Comparison of the Application of Three Methods for the Determination of Outdoor PM2.5 Design Concentrations for Fresh Air Filtration Systems in China

With the increasing popularity of fresh-air filtration systems, the methods of determining the outdoor PM2.5 design concentration have become more important. However, the monitoring of atmospheric fine particles in China started relatively late, and there are relatively few cities with complete data, with obvious regional differences, which led to many problems in the selection of air filters for fresh-air filtration systems. In this paper, three methods of determining outdoor PM2.5 design concentration were analyzed using the daily average concentration of PM2.5 in 31 provincial capital cities from 2016 to 2020. Six typical cities in different regions were also taken as examples. The advantages and disadvantages of the three existing statistical methods were compared and analyzed, as well as the corresponding differences in the selection of outdoor PM2.5 concentration value on the filter systems. The results showed that the method of mathematical induction was more accurate and reasonable for the calculation of outdoor PM2.5 design concentrations. The local outdoor PM2.5 design concentration could be quickly calculated using the recommended coefficient K and annual average PM2.5 concentration of the region, especially for small and medium-sized cities without monitoring data. However, the recommended coefficient K should be provided based on the specific region, and should be divided into values for strict conditions and normal conditions during use. This would provide a simple and effective way to select the correct air filters for practical engineering.


Introduction
With the increasing concentrations of particulate matter in the atmosphere, how to build a good indoor environment is a current focus of attention. Haze weather frequently appears in most cities in China [1][2][3], especially in winter [3]. In addition, PM 2.5 in the atmosphere might cause some diseases related to human lung disease, cardiovascular disease, etc. [4][5][6][7][8]. However, people spend as much as 80-90% of their time indoors [9], and the quality of the indoor environment was the most important to us. The relevant literature showed that~50% of indoor PM came from the outdoor environment [10]. Outdoor air pollution enters the room through the ventilation systems and gaps in building envelopes [11]. With the outbreak of Coronavirus Disease 2019 (COVID-19) [12], complex cross-contamination in the environment brought greater challenges and had an impact on the indoor environment. Meanwhile, with the continuous improvement in buildings' airtightness levels [13], indoor environments have become relatively closed. As a result, the fresh-air system has become an important way to effectively solve the indoor sanitary environment issue and provide fresh air [14]. accessed on 6 June 2021. The hourly average concentration values for PM 2.5 in the atmosphere from 31 January 2016 to 31 December 2020 were used.

The Method for No-Guarantee Days
The method for no-guarantee days used the outdoor calculated concentration of PM 2.5 to adopt an average daily mass concentration that was not guaranteed for specific days [6,15,22]. For example, if the daily mass concentration was not guaranteed for 5 days, the average mass concentration of PM 2.5 per year in each statistical year should be arranged in descending order, and then the highest 5 days in terms of average mass concentration of PM 2.5 should be removed. The average mass concentration of PM 2.5 on the sixth day is the daily mass concentration that is not guaranteed for 5 days [6,15,22]. No-guarantee for 10 days means that the average mass concentration of PM 2.5 per year in each statistical year should be arranged in descending order, and then the highest average mass concentration of PM 2.5 for 10 days should be removed. The average mass concentration of PM 2.5 on the 11th day is the daily mass concentration that is not guaranteed for 10 days.

The Method of Guarantee Rate
The method of guarantee rate creates a guarantee rate curve according to the corresponding outdoor PM 2.5 design concentration. The corresponding outdoor design concentration value is determined according to the required guarantee rate [22].
Firstly, the PM 2.5 concentration grouping distance is calculated, as well as the number of groups and upper limit between different groups of PM 2.5 concentration in different cities. The calculation formulas are as follows in Equations (1) and (2) [22]: where C n is the concentration group interval group distance; C max and C min are the maximum and minimum concentrations of PM 2.5 in a recorded time period, respectively, µg/m 3 ; n is the number of groups, generally 5-10; C is the integer value greater than and closest to C n ; C i is the upper limit of the group. Secondly, counting the frequency of different groups, and the number of days that the PM 2.5 concentration values appear between different groups, the relative frequency of each group is calculated using Equation (3) [22]: where f i is the relative frequency of group i, %, and N i is the frequency of group i. Thirdly, the cumulative relative frequency corresponding to each grouping interval is calculated in the grouping order from small to large using Equation (4) [22], which shows that the cumulative relative frequency of group i is equal to the sum of the relative frequencies of i groups: where F i is the cumulative relative frequency of group i, %. Finally, the cumulative relative frequency is shown as the horizontal axis, while the upper limit of PM 2.5 concentration group is shown as the vertical axis [22]. The cumulative relative frequency is the guarantee rate. The PM 2.5 concentration corresponding to the guarantee rate is the outdoor PM 2.5 design concentration value.

