-
The mixing ratios of HFCs at LAN from 4 January 2012 to 28 December 2016 are illustrated in Table 1 (except for the observation of HFC-365mfc, which started from 6 August 2014), and five-year time series for HFC mixing ratios are presented in Fig. A2. The background mixing ratios of both HFC-23 and HFC-32 accounted for less than 5%, while the background values of HFC-125, HFC-134a, HFC-152a and HFC-245fa took percentages of 10%–20%. Therefore, the background mixing ratios of all 10 HFCs at XGL were instead considered. The relative errors between the background mixing ratios of HFCs at LAN and at XGL were 3.8% (HFC-23), 23.0% (HFC-32), 3.9% (HFC-125), 5.2% (HFC-134a), 6.0% (HFC-143a), 31.8% (HFC-152a), 0.0% (HFC-227ea), 7.1% (HFC-236fa), 10.3% (HFC-245fa) and 10.6% (HFC-365mfc), which could be mainly explained by the asynchronous sampling time of the background samples at these two stations. Also, they were much lower than the relative differences between background mixing ratios and pollution mixing ratios of HFCs at LAN, except for HFC-152a, for the mixing ratio of HFC-152a probably has a non-negligible latitudinal gradient (Zhang et al., 2017). Overall, it was feasible to employ the background mixing ratios of HFCs at XGL to represent those at LAN to calculate the mixing ratio enhancements.
Compound Observing period
(YYYY.MM.DD)LAN XGL GWP100 (Forster et al., 2007) Number of air samples Percentage of background Mixing ratio enhancement (ppt) Number of air samples Percentage of background Background mixing ratio(ppt) HFC-23 2012.01.04–
2016.12.28230 3.9% 12.3 176 90.9% 27.8 14800 HFC-32 2012.01.04–
2016.12.28233 2.6% 19.0 184 75.0% 10.6 675 HFC-125 2012.01.04–
2016.12.28223 12.1% 10.4 179 95.5% 17.3 3500 HFC-134a 2012.01.04–
2016.12.28180 16.7% 32.4 177 84.7% 81.1 1430 HFC-143a 2012.01.04–
2016.12.28227 44.1% 5.3 187 96.3% 16.7 4470 HFC-152a 2012.01.04–
2016.12.28227 17.2% 7.2 158 80.4% 7.1 124 HFC-227ea 2012.01.04–
2016.12.28219 21.5% 0.95 161 94.4% 1.1 3220 HFC-236fa 2012.01.04–
2016.12.28232 41.8% 1.2 181 90.6% 0.13 9810 HFC-245fa 2012.01.04–
2016.12.28235 17.4% 0.79 188 74.5% 2.2 1030 HFC-365mfc 2014.08.06–
2016.12.28107 43.0% 0.44 96 79.2% 1.0 794 Table 1. Mixing ratios of HFCs measured at LAN and XGL during 2012–16.
The pollution mixing ratios of HFCs at LAN were much higher than the background values observed at XGL over the period 2012–16. The enhancements between mixing ratios measured at these two stations were 30%–50% (HFC-23, HFC-134a, HFC-143a, HFC-245fa and HFC-365mfc), 61% (HFC-125), 87% (HFC-227ea), 102% (HFC-152a), 179% (HFC-32) and 931% (HFC-236fa).
-
The interspecies correlations of enhanced mixing ratios of the 10 HFCs versus CO are shown in Table 2 and Fig. 2. HFC-23, HFC-32, HFC-125, HFC-152a and HFC-245fa presented significant correlations with the tracer CO (p <0.001), showing Pearson correlation coefficients larger than 0.31. HFC-134a and HFC-227ea also displayed significant but slightly worse correlations (p <0.025), with the correlation coefficients between 0.22 and 0.23. The enhanced mixing ratios of HFC-143a, HFC-236fa and HFC-365mfc did not significantly correlate with those of CO, which might be partly explained by less polluted samples compared to other HFCs (Table 2), and China’s emissions of these three HFCs were relatively small among all HFCs in previous studies (Kim et al., 2010; Yao et al., 2019). Therefore, the emissions of HFC-143a, HFC-236fa and HFC-365mfc are not discussed in this paper.
Compound Number of air samples R S (ppt/ppm) HFC-23 125 0.367(*) 35.3(19.4−51.3) HFC-32 128 0.557(*) 56.6(41.7−71.5) HFC-125 114 0.377(*) 18.8(10.1−27.4) HFC-134a 85 0.227(**) 48.9(3.0−94.7) HFC-143a 72 0.053 − HFC-152a 105 0.311(*) 13.5(5.45−21.6) HFC-227ea 107 0.220(**) 2.00(0.28−3.71) HFC-236fa 82 −0.117 − HFC-245fa 115 0.332(*) 2.10(0.99−3.21) HFC-365mfc 30 0.237 − *Correlation is significant at the 0.001 level (one-tailed).
**Correlation is significant at the 0.025 level (one-tailed).Table 2. Pearson correlation coefficient (R) and regression slope (S) of mixing ratio enhancements between HFCs and CO, with 95% confidence bounds, over the period 2012–16.
