Comparison of PMtotal, PM10, PM2.5, NOx, and SO2 Emission Factors from Coal-fired Power Plants per Load Change

For two bituminous coal-fired power plants with 500 MW and pulverized coal combustion type, the concentration of PMtotal, >PM10, PM2.5-10, PM2.5, NOx, and SO2 was measured, and their emission factors were calculated through field measurement. The measurement points started from the boiler downstream and continued to the air pollution control devices (APCDs) that are installed in series, namely, the selected catalytic reduction system (SCR), air preheater (APH), electrostatic precipitator (ESP) and wet flue gas desulfurization system (WFGD). The measurement was performed at one point for more than three times by using the Korean standard method for air pollutants. However, all measurement points, except for the stacks, were not representative of the standard test method. In addition, the PM concentration was too high to reduce the collection time due to isokinetic sampling. There is a limitation of how representative the measurement results can be. During the field measurement period, the power production rate of the two coal-fired power plants was 91.6% and 79.2% in the P-1 and P-2, respectively. Moreover, in the P-2, with a low power production rate, the concentration of PMtotal, PM10, PM2.5, and NOx was found to be low, and the emission factor calculated by dividing the measured concentration value by the fuel usage was also estimated to be low. Such results are due to the coal combustion chamber and various types of APCD being operated at a lower-load condition than the design capacity. In turn, the number of pollutants generated was less, and the removal efficiency of the pollutant became high. However, it was found that the concentration of SO2 generated and the emission factor are more significantly affected by the sulfur content of the coal than the load factor change. To this end, reducing the operation load of the coal-fired power plant improves the combustion efficiency and APCDs performance and decreases the emission factor, resulting in more reduction of the air pollutants than that based on the simple calculation.


INTRODUCTION
The concentration of fine particulate matter (PM) is increasing, and in particular, the number of days in which high-concentration fine particulate matter occurred is also increasing (Wu et al., 2020;EEA, 2019).Fine particulate matter has an adverse Comparison of PM total , PM 10 , PM 2.5 , NOx, and SO 2 Emission Factors from Coal-fired Power Plants per Load Change effect on human health and visibility of the atmosphere and also induces climate change (Watts et al., 2015).
The ambient air quality standards for PM 2.5 were strengthened in 2018 to 35 μg/m 3 (daily average) and 15 μg/m 3 (annual average) in Korea.To satisfy ambient air quality standards, governments, businesses, and the private sector are making air quality changes (Kim et al., 2017).It is important to identify the air pollution sources to effectively manage air quality (Ma and Gang, 2019;Han et al., 2018).Air quality management strategies are forced to focus on high emissions as a priority.
The capacity of the coal-fired power plants in South Korea is 34.7 GW (as of 2019) from 60 power production facilities, and approximately 98% of them use bituminous coal as fuel.The largest number of coal-fired power plants has a 500 MW class capacity (35 units).Recently, the six coal-fired power plants with a capacity of 1000 MW have been constructed.There are two main types of combustion for coal-fired power plants in Korea.One is the use of a pulverized coal combustion (PC) boiler, in which coal of 3 mm or less is mixed with air and injected into a combustion chamber.Another is the circulating fluidized bed combustion (CFBC) boiler, which burns by circulating particles (sand and limestone) and coal in the combustion chamber.Most coal-fired power plants use PC combustion systems in Korea.
The air pollutants emitted from the coal-fired power plants are PM, SOx, and NOx alongside hazardous air pollutants such as Hg and As.These air pollutants are removed by installing various types of air pollution control devices, and the most highly used air pollution prevention system in coal-fired power plants in Korea is the selected catalytic reduction (SCR) -electrostatic precipitator (ESP) -wet flue gas desulfurization (WFGD).The selected catalytic reduction is able to remove NOx.Furthermore, the low-NOx combustion burner is installed for suppressing the generation of NOx from the combustion process.The electrostatic precipitator removes PM and heavy metals.The wet flue gas desulfurization system is able to remove SOx, HCl, HF, and PM (Liang et al., 2020;IPPC, 2017;Saarnio, 2014).
In this study, when two coal-fired power plants with the same capacity and the same APCDs are installed to operate at different power production rates, the concentration of PM (PM total , >PM 10 , PM 2.5-10 , PM 2.5 ), NOx, and SO 2 in the air pollution prevention device installed based on the combustion exhaust gas flow was measured, and their emission factors were calculated and compared.

