Numerical simulation of combustion characteristics at different coal concentrations in bituminous coal ignition in a tiny-oil ignition burner

Abstract

With the objective of producing a full-scale tiny-oil ignition burner, identical to the burner used in an 800 MWe utility boiler, numerical simulations were performed using Fluent 6.3.26 to study the progress of ignition for four coal concentration settings covering sub-operation conditions prevailing during the experiments performed with the burner. The numerical simulations conformed to the experimental results, demonstrating the suitability of the model used in the calculations. Simulations for a coal concentration of 0.40 kg/kg corresponding to a single burner operating at its rated output were also conducted, which indicated that gas temperatures along the burner centerline were high. As gas flowed to the burner nozzle, the high-temperature region expanded, ensuring a successful pulverized-coal ignition. With increasing coal concentration (0.08–0.40 kg/kg), the gas temperature along the burner centerline and at the first and second combustion chamber exits decreased at the equivalent radial points. At the center of the second combustion chamber exit, the O₂ concentrations were almost depleted for the five coal concentrations, while the CO concentrations peaked.

Keywords

numerical simulation / tiny-oil ignition burner / pulverized coal / temperature field

Introduction

Because of the use of oil guns with high oil-flow rates, the traditional firing-up and partial-load operation of boilers consume large quantities of oil. At present, oil is consumed in great amounts in sectors such as transportation and chemical industry around the world, which threatens the world with the exhaustion of oil resources. As energy efficiency is closely related to different heating technologies, baselines have been determined for boiler plants and thermal power plants respectively [1]. Three technical ways have been studied for efficient use of thermal power [2]. In addition, an increase in the efficiency and control of pollutant emissions are the most significant utility boiler projects worldwide [3-7].

Tiny-oil ignition technology is a mature and widely used oil-saving technology. Against the background of limited energy resources, tiny-oil ignition centrally fuel-rich swirl burners have been proposed, and experiments have been conducted to determine their ignition characteristics [8,9]. In the traditional firing-up of a boiler, oil guns are arranged at the secondary air nozzles to pre-heat the combustion chamber of a furnace to its normal operating temperature. In tiny-oil ignition burners, two oil guns are arranged in the central primary air duct so as to use the heat from the oil combustion to directly ignite pulverized coal. Oil energy is used more efficiently by employing tiny-oil technology, and thus, oil guns with smaller delivery capacity suffice in the firing-up and partial-load operations of boilers. As a result, oil consumption is reduced.

In this paper, numerical simulations are performed to study the combustion characteristics in tiny-oil burners, in which the fuel is comprised of pulverized coal and oil. A probability density function (PDF) model of the two fractions was used to describe gaseous combustion. Numerical simulations were performed at coal concentrations of 0.08-0.21 kg/kg, corresponding to the condition prevailing in the experiment conducted earlier. Numerical simulations for a coal concentration of 0.40 kg/kg, corresponding to actual operating conditions, were also conducted to determine the progressive evolution of ignition. The temperature field and distributions of the O₂ and CO concentrations were also obtained from the numerical simulations. The results could provide reference and guidance for laboratory studies and engineering adaptations.

The full-scale tiny-oil burner is identical to the burner used in an 800 MWe utility boiler. Eight MBC-260 roll-type medium speed mills whose total rated output was 57 t/h were used in this utility boiler to provide pulverized coal for the six burners. Therefore, the rated output of a single burner was 9.5 t/h (the corresponding coal concentration is 0.40 kg/kg). In tiny-oil burners, the atomized oil from the main oil gun was ignited and burnt in an adiabatic chamber. Subsequently, the oil flame from the auxiliary oil gun ignited the atomized oil. Cone separators were installed in the primary air-coal mixture duct to concentrate the pulverized coal into the central zone of the burner. The fuel-rich primary air-coal mixture passed into the first combustion chamber whereby the fuel-rich primary air-coal mixture was ignited by a high-temperature oil flame formed by the main and auxiliary oil guns. Next, the burning pulverized coal and oil flame from the first combustion chamber were directed into the second combustion chamber where the coal was ignited. In bituminous coal ignition experiments performed with tiny-oil ignition burners, the pulverized coal could be ignited at coal concentrations of 2-5 t/h with a primary air temperature of 15°C, primary air velocity of 23 m/s, and total oil-flow rate of 100 kg/h. Along the centerline of the burner in the direction of primary air flow, the gas temperature gradually increased. Wall temperatures of the first and second combustion chambers were low. Under these low temperatures, the burner wall was safe during operations.

