Experimental and Numerical Simulation Research on Combustion and NOx Emission of Low Calorific Value and High-Alkaline Coal Based on Jet Characteristics of Flow Field

. A numerical model of a tower boiler burning low calori ﬁ c value and high-alkali coal was established; based on the jet rigidity of the ﬂ ow ﬁ eld in the furnace, the e ﬀ ects of air velocity and air volume ﬂ ow on the combustion of high-alkali coal and NOx emissions were studied. The optimal parameters obtained by numerical simulation are applied to a typical experimental boiler. The experimental results show that when SOFA operates at about 65 m/s, the rigidity of the jet is obviously increased, the ﬂ ow ﬁ eld and temperature ﬁ eld in the furnace are uniform, the temperature of the ﬂ ue gas at the outlet is reduced by 53 K, the CO concentration at the furnace outlet was reduced from 3713 ppm to 57ppm, the exhaust gas temperature was reduced by 6 K, and the concentration of NOx is reduced to 163 mg/m 3 . On the other hand, the combustion e ﬃ ciency is increased by 0.86%, which translates into 2g/kwh of standard coal, and the problem of easy coking has also been e ﬀ ectively solved. The collaborative optimization of high-alkali coal combustion and NOx emission is realized.


Introduction
High-alkali coals are widely distributed around the world and are characterized by low calorific value and high content of alkali metal elements and alkaline earth metal elements [1].If high-alkali coal is directly burned, it will bring about a series of problems such as difficult combustion, ash accumulation on the heating surface, fouling, and slagging [2,3].An unstructured coal field with huge reserves has been discovered in the Zhundong area of Xinjiang province, with an estimated reserve of 390 billion tons.According to China's current coal consumption rate, it can be used for about 100 years, which can provide huge support for China's growing energy demand [4][5][6].However, the high content of AAEMs (alkali and alkaline earth metals) in ash as well as its relating serious ash fouling slagging problems is the main reason for the boiler, which cannot absolutely fully use high-alkali coal [7][8][9][10].
Many scholars have studied the migration characteristics of alkali metal and alkaline earth metal elements in highalkali coal [11,12].The study of Wang et al. [13] focused on the characteristics of ash deposition during oxy-fuel combustion of high-alkali coal, especially the differences in morphologies and chemical compositions of ash deposits using a drop-tube furnace, with the ash deposition mechanisms being further elucidated.Ji et al. [14] proposed a sodium migration model of high-alkali coal combustion.This model included submodels of sodium species release, chemical reactions, vapor condensation, particle deposition, and shedding.It was verified on a circulating fluidized bed boiler, and a relatively accurate prediction result was obtained.Ji and Cheng [15] used the numerical simulation method to study the ash and slagging problems of high-alkali coal in circulating fluidized bed boilers.The results show that the calculated results of gas-solid flow rate, furnace temperature, and gas type are in good agreement with the measured data.Rokni and Levendis [16] explored the use of ashes of a low-sulfur, high-alkali lignite coal for partially capturing the sulfur dioxide emissions from combustion of a high-sulfur bituminous coal.The result showed that the alkali-rich ashes of the lignite coal acted as sulfur sorbents for the abundant SO 2 emissions of the bituminous coal.Narukawa [17] evaluated the formation and mechanism of NOx in the pulverized coal combustion process by the determination of nitrogen isotope content.Kamal [18] conducted a cocombustion experiment of pulverized coal gas and premixed gas in a two-swirl burner.The results showed that a double-flame cladding under the condition of high temperature and rich fuel was formed between the two reaction zones by efflux gas mixture into the fuel, which accelerated the pyrolysis reaction of N 2 .If the air is injected into the opposite direction and the air and gas mixture is disturbed, the denitrification efficiency is greatly improved, and the NOx and CO concentrations are reduced to 310 ppm and 480 ppm, respectively.
A comprehensive analysis of previous research literature on high-alkali coal shows that most of the researchers only pay attention to its characteristics such as ash deposition and slagging but ignore the effects of temperature field and flow field characteristics on combustion and pollutant emission during combustion.
SOFA (separated overfire air) as an effective lownitrogen combustion technology has attracted the attention of many researchers in recent years; its diagram is shown in Figure 1.The SOFA is to separate a portion of air from the main combustion area and then inject it through the burnout area to encourage further combustion of unburned pulverized coal.After air classification, the injected air in the main combustion zone is reduced, the combustion temperature is reduced accordingly, and the generation rate of NOx will be effectively inhibited.However, the characteristics of SOFA jet have great influence on the stability of temperature field and flow field in the combustion area.In this study, a numerical model of a typical tower boiler burning with low calorific value and high alkalinity coal was established and the effects of pendulum angle, wind speed, air volume flow, and other parameters are simulated.The rigidity of SOFA jet is rarely paid attention to by previous researchers.In this paper, the influence of the rigidity of the jet on the temperature field and the flow field is emphatically studied, and the change law of the rigidity of the jet on combustion and NOx emission is obtained, which has an illuminating significance in the field of combustion.The results of numerical calculation are applied to a real experimental boiler.

