Sage Journals: Discover world-class research

Abstract

A three-dimensional numerical simulation was carried out to study the pulverized-coal combustion process in a tangentially fired ultra-supercritical boiler. The realizable k-ε model for gas coupled with discrete phase model for coal particles, P-1 radiation model for radiation, two-competing-rates model for devolatilization, and kinetics/diffusion-limited model for combustion process are considered. The characteristics of the flow field, particle motion, temperature distribution, species components, and NO_x emissions were numerically investigated. The good agreement of the measurements and predictions implies that the applied simulation models are appropriate for modeling commercial-scale coal boilers. It is found that an ideal turbulent flow and particle trajectory can be observed in this unconventional pulverized-coal furnace. With the application of over-fire air and additional air, lean-oxygen combustion takes place near the burner sets region and higher temperature at furnace exit is acquired for better heat transfer. Within the limits of secondary air, more steady combustion process is achieved as well as the reduction of NO_x. Furthermore, the influences of the secondary air, over-fire air, and additional air on the NO_x emissions are obtained. The numerical results reveal that NO_x formation attenuates with the decrease in the secondary air ratio (γ_2nd) and the ratio of the additional air to the over-fire air (γ_AA/γ_OFA) was within the limits.

Keywords

Pulverized-coal boiler coal combustion process air staging

Introduction

Tangentially fired pulverized-coal boilers are widely applied in power plants. During the coal combustion processes, nitrogen oxide (NO_x) is one of the main pollutants which must be reduced. Nowadays, there exist two main techniques to reduce the NO_x emission in pulverized-coal boilers.^1,2 The first one is the low NO_x combustion technology, while the other is based on the flue gas denitrification technology including selective non-catalytic reduction (SNCR) and selective catalytic reduction (SCR). In this article, the attention is focused on the low NO_x combustion technology applied on the tangentially fired coal boiler.

Generally, the low NO_x burner is one of the most effective methods used in the low NO_x combustion technology. In recent decades, lots of experimental or numerical researches^3–5 have been conducted on the coal combustion boiler to investigate the flow field, coal combustion, and NO_x emission. Compared with the expensive experiments, computational fluid dynamics (CFD) is an indispensable tool to optimize the low NO_x technology. For example, Fan et al.^4,5 made attempts to improve the accuracy of numerical simulation of combustion process inside tangentially fired furnaces, and proved that the RNG k-ε model was more available than the standard k-ε model. Zhang et al.⁶ also investigated the performance of a traditional tangentially fired pulverized-coal boiler and proved that the simulation was pretty reasonable by validating the operation results. Due to the complexity of the combustion process and heat transfer, at least now, it may not be possible to involve all the influence factors in one model. However, several available simplified models have been proposed^7–11 to adapt the unique features of pulverized-coal boilers, and many useful results have been obtained. Hence, the relatively well-developed mathematical models of coal combustion process and NO_x emission, especially for commercial-scale applications, have gained lots of popularity and been put forward in commercial software package such as Ansys 16.0.

As known, the NO_x emission in coal boilers could be affected by many factors, such as the excess air ratio,¹² the type of air distribution,^11,13,14 the combustion burner,^6,15 pulverized-coal characters; Choi and Kim¹¹ emphasized the influence of the over-fire air (OFA) distribution on the fuel and thermal NO_x emissions in a 500 MWe tangentially fired pulverized-coal boiler using commercial comprehensive models. Zhang et al.⁶ studied the effects of horizontal bias combustion (HBC) and air staging combustion technologies on the NO_x emission of a 200 MWe tangentially fired pulverized boiler. Besides, the effects of deflection angles of pulverized-coal nozzle flow on NO_x emission in a 600 MWe tangentially coal boiler were researched by Zhan et al.¹⁵ Furthermore, some studies^16–19 have focused on the effect of the wall-fired burner on the optimization of NO_x emission.

In this article, an unconventional tangential-fired burner and a novel air staging technology are proposed based on the 660 MWe tangentially fired ultra-supercritical boiler built in Nanjing Datang powder plant. The purpose of this numerical research is to investigate the performance of the boiler, gas flow, particle motion, coal combustion, and NO_x emission.

