Influence of mass air flow ratio on gas-particle flow characteristics of a swirl burner in a 29 MW pulverized coal boiler

Abstract

In a gas/particle two-phase test facility, a three-component particle-dynamics anemometer was used to measure the characteristics of gas/particle two-phase flows in a 29 megawatt (MW) pulverized coal industrial boiler equipped with a new type of swirling pulverized coal burner. The distributions of three-dimensional gas/particle velocity, particle volume flux, and particle size distribution were measured under different working conditions. The mean axial velocity and the particle volume flux in the central region of the burner outlet were found to be negative. This indicated that a central recirculation zone was formed in the center of the burner. In the central recirculation zone, the absolute value of the mean axial velocity and the particle volume flux increased when the external secondary air volume increased. The size of the central reflux zone remained stable when the air volume ratio changed. Along the direction of the jet, the peak value formed by the tertiary air gradually moved toward the center of the burner. This tertiary air was mixed with the peak value formed by the air in the adiabatic combustion chamber after the cross-section of x/d = 0.7. Large particles were concentrated near the wall area, and the particle size in the recirculation zone was small.

Keywords

industrial pulverized coal boiler / swirl burner / air/particle flow / particle dynamic analyzer (PDA)

Introduction

In China, coal-fired industrial boilers are a primary type of coal-fired equipment, with the exception of utility boilers. There are almost 600000 such industrial boiler units, consuming nearly 640 million tons of coal a year. In many developed countries, coal-fired industrial boilers mainly use pulverized coal; however, those in China primarily use lump coal. The main problems associated with China’s coal-fired industrial boilers are [¹] that they have a low actual average thermal efficiency of approximately 65% and they produce excessive pollutant emissions which generally cannot meet the national and local emission standards.

Medium and small-capacity pulverized coal industrial boilers became mature products in developed countries in the mid-to-late 1990s, especially in Germany and France [²]. Compared with the layered coal burning boiler, pulverized coal industrial boilers use a combustion mode for chamber combustion. The pulverized coal particles are sprayed into the furnace with air to facilitate suspension combustion. This makes the boiler’s thermal efficiency reach more than 90%, and pollutants effectively controlled. In the 1960s, Feng [³] successfully transformed a hand-fired boiler into a boiler that burned pulverized coal; however, the pulverized coal industrial boiler was not advertised, leading to a poor adoption.

In small-capacity pulverized coal industrial boilers, the swirling pulverized coal combustion technology has been widely adopted, because it adapts to frequent starts and stops and adjusts to wide load ranges. Scholars have been interested in the swirling pulverized coal combustion technology since it was introduced. However, most research on this technology has focused on utility boilers [⁴–¹¹]. Few studies have examined the swirling pulverized coal combustion technology in coal-fired industrial boilers.

The structure of the swirling burner and the ratio of the winds at each stage determine the air/particle flow characteristics within a furnace. Airflow characteristics of the pulverized-coal particles critically affect coal ignition and combustion characteristics. However, it is not feasible to measure the actual two-phase flow in practical situations. Generally, it is believed that the air/particle flow fields acquired in a small-scale boiler can quantitatively reflect the activities in a full-scale version, based on specific similarities in modeling criteria. Researchers widely use a three-dimensional laser particle dynamic analyzer (PDA) to measure the air/particle flows in a laboratory-scale model boiler. Sommerfeld [¹²] studied the particle behavior in a confined swirling flow by applying a PDA and by using numerical simulation, respectively, and proved that the experimental calculations for both the airflow and the particle phase conform well with the numerical simulation. Jing et al. [¹³] studied the influence of the mass flow rate of secondary air on the air/particle flow characteristics of the double swirl flow burner by PDA measurements. Zhou et al. [¹⁴] investigated the influence of the air particle flow characteristics on the formation of high temperature corrosion with the PDA technology. Wang et al. [¹⁵] confirmed the validity and the progressiveness of the eccentric-swirl-secondary-air combustion technology by comparing and analyzing air/particle flow characteristics of the Babcock and Wilcox combustion technology and the eccentric-swirl-secondary-air combustion technology using a PDA measurement system. It is of great significance to grasp the flow laws in industrial boilers under variable operation conditions, to predict combustion performance and promote the safe and economic operation of industrial boilers.

