Numerical Study on Layout of Refractory Belt and Fouling ...

$Page 1: Numerical Study on Layout of Refractory Belt and Fouling ...$
Numerical Study on Layout of Refractory Belt and Fouling Deposition in Tangentially Fired Boiler

Qiongliang Zha a, Kai Chen a, Jianwen Zhang b, Jiangtao Li b, Chang’an Wang a, Defu Che a*

a State Key Laboratory of Multiphase Flow in Power Engineering, School of Energy and Power Engineering, Xi’an Jiaotong University, Xi’an 710049, P.R.China

b Shanghai Boiler Works Co., Ltd., 250 Huaning Road, Minhang, Shanghai 200245, P.R.China * Tel: +86-029-82665185, Fax: +86-029-82668703

Email: [email protected] ABSTRACT

The refractory belt installed in primary combustion zone provides simplest and most effective solution to suppress ignition delay and enhance combustion stability for low volatile anthracite and lean coal. The fouling deposition generally formed on radiative refractory lined wall of the boiler due to a high surface temperature. The growth of deposition thickness is mainly dependent on the parcile impact on the surface of water wall. A particle capture submodel was used to determine whether a particle was captured to form deposition or not when it reached the furnace wall, and the particle capture criterion was based on the particle’s viscosity and the temperature of the furnace wall. A reduced fouling deposition model was implemented in a three dimensional simulation of a tangentially fired boiler. The numerical investigation was conducted to assess the performance of different layouts of refractory belt. Furnace temperature, surface temperature of refractory belt, and deposition distributions on the furnace wall should be taken into account when layouts of refractory belt are optimized. Based on

this, three layouts of refractory belt were proposed for tangentially fired boilers. A numerical investigation was conducted to assess the performance of different layouts of refractory belt and the results showed that the temperature in furnace was increased, and the ignition and combustion processes were stabilized when refractory belts were installed. The reasonable arrangement of refractory belt could reduce the possibility of fouling deposition in furnace. Key words: Layout of refractory belt; simulation; fouling deposition; tangentially fired boiler. 1. INTRODUCTION

In China, the reserves and consumption of anthracite and lean coal are the largest in the world and about 30% of the electricity is generated by burning these coals [1, 2]. Anthracite and lean coal, characterized by low volatile matter and poor reactive activity, present difficulties in achieving ignition, maintaining stable combustion, and completing burnout when industrially fired in furnaces. The refractory belts installed in primary

Proceedings of the ASME 2016 Power Conference POWER2016

June 26-30, 2016, Charlotte, North Carolina

POWER2016-59570

1 Copyright © 2016 by ASME

combustion zone provide simplest and most effective solution to improve ignition and burnout efficiency of faulty coal. The refractory belt can reduce heat absorption capacity of the water wall to improve the temperature in furnace, enhance ignition and combustion stability, and increase combustion efficiency.

The ash deposition generally forms at high temperature on radiant refractory lined wall of the boiler. The progression of slagging is dependent both on the surface temperatures of refractory belt and the impact of pulverized coal flow [3, 4]. For instance, the flue gas and coal particle trajectory for a tangentially fired boiler is shown in Figure.1. Relative low pressure regions adjacent to the burners are formed as the high-speed pulverized coal stream is injected into the furnace. The transverse push by the adjacent upstream flame will force the jet to deflect to the back side of the burner. In addition, the upstream flame is inclined to be involved to the face side of the burner by the high speed jet. These two effects results in the flame to impinge or stick to the entire wall. For instance, the pulverized coal in the upstream (injected from “B”) would be easily involved into the fire-facing side of “C”; this would cause more ash deposit and slag formation in the region of “CE”. On the other hand, the pulverized coal injected from burner “C” would be involved into the back side of the burner, i.e., CF region.