1 Facility
For the measurement targets, two PC type coal-fired power plants with the capacity of 500 MW were selected, which are the type and capacities that are used the most in the domestic coal-fired power plants.Based on the exhaust gas flow, the measurement was conducted at five points, as presented in Fig. 1.Specifically, the measure-  ment was carried out at the SCR for De-NOx, an electrostatic precipitator that removes particulate matter; air preheater (APH) that reduces the temperature of the electrostatic precipitator; and the WFGD.Table 1 and Fig. 1 show a simple summary of the descriptions of the measurement target coal-fired power plants and the measurement points.

2 Test Method
The measurement items are particulate matter, sulfur oxides, and nitrogen oxides.For the particulate matter, the concentration of total particulate matter (PM total ) and particulate matter of different sizes (PM 2.5 , PM 2.5-10 , PM 10 or more) was measured.The measurement equipment of the ES 01301.1 method (NIER, 2014) was used to measure the concentration of the total particulate matter, while the measurement equipment of the KS ISO 23210 method (KATS, 2012) was utilized to measure the concentration of the particulate matter with different sizes.For the field measurement, the PM sampler and cascade impactor were placed at the same point and conducted the measurement at the same time.The PM total was calculated based on the values measured with the PM sampler, the PM 2.5 /PM total and PM 10 /PM total ratios were calculated based on the concentration values obtained using the cascade impactor.The PM 10 and PM 2.5 concentrations were estimated by applying these ratios to the total particulate matter concentration.The measurement equipment for each item is tabulated in Table 2.The measurement at each point could not be performed simultaneously for various reasons (multiple PM samplers and manpower required), and hence, it was conducted for the same facil ity for 1-3 days and repeated more than three times for each point.
The field measurement conducted in this study has limitations.In the boiler, SCR system, and APH points except for the stack, there were some points where the regulation of test methods for sampling of flue gas (ES 01114) could not be followed, i.e., where the fluid flow was unstable, according to the circumstances of the field.However, we measured the fluid velocity and static pressure using the pitot tube, selected the most stable point, and collected the sample in the field.Moreover, because the particulate matter concentration was high in the boiler, SCR system, and APH measurement points, the flow was lowered for the isokinetic sampling, and the sample collection time was shortened.Therefore, there is a possibility of issues regarding the representativeness of the measurement results.