In this study, a 1: 6 scale model boiler and a PDA measurement system were used to measure three dimensional velocities, volume fluxes, and particle sizes of a two-phase flow in a 29 MW pulverized coal industrial boiler. The experimental results can provide a useful reference for the design and operation of similar pulverized coal industrial boilers.

New burner

Figure 1 shows the structure of the swirling burner for a pulverized coal industrial boiler. For these experiments, the primary air was introduced through the primary channel in the center of the burner. The inner secondary air and the external secondary air were arranged in sequence outside the primary air. The angle of the tangential blades at the internal secondary air duct and the external secondary air duct were both set at 60°. The air passed through the secondary air blades, creating an internal central recirculation zone (ICRZ) at the center of the burner. This ICRZ allowed the high-temperature flue gases to flow backward, providing a heat source for the pulverized coal ignition. A cylindrical adiabatic combustion chamber (ACC) was cast outside the external secondary air nozzle. The space inside the ACC was narrow, causing the temperature in the ACC to rise rapidly and maintain a high temperature. This gave the burner good ignition and a combustion performance. Most tertiary air nozzles were divided into two layers and were evenly distributed outside the pre-combustion chamber. The tertiary air was arranged outside the ACC, to keep it away from the center of the burner. This reduced the slagging in the furnace and led to reduced NO_x emissions, because of strengthened air-staged combustion.

Experimental

Figure 2 demonstrates the air/particle test facility used in this paper, consisting of a feeder, a cyclone separator, a small-scale burner, and a boiler model. For these experiments, the sucker introduced air from the burner primary air duct, the inner secondary air duct, the external secondary air duct, and the tertiary air duct into the furnace. The different air flux levels of the burner were regulated using throttle valves and were measured using flow meters.

To ensure the similarity between the model and industrial prototypes, each model’s geometric dimensions and operational parameters were obtained using the scaling criteria [¹⁶–¹⁸]. These criteria included geometric similarity, model flow satisfying self-molding flows, unaltered momentum ratios under scale reduction, boundary condition similarity, and Stokes criterion.

Coal particles were not used in the experiment because of inability to meet the steradian and reflectance characteristics of measured particles for PDA. Instead, glass beads were determined to meet these demands and helped yield an accurate concentration measurement. At the same time, glass bead characteristics were similar to those of pulverized coal. First, the mean diameter of coal particles generally ranged from 30 to 60 μm. Figure 3 depicts the particle size distribution for the glass beads. Second, in general, the particle density of pulverized coal was approximately 2200 kg/m³ and the density of glass beads was 2500 kg/m³. Therefore, glass beads were selected as the model particle.

A glass window was installed in the front part of the boiler model which was constructed using the same material as the glass beads, resulting in little static electricity and a low adhesion. The glass also had a good light transmittance that benefitted PDA measurements. The glass beads were fed into the boiler model through the feeder. The glass beads were recovered by a cyclone separator, and could be continuously removed through a brief flapper. Part of the fine particles could not be separated; as such, the recovered glass beads thickened. During the experiment, new glass beads were continuously added to eliminate the old ones. This kept the particle size distribution of the glass beads entering the burner stable.

The PDA used in this study was a laser-optical measurement device, manufactured by Dantec Dynamics. This effectively and simultaneously measured three-dimension velocities, particle volume flux, and the mean particle diameter of the gas/particle two-phase jet flow. The PDA measurement system is a non-intrusive and real-time absolute measurement system, which uses the proven phase Doppler principle for simultaneous measurements. It has a very high accuracy and spatial resolution, and does not need to be calibrated before use. This has led to its wide use in difficult measurement environments.