Numerical Models
2.1.Research Objective.This article relies on a demonstration supercritical once-through boiler, which is one-time intermediate reheating, single furnace, tower layout, octagonal tangential combustion, balanced ventilation, and solid slag discharge.The pulverizing system is 8 sets of MB3600/ 1000/490 fan coal mill direct blowing pulverizing system.Each coal mill corresponds to a corner (five-layer) burner, which are layers A, B, C, D, and E from bottom to top. Figure 1 shows the distribution of different combustion zones in the furnace.It can be seen that SOFA nozzle is located at the top layer of the combustion zone.Figure 2 is the schematic diagram of the burner viewed from the top; the horizontal arrangement of the burner in the furnace is shown in the figure.There are a total of 8 burners, of which the odd numbered burner and the horizontal axis included angle is 22 °, the even numbered burner and the horizontal axis included angle is 31 °, and 8-burner jet collected in the furnace center, forming an imaginary tangent circle.Figure 3 is a single row of burner layout diagram, the left is the burner and wind nozzle layout in the main combustion area, and the right is located in the top SOFA nozzle layout.Figure 4 is the schematic diagram of the burner viewed from the side.It can be seen from the figure that the burner nozzle can swing up and down for a range of 25   International Journal of Energy Research 2.3.Physical Model.All the heating surfaces of the tower boiler are arranged in the furnace, such that the convection heating surface of the reheater and superheater has little effect on the combustion in the furnace.The numerical model in this paper is mainly used to study the combustion characteristics and NOx in the furnace, so the mathematical model does not include the heating surfaces such as the superheater and reheater at the top of the furnace, so the outlet of the first-stage superheater (i.e., the lower furnace outlet) is taken as the furnace outlet.The section size of the furnace, the size of the nozzle of the burner, and other related dimensions are all modeled at a 1 : 1 scale with fullscale modeling.The length of the burner nozzle is 1000 mm, the width is 400 mm, and the height is 350 mm; the diameter of the outer ring air duct is 500 mm.The physical model is shown in Figure 5.The grid is divided by Ansys structured grid partition, and the whole model is divided into cold ash hopper area, main combustion area, and burn-out area.The entire model is divided into 480,000 meshes, the minimum mesh size is 50 mm, and the maximum mesh size is 150 mm.The mesh division is shown in Figure 6.
2.4.Mathematical Model.In order to simulate the complex physical and chemical processes such as combustion, flow, and NOx generation in the tower boiler, the mathematical model in this paper adopts the following equations.
(1) The RANS model adopts realizable k-ε double equation model with swirl correction effect.The transport equations including the mass, momentum, and continuity equations are defined as follows.
Continuity equation: Equation ( 1) is the partial differential form of the continuity equation, which is based on the infinitesimal microblob model with fixed spatial position, applicable to compressible and incompressible fluids.The first term on the left is unsteady, the second term is convective, and the right is the source term.
Momentum conservation equation: momentum conservation equation is based on Newton's second law, indicating that the sum of various volume and surface forces on the scope element is equal to the time change rate of fluid element momentum.The general form of momentum conservation equation x i in the Cartesian coordinate system is where u is the average velocity of the fluid in the x direction, u ! is the velocity vector, ρ is the density, t is the time, P is the hydrostatic pressure, μ is the dynamic viscosity, and S xi is the source term; for Newton fluid, S xi is equal to F xi that is the volume force.In general, the volume force is gravity only, and in the three-dimensional Cartesian coordinate system, if the Z axis is vertical, then S x = 0, S y = 0, and S z = −ρg; g is the acceleration of gravity.
where E = h − P/ρ + u 2 i /2; effective thermal conductivity k eff = k + k t , k is the heat transfer coefficient of the fluid, and k t is the coefficient of turbulent heat conduction.The three items in parentheses on the right of the equation are energy transport due to heat conduction, component diffusion, and viscous dissipation; viscous dissipation is negligible in incompressible fluids, S h is the source term in the energy equation, and there is usually no bulk heat source in the fluid domain, so this term usually refers to the heat of a chemical reaction.The expression for S h is related to the component transport model and the expression for the rate of chemical reaction. k-equation: where K is the kinetic energy of turbulence, ε is the dissipation rate, G k is the turbulent kinetic energy caused by the average velocity gradient, G b is the turbulent kinetic energy caused by buoyancy, Y m is the effect of pulsating expansion on the total dissipation rate in compressible turbulence, C 1ε , C 2ε , G 3ε are empirical constants, and σ k and σ ε are the Prandtl numbers corresponding to K and ε, respectively.k and ε of the realizable k-ε model are given by the following equation: where C 1 = max ½0:43, η/ðη + 5Þ, η = Sðk/εÞ, S = 2 ffiffiffiffiffiffiffiffiffiffiffi ffi 2S ij S ij p , C 1ε , and C 2 are constants.
(2) Use standard wall functions near the wall (standard wall function).
(3) Radiation is dominant in the furnace, so the P 1 radiation model is used to calculate the heat transfer rates between gas and particles where G is the incident radiation, α is the absorption coefficient, C is the coefficient of linear anisotropic phase function, and δ s is the scattering coefficient.
(4) The trajectory tracking of particles adopts a random trajectory model (stochastic tracking).FLUENT predicts the orbit of a single discrete phase particle through the force balance of the integrating particle under the Lagrange reference frame.The particle force equilibrium equation can be expressed as follows: where u p is particle velocity, u is gas phase velocity, ρ p is particle density, ρ is gas phase density, F D ðu − u p Þ is the drag force per unit mass of a particle, and F x is additional mass force.
(5) The precipitation of volatile analysis adopts a singlestep reaction model (single-rate model).
where m p is the particle mass, f v,0 is the initial volatile mass fraction, f w,0 is the intrinsic moisture, m p,0 is the initial particle mass, and k is the reaction rate constant.
(6) Coke is the main combustible substance in coal, and its combustion process determines the combustion speed and burnout time of pulverized coal particles.
According to the theory of combustion, the combustion model of coke in pulverized coal is the dynamic/ diffusion controlled combustion model; two factors together control the combustion of coke.One is the diffusion rate of gas to particle surface, and the other is the chemical reaction kinetics between gas and particle pore surface [19][20][21].
Gas diffusion rate: Chemical kinetic rate of surface: Total burning rate of coke: where D 0 is the diffusion rate constant, R is the chemical kinetic rate of surface, N 1 is the diffusion rate constant, N 2 is the surface chemical kinetic rate constant, T p is the particle surface temperature, T͚ is the gas phase temperature, d p is the particle size, p ox is the partial pressure of the oxidizer, and E is the activation energy.Assumption: the particle size of pulverized coal particles remains unchanged during the reaction process, and the combustion process adopts a two-step reaction mechanism with CO as an intermediate product.The equation of chemical reaction is as follows: Solid particles obey the force balance and the six heat/ mass transfer correlation laws of CFD (computational fluid dynamics) in the entire flow, combustion, and heat transfer process.The calculation of NOx adopts postprocessing method; due to the small amount of fast-type NOx generated in the reaction, thermal-type NOx and fuel-type NOx are mainly selected for the generation model.The Zeldovich mechanism is adopted for thermal NOx, and the De Soete mechanism is adopted for fuel NOx.The SIMPLE algorithm is used to deal with the pressure-velocity coupling relationship, and the relaxation factor of some parameters is reduced to improve the convergence.

