
reveal the principal causes of this failure. However, in order to achieve the objective, this paper investigates
the effects of fuel type, swirl ratio of the burners, droplet size of heavy fuel oil, and air-fuel ratio on the
temperature distribution within the utility boiler.

RI
PT

2. Description of the utility boiler

There are many wall-fired utility boilers throughout the world working with various types of fuel. Some
instances can be found at Tous power plant in Iran [22], Hammond in The United States [23], Elektrnai in

SC

Lithuania [24], Majuba in South Africa [25], etc.

In a wall-fired boiler, burners are mounted in the boiler walls, producing discrete flames in the furnace.

M
AN
U

However, burners may be mounted in a single boiler wall (front-fired) [26] or in two opposing walls
(opposite walls-fired) [27]. The advantages of wall-fired furnaces are the possibility of the boiler operation
with low load, stable flame in a single swirl burner, easy ignition of fuel-air mixture, fast mixing of fuelprimary air mixture and secondary air, and considerable burn-out of carbon near the burners [28-30].
In comparison between front-wall-fired and apposite-walls-fired, the former has the ability to work in

TE
D

lower loads, convenient load control and also low maintenance costs. On the other hand, with apposite
arrangement of burners (apposite-walls-fired), fuel can be more efficiently burnt as a result of the intensive
turbulization of the flame in the main combustion zones, due to large velocity gradients between adjacent

EP

jets moving in opposite direction. Therefore, heat absorption by water walls in the furnace is much more

AC
C

uniform. Furthermore, the apposed firing is employed in high-capacity steam boilers in which the required
number of burners cannot be arranged on a single-wall [31].
Each of the four units at 600 MW Mashhad Touss Power Station consists of a natural circulation, dualfuel, suspension boiler with three staged superheating, two staged reheating and front-wall-fired furnace
with an installed capacity of 150 MW. These utility boilers are 40.9 m in height, 9.18 m in width, and 9.27
m in depth. Each unit has three burner elevations with three swirling burners per elevation (33 matrix) for a
total of nine burners installed on the rear wall (see Fig. 1). The boilers are designed for a peak main steam
mass flow rate of 525,000 kg/hr at 813 K and 160 bars. Table 1 shows the analysis of both fuels of the
boiler.

ACCEPTED
MANUSCRIPT
Conventional low-NOx swirling burners
used in the
utility boiler have a central nozzle for injecting a
mixture of fuel oil and primary air, with a circular register around the nozzle for supplying secondary air,
guiding vanes to swirl the combustion air, and also gas injectors around the guiding vanes (see Fig. 2).

3. Simulation and measurements

RI
PT

3.1. Numerical simulation

The CFD modeling involves with the numerical solution of the conservation equations. In this study the
commercial software, ANSYS FLUENT 14.5 [32] which solves the Reynolds-Averaged Navier-Stokes

SC

(RANS) and species transport equations, was used to build the model to simulate combustion, fluid and
particle flow, as well as heat and mass transfer inside the utility boiler. The time-averaged RANS equations

M
AN
U

are used because of the complex nature of the present phenomena and large range of turbulence scales. An
implicit pressure-based steady-state three-dimensional solver with standard wall functions is used for the
calculations. Also, turbulence is modeled with the realizable k-epsilon model. The realizable k-epsilon
model is relatively widely used for engineering applications and provides better performance in many
industrial turbulent flows than other k-epsilon models [11, 33]. Moreover, the pressure-velocity coupling

TE
D

was modeled with the semi-implicit method for pressure-linked equations (SIMPLE) scheme and the
turbulence-chemistry interaction with the eddy-dissipation concept. In the eddy-dissipation model, the total
space is subdivided into fine structures and the surrounding fluid. This model relates the rate of reaction to

EP

the dissipation rate of the reactant and product-containing eddies [34, 35]. In the present study, we used

AC
C

Four-step mechanism based on the one presented in [36, 37] for natural gas combustion:
CH 4 +0.5O 2 CO+2H 2

(1)

CH 4 +H 2 O CO+3H 2

(2)

CO+H 2O CO 2 +H 2

(3)

H 2 +0.5O 2 H 2 O

(4)

Since the real mechanism of fuel oil combustion is very complicated and has numerous steps, this
mechanism is not widely investigated. We simply utilize the global two-step mechanism as:
C19 H 30 +17O 2 19CO+15H 2 O

(5)

CO+0.5O2 CO2

(6)
4

MANUSCRIPT
The radiation was solved by using ACCEPTED
the discrete ordinate
(DO) model along with the weighted-sum-of-graygases model (WSGGM) used to calculate the absorption coefficient of the combustion products. Also, Soot
was modeled with the two-step Tesner [38] model. In addition, the discrete phase (heavy fuel oil droplets)
was modeled by using the EulerianLagrangian approach with the pressure-based solver. The energy,
species mass fractions, momentum, turbulent kinetic energy, and turbulent dissipation rate equations are

RI
PT

solved with second-order upwind discretization.

Conducting an investigation to ensure the independence of the simulated results from the grid size, the
3483 m3 utility boiler meshed into 2,300,000 tetrahedral control volumes including three computational grid

SC

zones: furnace zone (with 1,250,000 CVs), superheater zone (with 535,000 CVs), and economizer zone
(with 515,000 CVs). Relatively high grid densities are employed in the regions, like the furnace zone and

transfer were expected to occur (see Fig. 1).