The measurement equipment consisted of an argon ion laser, a Bragg cell, transmitting probes, fiber optics, optical receivers, signal processors, a computer, and a three-dimensional traverse mechanism. Figure 4 depicts the schematic diagram of the PDA measurement system. The maximum output power of the argon ion laser adopted in the test was 5 W, with an adjustable output power. The laser light emitted from the argon ion laser was split into six laser beams of green, blue, and violet colors through the Bragg cell. Then, the two blue and two green laser beams were delivered to the two-dimensional transmitting probe, and the two violet laser beams were delivered to the one-dimensional transmitting probe, through the 60 × fiber flow optics. The light scattered from the measurement point was received by 57 × 10 PDA optical receivers and was then conveyed to signal processors, where the flow field information was obtained. The signal processor used was a 58N50 PDA enhanced signal processor, which simultaneously analyzed and processed the velocity, size, and concentration of two-phase flow at the same time. The transmitting probes and optical receivers are fixed on the three-dimensional traverse mechanism. The computer controlled the three-dimensional traverse mechanism, making the laser light move in three dimensions. This enabled the measurement of the flow parameters in the furnace model at different positions.

Based on the characteristic timeframes of the tracer particles, the velocity data were divided into gas velocity and particle velocity. Particles with diameters up to 8 μm were used to trace the airflow, and particles with diameters of 10–100 μm were used to monitor the particle phase flow. The measurable range for the particle size was 0.5–1000 μm and the range for velocity was - 500 to 500 m/s. The measurement errors for mean velocity, particle diameter, and volume flux were 1%, 4%, and 15%, respectively. More specific details on the PDA measurement system are described in Refs. [¹⁹–²¹].

It can be observed from Fig. 2 that the center of the burner exit is set as the coordinate origin. The variable x represents the distance to the burner exit along the jet flow direction while r is the radial distance from the burner axis, and d is the outer diameter of external secondary air. Due to the symmetry of the flow field in the furnace, only half of the furnace was used as the measurement region. The air/particle flow characteristics were measured at cross sections at a dimensionless length-scale of x/d = 0.2, 0.3, 0.7, 1.0, 1.5, 2.0, 2.5, 3.0 and 3.5. Table 1 provides the specific parameters.

Results and discussion

Air and particle velocity distribution

Figure 5 exhibits the profiles of the air/particle mean axial velocities in different cross-sections of the furnace at the four different mass-flow ratios of the secondary and tertiary air. The overall velocity distributions for the air and particles were similar. The particles had more inertia compared to air. As such, there was a slip in the velocity between the air and particle velocity. It is seen from Fig. 5 that the particle axial velocity is slightly greater than the air axial velocity. This means that the particle velocity lagged slightly behind the air velocity, with relatively small slips in velocity. In the region x/d = 0.2–3.0, the mean axial velocity near the central axis of the burner was negative. This indicated that a stable central recirculation zone formed in the exit region of the burner at different mass-flow ratios. Near the center of the burner, the mean axial velocities for the air and particles for the burner were consistently negative, indicating that the air/particle mixture experienced a long residence time inside the central recirculation zone in the primary flow zone.

The range of the central recirculation zone was essentially the same under each working condition, suggesting that the mass-flow ratio of secondary and tertiary air had little effect on the size of the central recirculation zone. The total amount of the air supply was constant; as such, the tertiary air volume decreased as the external secondary air volume increased. As the amount of external secondary air increased from 0.0809 kg/s to 0.3225 kg/s, the absolute value of the mean axial velocity in the central recirculation zone gradually increased; the peak value of the mean axial velocity formed by the air in the ACC decreased; and the mean axial velocity at the tertiary air nozzles increased. Along the direction of the jet, the peak value formed by the tertiary air gradually moved toward the center of the burner. It was then mixed with the peak value formed by the air in the ACC after the cross-section of x/d = 0.7. The peak value of the mean axial velocity disappeared after the cross-section of x/d = 3.0, demonstrating that the stages were fully mixed at this cross section.

