Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum refinery in Pearl River Delta, China

STOTEN-21892; No of Pages 13 Science of the Total Environment xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum refinery in Pearl River Delta, China Zhijuan Zhang b,d, Hao Wang a, Dan Chen a,b, Qinqin Li a,b, Phong Thai c, Daocheng Gong a,b, Yang Li a,b, Chunlin Zhang a,b, Yinggang Gu b, Lei Zhou a, Lidia Morawska c, Boguang Wang a,b,⁎ a

Institute for Environmental and Climate Research, Jinan University, Guangzhou 511443, China Institute of Mass Spectrometer and Atmospheric Environment, Jinan University, Guangzhou 510632, China c International Laboratory for Air Quality and Health, Queensland University of Technology, GPO Box 2434, Brisbane, QLD 4001, Australia d Guangdong Provincial Engineering Research Center for On-line Source Apportionment System of Air Pollution, Guangzhou 510632, China b

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• The composite profiles of VOCs emitted from a petroleum refinery were obtained. • The VOC contributions to SOA formation related to petroleum refineries were estimated by three different methods. • The results demonstrated that the petroleum refinery is a potential important SOA source. • Toluene, benzene and o/m-ethyltoluene are of particular concern for SOA reduction in relation to petroleum refineries.

a r t i c l e

i n f o

Article history: Received 1 July 2016 Received in revised form 25 January 2017 Accepted 26 January 2017 Available online xxxx Keywords: Volatile organic compounds (VOCs) Petroleum refinery Secondary organic aerosol (SOA) Fractional aerosol coefficients (FAC) Secondary organic aerosol potential (SOAP) SOA yield

a b s t r a c t A campaign was carried out to measure the emission characteristics of volatile organic compounds (VOCs) in different areas of a petroleum refinery in the Pearl River Delta (PRD) region in China. In the refining area, 2methylpentane, 2,3-dimethylbutane, methylcyclopentane, 3-methylhexane, and butane accounted for N 50% of the total VOCs; in the chemical industry area, 2-methylpentane, p-diethylbenzene, 2,3-dimethylbutane, mdiethylbenzene and 1,2,4-trimethylbenzene were the top five VOCs detected; and in the wastewater treatment area, the five most abundant species were 2-methylpentane, 2,3-dimethylbutane, methylcyclopentane, 3methylpentane and p-diethylbenzene. The secondary organic aerosol (SOA) formation potential was estimated using the fractional aerosol coefficients (FAC), secondary organic aerosol potential (SOAP), and SOA yield methods. The FAC method suggests that toluene, p-diethylbenzene, and p-diethylbenzene are the largest contributors to the SOA formation in the refining, chemical industry, and wastewater treatment areas, respectively. With the SOAP method, it is estimated that toluene is the largest contributor to the SOA formation in the refining area, but o-ethyltoluene contributes the most both in the chemical industry and wastewater treatment areas. For the SOA yield method, aromatics dominate the yields and account for nearly 100% of the total in the three areas. The SOA concentrations estimated of the refining, chemical industry and wastewater treatment areas are 30, 3835 and 137 μg m−3, respectively. Despite the uncertainties and limitations associated with the three methods, the

⁎ Corresponding author at: Institute for Environmental and Climate Research, Jinan University, Guangzhou 511443, China. E-mail address: [email protected] (B. Wang).

http://dx.doi.org/10.1016/j.scitotenv.2017.01.179 0048-9697/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

2

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

SOA yield method is suggested to be used for the estimation of SOA formation from the petroleum refinery. The results of this study have demonstrated that the control of VOCs, especially aromatics such as toluene, ethyltoluene, benzene and diethylbenzene, should be a focus of future regulatory measures in order to reduce PM pollution in the PRD region. © 2017 Elsevier B.V. All rights reserved.