M
AN
U

regions close to the burners and air entrance channels, where most of the volatile combustion and heat/mass

To obtain the temperatures of all sections of the water-wall, besides using the existing experimental data
of the plant, a separate model was specifically created. Linking this model to full-scale model of the utility

TE
D

boiler, the temperatures of all sections of the water-wall were obtained through a time-consuming process of
trial and error. Through this process, after solving the full-scale model of the boiler, with assumed
temperatures of the water-wall sections, all parameters required to model the water-wall like: average

EP

temperature and velocity of combustion products, and average absorption coefficient close to each particular
water-wall section were obtained. Then, these obtained data were used in the separate water-wall model to

AC
C

achieve new temperatures of the sections. Using modified temperatures in the full-scale model of the boiler,
we obtained updated flow parameters to use in the water-wall model. Repeating this trend, corrected average
temperatures of all sections of the water-wall were calculated that are consistent with the existing
experimental temperatures in the plant. The final values of the temperatures and other related parameters
used in the boiler simulation are shown in Table 2.
Since each heat exchanger acts like a heat sink and also causes pressure drop, it is evident that solving
flow and thermal fields in boilers without considering the effects of heat exchangers includes immense error
and lack of precision in the results of simulation. In this study, the heat exchangers in the boiler are treated
as porous media, which allows the pressure drop and heat sink across the tube bundle. The pressure losses
5

MANUSCRIPT
and heat sinks of the heat exchangersACCEPTED
used in the present
study were calculated from the existed data in the
plant and are listed in Table 3.
To investigate the working conditions of the burners and atomization process, a sample nozzle by a wirecut process was cut in different cross sections and the geometric characteristics of the injector were carefully
analyzed. Also, a CAD model of the nozzle was created based on the measured dimensions of the sample

RI
PT

nozzle (see Fig. 2). Finally, the burner and its nozzle were simulated separately and obtained results were
utilized in the main model of the boiler. Since the focus of this paper is on the modeling of flow and
temperature distribution within the furnace zone and the region close to burners, the modeling of the burners

SC

was only described briefly here. More details can be found in [39]. The main operating conditions (for fullload) used in CFD simulation are shown in Table 4.

M
AN
U

3.2. Experimental measurements

The data of the utility boiler used for validation and comparison with numerical results are values at boiler
exit; gas temperature, CO2, H2O, SO2, and NO concentrations. The gas temperature was obtained from the
average temperature of 4 points at the boiler exit and species concentrations are measured with portable

TE
D

electrochemical-based emissions and combustion gas analyzers.

Gas temperatures just on the first pass supporting tubes were measured with the specially manufactured B
type thermocouples. The location of 136 thermocouples is shown in Fig. 3.

EP

The boiler has 33 bundles of supporting tubes with 12 tubes per bundle in three rows for a total number of
396 tubes. The thermocouples were installed just on the top side of the tubes of the top row at the height of

AC
C

24.68 m from the furnace bottom and to the depths of 0.38 m and 1.8 m from the rear wall (see Fig. 3). All
the mentioned experimental data were measured at full-load of the plant.

4. Results and discussion

4.1. Comparison between numerical results and field measurements

The simulation results and the measured values are shown in Table 5. The discrepancies between the
measured and predicted values are less than 10% in both cases of heavy fuel oil and natural gas except for
NO concentration, which has a difference of 14%.

There are three different chemicalACCEPTED

mechanisms ofMANUSCRIPT
NO formation in combustion, corresponding to three
distinctive names as thermal NO, prompt NO, and fuel NO. In a conventional furnace, the thermal NO
emission will be more significant when the temperature is over 1800 K. Also, the thermal NO formation is
highly dependent on temperature, but independent of fuel type. In fact, the thermal NO production rate
doubles for every 90 K temperature increase beyond 2200 K [40]. Therefore, with respect to the discrepancy

RI
PT

between the predicted and measured temperatures, the difference in simulated and measured values of NO
can be logical.

However, even with the relatively large difference in NO prediction, the features of flow and temperature

SC

distributions for both fuels described in the following sections are in an acceptable agreement with the plant
data, so this simulation is reasonable to be used in the present investigation.

M
AN
U

4.2. Effects of fuel type on predicted results

Needless to say, the type of fuel is the most important factor which affects the working characteristics of
utility boilers and industrial furnaces. In order to investigate the effects of fuel type on the boiler, CFD
simulation was performed at full load conditions (150 MW). It should be noted that the primary fuel of the

TE
D

boiler is natural gas, and heavy fuel oil is used as a secondary fuel.

Although the maximum temperature of the boiler for natural gas is more than 100 K higher than that of
heavy fuel oil, volumetric average temperatures of the boiler in both cases are nearly equal. Therefore, for

EP

heavy fuel oil, the temperature field in the boiler is more uniform, and the temperature gradient is less than
that of natural gas, Table 5.