Figure 6 displays the profiles of the air/particle root mean square (RMS) axial fluctuation velocities in different cross-sections of the furnace at the four different mass-flow ratios of the secondary and tertiary air. The RMS axial velocities at the outlet of the ACC and the tertiary air nozzles were higher than those in other areas. The region inside the peak zone at the outlet of the ACC served as the mixture zone between the primary air/particle mixture and the reverse flow while the region inside the peak zone at the outlet of the tertiary air nozzles served as the mixture zone between the secondary and tertiary air. The two peaks indicated there exists high axial turbulent diffusion in these regions. The distribution of RMS axial fluctuation velocity was essentially the same under different mass-flow ratios. In the region of x/d = 0.2–1.0, as the external secondary air volume increased, the RMS axial fluctuation velocity in the central region of the burner increased, and the RMS axial fluctuation velocity in the outlet region of the tertiary airflow decreased. As the jet developed, the radial distribution of the RMS axial fluctuation velocity smoothened. The peak value of the RMS axial fluctuation velocity disappeared after the cross-section of x/d = 1.5. The RMS axial fluctuation velocity near the wall decreased when the external secondary air volume increased.

Figure 7 shows the profiles of the air/particle mean tangential velocities in different cross-sections of the furnace at the four different mass-flow ratios of secondary and tertiary air. The tertiary air was non-swirling; as such, the mean tangential velocities for the burner were relatively small in the region near the wall. Upstream from the cross-section of x/d = 1.5, the peak mean tangential velocities for the burner moved toward the chamber axis, indicating that the air/particle mixture near the chamber axis began to swirl, driven by the secondary air. The region near the wall in each cross section satisfied the fact that as the external secondary air volume increased, the mean tangential velocity for the air and particles increased. In the region of x/d = 0.2–0.5, the mean tangential velocity was negative between r = 125 mm and 165 mm, when the mass-flow ratios of secondary and tertiary air were 20: 75 and 45: 55, respectively. In the region of x/d = 0.2–0.7, as the external secondary air volume increased, the mean tangential velocity for the air and particles increased. This is because there was an increased swirling of the burner with a larger secondary air volume. After the cross-section of x/d = 1.5, the radial distribution of the mean tangential velocity tended to be smooth.

Figure 8 presents the profiles of the air/particle RMS tangential fluctuation velocities in different cross-sections of the furnace at the four different mass-flow ratios of secondary and tertiary air. In the region of x/d = 0.2–0.3, the air/particle RMS tangential fluctuation velocity in the central recirculation zone and the tertiary air nozzle zone was large, indicating there exists a high turbulent diffusion in the two zones. Along the direction of the jet, the radial distribution of the RMS tangential fluctuation velocity tended to be smooth. In the regions of x/d = 0.2–0.3 and x/d = 1.5–2.5, the RMS tangential fluctuation velocity was large when the external secondary air volume was large. After the cross-section of x/d = 3.0, the radial distribution of the mean tangential velocity remained essentially the same at different mass-flow ratios.

Figure 9 shows the profiles of the air/particle mean radial velocities in different cross-sections of the furnace at the four different mass-flow ratios of secondary and tertiary air. The air/particle mean radial velocity in the central recirculation zone was essentially zero. From the burner jet to the cross-section of x/d = 0.7, the profiles peaked twice: the peak near the chamber axis was the secondary air flow zone while that near the wall was the tertiary air flow zone. As the jet flow developed, the profiles of the air/particle mean radial velocities for the burner flattened. In the region of x/d = 0.2–0.7, the mean radial velocity in the region near the wall was negative, indicating that the flows moved toward the chamber axis. However, after the cross-section of x/d = 1.0, the mean radial velocity in the region near the wall was positive. After the cross-section of x/d = 3.0, the mean radial velocity decreased as the external secondary air volume increased.

Figure 10 demonstrates the profiles of the air/particle RMS radial fluctuation velocities in different cross-sections of the furnace at the four different mass-flow ratios of secondary and tertiary air. The air/particle RMS radial fluctuation velocity increased as the external secondary air volume increased. In the region of x/d = 0.2–0.5, the RMS radial fluctuation velocity peaked twice in the radial direction. As the jet flow developed, the radial distribution of the RMS radial fluctuation velocity became smoother. After the cross-section of x/d = 1.0, the RMS radial fluctuation velocity remained essentially the same under different working conditions, showing that the mass-flow ratio of secondary and tertiary air had little effect on the distribution of the RMS radial fluctuation velocity.