1. Introduction During the last decade, formation of atmospheric secondary organic aerosol (SOA) from gas phase precursors has received considerable attention (Emanuelsson et al., 2013; Feng et al., 2016; Hallquist et al., 2009; Hodzic et al., 2016; Huang et al., 2015; La et al., 2016; Li et al., 2016; Zhang et al., 2015). Traditionally, gas-phase oxidation of volatile organic compounds (VOCs) has been considered to be a major source of SOA in urban areas. Depending on location, time and specific source regions, SOA can be produced from both anthropogenic and biogenic VOCs. On a global scale anthropogenic VOCs, typically substituted aromatics (mainly toluene and xylene), are estimated to contribute between 1.4 and 8.6 TgC/yr to SOA (Hallquist et al., 2009; Henze et al., 2008), and other anthropogenic precursors such as the first-generation anthropogenic VOCs (AVOCs) are likely to generate several-fold more anthropogenic secondary organic aerosols (ASOA) (de Gouw et al., 2008; Kleinman et al., 2008; Shrivastava et al., 2008; Volkamer et al., 2006). State-of-the-art SOA models suggest that these AVOCs is only a minor contributor to SOA compared to biogenic VOCs (BVOCs). However, a significant fraction of the excess ASOA is formed from the first-generation oxidation products of AVOCs (Volkamer et al., 2006), and locally and regionally ASOA can supersede the biogenic secondary organic aerosols (BSOA). For example, in Los Angeles, the contribution of individual VOC groups to aerosol concentrations was estimated to be approximately 62% from aromatics, 13% from biogenic monoterpenes, 18% from alkanes, and 7% from alkenes (Grosjean and Seinfeld, 1989). Generally, VOCs are oxidized by OH radicals, ozone, and NO3 radicals into reaction products with low vapor pressure, which eventually form SOA in the atmosphere. Many studies have been carried out to investigate the importance of VOCs in the formation of urban SOA (Dechapanya et al., 2004; Dusek et al., 2013; Goodman-Rendall et al., 2016; Guo et al., 2017; Ortega et al., 2016; Pereira et al., 2015; Sun et al., 2016), especially VOCs from vehicle exhaust (Cohan et al., 2013; Ensberg et al., 2014; Gentner et al., 2012; Gordon et al., 2014; Huang et al., 2015; Kleindienst et al., 2002; Liu et al., 2015; Nordin et al., 2013; Ots et al., 2016; von Stackelberg et al., 2013). A recent study by Huang et al. (2015) concluded that high aromatic contents in the gasoline exhaust was responsible for the larger SOA yield and the intermediate-volatile organic compounds (IVOCs) in vehicle exhausts greatly contributed to the SOA formation in the urban atmosphere of Shanghai in China. Zhao et al. (2014) also found that primary IVOCs produced about 30% of newly formed SOA in the afternoon during the CalNex campaign in Pasadena, California. Kleindienst et al. (2002) investigated the SOA formation from the irradiation of simulated automobile exhaust and showed that 75–85% of the SOA was from reaction products of C6– C9 light aromatic compounds. However, there have been relatively few efforts on the SOA formation from VOCs emitted by petroleum refineries. Ambient gas-phase VOC concentrations largely depend on the source region and reactivity of the compounds. Researchers have found that, besides vehicular exhaust, industrial emissions (particularly from petroleum industries) have also become a major source of anthropogenic VOCs in urban areas (Bo et al., 2008; Cetin et al., 2003; Chen et al., 2006; Mo et al., 2015; Watson et al., 2001; Wei et al., 2014a; Wei et al., 2014b; Wei et al., 2008; Yen and Horng, 2009). With the rapid economic development, and the increasingly stringent vehicle emission standards, industrial sources are projected to be the second largest contributor of VOCs in China in 2020 (Wei et al., 2011). The VOCs emitted from petroleum refineries not only cause air pollution problems locally

but also play an important role in regional air pollution, contributing to production of photochemical ozone and SOA (Kalabokas et al., 2001; Song et al., 2008; Volkamer et al., 2006). For example, Yanshan petroleum complex (including oil refining section and chemical industry section) in Beijing, China has been identified as one of the largest contributors of photochemical ozone with the contribution of 15–28% in summer (Song et al., 2008). Several studies have monitored ambient VOC concentrations near the petroleum refineries (Cetin et al., 2003; Gariazzo et al., 2005; Hoyt and Raun, 2015; Kalabokas et al., 2001; Lin et al., 2004; Pandya et al., 2006; Rao et al., 2005; Sonibare et al., 2007; Wei et al., 2016). However, very little is known about the VOC emissions by different processes in a petroleum refinery. This not only makes it difficult to identify key emission sources in a petroleum refinery for emission control in the future, but also hinders the assessment of regional VOC emissions and their contributions to particulate pollution. Guangzhou is one of the most urbanized cities in the Pearl River Delta (PRD) region in China, with a heavy petroleum refining industry. The city has been suffering high levels of particulate pollution as well as high levels of ozone and other air pollutants. However, the role of the VOCs from petroleum refineries in the PRD remains poorly understood. Understanding the contribution of VOC emissions from petroleum refineries to the SOA formation will help identify the sources of PM pollution in the region as well as in China. Therefore, we conducted a field campaign in a petroleum refinery in Guangzhou during 2014, carrying out VOC measurements inside the refinery as well as in its surrounding areas. The goal of the study was to understand the fugitive VOC emissions from key zones inside the refinery and investigate the characteristics of alkanes, alkenes and aromatics emitted from these areas. In addition, the SOA formation potential of the measured VOCs was estimated using three popular methods based on the fractional aerosol coefficients (FAC), secondary organic aerosol potential (SOAP) and SOA yield, respectively. To the best of our knowledge, this is the first effort in which the SOA formation from alkanes, alkenes, and aromatics emitted from a petroleum refinery was estimated, and the results of this study would be very valuable for better understanding the characteristics of VOC emissions from the petroleum refining industry and for supporting policy-making with respect to particulate matter (PM) reduction. 2. Experiments and methods 2.1. VOCs sampling A petroleum refinery located in Huangpu district of Guangzhou city was selected for this study. It is one of the largest modern petroleum refineries in south China. And, it has an annual capacity of refining 13.2 million tons of crude oil and producing 0.22 million tons of ethylene. Furthermore, there are N 50 facilities for oil refining and chemical processes, including multiple typical facilities emitting VOCs. This study was conducted from August 5th to August 15th, 2014. Although many pollution control devices were used throughout the petroleum refinery, significant VOC emissions were reported for the oil refining processes (e.g., catalytic cracking) and wastewater treatment (Mo et al., 2015; Pandya et al., 2006; Wei et al., 2014b). In this study, air sampling was carried out at the refining area, chemical industry area and wastewater treatment area. The refining area consists of wax oil catalytic cracking unit (CCU-wax oil), atmospheric and vacuum distillation unit (AVDU), desulfurization unit (DU) and hydrogenation unit (HU). The chemical