The Method of Mathematical Induction
The concentrations of particulate matter in the outdoor atmosphere are affected by many factors, and the law of changes is also relatively complicated. Therefore, professionals in Japan conducted an inductive analysis, and summarized this in six levels of "risk rates" (non-guaranteed rates). The PM 10 concentration values matched the risk rates, which were 2.5%, 5%, 10%, 15%, 20%, and 50%, respectively [16,24]. The regression correlations between the concentration values of atmospheric particulate matter and the annual average values at different guarantee rates were obtained [16,24]. The design concentrations of atmospheric particulate matter C D in each area were calculated using Equation (5) [16,24]: where C y is the annual average concentration of suspended particulate matter in the atmosphere in the area; K is the recommended coefficient. The recommended coefficient K is the ratio of the annual average value to the outdoor PM 2.5 design concentration value corresponding to different non-guaranteed rates [16,24]. This method could be used for situations where the monitoring data are relatively few, equipment is relatively imperfect, and data are lacking. Therefore, it is more in line with the current basic national conditions in China [16].

The Method of Air Filter Selection
Particulate concentrations in indoor environments are determined by filtration efficiency, and air filters are selected according to their filtration performance and the different grades of fresh-air filtration systems. The efficiency of air filters can be calculated by Equation (6) [16]: where η is the filtration efficiency of air filters, %; C 1 is the outdoor PM 2.5 design concentration, µg/m 3 ; C 2 is the indoor PM 2.5 design concentration, µg/m 3 . The series' combined efficiency could be calculated using Equation (7): where η is the series combined efficiency of air filters, %; η 1 to η n is the filtration efficiency of each filter grade, %.

Outdoor Concentration of PM 2.5
The PM 2.5 values in 31 major cities in China from 2016 to 2020 are shown in Figure 1. Figure 1 shows that the annual average concentration of PM 2.5 in 31 cities in China from 2016 to 2020 was 17~73 µg/m 3 . The cities with the smallest 5-year average concentration of PM 2.5 were Lasa and Haikou, while the largest was Shijiazhuang. Among the 31 cities, the city with the highest monthly average concentration of PM 2.5 was Shijiazhuang, which appeared in December 2016, with 254 µg/m 3 . This was followed by Urumqi, Xi'an, Hohhot and Zhengzhou, where the highest concentrations appeared in January 2017, January 2017, January 2020, and December 2016, respectively. The city with the lowest monthly average concentration of PM 2.5 was Lasa, at 6 µg/m 3 . The outdoor concentrations of PM 2.5 in China were unevenly distributed: the PM 2.5 concentration in southern coastal cities was relatively low, while that in northern cities was relatively high. Therefore, an in-depth study of the method used to determine the outdoor PM 2.5 concentration was significant to provide a reference for the accurate and reasonable selection of fresh-air filtration systems. The concentration control limit for indoor PM 2.5 could be determined according to the relevant air-quality control standards [27]. cities was relatively low, while that in northern cities was relatively high. Therefore, an in-depth study of the method used to determine the outdoor PM2.5 concentration was significant to provide a reference for the accurate and reasonable selection of fresh-air filtration systems. The concentration control limit for indoor PM2.5 could be determined according to the relevant air-quality control standards [27].

The Change in PM2.5 Concentration Using the Method of No-Guarantee Days
The monitoring data for 31 provincial capitals in China, from 2016 to 2020, were counted and analyzed based on the existing relevant research and conclusions. The outdoor atmospheric PM2.5 concentration values, corresponding to the different not-guaranteed day parameters, are given in Table 1 [6,15].