Air-particle mixture characteristic

Figure 11 exhibits the mean particle diameter profiles of 0–100 mm in different cross-sections of the furnace at the four different mass-flow ratios of secondary and tertiary air. The mean particle diameter is the arithmetic mean of the diameter measurements. In the region of x/d = 0.2–0.3, in the central recirculation zone, the mean particle diameter decreased as the external secondary air volume increased. The opposite was also true: outside the central recirculation zone, the particle mean diameter increased as the external secondary air volume increased. Large particles were concentrated in the area near the wall, and the particle size near the burner center, especially in the recirculation zone, was small. The reason for this is that the momentum of the large particles made them too large to be carried into the recirculation zone. As the jet flow developed, the radial distribution of the mean particle diameters became smoother. After the cross-section of x/d = 2.0, the distribution of the mean particle diameters for each cross-section remained essentially the same.

Figure 12 shows the profiles of the particle volume flux at sizes of 0–100 μm in different cross-sections of the furnace at the four different mass-flow ratios of secondary and tertiary air. Particle volume flux refers to the volume of the particle passing through the unit area of the measurement body per unit time. This study only considered the axial volume flux. The particle volume flux near the burner center was negative, indicating that a recirculation zone formed at the center of the burner. In the central recirculation zone, the absolute value of the particle volume flux increased as the external secondary air volume increased. Outside the central recirculation zone, the peak value of the particle volume flux increased as the external secondary air volume decreased. In the region of x/d = 0.2–1.0, there was an external recirculation zone near the wall, when the amount of external secondary air was large. After the cross-section of x/d = 2.0, the particle volume flux near the wall was small, indicating that the particles were not concentrated near the wall, which helped avoid the slagging of the water wall tube.

Conclusions

A PDA system was used to measure the three-dimensional air/particle velocity, the particle size, and particle volume flux at the four mass-flow ratios of secondary and tertiary air in a small-scale version of a 29 MW pulverized coal industrial boiler. The conclusions reached were as follows:

The mean axial velocity and the particle volume flux in the central region of the burner outlet were negative, indicating that the central recirculation zone formed in the center of the burner.

In the central recirculation zone, the absolute value of the mean axial velocity and the particle volume flux increased as the external secondary air volume increased. However, the size of the central reflux zone remained essentially unchanged when the air volume ratio changed.

Along the jet flow direction, the peak value formed by the tertiary air gradually moved toward the center of the burner, and mixed with the peak value formed by the air in the ACC after the cross-section of x/d = 0.7.

Large particles were concentrated in the area near the wall, and the particle size in the recirculation zone was small.

References

[1] Yu J. Status and transformation measures of industrial coal-fired boiler in China. Clean Coal Technology, 2012, 18(03): 89–91 (in Chinese)

[2] Ji R S, Wang N J, Xiao C W, Wang Y G. The situations and application of efficient pulverized coal fired industrial boiler technique. Clean Coal Technology, 2009, 15(05): 52–56 (in Chinese)

[3] Feng J K. Experience of industrial boilers burning pulverized coal. Boiler Technology, 1989, 11: 9–11 (in Chinese)

[4] Costa M, Azevedo J L T, Carvalho M G. Combustion characteristics of a front-wall-fired pulverized coal 300 MW_e utility boiler. Combustion Science and Technology, 1997, 129(1): 277–293

[5] Costa M, Silva P, Azevedo J L T. Measurements of gas species, temperature, and char burnout in a low-NO_x pulverized coal-fired utility boiler. Combustion Science and Technology, 2003, 175(2): 271–289

[6] Vikhansky A, Barziv E, Chudnovsky B, Talanker A, Eddings E, Sarofim A. Measurements and numerical simulations for optimization of the combustion process in a utility boiler. International Journal of Energy Research, 2004, 28(5): 391–401