1 Concentration of Air Pollutants
The measurement results of the PM (PM total , PM 10 , PM 2.5 ), NOx, and SO 2 concentrations for each coal-fired power plant and measurement point are presented in Fig. 2. The concentration of PM total generated at the boiler was 6,223.1±147.5 mg/Sm 3 and 4,252.9±403.8mg/ Sm 3 in the P-1 and P-2 facilities, respectively.After passing through the SCR, APH, ESP, and WFGD, it decreased to 5,233.5±63.1 mg/Sm 3 , 3,611.1±379.3mg/ Sm 3 , 14.4±0.8mg/Sm 3 , and 6.5±0.2 mg/Sm 3 in the P-1 facility and to 3,351.4±401.0mg/Sm 3 , 2,158.3±147.1 mg/Sm 3 , 6.5±0.6 mg/Sm 3 , and 3.2±0.39mg/Sm in the P-2 facility, respectively.The nitrogen oxide concentration in the boiler and stack was 187.2 mg/Sm 3 and 53.9 mg/Sm 3 in the P-1 facility and 186.9 mg/Sm 3 and 43.0 mg/Sm 3 in the P-2 facility, respectively.The sulfur dioxide concentration in the boiler and stack was reduced to 794.3 mg/Sm 3 and 36.5 mg/Sm 3 in the P-1 facility and 833.7 mg/Sm 3 and 43.9 mg/Sm 3 in the P-2 facility, respectively.Here, the concentrations were converted to the oxygen concentration 6% condition.Except for the sulfur oxide concentration, the concentration of the PM (PM total , PM 10 , PM 2.5 ) and NOx was found to be lower in the P-2 facility than in the P-1 facility.
The removal efficiency of the air pollution systems is shown in Table 3.The final removal efficiency of the air pollutants, namely PM total , PM 10 , PM 2.5 , NOx, and SO when they passed through the APCDs and were emitted to the outside of the stack was found to be 99.91%,99.76%, 99.67%, 74.11%, and 95.47%, respectively, which were obtained by taking an average of two coal-fired power plants.The removal efficiency of the PM total , PM 10 , PM 2.5 , and NOx appeared to be higher in the P-2 facility than in the P-1 facility, whereas the SO 2 removal efficiency was higher in the P-1 facility than in the P-2 facility.In terms of PM 2.5 , it was found to be generated as it passed through the SCR system, and this suggests that NH 3 sprayed as a reducing agent reacted with SO 2 , transformed into either (NH 4 ) 2 SO 4 or NH 4 HSO 4 particles, and generated fine particulate matter (Ruan et al., 2019;Li, 2017;Shi et al., 2016;Li et al., 2015).The concentration of NOx also increased as it passed through the WFGD system, probably because of the leakage of the exhaust gas since the leakage type Gas-Gas Heater (GGH), i.e., a heat exchanger for adjusting to the ade-quate operating temperature of the WFGD system for De-SOx, is installed (Seong and Lee, 2017).
Fig. 3 presents the overall particle collection efficiencies for PM total (>PM 10 + PM 2.5-10 + PM 2.5 ) at each measurement point.Although there is a slight difference between the P-1 and P-2 facilities, it was identified that the PM total is removed the most in the main collector, ESP.Moreover, it was found that PM total is removed in the SCR system for De-NOx, APH for the heat exchange, and WFGD system for De-SOx, and these results are in good agreement with the results of other studies (Ruan et al., 2019;Sui et al., 2016).
Fig. 4 shows the particle size distribution of the particulate matter at each measurement point.In terms of the size distribution of the particulate matter generated from  the coal combustion, the PM 2.5 , PM 2.5-10 , and > PM 10 accounted for 19.4%, 34.5%, and 46.1% in the P-1 facility and 22.5%, 33.2%, and 44.3% in the P-2 facility indicating a similar distribution in two facilities.Although the particulate matter size distribution was similar in the two facilities at the SCR system, APH, ESP, and WFGD system points, PM 2.5 fraction in the P-1 facility was slightly higher than that in the P-1 facility.
It was identified that the SCR system removes largesized particulate matter and generates small-sized particulate matter.It was observed that the concentration ratio of >PM 10 reduces, that of PM 2.5-10 does not significantly change, and that of PM 2.5 increases at the outlet of the SCR system.Such findings suggest that small-sized particulate matter is generated due to the reaction between SO 2 and NH 3 , i.e., a reducing agent, and the large-sized particulate matter (>PM 10 ) falls into the hopper below the catalyst layer by the soot blower and then is removed (Seo and Chang, 2012).APH is a heat exchanger for decreasing the temperature of the exhaust gas that passed through the SCR system (340°C) to the adequate operating temperature of the electrostatic precipitator (about 130°C).In general, the specific resistivity of the particulate matter generated from the coal-fired power plant decreases with decreasing temperature of the exhaust gas, and hence, the generation of the back corona is inhibited (Cooper and Alley, 2002).Furthermore, with the decreasing temperature, the volume flowrate decreases, and the residence time within the ESP increases.Therefore, APH plays a role in improving the particle collection efficiency of ESP.However, even in such a heat exchanger, the flow decreases during the process of decreasing the temperature, the effect of gravity increases, and >PM 10 is removed (Chen et al., 2017).In the measurement results of this study, the removal efficiency of >PM 10 was found to be 30.7%and 20.1% in the P-1 and P-2 facilities, respectively.
ESP is the main equipment that removes the particulate matter, and it removes 57.8% and 50.69% of the PM total generated during the coal combustion in the P-1 and P-2 facilities, respectively.Furthermore, for PM 2.5 , it removes almost all of the PM 2.5 generated during the coal combustion along with the one generated while passing through the SCR.The particle collection efficiency for PM 2.5 was 95.9% and 96.4% in the P-1 and P-2 facilities, respectively.Although the WFGD system is usually used for De-SOx, the wet-scrubbing technology can generally remove the gaseous substance and particulate matter simultaneously and changes the design and operating conditions depending on the substances to be treated.The particulate matter is removed by the inertial impaction, diffusion, electrostatic attraction, and thermophoresis between the sprayed droplet and particulate matter (Kim et al., 2014;Jaworek et al., 2013;Copper et al., 2002).Based on the measurement results of this study, it was identified that particulate matter is removed, and the removal efficiency for PM 2.5 , PM 2.5-10 , and >PM 10 was found to be 53.4%,55.9%, and 72.4%, in the P-1 facility and 47.5%, 60.8%, and 71.4% in the P-2 facility, which is in line with the results of other studies showing higher removal efficiency for larger particulate matter (Chen et al., 2019;Wu et al., 2019;Li et al., 2017;Sui et al., 2016).