Combustion test in the pulverized coal fired utility furnace is a costly and arduous task, since a large amount of sophisticated instruments are needed for measurements. And only a small quantity of deficient data sets with a large measurement uncertainty was gathered from limited combustion tests. Compared to experiments, numerical simulations are more flexible and cost-effective to generate detailed data sets in wide ranges of conditions. Therefore, the numerical method is more convenient and suitable to analyze the combustion, pollutant formation process, and slagging in a utility furnace. In the past few decades, the numerical method was widely applied to investigate the complex phenomena in pulverized coal fired utility boiler, such as turbulent flow, combustion, heat transfer, and NOx emission. But, limited attempts are taken to investigate the ash deposit and slag formation issue. It’s very important to know the amount and location of slag as well as its dynamics. It’s not possible to investigate the problems on site

unobtrusively and with reasonable effort. In recent years, a number of attempts have been made to utilize computational fluid dynamics (CFD) methods to model the ash formation process [5-8]. Richards et al. [9] defined a sticky ash particle as a particle with a viscosity that is lower than the ash melting viscosity. Thus, the melting temperature of the fly ash particles determines whether the condition for stickiness is satisfied. Li et al. [10] have found that there is a critical coal conversion of about 88% at which the molten ash is exposed to the surface, which makes a particle sticky. Yong et al. [11, 12] further proposed that the capture efficiency is also dependent on the particle’s Weber number, besides either or both of the particle and wall being sticky. All these studies have proven computational modeling is a powerful and feasible tool to investigate the performance of slagging for better understanding of this process.

Figure 1. Flue gas and coal particle trajectory in the burner block zone

In the present study, a reduced fouling deposition model is

implemented in a three-dimensional CFD simulation of the tangentially fired boiler in the form of user defined functions (UDFs). The numerical investigation is conducted to assess the performance of different layouts of refractory belt. The flame temperature, surface temperature of refractory belt, and slag behavior are examined.


2. THE TANGENTIALLY FIRED PULVERIZED-COAL BOILER

The boiler investigated in this study is a 330 MW tangentially fired pulverized-coal boiler. The height of the furnace is 55.158 m, and the horizontal cross section of the furnace is rectangular with a width of 14.140 m and a depth of 13.230 m. The single angle burner arrangement of the tangentially fired boiler is shown in Figure.2 (c). A total of 24 fuel rich/lean burners (A, B, C, D, E, F) provide the primary air/fuel mixture, and 24 side-secondary air burners are adopted in order to resist slagging on the water cool tubes. A total of 28 air burners (AA, BC, CD, DE, EF, FG1, FG2) provide the secondary air, while 4 over fired air (OFA) ports for air-staged combustion. In addition, a total of 8 fuel/air burners (G, H) provide the tertiary air /fuel mixture. The simulation focuses on boiler’s behavior under full load. The coal feeding rate is 148.8 ton/h. The total flux of primary air is 65.55 kg/s while the gas temperature is 210 oC. The total flux of secondary air is 259.00 kg/s while the gas temperature is 372.8 oC. The total flux of tertiary air is 72.70 kg/s while the gas temperature is 80 oC. The total flux of OFA is 27.20 kg/s occupying 6.85% of the total air amount.

Figure 2. Mesh system and burner arrangement of the tangentially fired

boiler

A partition meshing method was applied to generate high

quality mesh with hexahedral cells, and the mesh is refined in the burner regions for more accurate prediction. A grid independent is conducted on 3 different meshes consisting of 564880, 779864, and 938240 cells respectively. The predicted values of gas phase temperature distributions along the

centerline of the furnace were used as criteria. The partition with 779864 cells is finally adopted because of its similar prediction result with the finer mesh partition and a smaller computational cost. 3. COMPUTATIONAL MODELING APPROACH AND

OPERATING CONDITIONS The simulations were carried out using a commercial

computational fluid dynamics code Fluent. The mathematical model is based on an Eulerian description for the continuum phase while a stochastic Lagrangian description for the coal particles. Reasonable sub models were chosen to simulate turbulence, thermal radiation, gas phase combustion, char combustion in the boiler. The standard k-ε model was chosen in the present study to simulate turbulence because it had proven to be the most suitable model in terms of computational economy, stability, and reliability of results [13, 14]. Radiation was the dominant form of heat transfer in the furnace, and here the discrete ordinates (DO) model was chosen because of the high accuracy and small optical length of the panels [15]. The weighted sum of gray gases model (WSGGM) was applied to calculate the gas absorption coefficient, while the particle emissivity was assumed to be 0.9 [16, 17]. For a homogeneous reaction, fast chemistry, mixture fraction (MF), and probability density function (PDF) was chosen because of its high calculation efficiency, acceptable temperature, and species distribution prediction accuracy in boiler simulation [18]. The rate of volatile release was calculated using the single kinetic rate model, which was the most widely used empirical model [19, 20]. The process of char combustion sub-model was described by the kinetics/diffusion limited model of Field [14].