[7] Tsumura T, Okazaki H, Dernjatin P, Savolainen K. Reducing the minimum load and NO_x emissions for lignite-fired boiler by applying a stable-flame concept. Applied Energy, 2003, 74(3-4): 415–424

[8] Qi Li Z, Sun R, Wan Z X, Sun S Z, Wu S H, Chen L Z. Gas-particle flow and combustion in the near-burner zone of the swirl-stabilized pulverized coal burner. Combustion Science and Technology, 2003, 175(11): 1979–2014

[9] Xue S, Hui S E, Liu T S, Zhou Q, Xu T, Hu H. Experimental investigation on NO_x emission and carbon burnout from a radially biased pulverized coal swirl burner. Fuel Processing Technology, 2009, 90(9): 1142–1147

[10] Li Z Q, Jing J P, Liu G K, Chen Z, Liu C. Measurement of gas species, temperatures, char burnout, and wall heat fluxes in a 200-MW_e lignite-fired boiler at different loads. Applied Energy, 2010, 87(4): 1217–1230

[11] Li S, Li Z Q, Jiang B K, Chen Z, Zhang X. Effect of secondary air mass flow rate on the airflow and combustion characteristics and NO_x formation of the low-volatile coal-fired swirl burner. Asia-Pacific Journal of Chemical Engineering, 2015, 10(6): 858–875

[12] Sommerfeld M, Qiu H. Characterization of particle-laden, confined swirling flows by phase-doppler anemometry and numerical calculation. International Journal of Multiphase Flow, 1993, 19(6): 1093–1127

[13] Jing J P, Li Z Q, Wang L, Chen Z, Chen L, Zhang F. Influence of the mass flow rate of secondary air on the gas/particle flow characteristics in the near-burner region of a double swirl flow burner. Chemical Engineering Science, 2011, 66(12): 2864–2871

[14] Zhou H, Yang Y, Dong K, Liu H, Shen Y, Cen K. Influence of the gas particle flow characteristics of a low-NO_x swirl burner on the formation of high temperature corrosion. Fuel, 2014, 134: 595–602

[15] Wang Q X, Chen Z C, Chen Q W, Zeng L, Li Z. Experimental investigation of gas/particle two-phase flow characteristics in a down-fired boiler by PDA measurements. Experimental Thermal and Fluid Science, 2019, 107: 38–53

[16] Li Z Q, Wan Z X, Su R, Sun S Z, Chen L Z, Wu S H, Qin Y K. Influence of division cone angles between the fuel-rich and the fuel-lean ducts on gas-particle flow and combustion near swirl burners. Energy, 2002, 27(12): 1119–1130

[17] Chen Z C, Li Z Q, Zhu Q Y, Jing J. Gas/particle flow and combustion characteristics and NO_x emissions of a new swirl coal burner. Energy, 2011, 36(2): 709–723

[18] Kuang M, Li Z, Zhu Q, Wang Y, Chen L, Zhang Y. Experimental gas/particle flow characteristics of a down-fired 600 MW_e supercritical utility boiler at different staged-air ratios. Energy, 2012, 42(1): 411–423

[19] Fang N, Li Z Q, Wang J Q, Zhang B, Zeng L, Chen Z, Wang H, Liu X, Zhang X. Experimental investigations on air/particle flow characteristics in a 2000 t/d GSP pulverized coal gasifier with an improved burner. Energy, 2018, 165: 432–441

[20] Zeng L Y, Li Z Q, Zhao G B, Li J, Zhang F, Shen S, Chen L. The influence of swirl burner structure on the gas/particle flow characteristics. Energy, 2011, 36(10): 6184–6194

[21] Ren F, Li Z Q, Chen Z C, Xu Z, Yang G. Experimental investigations into gas/particle flows in a down-fired boiler: influence of the vent air ratio. Energy & Fuels, 2010, 24(3): 1592–1602

RIGHTS & PERMISSIONS

Higher Education Press