AC
C

As it is seen in Fig. 4, in comparison with the flame of heavy fuel oil, the flame due to natural gas forms
closer to the burners. Since the mixing rate of heavy fuel oil is lower than that of natural gas, combustion
occurs farther to burners for heavy fuel oil. Considering Fig. 5, this phenomenon could be described in
another approach, where the O2 mass fraction at far distances from the burners, in the case of heavy fuel oil,
is higher than the O2 mass fraction due to natural gas combustion. This means that the combustion process
occurs sooner, and closer to the burners for natural gas.
Fig. 6 shows the variation of temperature along the horizontal line which passes through the central point
of the central burner (burner 25 in Fig. 1). It is shown in this figure, and also was shown in Table 5, that the
maximum temperature of the boiler for natural gas is about 100 K higher than that of heavy fuel oil.
7

ACCEPTED
MANUSCRIPT
Furthermore, temperature discrepancy
in regions close
to burners is obvious in Fig. 6. For instance, at the
distance of 1 m from the rear wall, for natural gas, the temperature is about 600 K higher than the
temperature at the same position for heavy fuel oil. This discrepancy reaches to 730 K at 1.5 m from the rear
wall. However, by approaching the front wall (the furthest place from the burners) the temperature, for
heavy fuel oil, is a bit (54 K at 1 m to the front wall) higher than the temperature at the same position for

RI
PT

natural gas. This observation seems reasonable with respect to what was said about the difference between
mixing rates of natural gas and heavy fuel oil.

Furthermore, the average heat flux received by the rear wall, for natural gas, is about 0.25% higher than

SC

the received heat flux for heavy fuel oil combustion. In addition, the heat flux received by the burners of the
highest elevation (e.g. burner 37) is clearly higher than the heat flux received by the burners of the lower

M
AN
U

elevations (e.g. burner 13) (Fig. 7). These observations are completely consistent with the real situation
existed in the plant.

In accordance with foregoing, it can be concluded that, contrary to traditional beliefs, which consider
heavy fuel oil as the main cause of overheating damages in boilers, in the present case, the most damage to

TE
D

the burners and water-wall close to the burners occurs while the plant works with natural gas.
4.3. Effects of the swirl of secondary air

As it was shown in Fig. 2, in order to obtain better mixing of air and fuel, guiding vanes are used in the air

EP

path of the burners. Defining the swirl ratio (S.R.) as the ratio of tangential velocity to axial velocity and
simulating the utility boiler for various values of the swirl ratio, it was found that the boiler operating

AC
C

condition is highly dependent on this ratio. After analyzing the guiding vanes of a sample burner and also
comparing the predicted results with plant data, it was found that the present swirl ratio of the burners is 0.7
for both natural gas and heavy fuel oil.
S .R. =

Tangential velocity U
=
Axial velocity
Ux

(7)

Fig. 8 shows that the flame configuration is highly dependent on the swirl ratio. As the swirl ratio
increases, the flame loses its stable condition and approaches the rear wall. In high values of the swirl ratio,
all nine flames combine together and establish a flat flame on the rear wall. This trend in both cases of
natural gas and heavy fuel oil is qualitatively the same. Furthermore, as the swirl ratio rises, a flame hole,
8

MANUSCRIPT
characterized by a flame sheet with ACCEPTED
a circular extinguished
region in the center, forms. These observations
are consistent with the results of the experimental studies that were reported in [41] and [42].
The variation of the average temperature of a vertical plane close to the rear wall (at x=0.25 m) is shown
in Fig. 9. It is clearly evident as the swirl ratio rises, for both fuels, the average temperature of the rear wall
increases, too. In comparison, the boiler is more sensitive to variation of the swirl ratio, while working with

RI
PT

heavy fuel oil. Also, as the swirl ratio exceeds 4, there is no significant change in the average temperature of
the rear wall. For the present swirl ratio of the burners (0.7) the average temperature of the plane for natural
gas is more than 170 K higher than the average temperature of the plane for heavy fuel oil.

SC

Considering what was said, by decreasing the swirl ratio, the flame distances from the rear wall and as the
swirl ratio increases the flame approaches it, causing damage to the burners and the water-wall close to the

M
AN
U

burners. In addition, it should be noted that the swirl ratio cant be reduced to very small values. As the swirl
ratio decreases, in addition to the reduction of the temperature and combustion efficiency of the boiler, the
flame approaches the front wall and causes overheating damages. Therefore, more considerations should be
taken to obtain the optimal swirl ratio of the burners.

The variations in the average temperature of the planes close to the rear and front walls along with the

TE
D

combustion efficiency of the boiler with respect to the swirl ratio are shown simultaneously in Fig. 9. The
combustion efficiency was obtained from Eq. (8):

A+L
H

Where:

(8)

EP

AC
C

: Combustion efficiency of the boiler.

A: Absorbed heat flux by all heat exchangers and the water-wall.
L: Lost heat flux at the boiler outlet.
H: Higher heating value of the fuel.