The Change in PM 2.5 Concentration Using the Method of No-Guarantee Days
The monitoring data for 31 provincial capitals in China, from 2016 to 2020, were counted and analyzed based on the existing relevant research and conclusions. The outdoor atmospheric PM 2.5 concentration values, corresponding to the different not-guaranteed day parameters, are given in Table 1 [6,15].  Table 1 shows that the outdoor PM 2.5 concentration values corresponding to situations of no-guarantee for 5 days and no-guarantee for 10 days in each city within 5 years were very different. With an increase in the number of no-guarantee days, the corresponding outdoor PM 2.5 concentration values gradually decreased. The maximum concentration difference corresponding to no-guarantee for 5 days and no-guarantee for 10 days in the whole of 2020 was 54 µg/m 3 , and the corresponding cities were Tianjin and Harbin; the minimum concentration difference was 4 µg/m 3 , and the corresponding cities were Lasa, Lanzhou, Chengdu, and Kunming. The maximum concentration difference in 2019 was 42 µg/m 3 , and the corresponding city was Shenyang; the minimum concentration difference was 2 µg/m 3 , and the corresponding city was Lasa. The maximum concentration difference in 2018 was 33 µg/m 3 , and the corresponding city was Jinan; the minimum concentration difference was 4 µg/m 3 , and the corresponding cities were Guiyang and Kunming. The maximum concentration difference in 2017 was 70 µg/m 3 , and the corresponding city was Zhengzhou; the minimum concentration difference was 3 µg/m 3 , and the corresponding city was Lasa. The maximum concentration difference in 2016 was 152 µg/m 3 , and the corresponding city was Zhengzhou; the minimum concentration difference was 2 µg/m 3 , and the corresponding city was Nanning. Therefore, a more demanding living environment could be obtained using no-guarantee for 5 days. In addition, using data for many years led to more accurate results [6,15]. For further analysis, the comparison results of the outdoor concentration values of six typical cities, corresponding to no-guarantee for 5 days, are shown in Figure 2. The six representative cities, with obvious regional differences, were gave the same results [6,15], which again verified the correctness of this paper. However, a large amount of data are needed for analysis if the method of no-guarantee days uses the data for many years. In addition, the statistical analysis process is relatively cumbersome, requiring extensive time to complete, and errors occur in the data processing.

The Change in PM2.5 Concentration Using the Method of Guarantee Rate
Xi'an was taken as an example to draw a guarantee-rate curve to increase understanding of the guarantee rate method. The daily average PM2.5 concentration values for the whole year of 2020 were sorted, and the maximum and minimum values were 225 μg/m 3 and 6 μg/m 3 , respectively. The difference between them was 219 μg/m 3 , n was 10, Cn was 21.9 (Equation (1)), and C was 25. The group limits were 25, 50, 75, 100, 125, 150, 175, 200, 225, and 250 (Equation (2)), and the total number of days was 366. The calculation results for the relative frequency (Equation (3)) and the cumulative relative frequency (Equation (4)) are shown in Table 2. A guarantee rate curve for Xi'an for the whole of 2020 is shown in Figure 3.   Figure 2 shows that the outdoor concentration values for each year, corresponding to no-guarantee for 5 days, were quite different. The PM 2.5 concentration values would rebound to a certain extent; the pollution was heavier or lower than the previous year. Only the outdoor PM 2.5 concentration values for Guangzhou in 2020 and 2019 were lower than the standard (75 µg/m 3 ) [28], at 56 µg/m 3 and 67 µg/m 3 , respectively. The differences between the maximum value and the minimum value in 5 years for Harbin, Beijing, Xi'an, Changsha, Guangzhou, and Shanghai were 142 µg/m 3 , 133 µg/m 3 , 113 µg/m 3 , 57 µg/m 3 , 56 µg/m 3 and 36 µg/m 3 , respectively. The largest difference between the outdoor concentration in each year and the 5-year average concentration was found for Harbin, with 81.8 µg/m 3 . The smallest difference was found for Guangzhou, with 2.2 µg/m 3 . Therefore, it could be seen that the outdoor PM 2.5 design concentration, calculated using the method of no-guarantee days, still had certain fluctuations. The method of calculating no-guarantee days using the average data for many years was more reasonable. Related research also gave the same results [6,15], which again verified the correctness of this paper. However, a large amount of data are needed for analysis if the method of no-guarantee days uses the data for many years. In addition, the statistical analysis process is relatively cumbersome, requiring extensive time to complete, and errors occur in the data processing.