A particle capture submodel was used to determine whether a particle is captured to form deposition or not when it reached the wall. In the present study, the particle capture criterion (see Table 1) was based on the particle’s viscosity and the temperature of the wall. The wall was sticky when the particle or the wall temperature was above the ash temperature of criticla viscosity, whereas the particle was sticky when both the prticle temperature was above the ash temperature of critical viscosity. When both the particle and wall were sticky, the particle was always trapped. Conversely, when dealing with the


interaction of non-sticky particles and a non-sticky wall, it was assumed that all particles were rebounded. For the case in which the particle was sticky and the wall was non-sticky, or when the particle was non-sticky and the wall was sticky, the capture tendency depended on the Weber number of the colliding particle, which compared the kinetic energy of the particles and the interfacial surface tension energy between the particles and the slag surface. A particle was rebounded when its Weber number exceeded a critical value. This critical value was set to 0.1 [11]. These criteria were summarized in Table 1. Therefore, this particle capture sub-model provided a new alternative feature to the ‘‘trap’’ or ‘‘escape’’ options in the discrete particle phase boundary condition.

In order to simulate the layout of refractory belt conveniently, the furnace wall was divided into many small pieces. The every selected piece would be assigned given boundary when it was specified as refractory belt. In the present study, feasible layouts of refractory belt were obtained through simulation tests of six different arrangements. Six different layouts (i.e. case 2-7) are shown in Figure.3, and the shaded regions are assigned as refractory belt; to simplify, only front wall was shown in the figure. Meanwhile, case 1 is the base operation condition, in which no refractory belt is arranged.

Figure 3. Layouts of refractory belt in the primary combustion zone

The fouling deposition model was implemented in a three-dimensional CFD simulation of the tangentially fired boiler in

the form of UDFs. First, a converged steady state solution of coal combustion was reached; then the fouling deposition model was loaded and implemented into the CFD calculation, and the problem was calculated using the transient solver. In a discrete particle model (DPM) iteration step, the particle trajectories were calculated, and deposition rates and deposition thickness were saved in the user defined memories (UDMs) associated to the local volumetric cells.

Table 1 Particle capture criterion Sticky particle Non-sticky particle We<Wecr We>Wecr We<Wecr We>Wecr

sticky wall Fouling Fouling Fouling Reflect non-sticky wall Fouling Reflect Reflect Reflect

4. RESULTS AND DISCUSSION 4.1. Validation of the combustion simulation results

In order to obtain creditable and reasonable simulation results, the numerical calculation results were compared with the design/measured values at the furnace exit and boiler exit as well as in major cross sections, as shown in Table 2. The comparison between the measured values and predicted results shows good agreement. The predicted total heat flux to the furnace wall was 557 MW, while the design value was 555 MW. The predicted temperature at the rear plate SH exit was 1250 K, while the design value was 1244 K. The predicted temperature at the front plate SH exit was 1287 K, while the design value was 1270 K. The predicted concentration of O2 at the boiler exit was 4.01%, while the design value was 3.93%. Features in flow patterns and temperature distributions as shown in Figure.4 and Figure.5, have been qualitatively consistent with the results in the previous studies. Thus, the adopted numerical models in the present calculations were reasonable for combustion analysis in the boiler. Table 2. Comparison between the design/measured values and the

predicted results. Section name Design value Predicted

Heat flux [MW] Furnace wall 557 555

Temperature [K] Rear plate exit 1244 1250 Front plate exit 1270 1287

O2 (%, mole fraction) Boiler exit 3.93 4.01

In tangentially fired boilers, the jets coming out of the nozzles

of the burners located in the four corners are directed tangentially to an imaginary circle. The velocity distributions in


the cross sections along the furnace height are shown in Figure.4. An anticlockwise swirling flow formed via the air and coal particles injected through the burner ports was found in the center of the furnace. Deflection of the gas jet from the expected direction, i.e., the deflection from the geometrical axis of the nozzle, was found in Figure.4. The transverse push by the spinning gas would force the jet from the burner to deflect. A severe deflection would lead to the flame impingement on the water wall, and deposition on the water wall tubes. As shown in Figure.4, side-secondary air in the primary burners (i.e., A, C, E in Figure.4) would effectively alleviate the flame impingement on the water walls.