Gases were sampled using a water-cooled stainless steel probe and analyzed online with a Testo 350M instrument. The water-cooled stainless steel probe consisted primarily of a water-inlet pipe, a water-outlet pipe, a sampling tube, an outer pipe and supporting components. The high pressure cool water coming from the water-inlet pipe cooled the sampling tube, and after heat change, flowed out via the water-outlet pipe. The gas was sampled in the sampling tube. If the gas entered this tube, the temperatures deceased rapidly and the pulverized coal stopped burning. The samples are passed through the filtrating devices into a Testo 350 M gas analyzer for subsequent analysis. The O₂ concentrations at the exit of the burner were 0.01%-0.04% and the CO concentrations were more than 10000×10^-6. Further details are available in Ref. [8].

Numerical simulations of tiny-oil burners

Simulation objects

In the experimental setup to study tiny-oil ignition (Fig. 1), the pulverized coal was carried from the feeder to the tiny-oil ignition burner by the primary air from the blower. The oil was drawn from the oil tank and sent to the main and auxiliary guns which employed mechanical and air atomization. Administered as atomized air, compressed air entered the oil guns, providing O₂ for the initial combustion of oil. The main body of the air consumed in oil combustion was supplied by another blower.

In actual experiments, the gas temperatures inside the burner were measured with thermocouples, the ends of the bare thermocouples being exposed. Therefore, the temperature measured should be higher than the local gas temperature because of the high radiation from the flame. Meanwhile, because of the radiation between the bare thermocouple and the burner wall, which are in close proximity, the temperature measured should be lower than the local gas temperature. Taking these two effects into account, the measurement error should be approximately 8%.

Computational domain and mesh generation

Figure 2 shows the three-dimensional mesh generation of the whole burner based on the structural design of the burner.

Selection of model

At present, the numerical simulation of the combustion involving two fuels is restricted to coal blends [10,11]. The aim of this paper is to investigate the co-combustion of oil and coal based on methods pertaining to coal blends. The CFD modeling was applied to simulate a number of large-scale combustion facilities [12,13]. The gas turbulence was specifically taken into account by the so-called realizable k-ϵ model [14]. Lagrangian stochastic-tracking was applied to analyze the gas/particle flow field [15]. The radiation was described using the P-1 model [16], and the devolatilization was modeled using the two-competing-rate Kobayashi model [17]. For gaseous phase combustion, a two-fraction approach to calculate the mixture fractions and their variances, using transport equations, was employed [18]. The individual-species molar fraction, temperature, and density were derived from the predicted mixture fraction distribution. It was assumed that the volatile reaction is rapid enough to ensure chemical equilibrium and this assumption was taken with the PDF approach. At the injection plane of the mesh, two different types of particles were introduced. Coal was set as the primary stream and oil as the secondary stream. This method allows each fuel in the binary mixture to be described separately with different properties and initial conditions. The char combustion was described by a diffusion/kinetics model [19].

Operating parameters

Table 1 lists the parameter settings, which are consistent with the values in power plant operations. Table 2 presents the ultimate analysis results and other characteristics of the 0# light diesel oil used in the experiments. Table 3 gives the proximate analysis and ultimate analysis of the bituminous coal. The volatile matter and gross calorific value were high. The coal quality was excellent for ignition.

Results and Discussion

Distribution of gas temperature in burner

The gas temperature distributions along the burner centerline illustrated in Fig. 3 were obtained from the numerical simulation with coal concentrations of 0.08-0.40 kg/kg and for comparison with the experimental results of coal concentrations of 0.08-0.21 kg/kg during the ignition of pulverized coal. The gas temperature distributions were similar at different coal concentrations along the burner centerline in the simulations and experiments. In the numerical simulation, the gas temperature increased along the burner centerline with the primary air flow direction. Between x = 1000 mm and the burner nozzle, the gas temperature tended to be steady, and hence, the mean combustion became steady. As the coal concentration increased, the heat absorbed by the pulverized coal increased. The gas temperature decreased more rapidly than indicated by the experimental data at the equivalent points. At a coal concentration of 0.40 kg/kg, the gas temperature along the centerline remained high over this same range in the burner, which was advantageous in successful ignition of pulverized coal.

In numerical simulations, maximum temperature distributions (Fig. 4) were generated in various sections along the flow direction and in the radial direction at specified coal concentrations of 0.08-0.40 kg/kg. These distributions were similar to the gas temperature distribution along the centerline obtained from the simulations. These maximum temperatures were located at the side of the main oil gun. Along the flow direction, the maximum temperature moved gradually to the side of the auxiliary oil gun.

For a coal concentration of 0.40 kg/kg, corresponding to the actual operating conditions, the gas temperature field of the inner burner (Fig. 5) had separate high-temperature regions running along the centerlines of the main oil gun and the auxiliary oil gun. This followed the shift in the location of the maximum temperature from the main oil-gun side to the auxiliary oil-gun side.