-
As listed in Table 3, the estimated total emissions of HFCs in the YRD from 2012 to 2016 were 2.4±1.4 Gg yr−1 for HFC-23, 2.8±1.2 Gg yr−1 for HFC-32, 2.2±1.2 Gg yr−1 for HFC-125, 4.8±4.8 Gg yr−1 for HFC-134a, 0.9±0.6 Gg yr−1 for HFC-152a, 0.3±0.3 Gg yr−1 for HFC-227ea and 0.3±0.2 Gg yr−1 for HFC-245fa, respectively. As shown in Fig. 3, among these seven HFCs, HFC-134a, HFC-32, HFC-23 and HFC-125 were the four main HFCs in the YRD, with proportions of 35%, 21%, 18% and 16%, respectively, while the total emissions of the other three HFCs shared less than 11%.
Compound Emissions by CO tracer ratio Emissions by bottom-up method Emissions by top-down method YRD(2012–16) YRDa(2015) Chinaa(2015) Chinab(2012–14) Chinac(2012) Chinad(2014) Chinae(2012–16) HFC-23 2.4±1.4 2.3 5.2 − 9.9 12.5 − HFC-32 2.8±1.2 1.2f 4.0f 11.9 0.2 3.1 7.5 HFC-125 2.2±1.2 1.2f 4.0f 11.5 0.3 3.2 9.5 HFC-134a 4.8±4.8 4.1 19.1 33.1 28.8 41.9 25.6 HFC-152a 0.9±0.6 − − 16.4 0.2 0.2 4.8 HFC-227ea 0.3±0.3 − − 0.2 0 0.1 0.9 HFC-245fa 0.3±0.2 − − 0.1 0.1 0.2 1.0 a Data from CCGGWP(2019); b Data from Fang et al. (2016); c Data from National Development and Reform Commission (2016); d Data from MEE (2019); e Data from Yao et al. (2019); f Here, only the consumption of HFC-410A in the room air-conditioning sector was considered. Table 3. Emissions (Gg yr−1) of each HFC from the YRD estimated by the CO tracer ratio method from 2012 to 2016 and comparison with previous studies.
Figure 3. HFC emission proportions in the YRD over the period 2012–16: (a) mass emissions; (b) CO2-equivalent emissions.
In terms of CO2-equivalent emissions, HFC-23 contributed about two thirds to the total CO2-equivalent emissions due to its high GWP (GWP100=14800) among all 10 of the HFCs in this study. The contribution of HFC-125 or HFC-134a to total CO2-equivalent HFC emissions exceeded 10%, and HFC-32 emissions contributed 3.6%, while the other three HFCs in total contributed less than 3%. Thus, the major HFC compounds in the YRD were HFC-134a, HFC-23, HFC-125 and HFC-32, with both largest mass emissions and CO2-equivalent emissions, and HFC-23 was the primary HFC considering CO2-equivalent emissions. The total CO2-equivalent emissions of HFCs reached 53 Gg yr−1 in the YRD, accounting for 2.3% of the CO2 emissions without landuse, land-use change and forestry in this region (CCGGWP, 2019).
HFC emissions in this study were also compared with results reported by previous studies using the bottom-up method, as listed in Table 3. The HFC emissions of the whole of China are listed because studies of the YRD are limited. HFC-23 and HFC-134a emissions in this study and those using the bottom-up method (CCGGWP, 2019) differed relatively little, at 4.2% and 14.5%, respectively. HFC-23 is mainly emitted as a byproduct of HCFC-22 production (Miller et al., 2010). There are nine fluorine chemistry plants producing HCFC-22 in the YRD, of which three are located in Jiangsu Province (contributing 18% of HFC-23 emissions) and six in Zhejiang Province (contributing 82% of HFC-23 emissions). HFC-134a is mainly used as a substitute for CFC-12 in the automobile air-conditioning sector (Hu et al., 2010). Over the period 2012–16, the HFC-134a emissions in the YRD derived from the tracer ratio method comprised 18.6% of those in China by the inverse modeling method, which was consistent with the percentage (approximately 20%) of vehicle numbers in the YRD of the whole country (National Bureau of Statistics of China, 2013a, 2014a, 2015a, 2016a, 2017a). There were understandable differences in HFC-125 and HFC-32 emissions between this study and those using the bottom-up method (CCGGWP, 2019) because only their consumption as mixed-blend HFC-410A from the room air-conditioning sector was considered by the bottom-up method. Besides their primary usage, HFC-125 and HFC-32 are also used as R-32 or other mixed blends, such as R-404a and R-507a (Li et al., 2011). The ratios of HFC-410A emissions (sum of HFC-32 emissions and HFC-125 emissions) by the bottom-up method (CCGGWP, 2019) to the sum of HFC-32 and HFC-125 emissions by the tracer ratio method or top-down method were all around 0.5, both in the YRD and at the national scale, indicating that half of HFC-32 and HFC-125 might be used as HFC-410A in the room air-conditioning sector in China.