The SCR system installed to remove NOx is located downstream of the boiler, which is a good location to adjust the typical reaction temperature of the catalyst (300-400°C) (Li, 2017).The measurement target SCR of this study consists of three plate-type catalyst layers.The catalyst is composed of V 2 O 5 , and it is designed to use surface velocity of 3000-4,000 hr -1 , and NH 3 as a reducing agent.The NOx concentration was 187.2 mg/ Sm 3 and 51.3 mg/Sm 3 at the SCR inlet and outlet in the P-1 facility, respectively, and the removal efficiency was measured as 72.6%.The NOx concentration was 186.9 mg/Sm 3 and 39.0 mg/Sm 3 at the SCR inlet and outlet in the P-2 facility, respectively, and the removal efficiency was measured as 79.1%.These NOx removal efficiencies appeared to be slightly lower than those of the typical SCR system (Sorrels et al., 2019;Li., 2017).The emission limit value of the measurement target coal-fired power plant is 102.7 mg/Sm 3 (@ 6% O 2 ), and the emitted concentration can conform with the emission limit value, suggesting it is not processed at high efficiency.Furthermore, because the low-NOx burner (LNB) is installed at this measurement facility, the concentration of NOx generated in the boiler is low, leading to a low removal efficiency (Zhang et al., 2021;Ti et al., 2014).
The WFGD system that removes SO 2 is the desulfurization technique that is the most widely used in coalfired power plants.By spraying the limestone slurry (CaCO 3 ), SO 2 is usually converted to gypsum (CaSO 4 ) and then removed.The SO 2 concentration was 794.3 mg/Sm 3 and 36.5 mg/Sm 3 at the WFGD inlet and outlet in the P-1 facility, respectively, and the removal efficiency was measured as 95.4%.In the P-2 facility, the SO 2 concentration was 833.7 mg/Sm 3 and 43.9 mg/Sm 3 at the WFGD inlet and outlet, and the removal efficiency was measured as 94.7%.In this coal-fired power plant, the heat exchanger for the adequate operating temperature of the WFGD system, which is the leakage-type GGH, is installed upstream of the WFGD, and the leakage rate of GGH is between 3% and 5%.Therefore, the theoretical SO 2 removal efficiency of WFGD would be slightly higher considering the leakage rate.
Emission factor (kg/ton) = Concentration (mg/Sm 3 ) × amount of flue gas (Sm 3 /hr) × 10 -6 /amount of coal feeding (ton/hr) (1) Table 4 shows the information on the coal used in two coal-fired power plants during the measurement period.During this period, the P-1 facility generated 91.6% power, and the P-2 facility operated at 79.2%.During the measurement period of the P-2 facility, the power production rate of the power production facility was maintained at <80% due to the upper limit restrictions operated during the seasonal management system for highconcentration fine particulate matter.
During the measurement period, the power production was kept at a constant level (maximum variance of 0.9%) in the two coal-fired power plants, but the coal consumption was found to change (maximum variance of 4%).This change is because of the calorific value change of the coal, and the calorific value of the coal changes depending on its composition, including carbon, moisture, and ash contents.For the stable operation of the power production facility, constant calorific value and quality of the coal are favorable; however, the calorific value is known to vary in the field.During the measurement period, the change in the ash and sulfur contents of the coal was 10.84-14.69%and 0.33-0.46%,respectively.
Using Eq. ( 1) and information tabulated in Table 4, the emission factors of PM (PM total , PM 10 , PM 2.5 ), NOx, and SO 2 at each spot of the coal-fired power plant were calculated based on the measurement data for each point and the same coal information data, and the results are presented in Table 5.The particulate matter (PM total , PM 10 , PM 2.5 ) emission factors were lower in the P-2 facility with a lower load factor than in the P-1 facility with a higher load factor.The load factor of the P-2 is 84.4% of that of the P-1, but the emission factors of PM total , PM 10 , PM 2.5 , and NOx were found to be 56.0%,56.7%, 58.8%, and 78.0%, respectively, suggesting that the reduction of the PM total , PM 10 , PM 2.5 , and NOx is more significant than the reduction of the coal consumption.However, although the load factor of SO 2 decreased, its emission factor increased by about 1.5 times.
Under the same operating conditions in the facility with the same capacity, if the coal consumption decreases, the exhaust gas flow (Q) becomes lower than the design value, and as a result, the fluid velocity of the combustion chamber and APCDs reduces and the retention time increases.In general, if the passage velocity decreases and, in turn, retention time increases, the removal efficiency of the SCR, ESP, and WFGD system is improved (Chen et al., 2019).
Therefore, the upper limit restrictions operated during the high-concentration fine dust generation period not only reduces the operating load of the power production facility and air pollutant emissions but also improves the efficiency of APCDs and reduces the dust emission factor.However, in the case of SO 2 , it is important to use coal with low sulfur content.
The LNB is installed in the two coal-fired power plants.Although the performance of LNB shows 80% efficiency in the two facilities considering NAER's "uncontrolled" emission factor (7.5 kg/ton) (NAER, 2020), the efficiency of the SCR system was 58.2% and 86.0% in the P-1 and P-2 facilities, respectively, indicating lower emission factor in the facility with a lower load factor.
An SO 2 emission factor is determined by the sulfur content of the fuel.Based on the measurement results, it was identified that the difference due to the different sulfur contents of the coal is more significant than the difference in the emission factor due to the load factor.In particular, the emission factor of the WFGD can be affected by the leakage rate of the GGH.
By comparing the "uncontrolled" emission factor of PM total , PM 10 , and PM 2.5 with the existing emission factor, the "uncontrolled" emission factor of PM total , PM 10 , and PM 2.5 used nationally is 50 kg/ton, 29.1 kg/ton, and 10.14 kg/ton, respectively.In the research of AP-42 (U.S. EPA) and Zhao et al., it is calculated as a function of the ash content included in coal.The emission factor of PM total , PM 10 , and PM 2.5 calculated in this study was similar to that of AP-42 but lower than the PM total emission actor of Zhao et al.In terms of the emission factor for each prevention device, the emission factor of ESP and ESP + WFGD appeared to be lower than that of the other stud-ies, namely AP-42 and Zhao et al (Wu et al., 2018;Zhao et al., 2010;U.S EPA, 1998).