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

industry area includes cracking and quenching zone (CQZ), separate cooling zone (SCZ), compression in alkaline washing zone (CAW), hydrogenation plant, hot zone, butadiene unit, aromatics extraction unit (AEU) and aromatic distillation unit (ADU). The wastewater treatment area comprises of oil separation tank (OST), regulation tank (RT), flotation tank, odor treatment unit, biologic treatment unit, flocculation and sedimentation unit (FSU), adsorption and sedimentation unit (ASU) and biological aerated filter (BAF). To characterize the fugitive emissions of VOCs within the petroleum refinery, we conducted sampling close to the major inner devices, which mostly represented the manufacturing technology of the petroleum refinery. Sampling was carried out at 35 different locations in the refinery (Fig. 1), and in total 15, 20, and 19 air samples were collected for the refining, chemical industry, and wastewater treatment areas, respectively. The air samples were collected using 3.2 L silica vacuum SUMMA canisters equipped with flow-limiting valves, and analyzed at the laboratory. The canisters had been precleaned with high purity nitrogen and evacuated with an automated canister cleaner. This allowed air sampling to be carried out on multi-sites at the same time, assuring the comparability of the VOCs concentrations at various locations. During the sampling, the canisters were held at about 1.5 m above the ground and close to the processing units for 15 to 20 min. To minimize the influence from the background and the nearby facilities, the sampling was conducted on days with calm wind conditions (wind speed was lower than 0.5 m s−1).

3

Technologies, USA) to introduce the effluent into a DB-624 column (60 m × 250 μm × 1.4 μm, Agilent Technologies Inc., USA) with an MSD to separate and analyze C4–C12 hydrocarbons and a PLOT (Al/ KCl) column (20 m × 320 μm × 3 μm, Dikma Technologies Inc., USA) with an FID to measure the C2–C4 hydrocarbons. The GC oven temperature was programmed initially at 38 °C for 3.5 min, increasing to 180 °C at 6 °C/min and holding for 15 min. The entire process took about 43 min. For QA/QC, the cryogenic pre-concentrator was baked after each analysis, and the GC column was also baked after analysis of every 20 samples. During the analysis, a standard gas including Photochemical Assessment Monitoring Stations (PAMS) standard mixture (55 NMHCs) and TO-15 standard mixture (65 compounds, from Spectra Gases Inc., NJ, USA) was used to calibrate the C2–C11 VOCs. The PAMS certified gas was diluted to 6 different concentrations (5, 10, 15, 20, 30 and 50 ppb) by a dilution system (Nutech 2202B, USA) to make the calibration curve. For the quantification of each target compound, the calibration curve was based on the relationship between the integrated peak area and the corresponding concentration (5 levels) with the correlation coefficient values N0.99. Furthermore, three VOCs (i.e. bromochloromethane, 1,4-difluorobenzene, and 1-bromo-3fluorobenzene) with known concentrations were used as internal standards for each sample to calibrate the system. In this study, a total of 61 VOC compounds were detected, including 33 alkanes, 12 alkenes, and 16 aromatics. The precision of each species was within 5%. The VOCs species measured by the GC–MS/FID system were listed in Table 1.