It should be noted that up-to-date HFC emissions in China have been reported by a few different studies in the past two years (MEE, 2018; CCGGWP, 2019; Yao et al., 2019). However, there were significantly large differences (as large as 2 to 90 times, depending on specific HFC species) among the different studies, which could hardly be explained by their different time scales. Further studies are needed to reduce these discrepancies.
-
The proportion of HFC emissions in the YRD to national totals reflects the regional significance. Here, the emissions using the top-down method for the same period (Yao et al., 2019) and HFC-23 emissions in 2015 by CCGGWP (2019) were chosen to represent national levels. Therefore, the HFC emissions in the YRD counted for 46% (HFC-23), 38% (HFC-32), 23% (HFC-125), 19% (HFC-134a), 18% (HFC-152a), 36% (HFC-227ea) and 27% (HFC-245fa) of national totals. The total CO2-equivalent emissions of these seven HFCs in the YRD contributed 34% of the total HFC emissions in China. In comparison with previous studies that were carried out in other regions of China, the HFC-134a in the PRD accounted for only about 8% of national total emissions in 2010 (Wu et al., 2014), and the per capita HFC-134a emissions rate was 12.5 g yr−1, which is much lower than that in the YRD at 18.7 g yr−1. Therefore, the YRD constitutes a significant emissions source of HFCs in China.
The CO2-equivalent emission intensities of HFCs are illustrated in Fig. 4. Here, global total CO2-equivalent emissions were based on emissions derived from observations at remote AGAGE stations (Rigby et al., 2014; Engel et al., 2018), and the global population and GDP were derived from the International Statistical Yearbook (National Bureau of Statistics of China, 2013b, 2014b, 2015b, 2016b, 2017b). The national population and GDP data were from the China Statistical Yearbook (National Bureau of Statistics of China, 2013a, 2014a, 2015a, 2016a, 2017a). Compared with either the national or global emissions intensity, the YRD displayed as a higher HFC emitter. The per capita HFC CO2-equivalent emissions intensity of the YRD was nearly twice the national or global levels, at 240 kg per capita per year. In terms of emissions per unit area, the emissions intensity of the YRD was about 24 times the global value and 8 times the national value, amounting to more than 150 Mg km−2 yr−1. The emissions rate per million GDP in Chinese Yuan (CNY) was 3500 kg yr−1, twice the global value. In conclusion, the emissions intensity of HFCs in the YRD was notably high, whether in terms of per capita or per unit area or per unit GDP. It should be noted that HFC-23 emissions in the YRD contributed around two thirds to total HFC emissions in the YRD, or more than 1/5 to the national HFC emissions, in terms of CO2-equivalent. Hence, if HFC-23 emissions were totally eliminated, the emissions intensity of the YRD would be lower than national and global levels, and thus HFC-23 is the key HFC species when considering the mitigation potential of HFCs in the YRD, or even in China as a whole.
Figure 4. Comparison of total HFC CO2-equivalent emission intensities of the YRD, China and global levels from 2012 to 2016. National CO2-equivalent emissions of HFC-23 were derived from the bottom-up method (CCGGWP, 2019), and other HFC emissions were from the inverse modeling method (Yao et al., 2019). Global total CO2-equivalent emissions were based on emissions derived from observations at remote AGAGE stations (Rigby et al., 2014; Engel et al., 2018).
Compound | Observing period (YYYY.MM.DD) | LAN | XGL | GWP100 (Forster et al., 2007) | |||||
Number of air samples | Percentage of background | Mixing ratio enhancement (ppt) | Number of air samples | Percentage of background | Background mixing ratio(ppt) | ||||
HFC-23 | 2012.01.04– 2016.12.28 | 230 | 3.9% | 12.3 | 176 | 90.9% | 27.8 | 14800 | |
HFC-32 | 2012.01.04– 2016.12.28 | 233 | 2.6% | 19.0 | 184 | 75.0% | 10.6 | 675 | |
HFC-125 | 2012.01.04– 2016.12.28 | 223 | 12.1% | 10.4 | 179 | 95.5% | 17.3 | 3500 | |
HFC-134a | 2012.01.04– 2016.12.28 | 180 | 16.7% | 32.4 | 177 | 84.7% | 81.1 | 1430 | |
HFC-143a | 2012.01.04– 2016.12.28 | 227 | 44.1% | 5.3 | 187 | 96.3% | 16.7 | 4470 | |
HFC-152a | 2012.01.04– 2016.12.28 | 227 | 17.2% | 7.2 | 158 | 80.4% | 7.1 | 124 | |
HFC-227ea | 2012.01.04– 2016.12.28 | 219 | 21.5% | 0.95 | 161 | 94.4% | 1.1 | 3220 | |
HFC-236fa | 2012.01.04– 2016.12.28 | 232 | 41.8% | 1.2 | 181 | 90.6% | 0.13 | 9810 | |
HFC-245fa | 2012.01.04– 2016.12.28 | 235 | 17.4% | 0.79 | 188 | 74.5% | 2.2 | 1030 | |
HFC-365mfc | 2014.08.06– 2016.12.28 | 107 | 43.0% | 0.44 | 96 | 79.2% | 1.0 | 794 |