Physical and mathematical models

Boiler specification

The tangentially fired pulverized-coal boiler considered in this research is a 660 MWe ultra-supercritical boiler as schematically shown in Figure 1. The height of the furnace exit is approximately 50.4 m, and the cross-section of the combustion zone has a width of 19.68 m and a depth of 19.23 m. The low NO_x concentric firing system is applied in the primary coal burner, as also presented in Figure 1. In all, 20 coal burner sets are installed at the four corners ranging from the lower group A to the upper group E. On the other hand, the air staging is made up of secondary air, OFA, and additional air (AA). Emerging from Figure 2(c) is that every four secondary air burner sets are installed at the same height with the corresponding lean/rich coal burner sets. Apparently, the tangential circle formed by the secondary air is five times larger than that formed by the coal burners. Besides, OFA ports and four groups of AA ports are placed at the corners above the group E coal burners in sequence. Then, the platen superheater (SH), the secondary SH, the final SH, and the final re-heater are installed horizontally in the crossover pass upon the furnace arch. The primary SH and re-heater are placed side by side in the rear pass zone. The furnace walls from the hopper inlet to the furnace exit are vertical water walls, which are separated from the SHs by the steam-water separator.

Figure 1.

(a) Grid scheme at furnace cross-section and (b) grid independency tests.

Figure 2.

(a) Schematic configurations of the tangentially fired pulverized-coal boiler, burner sets distribution on, (b) the longitudinal section and (c) the cross section.

Domain and mesh system

The whole computational domain includes the furnace, crossover pass, and rear pass, as shown in Figure 1. In order to improve the mesh quality and decrease the mesh quantity, the grid domain is divided into six parts, including the downer combustion zone, combustion zone, upper combustion zone, platen SH zone, crossover pass, and rear pass. The grid of the combustion zone is to be refined because it has great influence on the calculation of the coal combustion. Due to the special installment of the burner sets applied in this boiler, the tetrahedral cells are used here to reduce the pseudo-diffusion in the combustion zone, as shown in Figure 2(a). Besides, the downer combustion zone and the upper combustion zone are discretized by the structured hexahedral and wedge cells for saving calculation resources. Then, the tetrahedral cells are used in the platen SH zone, crossover pass, and rear pass because of the complex structure. Grid independency tests are carried out among three grid systems of 1.5 million, 3 million, and 5 million, which is mainly resulted from the three different mesh scales of the furnace zone. Obviously, emerging from Figure 2(b), the grid system of 3 million is suitable in this simulation compared with the other two systems.

Numerical models

Turbulent, discrete phase, and radiation models

The numerical simulation in this study is supported using a commercial software package, Ansys Fluent 16.0. The adopted mathematical modeling for pulverized-coal boiler has been verified by many experts all over the world.^3–7

The species transport model is applied on the continuous phase equations. The realizable k-ε model is employed to consider the effect of swirling turbulent flow in the furnace due to its superiority than the standard k-ε model.^20,21 The continuum and momentum equations are written as follows

Continuum equation : \frac{\partial u_{i}}{\partial x_{i}} = 0

(1)

Momentum equation : \frac{\partial u_{i}}{\partial t} = - u_{j} \frac{\partial u_{i}}{\partial x_{j}} - \frac{1}{ρ} \frac{\partial p}{\partial x_{i}} + ν \frac{\partial^{2} u_{i}}{\partial x_{j} \partial x_{j}} + f_{p}

(2)

\begin{matrix} Realizable k - ε model : \\ \frac{\partial (ρ k)}{\partial t} + \frac{\partial (ρ u_{i} k)}{\partial x_{i}} = \frac{\partial}{\partial x_{j}} ((μ + \frac{μ_{t}}{σ_{k}}) \frac{\partial k}{\partial x_{j}}) + G_{k} - ρ ε \\ \frac{\partial (ρ ε)}{\partial t} + \frac{\partial (ρ u_{i} ε)}{\partial x_{i}} = \frac{\partial}{\partial x_{j}} ((μ + \frac{μ_{t}}{σ_{ε}}) \frac{\partial ε}{\partial x_{j}}) \\ + ρ C_{1} S ε - ρ C_{2} \frac{ε^{2}}{k + \sqrt{ν ε}} \end{matrix}

(3)

where $C_{1} = max (0.43, (η / η + 5))$ , $η = (2 S_{ij} \cdot S_{ij})^{1 / 2} (k / ε)$ , $C_{2} = 1.9$ , $σ_{k} = 1$ , $σ_{ε} = 1.2$ , $C_{μ} = (1 / (A_{0} + A_{S} U^{*} k / ε))$ , $A_{0} = 4.0$ , $A_{S} = \sqrt{6} \cos ϕ, ϕ = (1 / 3) \cos^{- 1} (\sqrt{6} W)$ , $W = (S_{ij} S_{jk} S_{kj}) / ((S_{ij} S_{ij})^{1 / 2})$ , $S_{ij} = (1 / 2) ((\partial u_{i} / \partial x_{j}) + (\partial u_{j} / \partial x_{i}))$ , $U^{*} = \sqrt{S_{ij} S_{ij} + {\tilde{Ω}}_{ij} {\tilde{Ω}}_{ij}}$ , ${\tilde{Ω}}_{ij} = Ω_{ij} - 2 ε_{ijk} ω_{k}$ , and $Ω_{ij} = {\bar{Ω}}_{ij} - ε_{ijk} ω_{k}$ . In these equations, G_k represents the generation of turbulence kinetic energy due to the mean velocity gradients; σ_k and σ_ε are the turbulent Prandtl numbers for k and ε, respectively.

The trajectories of the pulverized-coal particles are calculated under the Lagrangian framework. The discrete phase model (DPM) is employed to consider the mass, momentum, and heat exchange between the discrete particles and turbulent gas. The particle motion equations are written as follows

\frac{d u_{p}}{dt} = F_{D} (u - u_{p}) + \frac{g_{i} (ρ_{p} - ρ)}{ρ_{p}}, F_{D} = \frac{18 μ}{ρ_{p} d_{p}^{2}} \frac{C_{D} Re}{24}

(4)

The P-1 radiation model is selected for solving the radiation between boiler furnace wall and coal particle surface.²⁰ The P-1 radiation model is based on the assumption that the expansion of radiation intensity into an orthogonal series of spherical harmonics²² is suitable for the complex combustion facilities and saving computational resource. The energy equation can be described as

q_{r, w} = - \frac{ε_{w}}{2 (2 - ε_{w})} (4 n^{2} σ T_{w}^{4} - G_{w})

(5)

where ε_w is the emissivity of the wall.

The incompressible ideal gas, mass-weighted mixing law, and weighted-sum-of-gray-gases model²³ are chosen to define the density, viscosity, and absorption coefficient of gas phase mixture, respectively.

Coal combustion model

The coal combustion process consists of three stages: dehydration, devolatilization, and char combustion. The first stage is ignored because the coal particles have been heated to 368 K for dehydration before being carried into boilers. The two-competing-rates model is selected to predict the devolatilization process.²⁴ The devolatilization is controlled by two competing rates, R₁ and R₂, over different particle temperatures T₁ and T₂. The expression for the mass and energy of the coal particles are shown as

\frac{m_{V} (t)}{(1 - f_{W, 0}) m_{a}} = \int_{0}^{t} (α_{1} R_{1} + α_{2} R_{2}) \exp (- \int_{0}^{t} (R_{1} + R_{2}) dt) dt

(6)

where $R_{1} = A_{1} e^{- (E_{1} / R T_{P})}, R_{2} = A_{2} e^{- (E_{2} / R T_{P})}$

m_{p} c_{p} \frac{d T_{p}}{dt} = h A_{p} (T_{\infty} - T_{p}) + \frac{d m_{p}}{dt} h_{fg} + A_{p} ε_{p} (θ_{R}^{4} - T_{p}^{4})

(7)

In this equation, the f_W,0 is the initial moisture mass fraction, R₁ and R₂ are the devolatilization rate of the two steps, and h_fg is the latent heat of vaporization.

The kinetics/diffusion-limited model is applied to consider the char combustion,²⁵ which is based on the assumption that the surface reaction rate is controlled either by the kinetic reaction rate or by the diffusion rate.^26,27 The diffusion rate coefficient D₀ and the kinetic reaction rate R are applied to achieve the char combustion rate

\frac{d m_{p}}{dt} = - A_{p} p_{OX} \frac{D_{0} R}{D_{0} + R}

(8)

where $D_{0} = C_{1} ([(T_{p} + T_{\infty}) / 2]^{0.75} / (d_{p}))$ , $R = C_{2} e^{- (E / R T_{P})}$ , p_OX is the partial pressure of oxidizer around the particle during combustion, and A_p is the surface area of particles.

The non-premixed model is applied in this research, where a mixture-fraction equation and chemical equilibrium are proposed to compute individual species. The interaction between the turbulence and chemistry is calculated using the probability density function (PDF).^28,29

NO_x model

NO_x model is calculated as a post-processing model based on the assumption that NO_x concentration is too low to impact the coal combustion in the furnace. NO_x is formed mainly by thermal NO_x, fuel NO_x, and prompt NO_x formation mechanisms, where the prompt NO_x is so minimal and thus ignored in this research.

Thermal NO_x is formed when nitrogen and oxygen, within the combustion air, combine at a relatively high temperature in fuel-lean environments. Here, the formation of thermal NO_x is modeled by the extended Zeldovich mechanism, which consists of three fundamental reactions ((O + N₂, N + O₂, and N + OH). The partial equilibrium model is used to calculate the concentrations of O, H, and OH.

Fuel NO_x is formed when the nitrogen, coming from both volatile and char, combines with the excess oxygen of the combustion air. The ratio of nitrogen from the volatiles to nitrogen from char is assumed as 7:3. The HCN and NH₃ are considered the intermediate species in the fuel NO_x mechanism. All char-N is transformed into HCN, while 90% of volatile-N is converted to HCN, and the rest is released as NH₃. The intermediates, HCN and NH₃, react to form either NO in fuel-lean regions or N₂ in fuel-rich regions according to DeSoete.³⁰

Simulation parameters

This study is based on the practical boiler maximum continuous rating (BMCR) operating condition. The total mass flow rate of pulverized coal is 274.9 t/h at 368 K, while the total mass flow rate of air is 2239.2 kg/h, consisting of primary air (24%) at 368 K, secondary air (50%) at 573 K, OFA (8%) at 573 K, and AA (18%) at 573 K. As mentioned above, the influence of air staging on coal combustion is investigated in this article. The pulverized coal used in this boiler is mine in Datong, China. Table 1 shows the main parameters of this coal, including the proximate analysis, ultimate analysis, and diameter distribution. The coal composition is 22.16% volatile matter, 50.29% fixed carbon, and 29.55% ash in terms of the as-received basis. The pulverized coal’s density is 1300 kg/m³ and the mean diameter is 50 µm. The Rosin–Rammler distribution model with a spread parameter of 1.189 is selected for the coal size distribution. The crossover pass and rear pass are considered as porous media³¹ because these heater transfer tubes are orderly and closely arranged in these zones. Meanwhile, appropriate viscous resistance and inertial resistance are used to affect the flue flow. The walls of the boiler are set as no-slip and stationary wall. The wall temperatures are based on the temperature of water tubes. The out pressure is set as −50 Pa, because the negative pressure occurs during coal combustion.

Table 1.

Main parameters of pulverized coal.

Proximate analysis (%, ar)	Ash	29.55
	Volatile	22.16
	Fixed carbon	40.29
	Moisture	8
Ultimate analysis (%, ar)	C	51.92
	H	3.53
	O	5.04
	N	1.1
	S	0.86
LHV (ar)	Qar,net (KJ/kg)	20,258
Particle diameter distribution	Min/Max diameter (µm)	6/192
	Mean diameter (µm)	50
	Rosin–Rammler spread	1.189

LHV: lower heating value.

The purpose of this work is to study the influences of the applications of OFA and AA on the performance of this unconventional 660 MWe boiler. Hence, three main cases named “None,”“OFA Only,” and “OFA + AA” are designed as listed in Table 2. Besides, the cases of “OFA + AA” are divided into two parts: Case I is to study the influence of the secondary air ratio, and Case II is to research the effect of the ratio of AA to OFA. As shown, all the cases have a same excess air ratio (α = 1.24) and primary air ratio (γ_pri = 24%). Just as the names imply, the case of “None” only consists of 24% primary air and 76% secondary air, while the case of “OFA Only” has 24% primary air, 68% secondary air, and 8% OFA. With the same γ_AA/γ_OFA, the secondary air ratio of Case I changes from 0.48 to 0.528. Similarly, the ratio of AA to OFA increases from 2.0 to 2.4 with the same secondary air ratio. Obviously, the air mass flux ratio depends on the corresponding gas velocity.

Table 2.

Main parameters for different simulation conditions.

Case			Excess air ratio (α)	Primary air ratio (γ_pri/%)	Secondary air ratio (γ_2nd/%)	OFA (γ_2nd/%)	AA (γ_2nd/%)	Ratio of AA to OFA(γ_AA/γ_OFA)
None			1.24	24	76
OFA Only			1.24	24	68	8.00
OFA + AA	I	1	1.24	24	48.0	9.03	18.97	2.1
		2	1.24	24	49.2	8.64	18.16	2.1
		3	1.24	24	50.4	8.26	17.34	2.1
		4	1.24	24	51.6	7.87	16.53	2.1
		5	1.24	24	52.8	7.48	15.72	2.1
	II	1	1.24	24	50.4	8.53	17.07	2
		2	1.24	24	50.4	8.26	17.34	2.1
		3	1.24	24	50.4	8.00	17.6	2.2
		4	1.24	24	50.4	7.76	17.84	2.3
		5	1.24	24	50.4	7.53	18.07	2.4

Results and discussion

Simulation results are presented and discussed in this section. First, the combustion simulation results for the cases with both OFA and AA are validated by the designed/measured values at the furnace exit. Then, the comparisons among the three main cases, “None,”“OFA Only,” and “OFA + AA,” are demonstrated to show the advantages of the new application, OFA and AA. Last but not least, the influence of the air staging technology on the NO_x emissions is displayed in section “NO_x emissions.”

Validation of the combustion simulation results

The results predicted by the numerical simulations are compared with the designed/measured values as presented in Table 3. The designed and predicted temperatures of the standard condition are 1238 and 1197 K, which shows a good agreement. The predicted concentrations of O₂ and CO₂ at the boiler exit are 1.058% and 16.17%, respectively, while the measured concentrations of O₂ and CO₂ at the boiler exit are 1.3% and 16.68%, respectively. As shown, the errors of temperature, O₂ and CO₂ are all within the simulation limits, which confirm that these above-selected models are reasonable for simulating the performance of the unconventional boiler.

Table 3.

Comparison between the measurements and prediction.

		Measurement/design	Prediction (OFA + AA/Case I-3)
Temperature (K)	Furnace exit	1238	1197
O₂ (%, mole fraction)	Boiler exit	1.3	1.058
CO₂ (%, mole fraction)	Boiler exit	16.68	16.17

Gas flow field and particle trajectory

The novel tangentially fired pulverized-coal boiler considered in this research consists of five groups of burner sets and each group of burner set could be divided into two layers. One layer consists of four rich pulverized-coal burners and four secondary air burners, and the other layer consists of four lean pulverized-coal burners and four secondary air burners. With no doubt, the flow field inside the furnace plays an important role in coal combustion and NO_x emissions.

The velocity distribution and vectors at horizontal cross-sections along furnace height are shown in Figure 3. A symmetrical anticlockwise swirling flow is found in the center of the furnace, as expected. Emerging from the velocity vectors (Figure 2(b)–(f)), the swirling flow gets stronger along the furnace height from groups A to E burner set. Then, the swirling intensity decreases remarkably due to the concentric counter-tangential flow of OFA and AA jets. Apparently, the performance of the unusual burner sets has a great coincidence with the flow field of the traditional tangential firing coal boiler.^5,32,33

Figure 3.

Velocity distribution and velocity vectors at horizontal cross-sections.

Figure 4 depicts the trajectories and velocities of the coal particles injected from one corner. Obviously, the particles are injected from the primary air burners and swirled by the combustion air in the furnace. The particles are accelerated at the inject air and moderated during the coal combustion, which is benefit to the coal complete combustion and the mixing of the fuel and oxygen.

Figure 4.

Trajectories and velocities of coal particle injected from one corner burners.

Temperature distribution

The temperature distributions of the three cases, “None,”“OFA Only,” and “OFA + AA,” are discussed in this section. Figure 5(a) presents the temperature distributions at several horizontal cross-sections in the combustion zone and Figure 5(b) shows the mass-weighted average temperature of cross-sections along the furnace height. Emerging from Figure 5(a) is that the flue gas temperature is getting higher from the boundary to the center with the mixing of fuel and air, and reaches higher along the furnace height. This phenomenon has been found in many other tangential firing boilers,^3,11,34 which indicates the current novel application is reasonable in pulverized-coal boilers.

Figure 5.

(a) Temperature distributions of case “OFA + AA” and (b) average temperature at horizontal cross-section along the furnace height for three cases.

Figure 5(b) shows the mass-weighted average temperature at horizontal cross-sections along the height in the combustion zone. For all the cases, it can be found that the temperature rises sharply first in the combustion zone, reaches the peak around the OFA region, and then decreases slowly along the furnace height. Also, the maximum temperature within both OFA and AA is a little lower than the other two cases, “None” and “OFA Only.” Compared with the traditional tangential firing ones,^11,34 it can be found that the highest average temperature 1758 K is a little lower, which is benefit to the NO_x formations. Obviously, with the application of OFA and AA, less secondary air is conveyed into the combustion zone, which leads to the lean-oxygen combustion and high residence of the flue gas. Meanwhile, the pulverized coals are still fired completely in the combustion zone. Hence, the flue gas temperature at the furnace exit is much higher with OFA and AA, which will contribute to the heat transfer of the SHs and re-heaters. From the comparison of the three temperature distributions, the air staging applied in this novel tangentially fired boiler is helpful for the pulverized-coal combustion.

Species components

Burnout and devolatilization

The coal burning degree depends on the coal burnout and devolatilization. Figure 6 presents the average fraction of coal burnout and devolatilization at horizontal cross-sections along the furnace height. As expected, the coal devolatilization fraction shows the contrary performance to coal burnout fraction. It is obvious to find the coal burnout and devolatilization fluctuate only in the burner sets region and be close to steady around the OFA region, even with the new coming OFA or AA. Compared with the cases without OFA or AA, the coal particles devolatilize more fully with the application of OFA and AA, which results from the higher temperature of flue gas and long residence time of coal particles. The coal burnout curves indicate that the char combustion is few influenced by the air staging. Meanwhile, the coal devolatilization decreases with more secondary air injection. Hence, it comes that the OFA and AA are helpful for the coal combustion within the air staging limits.

Figure 6.

Average fraction of coal burnout and devolatilization at horizontal cross-sections along the furnace height.

O₂, CO₂, and CO distribution

The concentration distributions of the main species, including O₂, CO₂, and CO, are used to analyze the difference among the three main cases. The mole fraction distributions of O₂, CO₂, and CO at the furnace diagonal section for the case “OFA + AA” are depicted in Figure 7(a)–(c), respectively. Apparently, the oxygen concentration decreases with the coal combustion and the carbon dioxide mole fraction increases conversely. With the application of OFA and AA, the lean-oxygen combustion happens in the combustion zone, which causes the carbon monoxide formation near burner sets region. Furthermore, the mass-weighted average mole fraction of species components at horizontal cross-sections along the furnace height is illustrated in Figure 7(d). For the case with OFA and AA applications, less secondary air is injected into the furnace; subsequently, more CO is formed due to the lean-oxygen combustion and more CO₂ is formed because of the more residence time of coal particles. Meanwhile, the fluctuations of CO₂ and O₂ concentration are much gentler comparing with the cases without OFA or AA. However, comparing with the other two cases, it is found that the concentration of CO₂ at the furnace exit is a little lower and the O₂ concentration is a little higher, which means the coal burnout fraction decreases if the secondary air ratio is low enough. From the discussion, it comes that the coal combustion in the furnace becomes steadier with the application of OFA and AA. On the other hand, the ratios of OFA and AA have to be controlled within the limits so that the pulverized coal can be burn completely, which is necessary for coal boilers.

Figure 7.

Mole fraction distribution of (a) O₂, (b) CO₂, and (c) CO at the furnace diagonal section, and (d) average mole fraction at horizontal cross-sections along the furnace height.

NO_x emissions

NO_x emission is an important parameter to measure the quality of coal combustion. As mentioned above, many similar researches have been conducted for low NO_x combustion. In this study, both the thermal and fuel mechanisms are taken into account for the NO_x formation, while prompt mechanism is neglected due to its insignificant amounts. The aim is to investigate the effect of secondary air, OFA, and AA on the NO_x emission.

The iso-surface (400 ppm) of NO_x concentration for three cases, “None,”“OFA Only,” and “OFA + AA,” is displayed in Figure 8(a)–(c), respectively. The volume concentration is calculated by the mass concentration, $2.24 \times 10^{6} (C_{N O_{x}} / M) \times ((273 + T) / 273) \times (101, 325 / p)$ . As is known that, the NO_x formation is strongly influenced by oxygen concentration and flue gas temperature. It is obvious to find that the nitrogen oxides occur in the near burner sets region due to the combination of the incoming oxygen and the rapid release nitrogen from fuels, and then the NO_x gathers swirls with the coal combustion. The NO_x concentration for “OFA + AA” at the furnace exit is much lower compared with the other two cases. For the case without OFA and AA (Figure 8(a)), more secondary air is conveyed inside the furnace, much more nitrogen oxides are formed due to the rich-oxygen combustion and high temperature. Furthermore, the NO_x is mal-distribution in the crossover pass and rear pass because of the large rotation of gas flue. Emerging from the NO_x distributions in Figure 8(a)–(c) is that with the application of OFA and AA, the nitrogen oxides are less formed during the coal combustion and are uniformly distributed in the flue gas. So, it is worth to emphasize that this novel air staging technology applied in this boiler has great advantages for reducing NO_x emissions.

Figure 8.

Iso-surface (400 ppm) of NO_x concentration for three cases: (a) None, (b) OFA Only, and (c) OFA + AA.

The average NO_x concentrations at horizontal cross-sections along the furnace height are illustrated in Figure 9. As mentioned above, the NO_x concentrations reach peak in the combustion zone for all the cases. Compared with the NO_x concentration curves of the two typical cases, “None” and “OFA Only,” it is easy to find that the average NO_x concentration is few influenced by the amounts of secondary air when the oxygen is enough in near burner sets region. However, with the application of OFA, the NO_x emission reduces close to 17% at the furnace exit. What is more, the effects of the air distributions on the NO_x emissions are discussed. For the cases with both OFA and AA, the NO_x concentration reduces sharply in the combustion zone because of the lean-oxygen combustion and long residence time of coal particles. Thus, a little nitrogen oxide is formed due to the combination of the flue gas and the new coming oxygen around OFA and AA region. For both cases with OFA and AA, the NO_x concentrations at the furnace exit are approximately 21% loss by comparing with the case “OFA Only.” Besides, as shown in Figure 9(a), when the secondary air ratio (γ_2nd) decreases from 0.528 to 0.480, the NO emissions at the furnace exit decrease from 262 to 244 ppm. Besides the effect of the secondary air ratio, the ratio of the AA to OFA is also an important parameter for reducing nitrogen oxides. Apparently, with the same primary air and secondary air ratios, the nitrogen formations are slowly attenuated when decreasing air ratio. From above discussion, the results come that, first, based on the steady coal combustion, lower NO_x emissions could be achieved when decreasing the secondary air within limits. Second, the application of AA has great advantages for both coal combustions and NO_x emission; however, it could reduce NO_x emissions by increasing the OFA and decreasing the AA within small limits.

Figure 9.

Average NO_x concentration at horizontal cross-sections along the furnace height for (a) different secondary air ratios and (b) different ratios of AA to OFA.

Conclusion

Based on the Eulerian–Lagrangian approach, a three-dimensional numerical simulation was carried out to study the pulverized-coal combustion process on a novel tangential firing ultra-supercritical boiler. The characteristics of the flow field, particle trajectory, temperature distribution, species components, and NO_x emissions were numerically investigated. The good agreements with the predicted results and the measurements imply that the adopted simulation models are appropriate to predict the coal combustion and NO_x emissions. From the comparisons of these cases, the most notable findings are listed as follows:

An ideal turbulent flow can be observed in this unconventional pulverized-coal furnace. The coal particles are accelerated and mixed with the injecting combustion air. The turbulent intensity grows first and then decreases sharply along the height and the swirling intensity decreases remarkably due to the concentric counter-tangential flow of OFA and AA injections.

The temperature of flue gas in the combustion zone decreases a little with the application of OFA and AA, which is benefit to the NO_x reduction. Meanwhile, the temperature at the furnace exit is getting higher for better heat transfer of the SHs and re-heaters.

As expected, with the injection of OFA and AA, less secondary air is conveyed into the furnace, which causes more carbon monoxides. Besides, the fluctuations of the CO₂ and O₂ concentration are gentler, which indicates that more steady coal combustion is achieved with the novel air staging.

NO_x formations could be reduced sharply when applying the OFA and AA. Furthermore, the NO_x emissions reduce when decreasing the secondary air ratio (γ_2nd) and the ratio of AA to OFA (γ_AA/γ_OFA) within the limits.

Footnotes

Appendix 1

Academic Editor: Ibrahim Abdalla

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was financially supported by NSFC (no. 51390492 and 51325601) and the Jiangsu Natural Science Funds for Distinguished Young Scholar (no. BK20130022).

References

Okiura

Akiyama

Terada

. Combustion process for reducing nitrogen oxides. Patent 4,403,941, USA, 13 September 1983.

Breen

Hura

HS.

Method and apparatus for NO_x reduction in flue gases. Patent 6,258,336, USA, 10 July 2001.

Yin

Caillat

Harion

. Investigation of the flow, combustion, heat-transfer and emissions from a 609 MW utility tangentially fired pulverized-coal boiler. Fuel 2002; 81: 997–1006.

Fan

Liang

. Numerical simulation of the flow and combustion processes in a three-dimensional, W-shaped boiler furnace. Energy 1997; 22: 847–857.

Fan

Qian

. Computational modeling of pulverized coal combustion processes in tangentially fired furnaces. Chem Eng J 2001; 81: 261–269.

Zhang

Zhou

Sun

. Numerical investigation of low NO_x combustion strategies in tangentially-fired coal boilers. Fuel 2015; 142: 215–221.

Cao

Sun

Liu

. Computational fluid dynamics modeling of NO_x reduction mechanism in oxy-fuel combustion. Energ Fuel 2010; 24: 131–135.

Vuthaluru

HB.

Modelling of a wall fired furnace for different operating conditions using FLUENT. Fuel Process Technol 2006; 87: 633–639.

Man

Gibbins

Witkamp

. Coal characterisation for NO_x prediction in air-staged combustion of pulverised coals. Fuel 2005; 84: 2190–2195.

10.

Park

Baek

Kim

. Numerical and experimental investigations on the gas temperature deviation in a large scale, advanced low NO_x, tangentially fired pulverized coal boiler. Fuel 2013; 104: 641–646.

11.

Choi

Kim

CN.

Numerical investigation on the flow, combustion and NO_x emission characteristics in a 500 MW_e tangentially fired pulverized-coal boiler. Fuel 2009; 88: 1720–1731.

12.

Houshfar

Skreiberg

Løvås

. Effect of excess air ratio and temperature on NO_x emission from grate combustion of biomass in the staged air combustion scenario. Energ Fuel 2011; 25: 4643–4654.

13.

Hui

. NO_x emission and thermal efficiency of a 300 MW utility boiler retrofitted by air staging. Appl Energ 2009; 86: 1797–1803.

14.

Zhou

Cen

Fan

Modeling and optimization of the NO_x emission characteristics of a tangentially fired boiler with artificial neural networks. Energy 2004; 29: 167–183.

15.

Zhan

Liu

Fang

. Effect of horizontal staging of secondary air on NOx emissions in a 600 MW tangentially coal-fired utility boiler. In: Proceedings of the 2013 international conference on materials for renewable energy and environment (ICMREE), Chengdu, China, 19–21 August 2014, vol. 3, pp.745–749. New York: IEEE.

16.

Khanafer

Aithal

SM.

Fluid-dynamic and NO_x computation in swirl burners. Int J Heat Mass Tran 2011; 54: 5030–5038.

17.

Jia

Investigation on combustion characteristics and NO formation of methane with swirling and non-swirling high temperature air. J Therm Sci 2014; 23: 472–479.

18.

Xue

Hui

Liu

. Experimental investigation on NO_x emission and carbon burnout from a radially biased pulverized coal whirl burner. Fuel Process Technol 2009; 90: 1142–1147.

19.

Sun

Che

Chi

ZH.

Effects of secondary air on flow, combustion, and NO_x emission from a novel pulverized coal burner for industrial boilers. Energ Fuel 2012; 26: 6640–6650.

20.

Choudhury

Introduction to the renormalization group method and turbulence modeling (Technical memorandum TM-107). Lebanon, NH: Fluent Incorporated, 1993.

21.

Launder

Spalding

DB.

Lectures in mathematical models of turbulence. London: Academic Press, 1972.

22.

Cheng

Two-dimensional radiating gas flow by a moment method. AIAA J 1964; 2: 1662–1664.

23.

Raithby

Chui

A finite-volume method for predicting a radiant heat transfer in enclosures with participating media. J Heat Trans: T ASME 1990; 112: 415–423.

24.

Kobayashi

Howard

Sarofim

AF.

Coal devolatilization at high temperatures. Symp Combust Proc 1977; 16: 411–425.

25.

Saxena

SC.

Devolatilization and combustion characteristics of coal particles. Prog Energ Combust 1990; 16: 55–94.

26.

Baum

Street

PJ.

Predicting the combustion behaviour of coal particles. Combust Sci Technol 1971; 3: 231–243.

27.

Field

MA.

Rate of combustion of size-graded fractions of char from a low rank coal between 1200 K–2000 K. Combust Flame 1969; 13: 237–252.

28.

Peters

AAF

Weber

. Mathematical modeling of a 2.25 MW_t swirling natural gas flame. Part 1: eddy break-up concept for turbulent combustion; probability density function approach for nitric oxide formation. Combust Sci Technol 1995; 110: 67–101.

29.

Smoot

Smith

PJ.

Coal combustion and gasification. New York: Springer Science & Business Media, 2013.

30.

De Soete

GG.

Overall reaction rates of NO and N2 formation from fuel nitrogen. Symp Combust Proc 1975; 15: 1093–1102.

31.

Tao

Numerical simulations of SNCR in a T-fired furnace. Beijing, China: North China Electric Power University, 2004.

32.

Zhou

Hui

. Experimental and numerical study on the flow fields in upper furnace for large scale tangentially fired boilers. Appl Therm Eng 2009; 29: 732–739.

33.

Asotani

Yamashita

Tominaga

. Prediction of ignition behavior in a tangentially fired pulverized coal boiler using CFD. Fuel 2008; 87: 482–490.

34.

Zhou

Lou

Cheng

. Experimental investigations on visualization of three-dimensional temperature distributions in a large-scale pulverized-coal-fired boiler furnace. P Combust Inst 2005; 30: 1699–1706.

Numerical investigation on the flow,combustion,and NO X emission characteristics in a 660 MWe tangential firing ultra-supercritical boiler

Abstract

Keywords

Introduction

Physical and mathematical models

Boiler specification

Domain and mesh system

Numerical models

Turbulent, discrete phase, and radiation models

Coal combustion model

NOx model

Simulation parameters

Results and discussion

Validation of the combustion simulation results

Gas flow field and particle trajectory

Temperature distribution

Species components

Burnout and devolatilization

O2, CO2, and CO distribution

NOx emissions

Conclusion

Footnotes

Appendix 1

Declaration of conflicting interests

Funding

References

NO_x model

O₂, CO₂, and CO distribution

NO_x emissions