2.2. VOCs analysis 2.3. SOA formation estimation The gas samples were analyzed by a gas chromatograph-mass spectrometry/flame ionization detection (GC–MS/FID) system (GC, HP7820A; MSD, HP-5977E; Agilent Technologies Inc., USA) to determine VOC concentrations, following EPA TO-15 method. Before a sample was injected into the GC–MS/FID system, it was pretreated to remove N2, O2, CO2, CH4, CO, and H2O and then further concentrated using a double channel cryogenic preconcentrator (Model TH-PKU 300B, Tianhong Co, Ltd., China). The carrier gas was pure helium (purity N 99.999%). This system used a Dean Switch™ (Agilent

The formation of SOA is a complex process, and its many chemical pathways and reaction products remain unknown or not well understood. Because of the complexity of the chemical matrix of organic aerosols and the lack of direct chemical analysis methods for the majority of compounds comprising the organic aerosols, estimation of SOA formation potential has been mostly limited to indirect methods (Hoffmann et al., 1996). There are several approaches or methods for calculating the SOA formation potential, such as the fraction aerosol coefficient (FAC) method, the secondary organic aerosol potential (SOAP) method, and SOA yield estimation (Gentner et al., 2012). In this study, we have used these three methods to estimate the SOA concentration from the measured VOC compounds of the petroleum refinery. 2.3.1. FAC method The FAC method estimates SOA formation potential using reported FAC values (the ratio of the aerosol yield from a specific compound to its initial concentration (Grosjean and Seinfeld, 1989; Grosjean, 1992)) and the measured VOCs concentrations (Huang et al., 2011; Martín-Reviejo and Wirtz, 2005; Na et al., 2006; Sato et al., 2011). The potential of SOA formation (or SOA concentration, μg m−3) of a species i can be calculated by the following equation: P FACi ¼ VOC i;0  FAC i




ð1Þ

where FACi is the fractional aerosol coefficient of species i (%), (VOC)i,0 is the initial concentration of species i (μg m−3). It should be noted that, since all the samples were obtained close to the major inner devices, the concentration of a measured VOC was regarded as the (VOC)i,0. The values of FAC for the SOA formation potentials are listed in Table S1 in the Supplementary information.

Fig. 1. Schematic of sampling location in this study. (CAW: Compression in Alkaline Washing; ADU: Aromatic Distillation Unit; AEU: Aromatics Extraction Unit; CCU: Catalytic Cracking Unit; AVDU: Atmospheric and Vacuum Distillation Unit; OST: Oil Separation Tank; RT: Regulation Tank; FSU: Flocculation and Sedimentation Unit; ASU: Adsorption and Sedimentation Unit.)

2.3.2. SOAP (toluene weighted mass contributions) method The SOAP (toluene weighted mass contributions) method has been used to estimate the contributions of different precursor sources to SOA formation. The SOAP represents the propensity for an organic compound to form SOA when an additional mass emission of that compound is added to the ambient atmosphere relative to the SOA formed

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

4

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Table 1 VOC species detected in this study. No.

Hydrocarbons

No.

Hydrocarbons

No.

Hydrocarbons

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Carbon disulfidea 1,2-Dichloroethane DL=0.008 1,2-Dichloropropane DL=0.013 Bromochloromethane DL=0.015 1,1,2-Trichloroethane DL=0.036 Tetrachloroethane DL=0.134 Ethane DL=0.014 Propane DL=0.014 Isobutane DL=0.016 Butane DL=0.014 Cyclopentane DL=0.014 Isopentane DL=0.009 Pentane DL=0.008 2,2-Dimethylbutane DL=0.004 2,3-Dimethylbutane DL=0.003 2-Methylpentane DL=0.005 3-Methylpentane DL=0.003 Hexane DL=0.007 2,4-Dimethylpentane DL=0.004 Methylcyclopentane DL=0.003 2-Methylhexane DL=0.004

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

2,3-Dimethylpentane DL=0.003 Cyclohexane DL=0.002 3-Methylhexane DL=0.002 2,2,4-Trimethylpentane DL=0.001 Heptane DL=0.004 Methylcyclohexane DL=0.003 2,3,4-Trimethylpentane DL=0.003 2-Methylheptane DL=0.011 3-Methylheptane DL=0.004 Octane DL=0.003 Nonane DL=0.006 Decane DL=0.004 Undecane DL=0.007 1,3-Butadiene DL=0.008 Trichloroethylene DL=0.080 Propylene DL=0.012 Acetylene DL=0.008 trans-2-Butene DL=0.011 1-Butene DL=0.008 cis-2-Butene DL=0.024 1-Pentene DL=0.004

43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62

trans-2-Pentene DL=0.011 cis-2-Pentene DL=0.024 Isoprene DL=0.004 1-Hexene DL=0.013 Benzene DL=0.002 Toluene DL=0.002 Ethylbenzene DL=0.002 m,p-Xylenes DL=0.010 o-Xylene DL=0.002 Styrene DL=0.004 Isopropyl benzene DL=0.001 Propylbenzene DL=0.004 m-Ethyltoluene DL=0.003 p-Ethyltoluene DL=0.003 1,3,5-Trimethylbenzene DL=0.003 o-Ethyltoluene DL=0.004 1,2,4-Trimethylbenzene DL=0.006 1,2,3-Trimethylbenzene DL=0.002 m-Diethylbenzene DL=0.017 p-Diethylbenzene DL=0.016