Figure 4. Velocity distribution in different cross sections

4.2. The effects of installing height of refractory belt on the furnace temperature and fouling deposition

Figure.5 shows the temperature distributions in the vertical cross section for case 1, case 2, case 3, and case 4. It can be seen that the temperature increased along the furnace height in the burner zone. Figure.6 shows the predicted mass-weighted average temperature along the furnace height under different installation height of refractory belt (i.e., case 2, case 3, and case 4). The temperature decreases obviously since a large amount of secondary air (FG-1, FG-2) with low temperature is injected into the furnace. Then, the temperature increases which could be attributed to the coal combustion when tertiary air (a mixture of coal and air, i.e., G and H) is injected into the furnace. At the OFA nozzles layer, the temperature decreases obviously due to the injection of OFA with relative low temperature. Above the

OFA nozzles, a temperature rise is found because of the combustion of residual CO and char. When the height further increases, the temperature gradually decreases due to the heat transfer from the flue gas to the heating surface.

Figure 5. Temperature distributions

Figure.6 (b) shows the rise of temperature along the furnace

height under different installation height of refractory belt. For case 2, a maximum rise of furnace temperature reaches 54 K at the end of the refractory belt, which corresponds to the height of lower primary air nozzles. And the lowest rise of furnace temperature is about 25 K within all primary air regions (i.e., A-F) for case 2. For case 3, a maximum rise of furnace temperature is about 48 K, which locates at the end of primary air region (i.e., F). In other words, only the later layers of primary air are benefit from the combustion stabilizing techniques. For case 4, hardly any rise of temperature is found in the primary air regions, and the stability of coal ignition is not effectively raised. Therefore, in order to raise the stability of both ignition and burning of air and pc mixture, the refractory belt should be installed at the height of lower primary air nozzles (i.e., A, B, C).


The ash deposition generally forms at high temperature radiant refractory lined wall of the boiler. Figure.7 shows the wall temperature distributions under different installation height of refractory belt; and for simplicity, only rear wall and right wall are analyzed in the present study. It is clearly shown that a remarkable high temperature zone is found on the surface of refractory lined wall. And the temperature in case 3 is higher than that of the other two cases, which can be attributed to the higher furnace temperature in the installing height of case 2. In

addition, the high temperature zone of refractory belt is partial to the back side to the burner when case 2 and case 3 are adopted. High temperature flue gas is easily involved by the side-secondary air with a high speed, which results in a higher radiation heat transfer to the back side than that for the fire side to the burner. A uniform wall temperature distribution for case 4 is shown in Figure.7 when refractory belt is installed above all primary air nozzles.

Figure 6. Average temperature profiles and temperature elevation along the furnace height under different installation height of refractory belt

The progression of ash deposition is dependent both on the

surface temperatures of furnace wall and the impact of pulverized coal jet. In tangentially fired boilers, the transverse push by the adjacent upstream flame will force the jet to deflect to the back side of the burner. In addition, the upstream flame is inclined to be involved to the face side of the burner by the high speed jet. These two effects results in the flame to impinge or stick to the entire wall, especially in the primary combustion zone. Figure.8 shows the particle impact distributions on the wall, and the furnace wall is almost impinged by the particles. Consequently, side-secondary air burners are often adopted in order to resist slagging on the water cool tubes in tangentially fired boilers. From the simulation results, particle impact distribution on the lower furnace wall is obviously less than that on the upper furnace wall, and this is mainly due to none side-secondary air in the tertiary burner (i.e., G and H).

Figure 7. Wall temperature distributions


Figure 8. Particle impact distributions on the rear wall and right wall

Figure.9 shows the deposition distributions on the rear wall

and right wall under different installation height of refractory belt. It can be concluded that the deposition distributions on the wall is different at different points. Deposition rate is proportional to the surface temperature of furnace wall, particle temperature, and the impact of pulverized coal jet. As can be seen from the simulation results, deposition is mainly distributed on the surface of refractory belt and on the water-cooled wall surface in the upper furnace. On the one side, a high surface temperature of refractory promotes the deposition rate. On the other side, particle temperature in the upper furnace is much higher due to a longer combustion process, and more particles in the tertiary air are impinged on the upper furnace wall.