From Figs. 4 and 5, it could be observed that the flame spread out to the burner wall and the high-temperature region expanded along the flow direction. This could be explained by the fact that the two oil guns were arranged at either side of the burner centerline. In the process of ignition, a high-energy igniter first ignited the atomized oil from the main oil gun. The initial heat of the ignition came from the oil flame at the side of the main oil gun; the maximum gas temperature was therefore initially on this side. With the primary air flowing into the burner nozzle, the oil from the auxiliary oil gun then ignited. Because the oil-flow rate of the auxiliary oil gun (65 kg/h) was greater than that of the main oil gun (35 kg/h), the oil combustion released more heat and ignited more pulverized coal. The maximum gas temperature thus gradually moved from the side of the main oil gun to the side of the auxiliary oil gun. In the overall ignition process, the flame spread to the burner wall and the high-temperature region expanded along the flow direction. As a result, given a coal concentration of 0.40 kg/kg, the gas temperature remained high in the central region of the burner. In addition, there was a large high-temperature region in the inner burner. The presence of both regions benefited coal ignition.

The calculated and measured temperature fields at the exits of the first and second combustion chambers (Fig. 6(a)) show little difference at coal concentrations of 0.08, 0.12, 0.16, and 0.21 kg/kg. Therefore, the gas temperatures were similar at these exits for different coal concentrations. The gas temperature was highest along the burner center and gradually decreased towards the burner wall. This temperature trend was caused by both the main and the auxiliary oil guns arranged in the central primary air-coal mixture duct where the high-temperature oil flame formed.

The cone separators installed in the primary air-coal mixture duct concentrated the pulverized coal into the central zone of the burner. The pulverized coal was then heated by the oil flame and ignited. Therefore, the pulverized coal combustion was intense in the central zone of the burner where the gas temperature was the highest. The flame spread to the burner wall along the direction of the primary air flow. Because the diameter of the burner duct was large, the unburned primary air-coal mixture continued to enter the flame. An air membrane, i.e., a cooling wall, was formed near the wall. Therefore, the gas temperature decreased gradually from the center to the wall. With low temperatures, the burner wall was safe during operations.

With pulverized coal burning along the direction of the primary air flow, the exit gas temperature of the second combustion chamber was higher than that of the first combustion chamber at the equivalent points. With increasing coal concentration, the exit gas temperatures of the first and second combustion chambers decreased at the equivalent points, and the width of the high-temperature zone decreased. The reason for this was that the momentum of the primary air-coal mixture increased as coal concentration increased, making it more difficult for the flame to expand in the radial direction. In addition, the O₂ supply was limited. An increase in the coal concentration resulted in a coal combustion that was more incomplete, and the gas temperature, therefore, decreased. With the coal concentration of 0.40 kg/kg, the exit gas temperature distribution of the first and second combustion chambers were similar to those for coal concentrations of 0.08-0.21 kg/kg, but were lower at the equivalent points.

Distribution of O₂ and CO concentrations at exits of combustion chambers

The distribution of the O₂ concentration along the radial direction at the first and second combustion chamber exits are depicted in Fig. 7 with different coal concentrations. In the numerical simulation, the O₂ concentration distribution correlated with the corresponding exit gas temperature distributions. As the gas temperature increased, more O₂ was consumed at the equivalent points. Therefore, the O₂ concentrations were the lowest at the burner center and increased towards the walls, near which the O₂ concentration was approximately 21%. Therefore, the pulverized coal was not combusting near the wall. For the first combustion chamber exit, the minimum O₂ concentration was 6%-10% at the center. However, at the center of the second combustion chamber exit, the O₂ concentration was almost zero irrespective of the coal concentration. The reason for this was that the O₂ was limited and pulverized coal combustion was intense in the central burner, thus consuming the available O₂ completely. In addition, the momentum of the primary air-coal mixture was large, thus entraining the O₂ along the burner center. The O₂ concentrations at the second combustion chamber exit were measured at 0.01%-0.04% in the experiments. The numerical results conformed to the experimental results. With coal concentrations at 0.40 kg/kg, the O₂ concentrations were almost zero along the centerline at the second combustion chamber exit, indicating that the combustion in the central burner was good.

Radial distributions of CO concentration at the first and second combustion chamber exits (Fig. 8) for different coal concentrations further indicated that there was insufficient O₂ at the center of the burner. The O₂ consumed in the combustion was divided into two parts: one part for oil combustion and the other for a large portion of the volatile matter and a small portion for carbon combustion. However, with entrainment stopping the O₂ from spreading to the burner center, the pulverized coal burned incompletely under oxygen-poor conditions, thus producing greater quantities of CO. The CO concentration peaks occurred at the second combustion chamber exits with concentrations of 7.9%, 9.9%, 11.3%, and 10.6% for the coal concentrations of 0.08-0.21 kg/kg. Thus as coal concentration increased, the CO formation increased. With a coal concentration of 0.40 kg/kg, the O₂ concentrations were almost zero along the centerline at the second combustion chamber exit, as for the other four coal concentrations. However, the incomplete coal combustion pushed the CO concentration to 13.1%.