Boundary Condition.
The numerical calculation mainly sets the boundary conditions according to the full load parameters.The velocity of pulverized coal particles is slightly lower than that of primary wind (empirical value), and the temperature of the pulverized coal particles is the same as the temperature of the primary air.Use PDF model to define inlet coal quality parameters [22,23].The particle size distribution of pulverized coal is set according to the Rosin-Rammler distribution [24].Assuming that the mass International Journal of Energy Research fraction of particles with diameter d and particles with diameter greater than d can be calculated using Equation (5), where w d is the mass fraction of pulverized coal particles, d is the diameter of the pulverized coal particle, d is the average diameter, n is the distribution index, and n is spread parameter.
According to R 90 = 45%, the pulverized coal is divided into 10 intervals according to the particle size.The maximum particle size, minimum particle size, and average particle size of pulverized coal particles are set, respectively, and the average particle size is calculated to be 102 μm according to the crushing formula and particle size distribution.The final particle size distribution of pulverized coal is shown in Table 3. R 90 is a standard for fineness of pulverized coal, and its calculation formula is shown in Equation ( 6): where A 90 is the mass of pulverized coal on a 90 μm sieve and M is the total mass of the pulverized coal sample.
2.6.Working Condition Setting.Referring to the long-term experience data of the experimental boiler, the oxygen content at the outlet which can guarantee the combustion is 3.5%, and the oxygen content used for calculation is also set as this value in this paper.The effect of the jet characteristics of SOFA on the combustion under 7 different working conditions is studied.Condition 1 (C 1 for short) is the verification condition, C 2 and C 3 are optimized conditions for SOFA vertical swing angle, C 4 and C 5 are optimized conditions for SOFA air volume, and C 6 and C 7 are the optimal conditions for SOFA wind velocity; the working conditions are set as shown in Table 4.