The Change in PM 2.5 Concentration Using the Method of Guarantee Rate
Xi'an was taken as an example to draw a guarantee-rate curve to increase understanding of the guarantee rate method. The daily average PM 2.5 concentration values for the whole year of 2020 were sorted, and the maximum and minimum values were 225 µg/m 3 and 6 µg/m 3 , respectively. The difference between them was 219 µg/m 3 , n was 10, C n was 21.9 (Equation (1)), and C was 25. The group limits were 25, 50, 75, 100, 125, 150, 175, 200, 225, and 250 (Equation (2)), and the total number of days was 366. The calculation results for the relative frequency (Equation (3)) and the cumulative relative frequency (Equation (4)) are shown in Table 2. A guarantee rate curve for Xi'an for the whole of 2020 is shown in Figure 3.   Figure 3 shows the change trend for the outdoor PM2.5 concentration of the guarantee rate and its specific values. If the guarantee rate was 95%, the outdoor PM2.5 design concentration was about 133 μg/m 3 . If the guarantee rate was 97.5%, the outdoor PM2.5 design concentration was about 163 μg/m 3 . The data for a total of 5 years, from 2016 to 2020, were   Figure 3 shows the change trend for the outdoor PM 2.5 concentration of the guarantee rate and its specific values. If the guarantee rate was 95%, the outdoor PM 2.5 design concentration was about 133 µg/m 3 . If the guarantee rate was 97.5%, the outdoor PM 2.5 design concentration was about 163 µg/m 3 . The data for a total of 5 years, from 2016 to 2020, were used for calculation using the method, and the outdoor PM 2.5 design concentrations of different guarantee rates for 31 major cities in China could be obtained. The outdoor PM 2.5 design concentration values corresponding to guarantee rates of 95% and 97.5% for six typical cities are shown in Table 3. The outdoor PM 2.5 concentration values also had a certain rebound phenomenon when the guarantee rates were 97.5% and 95%. The higher the guarantee rate, the greater the concentration rebound. Only the outdoor PM 2.5 concentration values corresponding to a guarantee rate of 95% in Guangzhou were lower than the standard (75 µg/m 3 ) [28]. Except for 2019, the outdoor PM 2.5 concentration values corresponding to a guarantee rate of 97.5% in Guangzhou were lower than the standard (75 µg/m 3 ) [28]. Except for 2016, the outdoor PM 2.5 concentration values corresponding to a guarantee rate of 95% in Shanghai were lower than the standard (75 µg/m 3 ) [28]. The largest difference between the outdoor concentration in each year and the 5-year average concentration was found in Beijing, at 87.6 µg/m 3 , while the smallest difference was found in Guangzhou, at 0.8 µg/m 3 . However, the selection of the specific guarantee rate was related to many factors, such as the types of buildings, their specific use requirements, the local environment, etc., as well as the needs of designers. The outdoor PM 2.5 design concentration of any guarantee rate could be obtained according to the drawn guarantee rate curve. Therefore, the required values could be easily found for the guarantee rate of 0-100% [22].

The Change in PM 2.5 Concentration Using the Method of Mathematical Induction
Beijing was taken as an example to facilitate understanding of the mathematical induction method. The annual average values of PM 2.5 concentration and their average concentration value, corresponding to non-guaranteed rates of 2.5% and 5.0%, were calculated using monitoring station data from 2016 to 2020. The relationship between the annual average values of PM 2.5 concentration and their average concentration value is shown in Figure 4, in which the average concentration values correspond to non-guaranteed rates of 2.5% and 5.0%. Figure 4 shows that the fitting curves had good fitting effects for the non-guaranteed rates of 2.5% and 5%. The outdoor PM 2.5 design concentration values for different cities in China also could be calculated using the method of mathematical induction. However, China is relatively vast, and the source distribution of PM 2.5 in different regions is quite unbalanced. The recommended coefficient K for Japan cannot be directly applied in China, and the same recommended coefficient K could not be adopted for the whole country [16]. The recommended coefficient K for Japan is usually divided into values for strict conditions and normal conditions. For strict conditions, the non-guaranteed rate was 2.5%, and the recommended coefficient K was 3.7. For normal conditions, the non-guaranteed rate was 5% and the recommended coefficient K was 2.7 [24]. Strict conditions refer to buildings with extremely high requirements for the concentration of indoor particulate matter, such as clean rooms and wards and dust-free workshops. Normal conditions refer to environments that do not need to strictly control the concentration of indoor particulate matter, such as houses, shopping malls, schools, and airports [16]. The differences between the recommended coefficient K for Japan and six typical cities in China are shown in Figure 5.