Conclusions

A tiny-oil ignition burner was numerically simulated for different coal concentrations of 0.08-0.21 kg/kg. The results of these simulations conformed to the experimental results, demonstrating the suitability of the model used in the calculation. Numerical simulations for a coal concentration of 0.40 kg/kg were conducted to compare and predict the combustion characteristics of a single burner operating at its rated output. For steady-state combustion, the gas temperature was high in the central zone of the burner. With primary air flowing to the burner nozzle, the high-temperature region expanded, allowing the ignition of the pulverized coal. With increasing coal concentration, specifically 0.08 kg/kg to 0.40 kg/kg, the gas temperature decreased at the equivalent points along the centerline as well as at the equivalent points at the first and second combustion chamber exits. At these exits, the O₂ concentrations were almost zero. In such oxygen-poor combustion conditions, the pulverized coal burnt incompletely, producing greater CO concentrations and an increasing reduction atmosphere.

References

[1] Dai D, Liu J. Tackling global electricity shortage through human power: Technical opportunities from direct or indirect utilizations of the pervasive and green human energy. Frontiers in Energy, 2012, 6(3): 210–226

[2] Chen G F, Zheng X J, Cong L. Energy efficiency and carbon dioxide emissions reduction opportunities in district heating source in Tianjin. Frontiers in Energy, 2012, 6(3): 285–295

[3] Franco A, Diaz A R. The future challenges for ‘‘clean coal technologies’’: Joining efficiency increase and pollutant emission control. Energy, 2009, 34(3): 348–354

[4] You C F, Xu X C. Coal combustion and its pollution control in China. Energy, 2010, 35(11): 4467–4472

[5] Liszka M, Ziebik A. Coal-fired oxy-fuel power unit- Process and system analysis. Energy, 2010, 35(2): 943–951

[6] Schaffel-Mancini N, Mancini M, Szlek A, Weber R. Novel conceptual design of a supercritical pulverized coal boiler utilizing high temperature air combustion (HTAC) technology. Energy, 2010, 35(7): 2752–2760

[7] Fan W D, Li Y Y, Lin Z C, Zhang M C. PDA research on a novel pulverized coal combustion technology for a large utility boiler. Energy, 2010, 35(5): 2141–2148

[8] Liu C L, Li Z Q, Zhao Y, Chen Z C. Influence of coal-feed rates on bituminous coal ignition in a full-scale tiny-oil ignition burner. Fuel, 2010, 89(7): 1690–1694

[9] Li Z Q, Liu C L, Zhao Y, Chen Z C. Influence of the coal-feed rate on lean coal ignition in a full-scale tiny-oil ignition burner. Energy & Fuels, 2010, 24(1): 375–378

[10] Sheng C, Moghtaderi B, Gupta R, Wall T F. A computational fluid dynamics based study of the combustion characteristic of coal blends in pulverized coal-fired furnace. Fuel, 2004, 83(11,12): 1543-1552

[11] Backreedy R I, Jones J M, Ma L, Pourkashanian M, Williams A, Arenillas A, Arias B, Pis J J, Rubiera F. Prediction of unburned carbon and NO_x in a tangentially fired power station using single coals and blends. Fuel, 2005, 84(17): 2196–2203

[12] Saario A, Oksanen A. Comparison of global ammonia chemistry mechanisms in biomass combustion and selective noncatalytic reduction process conditions. Energy & Fuels, 2008, 22(1): 297–305

[13] Yin C, Rosendahl L, Kær S, Clausen S, Hvid S L, Hille T. Mathematical modeling and experimental study of biomass combustion in a thermal 108 MW grate-fired boiler. Energy & Fuels, 2008, 22(2): 1380–1390

[14] Shih T H, Liou W W, Shabbir A, Shabbir A, Yang Z G, Zhu J. A new k-ϵ eddy viscosity model for high reynolds number turbulent flows-model development and validation. Computers & Fluids, 1995, 24(3): 227–238

[15] Gosman A D, Loannides E. Aspects of computer simulation of liquid-fuelled combustors. In: AIAA 19th Aerospace Science Meeting, St Louis, USA, 1981, Paper No. AIAA-81-0323

[16] Cheng P. Two-dimensional radiation gas flow by a moment method. AIAA Journal, 1964, 2(9): 1662–1664

[17] Smoot L D, Smith P J. Coal Combustion and Gasification. New York: Plenum Press, 1985

[18] Spalding D B. Combustion and Mass Transfer. New York: Pergamon Press, 1979

[19] Zhou L X. Theory and Numerical Modeling of Turbulent Gas-Particle Flows and Combustion. Boca Raton: CRC Press, 1993

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg