Keywords

1 Introduction

With the slowdown of economic growth and the rapid adjustment of the economic structure, China’s overall electricity demand has declined. The contradiction between power supply and demand has shifted from electricity shortage to relative surplus. Meanwhile, to face the increasingly severe environmental regulations, China continues to promote energy structure adjustment and develop the clean energy rapidly. These all lead to the decline of our country’s installed capacity of thermal power units while moving toward the development of large-capacity and high-parameter double-reheat unit (Tumanovskii and Shvarts 2017; Chang and Zhuo 2016). Besides, conditions of medium and low loads would become more universal during the unit operations due to the abundant hydropower or low power demand in late nights (Zhao and Fang 2017). However, most of the units lack experiences of long-term operation under medium and low loads, which makes the efficiency of the units decrease and the coal consumption increase. Therefore, it is necessary to carry out investigation on energy-saving operation optimization of large-capacity and high-parameter double-reheat units under conditions of wide loads to improve the efficiency and reduce pollutant emissions and power consumption.

Ultra-supercritical double-reheat units with high efficiency, high parameter and large capacity have become a main development tendency on the coal-fired power plants. But double-reheat units are still in the initial development stage in China. Several researches have analyzed the technical (Huo and Shi 2014) and economic characteristics (Jiang 2000; Yao and Zhu 2017), heat load distribution (Zhang and Yan 2016) and pollutant control (Guo and Zhang 2011) during the operation of double-reheat units. However, the sudies on ultra-supercritical double-reheat units are still limited.

Numerical simulation, which has been widely applied to investigate the combustion and pollutant formation process in pulverized coal fired boiler in the past few decades (Zhou and Mo 2011; Liu and Chen 2015), is more economical and convenient than experimental study. In this study, the effects of different loads and excess air ratios on combustion characteristics and NOx emission of an ultra-supercritical double-reheat boiler were studied. In NOx calculation, the revised NOx model is used to replace the default model to accurately predict the NOx emission. This work can provide reference for optimization of combustion organization for double-reheat tower boiler.

2 Boiler Decription

The simulations were based on a 660 MW double reheat tower boiler with bituminous coal, with the schematic of boiler geometry illustrates in Fig. 20.1a. The width, depth and height of the boiler are 18,150, 18,150 and 97,500 mm, respectively. Figure 20.1b shows the arrangement of nozzles in each corners. It includes the primary air (PA), secondary air (SA) and separate over fire air (SOFA). In order to control the NOx generation by air staging, two stage SOFA nozzles are arranged above the burners. Figure 20.1c presents the air injecting direction.

Fig. 20.1
figure 1

Schematic configuration and burner nozzle distribution of the tangentially pulverized coal fired boiler

Full size image

3 Modeling Methodology

3.1 Grid Independence

In this paper, a structured grid is used to simulate the coal combustion. Because of the intense combustion reaction in the burner area, the grid is refined locally to receive more accurate results in this area. The gird independence was verified with three different mesh systems, i.e., 1,565,078, 1,880,167 and 2,241,660 cells. Figure 20.2 decipts the temperature distribution in the three different mesh systems. By comparing the simulation results, the mesh with 1,880,167 cells is chosen to calculate.

Fig. 20.2
figure 2

Grid independence

Full size image

3.2 Numerical Models

The combustion process of pulverized coal furnace includes a series of complex physical and chemical processes (Hu and Liu 2013). The standard \(k - \varepsilon\) model was chosen in the present study. The radiation heat transfer was calculated by P-1 model (Drosatos and Nikolopoulos 2015). The WSGGM-domain based model was used to calculate the gas absorption coefficient. The volatile combustion is described by the two-step global reaction mechanism as follow. Coal particle trajectories were calculated using the Lagrangian approach.

$$ {\text{CH}}_{y} {\text{O}}_{x} + (0.5 + 0.25y - 0.5x){\text{O}}_{2} \to {\text{CO}} + 0.5y{\text{H}}_{{2}} {\text{O}} $$
(20.1)
$$ {\text{CO}} + {0}{\text{.5O}}_{2} \to {\text{CO}}_{{2}} $$
(20.2)

The NO is generally discussed when investigating NOx formation in coal-fired boilers. That’s because the NO accounts for 95% of total NOx. Three types of NOx are generated, named as prompt NOx, thermal NOx and fuel NOx. In large pulverized coal combustion system, the prompt NOx production is little, which is generally neglecte (Panagiotis and Athanasios 2016; Álvarez and Yin 2013). Thus, only thermal NOx and fuel NOx are taken into consideration in the calculation process. NOx generation is calculated by post-processing method (Park 2013).

The generation and reduction of thermal NOx are determined by the extended Zeldovich mechanism as follows:

$$ {\text{N}}_{{\text{2}}} + {\text{O}}\mathop {\mathop \rightleftarrows \limits^{{k_{1} }} }\limits_{{k_{{ - 1}} }}{\text{N}} + {\text{NO}} $$
(20.3)
$$ {\text{N}} + {\text{O}}_{2} \mathop {\mathop \rightleftarrows \limits^{{k_{2} }} }\limits_{{k_{{ - 2}} }}{\text{O}} + {\text{NO}} $$
(20.4)
$$ {\text{N}} + {\text{OH}}\mathop {\mathop \rightleftarrows \limits^{{k_{3} }} }\limits_{{k_{{ - 3}} }}{\text{H}} + {\text{NO}} $$
(20.5)

where k1, k2, k3 present the forward reaction rate constant, while k-1, k-2, k-3 present the reverse reaction rate constant in the Eqs. (20.3–20.5). The six reaction rate constants are determined by the experimental results acquired by Hanson. Thus, the thermal NO formation rate can be described as the Eqs. (20.6). Besides, the concentrations of O, H, and OH in the Eqs. (20.3–20.6) are computed by the partial equilibrium approach.

$$ \begin{gathered} \frac{{d[{\text{NO}}]}}{dt}{ = }k_{1} {\text{[O][N}}_{2} {]} + k_{2} {\text{[N][O}}_{2} {]} + k_{3} {\text{[N][OH]}} \hfill \\ \, - k_{ - 1} {\text{[NO][N]}} - k_{ - 2} {\text{[NO][O]}} - k_{ - 3} {\text{[NO][H]}} \hfill \\ \end{gathered} $$
(20.6)

Fuel NOx is usually released in the form of volatile nitrogen and char nitrogen. Char nitrogen is converted to NO directly. During the transformation process from volatile nitrogen to the fuel NOx, volatile nitrogen is first converted to HCN and NH3, and then oxidation and reduction reactions occur between HCN, NH3 and NO, O2, depicted as the Eqs. (20.7–20.10).

$$ {\text{HCN}} + {\text{O}}_{2} \to {\text{NO}} + \cdots $$
(20.7)
$$ {\text{HCN}} + {\text{NO}} \to {\text{N}}_{2} + \cdots $$
(20.8)
$$ {\text{NH}}_{3} + {\text{O}}_{2} \to {\text{NO}} + \cdots $$
(20.9)
$$ {\text{NH}}_{3} + {\text{NO}} \to {\text{N}}_{2} + \cdots $$
(20.10)

In this study, the User Defined Functions (UDF) was used to substitute the default models for NOx calculation, which has been verified by Zhang and Ito (2013) and Zha and Li (2017). Surface reaction of NOx reduction by char nitrogen was calculated according to Levy and Chana (1981) mechanism.

The properties of the bituminous coal are presented in Table 20.1. The basic simulation case is based on the design case of 100% turbine heat acceptance (THA) operation. The effect of different unit loads and excess air ratios on the characteristics of NOx emission are considered to study in the work. The three loads are 100% THA, 70% THA and 50% THA with the total excess air ratio of 1.15. The three excess air ratios are 1.15, 1.20 and 1.25 with 100% THA load. All detailed simulation cases are listed in Table 20.2.

Table 20.1 Basic information of the bituminous coal
Full size table
Table 20.2 Simulation cases of the 660 MW tower boiler
Full size table

4 Results and Disscusion

4.1 Validation of the CFD simulation

In order to obtain credible and reasonable simulation results, the numerical calculation results were verified by the thermal calculation values with three different loads. Table 20.3 shows the comparison between the CFD results and thermal calculated value of furnace outlet temperature. The maximum deviation between the CFD results and thermal calculated value is 18 K under 100% THA operation.

Table 20.3 Comparison of furnace outlet temperature (K) between CFD results and thermal calculated value
Full size table

4.2 Effects of Different Unit Loads on the Combustion Characteristics and NOx Emission

Figure 20.3 shows the temperature distribution along furnace height under different unit loads. The coal quantity decreases with the load decrease. This results in reduction of the temperature level in the furnace. Besides, both the amount of pulverized coal and air are reduced, which leads to a decreasing gas fullness in the furnace. It can be seen from Fig. 20.3 that under the condition of 50% THA load, the temperature distribution at the upper part of the furnace is uneven. This can be explained by that the low gas fullness under the condition of 50% THA load results in the uneven velocity distribution in the furnace. Thus, heat exchanges between the gas and wall are different in different positions, which eventually leads to uneven temperature distribution at the upper part of the furnace. The temperature distribution showed a similar tendency with the results of Liu and Zhang (2019).

Fig. 20.3
figure 3

Temperature distribution along the height direction under different loads

Full size image

Figure 20.4 presents the average temperature on furnace cross-section along the height direction under different loads. The temperature distribution is similar under three boiler loads, while combustion details are different in the main combustion zone ranging from 24 to 41 m. In the main combustion zone, the temperature has a sharp decrease due to the injection of a large amount of primary air with the lower temperature. For the 50% THA load, that coal mills A-B in the lower position are closed causes that the temperature sharp decrease occurs in a higher position when compared with the 100% THA load and the 70% THA load. There is another sharp temperature decrease resulted from the injection of SOFA in the burnout zone. It is evident that the area-weighted average temperature at the furnace outlet Tav decreases as Tav (100% THA) > Tav (70% THA) > Tav (50% THA) with the highest temperature appearing at 100% THA load.

Fig. 20.4
figure 4

Area-weighted average temperature along the height direction under different loads

Full size image

Figure 20.5 shows area-weighted average species distribution along the height direction under different loads. With the injection of the air and the combustion reaction, the O2 concentration CO2 changes drastically in the combustion zone. As shown in Fig. 20.5, the area-weighted average O2 concentration at the furnace outlet CO2 descends as CO2 (100% THA) > CO2 (70% THA) > CO2 (50% THA). During the combustion of the furnace, most of O2 is consumed to generate CO2. The CCO2 decreases with the increasing loads as CCO2 (50% THA) > CCO2 (70% THA) > CCO2 (100% THA) with the highest CCO2 appearing at the 50-THA load. In the furnace outlet, it is obviously seen that NO concentration decreases with increased loads as CNO (50% THA) > CNO (70% THA) > CNO (100% THA). So it can be conclueded that NOx concentrations in the flue gas often increases obviously with the decrease of the boiler operating load, which is consistent with the results of Zheng (Yang and Gao 2017).

Fig. 20.5
figure 5

Area-weighted average species distribution along the height direction under different loads

Full size image

4.3 Effects of Different Excess Air Ratios on the Combustion Characteristics and NOx Emission

Temperature distributions along the height direction at different excess air ratios are demonstrated as Fig. 20.6. The combustion reaction near the nozzles in the main combustion zone is severe and the average temperature of the furnace gradually increases along the height of the furnace. After the overfire air with lower temperature is injected, the temperature in the furnace is reduced, while a small amount of heat is released as the char combustion. Above the main combustion zone, due to the heat exchange with water-cooled walls and superheater, the gas temperature begins to decrease. Under the three conditions of different excess air ratios, the temperature distributions in the furnace have little differences. This is because the oxygen concentration in the main combustion zone increases as the excess air ratio increases, which leads to that the combustion reaction in the main combustion zone is fully carried out and releases more heat. On the other hand, to heat the air with lower temperature causes the furnace temperature to decrease.

Fig. 20.6
figure 6

Temperature distribution along the height direction at different excess air ratios

Full size image

Figure 20.7 shows the O2 distribution under different excess air ratios. As the excess air ratio increases, the oxygen concentration above the overfire air nozzles increases significantly. Figure 20.8 shows area-weighted average temperature and species distribution along the height direction under different air ratios. It can be seen from the Fig. 20.9 that the oxygen concentration is smaller in the main combustion zone. This is because the boiler adopts air-staging combustion technology, and the excess air ratio in the main combustion zone is less than 1.0, forming a strong reductive atmosphere, and the combustion reaction is insufficient. As the excess air ratio increases, the amount of oxygen concentration in the main combustion zone increases and combustion reaction in the furnace is more sufficient, so the high temperature region in the furnace gradually expands. With the injection of SOFA, the oxygen concentration increases significantly, and the char has almost burned out, so that the increase of excess air ratio will lead to an increase in NOx emissions at the furnace outlet.

Fig. 20.7
figure 7

O2 distribution along the height direction under different excess air ratios

Full size image
Fig. 20.8
figure 8

Area-weighted average temperature along the height direction under different excess air ratios

Full size image
Fig. 20.9
figure 9

Area-weighted average species distribution along the height direction under different excess air ratios

Full size image

5 Conclusions

In this study, the effects of different loads and excess air ratios on combustion characteristics were studied by numerical simulation in an ultra-supercritical double-reheat unit. In NOx calculation, the revised NOx model is applied to replace the default model to accurately predict NOx emission. The following conclusions are obtained:

  1. 1.

    As the load decreases, velocity distribution in the furnace are more uneven, which leads to more uneven temperature distribution at the upper part of the furnace. The NOx emission increases while the boiler operates under the low and medium loads.

  2. 2.

    Under three conditions of different excess air ratios, the temperature distributions in the furnace have little differences. With the injection of SOFA, the oxygen concentration increases significantly, and the char has almost burned out, so that the increase of excess air ratios will lead to an increase in NOx emission at the furnace outlet.

  3. 3.

    In the five simulation cases of the 660 MW tower boiler, the minimum NOx concentration at the furnace outlet is 308 ppm, appearing under 100% THA with excess air ratio of 1.15.