The Change in PM2.5 Concentration Using the Method of Mathematical Induction
Beijing was taken as an example to facilitate understanding of the mathematical induction method. The annual average values of PM2.5 concentration and their average concentration value, corresponding to non-guaranteed rates of 2.5% and 5.0%, were calculated using monitoring station data from 2016 to 2020. The relationship between the annual average values of PM2.5 concentration and their average concentration value is shown in Figure 4, in which the average concentration values correspond to non-guaranteed rates of 2.5% and 5.0%.  Figure 4 shows that the fitting curves had good fitting effects for the non-guaranteed rates of 2.5% and 5%. The outdoor PM2.5 design concentration values for different cities in China also could be calculated using the method of mathematical induction. However, China is relatively vast, and the source distribution of PM2.5 in different regions is quite unbalanced. The recommended coefficient K for Japan cannot be directly applied in China, and the same recommended coefficient K could not be adopted for the whole country [16]. The recommended coefficient K for Japan is usually divided into values for strict conditions and normal conditions. For strict conditions, the non-guaranteed rate was 2.5%, and the recommended coefficient K was 3.7. For normal conditions, the non-guaranteed rate was 5% and the recommended coefficient K was 2.7 [24]. Strict conditions refer to buildings with extremely high requirements for the concentration of indoor particulate matter, such as clean rooms and wards and dust-free workshops. Normal conditions refer to environments that do not need to strictly control the concentration of indoor particulate matter, such as houses, shopping malls, schools, and airports [16]. The differences between the recommended coefficient K for Japan and six typical cities in China are shown in Figure 5. Figure 5 shows that the average recommended coefficient K for each city in China was quite different, and there was a big difference with the K value recommended for Japan. Among the six typical cities in China, only the recommended K for Harbin was higher than that for Japan under strict conditions, while only the recommended K for Xi'an was higher than that for Japan under normal conditions. Therefore, the recommended K for Japan could not be directly applied in China. The maximum value of coefficient K was found in Harbin under the non-guaranteed rate, at 2.5%, while the average value was 4.07. The minimum value of coefficient K was found in Guangzhou; the average value was 2.02, and the difference between the two cities was 2.05. The maximum value of coefficient K was found in Xi'an under the non-guaranteed rate, at 5%, with an average value of 2.71.  The minimum value of coefficient K was still found in Guangzhou; the average value was 1.55, and the difference between Guangzhou and Xi'an was 1.16. The largest difference in the coefficient K value in the same city was found in Harbin, at 1.39, while the smallest difference was found in Guangzhou, at 0.463. Therefore, it could be seen that the same recommended coefficient K could not be used. The recommended coefficient K should be provided separately according to region. Different provinces and cities need different coefficient K recommendations. In addition, there is a new trend under way in the regional economies of China at present that differentiates north-south economic growth, and the development of the south's economy is rapid relative to that of the north [29]. Therefore, there will be greater gaps between the values of coefficient K for different cities. The recommended coefficient K values for six typical cities in China from 2016 to 2020 under strict conditions (non-guaranteed rate 2.5%) and normal conditions (non-guaranteed rate 5%) are shown in Table 4.

Strict conditions Normal conditions
The value differences in coefficient K 3.7 Figure 5. Differences in coefficient K between China and Japan. Figure 5 shows that the average recommended coefficient K for each city in China was quite different, and there was a big difference with the K value recommended for Japan. Among the six typical cities in China, only the recommended K for Harbin was higher than that for Japan under strict conditions, while only the recommended K for Xi'an was higher than that for Japan under normal conditions. Therefore, the recommended K for Japan could not be directly applied in China. The maximum value of coefficient K was found in Harbin under the non-guaranteed rate, at 2.5%, while the average value was 4.07. The minimum value of coefficient K was found in Guangzhou; the average value was 2.02, and the difference between the two cities was 2.05. The maximum value of coefficient K was found in Xi'an under the non-guaranteed rate, at 5%, with an average value of 2.71. The minimum value of coefficient K was still found in Guangzhou; the average value was 1.55, and the difference between Guangzhou and Xi'an was 1.16. The largest difference in the coefficient K value in the same city was found in Harbin, at 1.39, while the smallest difference was found in Guangzhou, at 0.463. Therefore, it could be seen that the same recommended coefficient K could not be used. The recommended coefficient K should be provided separately according to region. Different provinces and cities need different coefficient K recommendations. In addition, there is a new trend under way in the regional economies of China at present that differentiates north-south economic growth, and the development of the south's economy is rapid relative to that of the north [29]. Therefore, there will be greater gaps between the values of coefficient K for different cities. The recommended coefficient K values for six typical cities in China from 2016 to 2020 under strict conditions (non-guaranteed rate 2.5%) and normal conditions (non-guaranteed rate 5%) are shown in Table 4. Therefore, the outdoor design concentration values could be obtained according to the recommended coefficient K and the annual average concentration values. Based on the development status of China, there were relatively few atmospheric PM 2.5 concentration data values available for this analysis, so the method of mathematical induction was an applicable and simple method to calculate the outdoor PM 2.5 design concentrations.