As it is seen in Fig. 9, the optimal value of the swirl ratio is 0.55 for natural gas. Furthermore, using the
same method, the optimal value of the swirl ratio for heavy fuel oil was obtained 0.65, which seems
reasonable, with respect to slower combustion and lower mixing rate of heavy fuel oil. The difference
between the combustion efficiency of the present swirl ratio (0.7) and the combustion efficiency for the
9

MANUSCRIPT
optimal swirl ratio (0.55) is less thanACCEPTED
0.5%, so this reduction
in the swirl ratio doesnt affect the combustion
process. It should be considered that the swirl ratio could be changed by changing the angle of the guiding
vanes of the burners, and it cant be changed during the operation of the boiler. Therefore, the swirl ratio
should be selected with respect to the primary fuel (natural gas in the present study).
4.4. Effects of the variation of fuel droplet size

RI
PT

Particle size is one of the most important physical characteristics for the injection of liquid fuels. This
property influences not only the combustion process, but also the total efficiency of fuel based energy
conversion devices like boilers and industrial furnaces.

SC

Many distribution functions have been proposed to predict particle diameter and characterize the fraction
of materials as a function of the particle size. The most famous distribution functions are the traditional log-

M
AN
U

normal, Rosin-Rammler (used in this study) and the more recently developed log-hyperbolic distribution
[43]. Since the focus in this paper is on the effects of the fuel droplet diameter on the flow and temperature
distribution within the furnace zone and the region close to the burners, the effects of droplet size on the
temperature field are only described here. More details about the effects of fuel injection diameter on other

TE
D

working characteristics of the boiler can be found in [44].

Considering Fig. 10, by increasing droplet mean diameter, temperature decreases and the maximum
temperature peak gradually approaches the rear wall. Needless to say, a rise in droplet diameter causes an

EP

increase in droplet mass, and consequently, the droplet injection velocity decreases. Therefore, as the droplet
diameter increases, droplet cant survive through long paths and evaporates closer to injection point and

AC
C

participates sooner in the combustion process. In addition, as the droplet mean diameter increases, the
maximum temperature of the boiler reduces (Fig. 11). Considering Fig. 10, it is evident that the distance that
the flame approaches the rear wall is very small, in comparison with the boiler depth, and this approaching is
also associated with a reduction in the temperature of the boiler, Fig. 11. As a result, the variation in droplet
size of heavy fuel oil couldnt be a cause of the failure.
4.5. Effects of the variation of the air-fuel ratio
The equation for ideal combustion of heavy fuel oil with atmospheric air is:
C19 H30 + 26.5(O 2 + 3.76N 2 ) 19CO 2 + 15H 2O + 99.64N 2

1442 443 144444442 44444443

258

26.5(32 +3.7628)

10

(9)

MANUSCRIPT
In this equation, just enough oxygenACCEPTED
is supplied to break
apart the fuel molecules (C19H30) and create CO2
and H20. Also, the nitrogen in the air does not react. This minimum amount of air required to complete the
combustion process is called the stoichiometric air.
In addition, assuming that natural gas is made up of 100% methane, we have:
CH 4 + 2(O2 + 3.76N 2 ) CO2 + 2 H 2O + 7.52 N 2

(10)

1444442 444443
2(32+3.7628)

RI
PT

16

The ideal air-fuel ratio in the combustion of heavy fuel oil and natural gas can be calculated from Eqs. (9)
and (10), respectively. Moreover, using data from Table 4, mass flow rate and air-fuel ratio for all burners in

SC

the current operating mode of the boiler for both fuels is acquired, Table 6.

Considering the results shown in Table 6, it is evident that the amount of excess air in combustion of

M
AN
U

heavy fuel oil is less than that in combustion of natural gas. It was also observed that the combustion
efficiency of the boiler for fuel oil is less than that for natural gas. Since the boiler is designed for operation
with natural gas as the primary fuel, these observations seem reasonable. Given that the amount of excess air
in both cases is less than 15%, reduce the combustion air greater than this amount leads to incomplete
combustion and loss of combustion efficiency, consequently.

TE
D

According to Fig. 12, the current value of the air-fuel ratio for natural gas is nearly at its optimum value.
However, by increasing air-fuel ratio (by changing the amount of secondary air) maximum boiler
temperature initially rises and then decreases. The highest temperature difference due to the variation of the

EP

air-fuel ratio for both fuels is less than 110 K, which is a relatively small temperature discrepancy that could
not cause overheating damages. Furthermore, as seen in Fig. 12, by increasing excess air, the flame moves

AC
C

away from the rear wall that can counteract the effects of the temperature rise.
According to these observations, although changing the air-fuel ratio has dramatic effects on combustion
efficiency and other working parameters of the boiler, it can be concluded that the variation of the air-fuel
ratio does not cause a significant change in the temperature of the rear wall, and it cannot cause the damage.

5. Conclusions
In this paper, a CFD model was developed for a full-scale front-wall-fired 150 MW utility boiler of
Mashhad Touss power station. The model was used to simulate both natural gas and heavy fuel oil

11

combustion, focusing on the furnaceACCEPTED

zone. Based onMANUSCRIPT
the simulation results and the plant data, we have the
following conclusions:
The CFD modeling is an applicable tool to investigate the causes of damages in boilers, in which the
experimental measurements is impossible or expensive.
Comparison between the simulated results and measurements showed that the CFD model can predict

RI
PT

combustion, heat and mass transfer in the boiler precisely for both natural gas and heavy fuel oil.
The main cause of the investigated failure is the use of natural gas. In this situation the flame establishes
closer to the burners and it causes higher absorbed heat flux, higher average temperature and

SC

consequently, overheating damages for the rear wall and burners.

