Characteristics and source apportionment of atmospheric volatile organic compounds in Beijing, China.

This paper focused on VOCs and their source apportionment in urban Beijing. Our monitoring measured 52 VOCs in July 2014 and January 2015. The concentration of VOCs was in the range of 14.5~95.2 ppb in July and 2.1~93.1 ppb in January, with the top five compounds of toluene (10.7%), ethane (6.9%), ethylene (6.3%), n-butane (5.7%), and propane (5.6%) in July and ethylene (14.7%), n-butane (14.2%), ethane (9.6%), propylene (8.0%), toluene (7.9%), and benzene (6.9%) in January. The ratio of VOCs to CO reached 0.059 in July and 0.022 in January on average. These differences implied a potential seasonal difference in the VOC source contribution. Then, we conducted a source apportionment study based on 21 major VOCs and CO by using probabilistic matrix factorization (PMF) receptor model. According to the similarity between the PMF analysis profiles and the known source profiles, combustion sources, petrochemical industry sources, solvent utilization sources, and gasoline evaporation sources were identified. The correlation coefficient (R) between the PMF analysis profile and the source profile reached 0.68~0.87 in July and 0.53~0.92 in January. The better apportionment performance in July was mainly due to the use of intensive VOC observations at a 3-h resolution. When we conducted another PMF source apportionment for July based on 12-h resolved concentration input, the R values decreased to 0.47~0.73. Thus, the PMF model depends heavily on the sample number of concentration inputs, and intensive observation is more propitious. Our PMF apportionment results showed that combustion sources, petrochemical industry, solvent utilization, gasoline evaporation, and other sources contributed ambient VOCs in Beijing urban areas of 13.7 ppb, 5.1 ppb, 7.7 ppb, 12.8 ppb, and 3.3 ppb in July and 13.2 ppb, 2.0 ppb, 5.7 ppb, 6.6 ppb, and 1.0 ppb in January, respectively, on a monthly average. These apportionment results match well with the 2013 VOC emission inventory calculated by this study, but also presented significant seasonal differences in the petrochemical industry and gasoline evaporation, in which VOC emissions strongly respond to environmental temperature.

[Multidimensional Verification of Anthropogenic VOCs Emissions Inventory Through Satellite Retrievals and Ground Observations].

In this paper, a regional emissions inventory of anthropogenic VOCs was established based on the traditional emissions factor method for the Beijing-Tianjin-Hebei (BTH) region, followed by a multidimensional calibration study based on regional satellite remote sensing information for formaldehyde and typical urban ground VOCs. Inventory calculations showed that the VOCs emissions in BTH in 2013, 2015, and 2017 were 2026700, 2073400, and 1934200 tons, respectively, comprising alkanes (29.83% to 30.72%), unsaturated hydrocarbons (16.54% to 17.68%), aromatic hydrocarbons (27.14% to 27.51%), aldehydes (8.75% to 9.52%), ketones (8.13% to 9.04%), and aldehydes and ketones lipids (5.13% to 6.60%). During 2013-2017, the emission of VOCs in Zhangjiakou, Qinhuangdao, and Hengshui increased slightly (1.10% to 1.66% per year); emissions in Xingtai and Handan decreased slightly (-1.46% to -1.12% per year); and emissions in Chengde, Tangshan, Baoding, and Cangzhou were stable. There trends were consistent with the inter-annual trend of satellite-derived HCHO column concentrations. However, in Beijing, Tianjin, Langfang, and Shijiazhuang, annual variations in VOCs emissions (-6.51%, -3.30%, 2.16%, and 0.11% per year) and HCHO column concentrations (-1.17%, 7.19%, -0.24%, and 6.68% per year) were observed, respectively. In the regional VOCs inventory, a good linear correlation (R>0.5) was achieved between the grid emissions of VOCs and HCHO column concentrations in urban areas, while the correlation was only 0.33 in suburban areas. This is mainly due to the important influence of secondary conversion of VOCs originating from natural sources to HCHO in suburban areas. In addition, ground-level VOCs concentrations were observed in the urban areas of Beijing and Handan, where the emission ratios (ERs) of VOCs and CO were regressed. The ERs of most hydrocarbons were in good agreement with the regressed ERs, but the ERs of ethane were significantly lower (-156% to -73%) and the ERs of aromatic hydrocarbons above C8 were relatively high (54% to 74%). In general, the regional anthropogenic VOCs emissions inventory established in this paper offers high accuracy and reliability.

[Characteristics of VOCs Emitted from the Rubber Tire Manufacturing Industry Based on the Inverse-Dispersion Calculation Method].

This study selected a rubber tire manufacturing factory located in the North China Plain, and conducted ambient volatile organic compounds (VOCs) observation tests, and calculated the emission of VOCs based on the inverse-dispersion calculation method. The monitoring results found significant differences in both VOC concentrations and chemical composition between the up-wind (background) and the downwind receptors. The average concentrations of VOCs measured by the background and receptors were 53.8 µg·m-3 and 127.5 µg·m-3, respectively. Propane (7.2 µg·m-3), cetone (7.5 µg·m-3), nonanal (12.7 µg·m-3), n-butane (4.9 µg·m-3), and acetaldehyde (2.7 µg·m-3) were the dominant components of background VOCs, and nonanal (43.5 µg·m-3), propane (11.4 µg·m-3), acetaldehyde (7.4 µg·m-3), hexane (11.9 µg·m-3), and n-butane (7.3 µg·m-3) were the dominant components of receptor VOCs. The difference in VOCs between the background and receptors is considered to reflect contributions from the factory, the main components of which were of alkanes (31.39%) and oxygenated organic compounds (33.15%). Using the ISC3 model, the relation coefficient γ between the downwind VOCs increment and the emissions of the factory was calculated for each receptor of each test based on the hourly average meteorological conditions during the observation period. Combining the relation coefficient γ with the difference in VOCs between the receptor and the background, we calculated VOC emission amounts from this factory of 152.8±188.2 t·a-1 and a VOC emission factor (EF) for the rubber tire manufacturing industry of VOC 101.9 g·tire-1. Our estimated EF was loser to EF of U. S. AP42 (55 g·tire-1), but greatly lower than the EF of China's reference (900 g·tire-1). Although our calculations had a relatively higher standard deviation, these results are helpful for better understanding the emission of VOCs from the rubber manufacturing industry. Based on our calculated EF, China's national VOCs emissions from the rubber tire manufacturing industry would be approximately 62.13 kt·a-1, which represents a significant potential contribution to ozone formation (130.87 kt·a-1), but the organic aerosol formation potential is small (0.86 kt·a-1).