Model Validation.
In order to verify the accuracy of the numerical model, the boundary and initial values of the numerical model were set according to the actual experimental air volume, coal volume flow, and other parameters, and numerical calculation was carried out (C 1 ).The simulation calculation results of temperature field distribution, O 2 concentration field distribution, CO concentration distribution, and NOx generation amount in the experimental furnace are shown in Figures 7 and 8. Due to the limitation of the accuracy of measuring instruments and equipment, there is an observation error between the simulated quantity and the actual quantity.Various elements in the experimental furnace were measured under the same conditions, and the experimental results were compared with the numerical simulation results.The verification results are shown in Table 5.
The verification results show that the error between the calculated value and the experimental value is within 5%, and the simulation results have high reliability.
Combining the simulation results in Figures 4 and 5, it can be seen that ( 1) under the set conditions, the oxygen content in the central area of the furnace outlet section is obviously low, and the oxygen content is less than 0.1%, resulting in the CO concentration as high as 5000 ppm (2) the temperature at the outlet of the experimental furnace is relatively high, and there is still a large amount of CO reburning at the outlet of the furnace, which causes the flue gas temperature at the exhaust port to exceed the ash melting point by almost 100 °C, which greatly increases the possibility of coking at the exhaust port (3) SOFA jet rigidity is poor and decays too fast so that it cannot penetrate the center of the flame, but by the reverse thrust of the flue gas in the flow field and close to the surface of the water wall, resulting in oxygen deficiency in the center of the flame and

Simulation Optimization
3.1.Swing Angle Characteristic Optimization.Figure 9 shows the effect of SOFA swing angle of tower boiler on fluid jet characteristics and combustion results in the furnace.Previous studies have shown that a certain amplitude of downswing of the primary or secondary air nozzle can enhance the rigidity of the jet.As can be seen from Figure 7, SOFA has a small air volume flow and is located at the top of the combustion zone, which is greatly affected by the flow field in the main combustion zone.The effect of improving the jet rigidity by adjusting the pendulum angle of the burner is not obvious, nor can the height of the combustion center be controlled.On the contrary, the upward or downward swing of the SOFA swing angle will cause disturbance of the flow field and temperature field in the furnace, which will increase the temperature of the flue gas at the furnace outlet by 80 degrees and 58 degrees, respectively.Combining with the O 2 concentration distribution in Figure 7, it can be seen that the change of SOFA swing angle does not affect the distribution of O 2 concentration in the furnace, nor can it enhance the mixing degree of air and flue gas, so that a large amount of "floating oxygen" in the SOFA flows along the water wall."Floating oxygen" is the oxygen molecules located in the outermost layer of the flame and cannot contact with the fuel particles.They have no way to participate in the combustion, but can only be transported to the outside of the combustion zone along with the air flow, without any influence on the combustion process.They are transported directly to the furnace outlet along with the flue gas, resulting in a slight increase in the concentration of  oxygen in this area.The swing angle of SOFA is increased by 15 °from 0 °, the CO concentration at the furnace outlet is increased from 3713 ppm to 3805 ppm, the swing angle of SOFA is decreased by 15 °from 0 °, the CO concentration at the furnace outlet is decreased from 3713 ppm to 3409 ppm, and the CO concentration slightly changes, but the effect is not obvious.At the same time, because the height of the flame center in the furnace has not been effectively controlled, the burnout time of the highly alkaline particles is prolonged, and the flue gas in the SOFA area still has a high temperature, which leads to the regeneration of thermal NOx in the burnout area.On the contrary, it increases the NOx concentration at the furnace outlet.Combining the characteristics of tower boilers and high-alkali coal, it can be seen from the numerical calculation that the SOFA swing angle fluctuation amplitude is too large, which leads to the backward movement of the combustion area and the high-temperature flue gas flow field in the furnace, which leads to the increase of the outlet flue gas exhaust temperature and the increase of the heat loss in the furnace and at the same time increases the probability of high-alkali coal coking and NOx generation.It is not a wise choice to control the temperature field and flow field in the furnace through the SOFA jet angle and direction.