DL = detection limit (ppbv). a Carbon disulfide was not included in the detected VOCs species.

when the same mass of toluene is added (Derwent et al., 2010; Li et al., 2015). SOAPs are expressed as an index relative to toluene which is set to equal 100. Toluene is chosen as the base compound for the SOAP scale because it is widely recognized as an important anthropogenic SOA precursor and has been well studied for SOA formation (Hu et al., 2008; Johnson et al., 2006; Kleindienst et al., 2007). SOAPi is the SOA formation potential parameter for species i (unitless) and can be calculated by the following equation (Derwent et al., 2010): SOAPi ¼

Increment in SOA mass concentration with species; i  100 ð2Þ Increment in SOA with toluene

SOAP values have been reported for a number of VOCs by Derwent et al. (2010). There are 38 VOC species selected in this study, as shown in Table S2 in the Supplementary information. The potential of SOA formation (μg m−3) of a species i can be calculated by the following equation: P SOAPi ¼

∑ðVOC i  SOAPi Þ  FAC toluene 100

ð3Þ

where (VOC)i is the mass contribution of a VOC source to species i (μg cm−3) (linking with the molar mass of VOC species and based on the ideal gas law. We convert the unit of VOC species from ppbv to μg m− 3). FACtoluene is the fractional aerosol coefficient of toluene (%) and it is chosen as 5.4% in this study. 2.3.3. SOA yield method The SOA formation potential of VOC emissions in the petroleum refinery is also calculated based on the SOA yields of the measured VOCs. For this method, VOC species are apportioned into five categories: aromatics, cycloalkanes, branched alkanes, straight-chain alkanes, alkenes. The SOA yield of each category of the petroleum refinery (YSOAj, unitless) is calculated by the following equation: Y SOA j ¼

∑ X i; j  Y i; j X TVOC


ð4Þ

where j represents the five categories of VOC species; YSOAjis the SOA yield of the petroleum refinery (unitless); Xi,j is the weight percent (by carbon) of species i in each category j which can be identified by measurements (weight C%); Yi,j is the yield of species i in each category j (by carbon, unitless); XTVOC is the weight percent (by carbon) of total

identified SOA precursors and unidentified species (weight C%). Identified non SOA precursors were excluded from total VOC emissions. The yield for each SOA precursor was referenced from Gentner et al. (2012), which listed the yields of known compounds using a combination of measured SOA yields derived from laboratory-chamber experiments and approximate SOA yields based on box modeling. The SOA yields derived from Gentner et al. (2012) are presented in Table S3 in the Supplementary information. It should be mentioned that, since the average organic particle concentration in the Guangzhou petroleum refinery was considerably higher than that in chamber experiments (around 10 μg m− 3), the SOA yields of the VOCs were recalculated using the semi-empirical model based on absorptive gas-particle partitioning of two semi-volatile products introduced by Odum et al. (1997). After the recalculation, the SOA yields of the compounds C6– C8 aromatics increased by an average of 19% (Huang et al., 2015). The SOA concentration was further calculated using the following equation:   P Y SOA ¼ ∑ Y SOA j  TVOC j

ð5Þ

where TVOCj is the total mass concentration of each category j (μg m−3). 3. Results and discussions 3.1. Characteristics of VOCs emitted from the petroleum refinery Ambient gas-phase concentrations of VOCs depend largely on the source region and reactivities of the compounds. Here, specified VOC concentrations were measured at a number of sites, including refining area, chemical industry area and wastewater treatment area within the petroleum refinery. Fig. 2 shows the average VOCs concentrations measured at different sampling areas. It can be noted that for the refining area, 2-methylpentane, 2,3-dimethylbutane, and methylcyclopentane had the highest concentrations. The compound 2methylpentane, with a concentration close to 8000 μg m−3, was mainly emitted from the Hydrogenation plant; 2,3-dimethylbutane was mostly emitted from the Desulfurizer installations; and methylcyclopentane was largely came from CCU-wax oil and AVDU facilities. For the chemical industry area, the three VOCs with the highest concentrations were p-diethylbenzene, 2-methylpentane, and m-diethylbenzene. The compound p-diethylbenzene, with a

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

5

Fig. 2. Average concentrations of VOCs emitted from the petroleum refinery. (The number represents species number, and the corresponding species can be found in Table 1.)