Figure 9. Deposition distributions on the rear wall and right wall under

different installation height of refractory belt

4.3. The optimal layouts of refractory belt in the primary combustion zone

As can be seen from Figure.6 to Figure.9, furnace temperature, surface temperature of refractory belt, and deposition distributions on the furnace wall should be taken into account when a layout of refractory belt is optimized. As a consequence, three optimal layouts of refractory belt in the primary combustion zone are proposed for tangentially fired boilers (i.e., case 5, case 6, and case 7). For case 5, refractory belts are arranged on the two sides of the burner with trapezoid. For case 6, refractory belts are arranged on the back side to the burner. For case 7, the refractory belts are arranged on the fire side to the burner.

Figure.10 shows average temperature profiles and temperature elevation along the furnace height under three different optimal layouts of refractory belt in the primary combustion zone with case 1 as a reference condition, and a similar temperature profile is found for the three optimal cases. A higher rise of furnace temperature ranged from 45 K to 54 K is detected at the location of the refractory belt, which corresponds to the level of lower primary air nozzles (i.e., A, B, C). And the lowest rise of furnace temperature is about 25 K within all primary air regions (i.e., A-F) for all three optimal cases.

Figure.11 shows the rear and right wall temperature distributions under three optimal layouts of refractory belt. Remarkable lower temperature is found on the refractory lined wall surface than that of case 2-4. Especially, a uniform and low wall temperature distribution is detected for case 7

Figure.12 shows deposition distributions on the rear wall and right wall under different optimal layouts of refractory belt. Deposition tendency is generally relieved, especially on the surface of refractory belt, when layouts of refractory belt are optimized. Deposition rate on the surface of refractory belt in case 6 is higher than that in case 5 and case 7, which can be attributed to higher surface temperature in case 6. Therefore, the refractory belts arranged on the fire side to the burner would be a better design for the present tangentially fired boiler. In addition, side-secondary air can be added to fire side of the tertiary air burner in order to resist deposition on the water cooled tubes on the upper furnace wall.


Figure 10. Average temperature profiles and temperature elevation along the furnace height under different optimal layouts of refractory belt in the

primary combustion zone

Figure 11. Rear and right wall temperature distributions under three optimal layouts of refractory belt 5. CONCLUSIONS

A numerical study on layout of refractory belt and fouling deposition in tangentially fired boiler was conducted in the present study. In order to deal with different layouts of refractory belt, a novel research approach was proposed. A reduced fouling deposition model was implemented in a three-dimensional CFD simulation of the tangentially fired boiler in the form of UDFs. The numerical investigation was conducted to assess the performance of different layouts of refractory belt. The refractory belt should be installed at the height of lower primary air nozzles in order to raise the stability of both ignition and burning of air and pc mixture. Three optimal

layouts of refractory belt in the primary combustion zone were proposed for tangentially fired boilers. Side-secondary air should be added to the fire side of the tertiary air burner in order to resist deposition on the water cooled tubes on the upper furnace wall.

Figure 12. Deposition distributions on the rear wall and right wall under different optimal layouts of refractory belt

NOMENCLATURE CFD computational fluid dynamics UDFS user defined functions UDMS user defined memories DPM discrete particle model MF mixture fraction PDF probability density function


WSGGM weighted sum of gray gases model DO discrete ordinates

ACKNOWLEDGMENTS The financial support from the National Natural Science

Foundation of China (51506163) is gratefully acknowledged. This work was also supported by the Fundamental Research Funds for the Central Universities (xjj2014041).