CONCLUSIONS
For the two bituminous coal-fired power plants with 500 MW and PC, the concentration of PM total , >PM 10 , PM 2.5-10 , PM 2.5 , NOx, and SO 2 was measured through field measurement, and their emission factors were calculated.The measurement was carried out more than three times in the boiler downstream, SCR system, APH, ESP, and WFGD system (stack) by using the Korean standard test method for air pollutants.
During the field measurement period, the load factor was 91.6% and 79.2% in the P-1 and P-2 coal-fired power plants, respectively, and in the P-2 facility with a lower load factor, the measured concentration of PM total , PM 10 , PM 2.5 , and NOx was low.The emission factor obtained by dividing the measured concentration value by the fuel consumption was also calculated to be low.The load factor of the P-2 facility is only 84.4% of that of the P-1 facility, but the emission factor of PM total , PM 10 , PM 2.5 , and NOx was 56.0%, 56.7%, 58.8%, and 78.0%, respectively, suggesting the reduction of the PM total , PM 10 , PM 2.5 , and NOx emissions are more significant than the reduction of the coal consumption.However, although the load factor of SO 2 decreased, its emission factor increased about 1.5 times, indicating no effect of the reduced load factor.
Since the coal combustion chamber and various types of APCDs are operated at lower-load conditions than the design capacity, the number of pollutants generated decreases and the pollutant removal efficiency increases.However, it was identified that the SO 2 concentration and emission factor are more significantly affected by the sulfur content of the coal rather than the difference in the load factor.
To this end, reducing the operating load of the coal-fired power plant improves the combustion efficiency and performance of several APCDs, and reduces the emission factor.It also leads to further decrease of the pollutants rather than the simply calculated reduction which is based on the reduction of power production rate.However, to decrease the SO 2 emission, it is more effective to provide low-sulfur coal.

Fig. 3 .Fig. 4 .
Fig. 3. Overall particle collection efficiencies in APCDs of flue gas stream.(b) P-2 This study was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industry, & Energy (MOTIE) of the Republic of Korea (grant no.20161110100140), and Korea Environmental Industry & Technology Institute (KEITI) (grant Graduate School of Fine Particle Management Specialization).

Table 1 .
Description of coal-fired power plant.

Table 3 .
Removal efficiencies of air pollutants in APCDs of flue gas stream.

Table 4 .
Coal usage and contents of ash, sulfur in feed coal during sampling periods.

Table 5 .
PM, NOx, and SO 2 emission factor (kg/ton of coal) in APCDs of flue gas stream.