concentration as high as 14,000 μg m− 3 , was emitted from CAW area; 2-methylpentane mainly came from the Butadiene Unit and Hydrogenation plant; and m-diethylbenzene was also mainly emitted from CAW area. For the wastewater treatment area, the three most abundant VOC species were all alkanes: 2-methylpentane, 2,3-dimethylbutane, and methylcyclopentane. The compound 2methylpentane had a concentration of 10,000 μg m− 3 and was mainly emitted from Odor treatment unit; both 2,3-dimethylbutane and methylcyclopentane primarily came from Flotation tank and Biological treatment unit. Fig. 3 shows the characteristics of VOC emissions in the three sampling areas. It can be seen that 2-methylpentane had the highest proportion of all compounds at the three sampling sites; that is, 19.85%, 10.91% and 28.72%, respectively. For the refining area, 2-methylpentane, 2,3-dimethylbutane, methylcyclopentane, 3-methylhexane, and butane together

accounted for N 50% of the total VOCs emitted. However, for the chemical industry area, 2-methylpentane, p-diethylbenzene, 2,3dimethylbutane, m-diethylbenzene and 1,2,4-trimethylbenzene were the top five VOCs measured. And for the wastewater treatment area, the five most abundant species were 2-methylpentane, 2,3dimethylbutane, methylcyclopentane, 3-methylpentane and pdiethlbenzene. Furthermore, the abundant alkanes in the petroleum refinery might be caused by the special raw materials and products of the processes. The alkanes, alkenes, and aromatics emitted in the petroleum refinery in this work were also compared with other studies (Cheng et al., 2013; Liu et al., 2008; Mo et al., 2015), as presented in Fig. 4. It can be clearly seen that for the refining area, the alkanes measured in this work had the highest proportion of all compounds in the studied petroleum refineries, with a few differences due to the different production processes. For the chemical industry area, the aromatics accounted for

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

6

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Fig. 3. Weight percent of VOCs emitted from the petroleum refinery. (a) Refining area; (b) Chemical industry area; (c) Wastewater treatment area. (The number represents species number, and the corresponding species can be found in Table 1.)

N50% of all the VOCs detected, which is very similar to the results by Liu et al. (2008). 3.2. Estimation of SOA formation using FAC approach In this study, the SOA formation potential was estimated using FAC values and the measured concentration of each gaseous organic compound. Fig. 5 shows the calculated SOA formation potential for the refining area, chemical industry area, and wastewater treatment area. It can

Fig. 4. Comparison of proportion of alkanes, alkenes and aromatics in different petroleum refineries (by weight).

be seen that the largest contributors to SOA formation predicted by FAC is toluene, p-diethylbenzene, and p-diethylbenzene in the refining, chemical industry, and wastewater treatment areas, respectively. Furthermore, the top five contributors to the SOA formation in the refining area are found to be toluene, benzene, methyl cyclohexane, ethylbenzene and m,p-xylene. However, for the chemical industry and wastewater treatment areas, the five largest contributors to SOA are all aromatics due to their high concentrations at these two sampling areas. Fig. 6 shows the total SOA concentrations predicted using FAC and the measured VOCs concentrations for the three sampling areas. It can be noticed that the total SOA formed have the highest concentration in chemical industry area, followed by the wastewater treatment area and then the refining area. The total concentrations of SOA calculated from the measured VOCs are approximately 183, 3904 and 386 μg m−3 for the refining, chemical industry and wastewater treatment areas, respectively (Fig. 6a). Furthermore, the concentrations of SOA formed from these three groups of VOCs follow the order of aromatics, alkanes, and alkenes from high to low. The highest SOA concentration is predicted to be in the chemical industry area and is in part due to the higher concentrations of the aromatics; the aromatic compounds is estimated to contribute N90% of the SOA. The annual SOA formed in the three sampling areas of this petroleum refinery is estimated to be 1633 mg m−3 per year, a very high level for such a petroleum refinery. Since the Guangzhou petroleum refinery is very typical in the PRD region, this suggests that there is a huge threat from VOC emissions from such petroleum refineries to the local environment and more

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Fig. 5. SOA concentrations (in μg·m−3) predicted using FAC approach with measured VOCs concentrations.