Air Volume Flow Characteristic
Optimization. Figure 10 is the numerical simulation result of the influence of SOFA volume flow on combustion and NOx generation in the furnace.In order to reduce the heat loss of incomplete combustion of gas, it is intuitive to reduce the CO concentration at the outlet of the furnace (the CO concentration at the outlet of the furnace is the characterization of the heat loss of incomplete combustion of gas).Based on the tower distribution air supply mode, the numerical simulation in this part  With the increase of oxygen content in the main combustion area, the temperature of this area increases significantly, the burnout distance and time of pulverized coal become shorter, and the CO concentration in the SOFA area decreases significantly.Compared with C 1 , the SOFA volume flow in C 4 and C 5 decreased by 9% and 20.5%, respectively, and the CO concentration at the furnace outlet decreased by 3181 ppm and 3714 ppm, respectively.Due to the reduction of SOFA volume flow, the position of the flame center moves up, which leads to a significant increase in the flue gas temperature at the furnace outlet.Compared with C 1 , the SOFA volume flow of C 4 is reduced by 9%, and the flue gas temperature at the furnace outlet is increased by 126 °C.When the SOFA volume flow of C 5 was 20.5% lower than that of C 1 , the furnace outlet temperature increased by 230 °C, exceeding the ash melting point of 126 °C of high-alkali coal, which significantly increased the risk of coking in the flue gas vent.In addition, reducing the volume flow of SOFA will weaken the effect of air classification, which will greatly increase the amount of NOx generated.Compared with C 1 , the SOFA volume flow of C 4 and C 5 decreased by 9% and 20.5%, respectively, and the NOx concentration at the furnace outlet increased by 378 mg/m 3 and 448 mg/m 3 , respectively.Therefore, although reducing the proportion of SOFA volume flow can greatly reduce the concentration of CO at the furnace outlet and improve the combustion efficiency, it will also greatly increase the flue gas temperature in the main combustion area and the furnace outlet, and the amount of NOx generated will increase significantly.From the results of the numerical simulation, it is not recommended to use the method of reducing the SOFA ratio to improve the combustion efficiency, and the better SOFA ratio should be kept at about 21%.

Wind Velocity Characteristic Optimization.
Figure 11 shows the effect of SOFA wind velocity on combustion and the results of optimization.In the graded burnout air system, the SOFA nozzles of the lower two layers of C 6 are fully opened, the nozzles of the upper two layers are fully closed, and the SOFA wind velocity is 65 m/s.The SOFA vents on the bottom layer of C 7 are fully open, the SOFA vents on the upper three layers are fully closed, and the wind velocity is 100 m/s.The SOFA wind velocity increased from 50 m/s to 65 m/s, the flue gas temperature at the furnace outlet decreased by 53 K, the CO concentration decreased by 3661 ppm, and the NOx concentration increased by 22 mg/m 3 .It can be seen from the above data that the increase of wind velocity can significantly improve the rigidity of SOFA jet and enhance the jet intensity, and the control effect of SOFA on the flame center is more obvious.The rigidity of SOFA jet is enhanced, the mixing effect of SOFA and flue gas is improved, the CO concentration at the furnace outlet is reduced, and the combustion efficiency is improved.The content of alkali metals (K, Na) and alkaline earth metals (Ca, Mg) in high-alkaline coal is very high.These substances have low ash melting point and are easy to coking.The oxidizing atmosphere helps the two metal elements solidify into dense oxides with high ash melting point, thus reducing the risk of coking at the flue gas outlet simultaneously.Numerical calculation data display that the SOFA wind velocity increased from 50 m/s to 100 m/s, the flue gas temperature at the furnace outlet increased by 32 K, the NOx concentration increased by 75 mg/m 3 , and the CO concentration decreased by 1338 ppm.The reason is that simply increasing the SOFA wind velocity actually reduces the SOFA jet source area, which weakens SOFA's ability to resist decay and the SOFA wind rigidity is not substantially enhanced, resulting in a slight increase in the furnace outlet flue gas temperature and NOx generation.Therefore, in order to improve the rigidity of SOFA jet, on the one hand, it is necessary to increase the SOFA wind velocity as much as possible, and on the other hand, it is necessary to ensure that the SOFA wind has sufficient jet source area.
3.4.Final Optimization Effect.The numerical simulation results show that C 6 is the optimal condition compared to other conditions.In order to verify the accuracy of the numerical simulation results, the experimental furnace was set as follows according to the boundary conditions of C 6 : (1) Maintain the vertical swing angle of SOFA at 0 °(approximately horizontal) (2) The air volume flow is maintained at about 21%.At the same time, the dampers of the upper 2 layers of SOFA are fully closed, and the dampers of the lower 2 layers of SOFA are fully opened (3) The wind velocity is stable at about 65 m/s Previous studies have shown that the higher the temperature, the more likely the coke material is to reach the ash melting point and bond around the burner.In this paper, the flue gas parameters at the outlet of the experimental furnace are tested.The test results show that the flue gas temperature at the outlet of the experimental furnace is reduced by 53 K, and the coking state at the outlet is observed, which is also effectively alleviated.The CO concentration is reduced from 3713 ppm to 57 ppm; the concentration of NOx is reduced to 163 mg/m 3 .On the other hand, the heat loss of chemical incomplete combustion is reduced by 0.46%, the exhaust gas temperature is reduced by 6 K, the heat loss of exhaust gas is reduced by 0.3%, the heat loss of mechanical incomplete combustion is reduced by 0.1%, and the combustion efficiency is increased by 0.86%, which translates into 2 g/kwh of standard coal.