Although the highest temperature of the boiler for natural gas is more than 100 K higher than that of

boiler is less than that of natural gas.

M
AN
U

heavy fuel oil, the temperature field is more uniform for heavy fuel oil, and the temperature gradient in the

The temperature and velocity fields within the boiler are highly dependent on the swirl ratio. By
increasing the swirl ratio the flame approaches the burners and the average temperature of the rear wall

TE
D

increases. Therefore, this can be another cause of damage to the burners. It was also observed that the
boiler sensitivity to the variation of the swirl ratio for heavy fuel oil is higher than that of natural gas.
The optimal value of the swirl ratio of the burners was found 0.55 for natural gas and 0.65 for heavy fuel

EP

oil. The current value is 0.7 for both fuels.

By increasing the mean diameter of the droplets of heavy fuel oil injected from the nozzles, the

AC
C

temperature of the boiler decreases and the flame approaches the rear wall. However, in comparison with
the boiler depth, the distance that the flame approaches the rear wall is not significant, and this
approaching is also coupled with the decrease in the temperature of the boiler. Therefore, the variation of
fuel droplet size cannot be a cause of the failure.
For both natural gas and heavy fuel oil, as the air-fuel ratio increases (by increasing the secondary air), the
temperature of the boiler initially rises and then decreases. However, increasing the mass flow rate of the
secondary air leads the flame to get distance from the burners.

12

ACCEPTED MANUSCRIPT

Acknowledgements

The authors are grateful to the management and staff at Touss Power Generation Management Company
(TPGMC) for their great technical support and cooperation in this work.

References

wall- fired boiler, Applied Thermal Engineering, 42 (2012) 90-100.

RI
PT

[1] M.B. Gandhi, R. Vuthaluru, H.B. Vuthaluru, D. French, K. Shah, CFD based prediction of erosion rate in large scale

[2] J.B. Liierup, K. Dam-Johansen, P. Glarborg, Characterization of a full-scale, single-burner pulverizedcoal boiler:
temperatures, gas concentrations and nitrogen oxides, Fuel, 73 (1994) 492-499.

SC

[3] P.M. Walsh, J.Y. Xie, R.E. Douglas, J.J. Battista, E.A. Zawadzki, Unburned carbon loss from pulverized coal
combustors, Fuel, 73 (1994) 1074-1081.

performance, Fuel, 77 (1998) 1377-1328.

M
AN
U

[4] R.P. Van der Lans, P. Glarborg, K. Dam-Johnson, G. Hesselmann, P. Hepburn, Influence of coal quality on combustion

[5] H.C. Zhou, C. Lou, Q. Cheng, Z. Jiang, J. He, B. Huang, Z. Pei, C. Lu, Experimental investigations on visualization of
three-dimensional temperature distributions in a large-scale pulverized-coal-fired boiler furnace, Proceedings of the
Combustion Institute, 30 (2005) 1699-1706.

TE
D

[6] V. Tanetsakunvatana, V.I. Kuprianov, Experimental study on effects of operating conditions and fuel quality on thermal
efficiency and emission performance of a 300-MW boiler unit firing Thai lignite, Fuel Processing Technology, 88
(2007) 199-206.

[7] D. Yang, J. Pan, C.Q. Zhou, X. Zhu, Q. Bi, T. Chen, Experimental investigation on heat transfer and frictional

(2011) 291-300.

EP

characteristics of vertical upward rifled tube in supercritical CFB boiler, Experimental Thermal and Fluid Science, 35

AC
C

[8] P. Edge, M. Gharebaghi, R. Irons, R. Porter, R.T.J. Porter, M. Pourkashanian, e. al, Combustion modelling opportunities
and challenges for oxy-coal carbon capture technology, Chem. Eng. Res. Des., 89 (2011) 1470-1493.
[9] L. Chen, S.Z. Yong, A.F. Ghoniem, Oxy-fuel combustion of pulverized coal: characterization, fundamentals,
stabilization and CFD modeling, Prog. Energy Combust. Sci., 38 (2012) 156-214.
[10] H.Y. Park, M. Faulkner, M.D. Turrell, P.J. Stopford, D.S. Kang, Coupled fluid dynamics and whole plant simulation of
coal combustion in a tangentially-fired boiler, Fuel, 89 (2010) 2001-2010.
[11] N. Modlinski, Computational modeling of a utility boiler tangentially-fired furnace retro fi tted with swirl burners, Fuel
Processing Technology, 91 (2010) 1601-1608.
[12] M.y. Hwang, S.m. Kim, G.b. Kim, B.h. Lee, J.h. Song, M.s. Park, C.h. Jeon, Simulation studies on direct ash recycling
and reburning technology in a tangentially fired 500 MW pulverized coal boiler, Fuel, 114 (2013) 78-87.

13

ACCEPTED
MANUSCRIPT
[13] E.H. Chui, H. Gao, A.J. Majeski, G.K.
Lee, Performance
improvement and reduction of emissions from coal-fired
utility boilers in China, Energy for Sustainable Development, 14 (2010) 206-212.
[14] S.R. Gubba, D.B. Ingham, K.J. Larsen, L. Ma, M. Pourkashanian, H.Z. Tan, A. Williams, H. Zhou, Numerical
modelling of the co-firing of pulverised coal and straw in a 300 MWe tangentially fired boiler, Fuel Processing
Technology, 104 (2012) 181-188.
[15] A. Leppnen, H. Tran, R. Taipale, E. Vlimki, A. Oksanen, Numerical modeling of fine particle and deposit formation

RI
PT

in a recovery boiler, Fuel, 129 (2014) 45-53.