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

7

8

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

dinitrophenol and nitrocatechol compounds, providing an important source of non-volatile SOA precursors (Derwent et al., 2010). The other three compounds are all aromatics: ethylbenzene, propylbenzene, and m-ethyltoluene. The remaining 5 aromatics have SOAPs ranging from 75.8 for m,p-xylenes to 13.5 for 1,3,5trimethylbenzene. As a class, the aromatics exhibit the greatest propensity to form SOA. Fig. 7 shows SOA concentrations estimated using the SOAP approach for the three areas within the petroleum refinery. It can be seen that the calculated concentrations of SOA formed from the VOCs for the chemical industry area are significantly higher than those for the other two areas because of their high VOC concentrations, especially the aromatics. The top five SOA contributors in the chemical industry area follow the order of o-ethyltoluene, propylbenzene, m-ethyltoluene, toluene, and ethylbenzene, with the SOA concentration contributed by o-ethyltoluene estimated to be 433 μg m−3. The five most abundant SOA contributors in the refining area are toluene, ethylbenzene, benzene, m,p-xylene, o-xylene, with the contribution from toluene estimated to be 58 μg m−3. The top five SOA contributors in the wastewater treatment area are similar to those in the chemical industry area, including o-ethyltoluene, propylbenzene, m-ethyltoluene, toluene, and oxylene. The total SOA concentrations estimated using SOAP approach for the three areas of the refinery are shown in Fig. 8a. It can be seen from the figure that the chemical industry area has the highest SOA concentration, up to 2245 μg m−3. For both the wastewater treatment and refining areas, the total SOA concentrations are around 220 μg m−3. Fig. 8b details the contributions of individual groups to the total SOA concentrations (%). As can be seen from the figure, the aromatics are the major contributor to SOA in all three areas of the petroleum refinery, accounting for N96% of the total VOCs. This suggests that the control of aromatics from refineries should be given the greatest attention among all VOCs in the PRD region.

Fig. 6. (a) Total potential SOA concentrations predicted using FAC with measured VOCs concentrations. (b) Contributions of each VOCs species to the total SOA predicted using FAC with measured VOCs concentrations.

stringent VOC control measures should be considered for petroleum refineries. It should be noted that the FAC approach does not take into account possible changes in chemistry (e.g. in photochemical reactions or gas/ particle partitioning) under different meteorological conditions. In addition, the SOA formation in the petroleum refinery may be under-predicted since a number of VOC groups that are potentially important contributors to SOA were not measured in this study, such as oxygenated-volatile organic compounds (OVOCs), intermediate-volatile organic compounds (IVOCs) or semi-volatile organic compounds (SVOCs). In this study, only 30 VOCs (including 11 alkanes, 3 alkenes and 16 aromatics) were considered during the SOA estimation using FAC approach. 3.3. Estimation of SOA formation using SOAP approach In the SOAP approach, because the SOA increments are referenced to toluene (Gentner et al., 2012), much of the influence of the uncertainties in absolute SOA concentrations have been removed. Table S2 listed the SOAPs of the measured compounds expressed relative to toluene (= 100). There are a total of 9 VOCs that have SOAPs over 90 and are all aromatics. They show varying OH-reactivities, ranging from unreactive benzene to highly reactive o-xylene. Four of these aromatics have SOAPs higher than toluene, with the SOAP of styrene ranked the highest. Styrene degrades rapidly to generate phenoxy radicals, which form nitrophenol and subsequently

3.4. SOA yield approach estimation To properly model the formation of SOA from the complex VOC species, detailed knowledge of molecular structures is needed, such as carbon number and alkane branching positions, which have been experimentally shown to significantly affect final yields (GoodmanRendall et al., 2016). During the SOA yield approach, the VOC species of refining area, chemical industry area, and wastewater treatment area in the petroleum refinery were apportioned into five categories according to their chemical classes. Fig. 9a presents their weight percent (by carbon) distributions of mass by chemical class in carbon number of different VOCs emissions. Obviously, it can be seen that the carbon numbers of VOCs in the refining area concentrated mainly within the intervals from C6 to C8, especially in C6, and most of them were branched alkanes. In comparison, VOCs from chemical industry area had a wider distribution of carbon number, ranging from C6 to C10 . More than half of the species were aromatics but branched alkanes also accounted for a high fraction. The carbon numbers of VOCs from the wastewater treatment area were mainly distributed within the range from C 6 to C 10 , especially in C6 , with branched alkanes being the top contributor. Fig. 9b shows the estimated SOA mass yields of different chemical classes. Clearly, the aromatics dominate the SOA mass yields and account for nearly 100% of the total yields in the three areas. Among the three sampling areas, the chemical industry area has the highest SOA yield of VOCs. The SOA yield of the refining area VOC emissions is larger than the yield of gas evaporation reported by Huang et al. (2015). Furthermore, the SOA yield of the VOC emissions from the chemical industry area is higher than the yield of gasoline exhaust reported by Gentner et al. (2012) and Huang et al. (2015). And, the SOA yield of the wastewater treatment area VOC emissions is

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Fig. 7. SOA concentrations estimated from the VOCs emitted from the petroleum refinery using SOAP approach.

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

9

10

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

comparative to the yield of diesel exhaust which combined the inferred S/IVOCs (Huang et al., 2015). In order to quantitatively compare the SOA formation from the petroleum refinery, the SOA concentrations from refining, chemical industry and wastewater treatment areas were further estimated, as indicated in Fig. 9c. Evidently, it can be noted from the figure that the chemical industry area presents the highest SOA concentration, up to 3835 μg m−3. The SOA concentrations obtained from the wastewater treatment area is up to 137 μg m−3. For the refining area, the SOA concentration is around 30 μg m−3.