Conclusions
(1) The swing angle of SOFA is not obvious to improve the rigidity of wind.At the same time, the upper or lower swing of the swing angle will increase the center of the flame, and the temperature of the flue gas at the outlet of the experimental furnace will increase, which will cause a reduction in the effective heat capacity in the furnace.Reducing the air volume flow of SOFA can greatly reduce the generation of CO, but on the other hand, it will also destroy the air classification of SOFA, shorten the NOx reduction distance, and lead to a significant increase in the NOx content at the furnace outlet.Therefore, neither changing the SOFA swing angle nor increasing the air volume flow is the best choice for reducing NO X concentration (2) Increasing the SOFA velocity can effectively improve the rigidity of SOFA jet, enhance the mixing effect of wind and flue gas, and reduce the concentration of NOx.However, if the SOFA volume flow is a fixed value, simply increasing the wind velocity will inevitably reduce the jet source area of the SOFA, and the reduction of the jet source area will in turn weaken the attenuation resistance of the airflow.Therefore, in order to enhance the rigidity of SOFA jet, the air volume flow and wind velocity should be optimized together to obtain the best effect (3) For the experimental tower furnace, it uses low calorific value and high-alkali coal that is easy to burn but hard to burn out.The rigidity of SOFA jet should be improved as much as possible to enhance the mixing effect of flue gas and air in the later stage of combustion; it enables synergistic optimization of reduced combustion heat loss and NOx emission concentration.Numerical simulation can provide a visual research method for combustion optimization.The combination of numerical simulation and experimental measurement can provide refined predictions Temperature, K.

Figure 3 :
Figure 3: Longitudinal arrangement of single row burner.AA-bottom auxiliary air; A, B, C, D, E, F-each layer burner nozzle; AB, BC, DE-oil gun layer auxiliary air; DC-intermediate layer auxiliary wind; EF-top auxiliary air.

Figure 7 :
Figure 7: Temperature field and O 2 concentration distribution within the numerical model.

Figure 8 :
Figure 8: CO and NOx concentration distribution within the numerical model.
NOx concentration in the furnace

Figure 9 :
Figure 9: Effect of SOFA swing angle on combustion.
NOx concentration in the furnace

Figure 10 :
Figure 10: Effect of SOFA volume flow on combustion.

Figure 11 :
Figure 11: Effect of SOFA wind velocity on combustion.
fluid dynamics ppm: Parts per million SOFA: Separated overfire air T: The experiment in this paper is divided into two parts: assay and measurement.The experiment is completed in the tower boiler, and the assay is completed in the laboratory.The instruments used in the test are shown in Table1, and the uncertainty analysis of the test parameters is summarized in Table2.

Table 1 :
Instruments used in experiments.

Table 2 :
Test parameter uncertainty calculation table in the experiment.

Table 3 :
Distribution of pulverized coal particle size.

Table 5 :
Comparison of experimental and numerical results.