[16] R. Ben-Mansour, M.A. Habib, H.I. Abualhamayel, Influence of Inlet Conditions and Burner Tripping on NO and
Temperature Characteristics in a Tangentially Fired Furnace, Environmental Application & Science, 4 (2009) 1-21.

SC

[17] E.H. Chui, H. Gao, Estimation of NOx emissions from coal-fired utility boilers, Fuel, 89 (2010) 2977-2984.

[18] M. Kuang, Z. Li, Review of gas/particle flow, coal combustion, and NOx emission characteristics within down-fired
boilers, Energy, 69 (2014) 144-178.

M
AN
U

[19] M. Rahimi, A. Khoshhal, S.M. Shariati, CFD modeling of a boilers tubes rupture, Applied Thermal Engineering, 26
(2006) 2192-2200.

[20] H.Y. Park, S.H. Baek, Y.J. Kim, T.H. Kim, D.S. Kang, D.W. Kim, Numerical and experimental investigations on the
gas temperature deviation in a large scale, advanced low NOx, tangentially fired pulverized coal boiler, Fuel, 104
(2013) 641-646.

TE
D

[21] H.B. Vuthaluru, R. Vuthaluru, Control of ash related problems in a large scale tangentially fired boiler using CFD
modelling, Applied Energy, 87 (2010) 1418-1426.

[22] Touss Power Generation Company website, Available from: www.tousspowerstation.ir, Last access date: January 28,

EP

2015.

[23] 500-MW Demonstration of Advanced Wall-Fired Combustion Techniques for the Reduction of Nitrogen Oxide (NOx)

[24]

AC
C

Emissions from Coal-Fired Boilers U. S. Department of Energy, National Energy Technology Laboratory, (2004).
Elektrnai

Power

Plant

web

page,

Available

from:

http://www.elektrenai.lt/

,Also,

http://en.wikipedia.org/wiki/Elektr%C4%97nai_Power_Plant, Last access date: January 28, 2015.

[25]

Eskoms

Majuba

Power

Plant

web

page,

Available

from:

http://www.eskom.co.za/Whatweredoing/ElectricityGeneration/PowerStations/Pages/Majuba_Power_Station.aspx, Last
access date: January 28, 2015.
[26] E. Bar-Ziv, R. Saveliev, E. Korytnyi, M. Perelman, B. Chudnovsky, A. Talanker, Evaluation of performance of AngloMafube bituminous South African coal in 550 MW opposite-wall and 575 MW tangential-fired utility boilers, Fuel
Processing Technology, 123 (2014) 92-106.

14

ACCEPTED
MANUSCRIPT
[27] Y. Hu, H. Li, J. Yan, Numerical investigation
of heat transfer
characteristics in utility boilers of oxy-coal combustion,
Applied Energy, (In Press) (2014).
[28] A.P. Mann, B. Moghtaderi, J.H. Kent, Computational Modelling of a Slag-reduction Strategy in a Wall-fired Furnace,
J. of the Inst. of Energy, 68 (1995) 193-198.
[29] M. Costa, J.L.T. Azevedo, M.G. Carvalho, Combustion Characteristics of a Front-Wall-Fired Pulverized-Coal 300
MW UtilitY Boiler, Combustion Science and Technology, 129 (1997) 277-293.

RI
PT

[30] S. Park, J.A. Kim, C. Ryu, T. Chae, W. Yang, Y.-J. Kim, H.-Y. Park, H.-C. Lim, Combustion and heat transfer
characteristics of oxy-coal combustion in a 100 MW front-wall-fired furnace, Fuel, 106 (2013) 718-729.
[31] P.K. Nag, Power Plant Engineering, 3rd ed., Tata McGraw-Hill, New Delhi, 2008, pp. 270-290.

SC

[32] in, ANSYS Inc. ANSYS Academic Research, USA, 2013.

[33] FLUENT users guide, in, FLUENT Incorporated, Lebanon, New Hampshire, USA, 2006.

[34] B.F. Magnussen, B.H. Hjertager, On mathematical models of turbulent combustion with sp ecial emphasis on soot

M
AN
U

formation and combustion, in: 16th Symp. (Int'l.) on Combustion, The Combustion Institute, 1976, pp. 719-729.
[35] I.R. Gran, M.C. Melaaen, B.F. Magnussen, Numerical simulation of local extinction effects in turbulent combustor
flow of methane and air, in: Proceedings of 25th Symposium on Combustion, Combustion Institute, 1994, pp. 12831291.

[36] C.K. Law, C.J. Sung, J.Y. Chen, An augmented reduced mechanism for methane oxidation with comprehensive global

TE
D

parametric validation, in: 27th Symposium (Int.) on Combustion, Vol. 62, The Combustion Institute, 1998, pp. 219235.

[37] W.P. Jones, R.P. Lindstedt, Global Reaction Schemes for Hydrocarbon Combustion Combustion and Flame, 73 (1988)

EP

233-249.