Analysis and Suggestions Regarding the Calculation Results of Three Methods
The outdoor PM 2.5 design concentrations, calculated according to three existing methods, are summarized in Table 5. Table 5. Outdoor PM 2.5 design concentrations using three methods. Table 5 shows that the calculation results for the three methods were very different, and the concentration value obtained using the method of no-guarantee days was the highest. The 5-year average outdoor concentrations of Harbin, Beijing, Xi'an, Shanghai, Changsha and Guangzhou, calculated using the method of no-guarantee days, were 23.2 µg/m 3 , 16.6 µg/m 3 , 17 µg/m 3 , 9.2 µg/m 3 , 6.6 µg/m 3 and 18 µg/m 3 higher than those obtained using the other two methods, respectively. The largest difference was found in Harbin, and the smallest difference was found in Changsha. The difference in Beijing in 2017 was the largest when using a single year's data for calculation and comparison, at 57 µg/m 3 . The second largest difference was found in Xi'an in 2016, at 53 µg/m 3 . It is recommended to use the average value of previous years for calculation, to ensure results are more stable and accurate. Xi'an was taken as an example for a comparative analysis of the differences using the three methods. The outdoor PM 2.5 concentrations were calculated using the data in Table 5, and the indoor concentration was 75 µg/m 3 [28]. The comparison of methods for the determination of outdoor PM 2.5 concentrations for fresh air filtration methods in Xi'an is shown in Figure 6. μg/m , 16.6 μg/m , 17 μg/m , 9.2 μg/m , 6.6 μg/m and 18 μg/m higher than those obtained using the other two methods, respectively. The largest difference was found in Harbin, and the smallest difference was found in Changsha. The difference in Beijing in 2017 was the largest when using a single year's data for calculation and comparison, at 57 μg/m 3 . The second largest difference was found in Xi'an in 2016, at 53 μg/m 3 . It is recommended to use the average value of previous years for calculation, to ensure results are more stable and accurate. Xi'an was taken as an example for a comparative analysis of the differences using the three methods. The outdoor PM2.5 concentrations were calculated using the data in Table 5, and the indoor concentration was 75 μg/m 3 [28]. The comparison of methods for the determination of outdoor PM2.5 concentrations for fresh air filtration methods in Xi'an is shown in Figure 6.  Figure 6 shows that the filtration efficiency of air filters ranged from 60.7% to 75.3% using the method of no-guarantee days, and the average efficiency was 69.0%. The filtration efficiency of air filters ranged from 54.0% to 74.1% using the methods of guarantee rate and mathematical induction, and the average efficiency was 66.7%. The results obtained using the method of no-guarantee days were higher than those obtained using the  Figure 6 shows that the filtration efficiency of air filters ranged from 60.7% to 75.3% using the method of no-guarantee days, and the average efficiency was 69.0%. The filtration efficiency of air filters ranged from 54.0% to 74.1% using the methods of guarantee rate and mathematical induction, and the average efficiency was 66.7%. The results obtained using the method of no-guarantee days were higher than those obtained using the methods of guarantee rate and mathematical induction. The relevant parameters of existing air filters of different grades, according to previous market research and related test research, are given in Table 6 [16,30,31]. The series combination of air filters is given in Table 7 based on the performance parameters, which met the indoor air quality standards of Xi'an from 2016 to 2020. The relevant filter equipment could be quickly adjusted and selected according to this table.
According to a comprehensive comparative analysis of the above three methods, the method of no-guarantee days showed relatively large changes, which might increase the initial investment in equipment. The variances in outdoor PM 2.5 concentration obtained using the guarantee rate and mathematical induction methods were relatively small. The basic principles of the two methods of no-guarantee days and guarantee rate are the same. The outdoor PM 2.5 design concentration of any guarantee rate could be obtained according to the guarantee rate curve drawn using the guarantee rate method. In addition, the guarantee rate method is more complicated than the method of no-guarantee days, with relatively more steps. Although both methods have been applied in practice, the guarantee rate method was used more often, for more days [32].