REFERENCES [1] Kuang, M., Li, Z., Ling, Z., and Zeng, X. 2014. "Evaluation of staged air and overfire air in regulating air-staging conditions within a large-scale down-fired furnace." Appl. Therm. Eng., 67(1–2). pp. 97-105. [2] Kuang, M., and Li, Z. 2014. "Review of gas/particle flow, coal combustion, and NOx emission characteristics within down-fired boilers." Energy, 69. pp. 144-178. [3] Bhuiyan, A. A., and Naser, J. 2015. "Modeling of Slagging in Industrial Furnace: A Comprehensive Review." Procedia Engineering, 105. pp. 512-519. [4] Fan, J. R., Zha, X. D., Sun, P., and Cen, K. F. 2001. "Simulation of ash deposit in a pulverized coal-fired boiler." Fuel, 80(5). pp. 645-654. [5] Chen, L., and Ghoniem, A. F. 2013. "Development of a three-dimensional computational slag flow model for coal combustion and gasification." Fuel, 113. pp. 357-366. [6] Yong, S. Z. 2010. "Multiphase Models of Slag Layer Built-up in Solid Fuel Gasification and Combustion." [7] Ma, Z., Iman, F., Lu, P., Sears, R., Kong, L., Rokanuzzaman, A. S., McCollor, D. P., and Benson, S. A. 2007. "A comprehensive slagging and fouling prediction tool for coal-fired boilers and its validation/application." Fuel Process. Technol., 88(11–12). pp. 1035-1043. [8] Wang, H., and Harb, J. N. 1997. "Modeling of ash deposition in large-scale combustion facilities burning pulverized coal." Prog. Energy Combust. Sci., 23(3). pp. 267-282. [9] Richards, G. H., Slater, P. N., and Harb, J. N. 1993. "Simulation of ash deposit growth in a pulverized coal-fired pilot scale reactor." Energy & Fuels, 7(6). pp. 774-781. [10] Li, S., Wu, Y., and Whitty, K. J. 2010. "Ash Deposition Behavior during Char−Slag Transition under Simulated Gasification Conditions." Energy & Fuels, 24(3). pp. 1868-1876. [11] Yong, S. Z., Gazzino, M., and Ghoniem, A. 2012. "Modeling the slag layer in solid fuel gasification and combustion – Formulation and sensitivity analysis." Fuel, 92(1). pp. 162-170. [12] Yong, S. Z., and Ghoniem, A. 2012. "Modeling the slag layer in solid fuel gasification and combustion – Two-way coupling with CFD." Fuel, 97(0). pp. 457-466. [13] Zhang, J., Chen, K., Wang, C. a., Xiao, K., Xu, X., and Che, D. 2013. "Optimization of separated overfire air system for a utility boiler from a 3-MW pilot-scale facility." Energy & Fuels, 27(2). pp. 1131-1140. [14] Liu, H., Liu, Y., Yi, G., Nie, L., and Che, D. 2013. "Effects of air staging conditions on the combustion and NOx emission characteristics in a 600 MW wall fired utility boiler using lean coal." Energy & Fuels, 27(10). pp. 5831-5840. [15] Alvarez, L., Gharebaghi, M., Jones, J. M., Pourkashanian, M., Williams, A., Riaza, J., Pevida, C., Pis, J. J., and Rubiera, F. 2013. "CFD modeling of oxy-coal combustion: Prediction of burnout, volatile and NO precursors release." Applied Energy, 104. pp. 653-665.

[16] Yin, C. 2015. "On gas and particle radiation in pulverized fuel combustion furnaces." Applied Energy, 157. pp. 554-561. [17] Yin, C., Johansen, L. C. R., Rosendahl, L. A., and Kær, S. K. 2010. "New Weighted Sum of Gray Gases Model Applicable to Computational Fluid Dynamics (CFD) Modeling of Oxy−Fuel Combustion: Derivation, Validation, and Implementation." Energy & Fuels, 24(12). pp. 6275-6282. [18] Sivathanu, Y. R., and Faeth, G. M. 1990. "Generalized state relationships for scalar properties in nonpremixed hydrocarbon/air flames." Combust. Flame, 82(2). pp. 211-230. [19] Zhang, X., Zhou, J., Sun, S., Sun, R., and Qin, M. 2015. "Numerical investigation of low NOx combustion strategies in tangentially-fired coal boilers." Fuel, 142. pp. 215-221. [20] Vascellari, M., Arora, R., Pollack, M., and Hasse, C. 2013. "Simulation of entrained flow gasification with advanced coal conversion submodels. Part 1: Pyrolysis." Fuel, 113. pp. 654-669.


Numerical Study on Layout of Refractory Belt and Fouling ...

Documents