[38] A.P. Tesner, T.D. Snegiriova, V.G. Knorre, Kinetics of Dispersed Carbon Formation, Combustion and Flame, 17

AC
C

(1971) 253-260.

[39] I. Mirzaii, H. Sahabi, M. Pasandideh-fard, N. Shale, Simulation of liquid fuel atomization in an industrial spray nozzle
of a powerplant boiler, in: Proceedings of the ASME 2010 3rd Joint US-European Fluids Engineering Summer
Meeting and 8th International Conference on Nanochannels, Microchannels, and Minichannels, Montreal, Canada,
2010.

[40] ANSYS FLUENT. ANSYS FLUENT Theory Guide, in, ANSYS Inc, USA, 2011.
[41] V. Nayagam, A. Williams, Pattern Formation in Diffusion Flames Embedded in Von Karman Swirling Flows,
NASA/CP 2001-21082, (2001).
[42] V. Nayagam, A. Williams, Pattern Formation in Diffusion Flames Embedded in Von Karman Swirling Flows,
NASA/CR 2006-214057, (2006).

15

MANUSCRIPT
[43] E. Movahednejad, F. Ommi, S.M. ACCEPTED
Hosseinalipour, Prediction
of droplet size and velocity distribution in droplet
formation region of liquid spray, Entropy, 12 (2010) 1484-1498.
[44] M. Pourramezan, M. Kahrom, Numerical Investigation of the Effects of Fuel Injection Diameter on the Main Working
Characteristics of a Power Plant Boiler, in: The 8th International Chemical Engineering Congress & Exhibition, Kish,

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

Iran, 2014.

16

ACCEPTED MANUSCRIPT

Tables:
Table 1. Composition and analysis of the Fuels.

Table 3. Parameters of the heat exchangers as porous media.

Table 4. Operating conditions for CFD simulation.

Figures:

M
AN
U

Table 6. Comparison between the real and identical air-fuel ratio.

SC

Table 5. Comparison between the predicted and measured results.

RI
PT

Table 2. Average temperatures of all sections of the water-wall.

Fig. 1. a) Boiler geometry; b) Computational grid.

Fig. 2. a) Burner; b) Longitudinal cross section and the 3D CAD model of the heavy fuel oil nozzle.

TE
D

Fig. 3. Location of the thermocouples.

Fig. 4. Temperature distribution for both fuels.

Fig. 5. Distribution of O2 mass fraction for both fuels.

EP

Fig. 6. Temperature variation along the axis of the central burner (burner 25).
Fig. 7. Heat flux distribution on the rear wall.

AC
C

Fig. 8. Temperature and velocity distributions (for heavy fuel oil).

Fig. 9. Average temperature at vertical planes close to the rear and front walls (at x=0.25 and x=9.02 m,
respectively) along with the combustion efficiency of the boiler.
Fig. 10. Temperature variation along the axis of the central burner (burner 25).
Fig. 11. The maximum temperature of the boiler.
Fig. 12. Maximum temperature of the boiler along with Location of this maximum temperature.

17

Composition and analysis of the Fuels.

Heavy fuel oil
Proximate analysis (wt.%)
Volatile

0.5
96
Ultimate analysis (wt.%)
C
H
84.6
11.1
Composition (wt. %)
H2O Ash Res. carbon
0.6
0.1
0.14

Moisture

Fixed carbon

2.5

N
0.3

S
2.7

Sediments Asphal. Sulfur

0.16
11
11

Natural gas
Proximate analysis (wt.%)
Volatile

Moisture

--100
Ultimate analysis (wt.%)
C
H
74
25

--N
0.75

Composition (vol. %)
CO2
0.28

---

Heating
value
(kJ/g)
53

S
---

O
0.25

Fixed carbon

Nitrogen Heavier hydrocarbons

1.27
0.42

EP

Propane
0.48

Vanad.
0.08

AC
C

Ethane
0.6

O
1.3

TE
D

ash

M
AN
U

ash

Heating
value
(kJ/g)
41

SC

Table 1

RI
PT

ACCEPTED MANUSCRIPT

18

ACCEPTED MANUSCRIPT

Average temperatures of all sections of the water-wall.

Average velocity Average temperature

Wall
Absorption

Zone

Wall

of combustion

of combustion

temperature
Coefficient

Eco.

Front

10

1550

0.6

Rear

12

1450

0.65

Right

12

1650

Left

13

1550

Bottom

7.2

1050

Front

18

1550

Rear

17

1450

Right

19

Left

19

Roof

16

Front

SC

(K)

632.2

630.0

M
AN
U

products (K)

636.0

0.6

633.4

1.0

620.5

0.6

630.8

0.65

625.4

1600

0.55

633.1

1550

0.6

631.7

1400

0.8

612.3

20

1350

580.4

TE
D

0.5

0.4

Right

EP

Superheater

products (m/s)

21

1300

0.3

583.2

Left

21

1300

0.2

583.1

Roof

16

1350

0.3

612.3

Bottom

27

1700

0.6

638.3

AC
C

Furnace

RI
PT

Table 2

19

RI
PT

ACCEPTED MANUSCRIPT

Table 3

Heat exchanger

SC

Parameters of the heat exchangers as porous media.

13CrMo44

Heat
absorption
(MW)
62

inertial
Solid
resistance

Porosity

material

0.19

0.21

Superheater 3 + Reheater 2

0.55

0.23

4.43

15Mo3

51

Superheater 1

0.31

0.22

2.98

15Mo3

37

Superheater 2

0.38

0.22

3.65

15Mo3

31

Reheater 1

0.38

0.25

1.96

15Mo3

17

Economizer 2

0.26

0.27

1.8

15Mo3

11

Economizer 1

0.26

0.27

1.8

15Mo3

14

Supporting tubes 2nd pass

0.08

0.29

1.6

15Mo3

18

AC
C

EP

TE
D

M
AN
U

Supporting tubes 1st pass

(1/m)
5.75

20

RI
PT

ACCEPTED MANUSCRIPT

Table 4

Primary air inlet temperature (K)

561

Secondary air mass flow rate (kg/s)

140

Secondary air inlet temperature (K)

561

Atomizing steam temperature (K)

493

Atomizing steam pressure (atm)

12

G. R. Fan mass flow rate (kg/s)

23

Heavy fuel oil mass flow rate (ton/h)

Heavy fuel oil inlet temperature (K)

M
AN
U

12

TE
D

Primary air mass flow rate (kg/s)

SC

Operating conditions for CFD simulation.

37.01
398

935

Natural gas volumetric flow rate (m3/h)

42422

Inlet temperature (K)

300

EP

Heavy fuel oil density at 398 K (kg/m3)

0.656

Boiler inside pressure (mm Hg)

370

AC
C

Natural gas density at 398 K (kg/m3)

21

Table 5
Comparison between the predicted and measured results.
Simulation

CO2 M. fraction at the outlet (%)

N.G. F.O.
15.41 21.1

N.G.
15.9

F.O.
22.9

N.G.
3.1

F.O.
7.8

H2O M. fraction at the outlet (%)

11.02 7.10

10.3

6.4

6.9

9.8

SO2 M. fraction at the outlet (%)

----

0.98

----

1.03

----

4.8

NO Con. at the outlet (ppm)

305

368

268

326

13.8

12.8

Max. temp. of the boiler (K)

2146

2039

----

----

----

----

Vol. ave. temp. of the boiler (K)

1392

1383

----

----

----

----

Total heat flux (MW)

394

421

370

393

6.4

7.1

M
AN
U

785

729

753

701

4.2

3.9

1395

1468

1353

1410

3.1

4.1

89

84

85

79

4.7

6.3

AC
C

EP

Combustion efficiency (%)

TE
D

Ave. temp. at height of 24.7 m

(plane close to sup. tubes) (K)

Error (%)

SC

Parameter

Ave. temperature at outlet (K)

Measure.

RI
PT

ACCEPTED MANUSCRIPT

22

Table 6

14.1

Current air-fuel ratio (for heavy fuel oil)

14.78

Percentage of excess air (for heavy fuel oil) (%)

4.8

Ideal air-fuel ratio (for natural gas)

17.16

Current air-fuel ratio (for natural gas)

19.66

Percentage of excess air (for natural gas) (%)

14.6

AC
C

EP

TE
D

M
AN
U

Ideal air-fuel ratio (for heavy fuel oil)

SC

Comparison between the real and identical air-fuel ratio.

RI
PT

ACCEPTED MANUSCRIPT

23

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

Fig. 1. a) Boiler geometry; b) Computational grid.

24

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

Fig. 2. a) Burner; b) Longitudinal cross section and the 3D CAD model of the heavy fuel oil nozzle.

25

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

Fig. 3. Location of the thermocouples.

26

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

Fig. 4. Temperature distribution for both fuels.

27

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

Fig. 5. Distribution of O2 mass fraction for both fuels.

28

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

Fig. 6. Temperature variation along the axis of the central burner (burner 25).

29

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

Fig. 7. Heat flux distribution on the rear wall.

30

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

Fig. 8. Temperature and velocity distributions (for heavy fuel oil).

31

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

Fig. 9. Average temperature at vertical planes close to the rear and front walls (at x=0.25 and x=9.02 m,

AC
C

respectively) along with the combustion efficiency of the boiler.

32

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

TE
D

Fig. 10. Temperature variation along the axis of the central burner (burner 25).

33

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

EP

TE
D

Fig. 11. The maximum temperature of the boiler.

34

EP

TE
D

M
AN
U

SC

RI
PT

ACCEPTED MANUSCRIPT

AC
C

Fig. 12. Maximum temperature of the boiler along with Location of this maximum temperature.

35

ACCEPTED MANUSCRIPT

RI
PT

Mehdi Pourramezan, Ph.D. student

Department of Mechanical Engineering,
Faculty of Engineering, Ferdowsi
University of Mashhad, Mashhad, Iran
February 14, 2015

SC

Dear Editor,

We have revised the highlights and provide them as following:

M
AN
U

A dual-fuel front-wall-fired utility boiler is modelled by CFD.

There is a good agreement between the simulated and experimental results.

Overheating damages occur when the flame establishes closer to the rear wall.
The main causes of the investigated failure are the swirl ratio and natural gas.

TE
D

Other investigated parameters have no significant effect on the failure.

EP

Sincerely,

AC
C

Mehdi Pourramezan