3.5. Comparison of three SOA formation potential methods Table 2 lists the contributions of VOCs to the SOA concentrations for the three sampling areas. It can be seen that the aromatics contribute the largest proportion of the SOA concentrations, regardless of the approach used. Generally, alkanes contributed more than alkenes due to their large fraction in the measured VOCs. Table 3 is a summary of the total SOA concentrations formed and the top five contributors for the petroleum refinery using different approaches. It shows that special attention should be paid to the chemical industry area because of its highest SOA concentration no matter which method is used. Moreover, o-ethyltoluene, m-ethyltoluene, and toluene are all top contributors in the chemical industry and wastewater treatment areas based on the calculated results using FAC and SOAP approaches. Toluene, benzene, ethyl benzene, and m,p-xylene are of particular concern for the refining area. And, these VOCs have special health risk toward human beings (Durmusoglu et al., 2010), therefore, their control should be a priority for petroleum refineries. The results show some similarities because the parameters used in these methods are either from smog chamber experiments or from box modeling. However, it should be noted that there can be significant uncertainties when using these methods with measured concentrations to estimate the SOA formation potential. For the FAC method, there is lack of measurements for some VOC species that are potentially important in the aerosol formation process. In this study, there are only 30 VOCs having FAC parameters. Moreover, there is loss of organic vapors onto chamber walls during the chamber experiments (La et al., 2016; Zhang et al., 2014). As a consequence, this method may lead to significantly under-estimate ambient SOA concentrations. Furthermore, typically smog chamber conditions represent highly concentrated mixtures (e.g. high NOx and oxidant concentrations) relative to the real atmosphere and the experiments are conducted under conditions of high radiative flux and high temperatures. Thus, these parameters may represent optimal aerosol yields. For the SOAP method, the SOAP parameters are calculated by running the photochemical trajectory model (PTM) at conditions typical of northwest Europe (NOx = 15 μg m−3) (Derwent et al., 1998). And, the SOAP parameters vary with background environmental conditions, particularly NOx levels (Derwent et al., 2010), and it use toluene as the index compound for the SOA formed. Furthermore, for our study, there are only 38 VOCs having SOAP parameters, including 13 alkanes, 11 alkenes and 14 aromatic hydrocarbons. Thus, the SOA concentrations estimated from this method may also be under-estimated although correction has been made. For the SOA yield method, the model's source profiles use gasoline and diesel exhaust as a priori information. The SOA yield parameter is also dependent on the background environmental conditions, especially NOx levels. Moreover, in this work, there are only 27 VOCs considered when using SOA yield method to calculate the SOA concentration. Therefore, this method may also lead to formed SOA concentrations under-estimated. However, compared with FAC and SOAP methods, the SOA yield method considers the effect of molecular structures of VOCs species on the formation of SOA, and the initial conditions were used as a value relevant to chamber studies, urban areas, and downwind

Fig. 8. (a) Total SOA concentrations estimated using SOAP approach (in μg·m−3). (b) Contributions of each VOCs species to the total SOA concentrations using SOAP approach (%).

urban areas. Therefore, SOA yield method would be a good choice for estimating the SOA formation from petroleum refinery VOC emissions. 4. Conclusions In this study 61 VOCs were measured in samples collected in a petroleum refinery in the PRD region, China, including alkanes, alkenes and aromatics. Results showed that 2-methylpentane, 2,3-dimethylbutane were among the most emitted VOCs from the refining, chemical industry, and wastewater treatment areas. For the refining and wastewater treatment areas, the three most abundant VOC species were all alkanes, and for the chemical industry area, p-diethylbenzene had the highest concentration, close to 14,000 μg m−3. The FAC, SOAP, and SOA yield approaches were used to estimate the SOA concentrations formed from the VOC emissions. Results of all three approaches show that the aromatics contributed the largest proportion to the SOA formation potential. For the chemical industry area and wastewater treatment area, o-ethyltoluene, m-ethyl toluene and toluene were also among the top contributors, and for the refining area, toluene, benzene, ethyl benzene, and m,p-xylene were important as well. Overall, as in comparison with FAC and SOAP methods, the assumption conditions of SOA yield method are much more close to the real conditions, and the effect of molecular structures of VOCs species on the SOA formation is also considered. Thus, for the petroleum refinery the SOA yield method appears to yield a better results and it is suggested to be used for the estimation of SOA formed from the VOC emissions. Despite the uncertainties and limitations associated with the assessment of the SOA formation potential using the three approaches, this study demonstrates that VOCs, especially aromatics such as toluene,

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

11

Fig. 9. (a) Distributions of mass by chemical class in carbon number of different VOCs emissions; (b) calculated SOA yields based on C2-C12 VOCs measured in this study; (c) calculated SOA concentrations based on C2-C12 VOCs measured in this study.

benzene, o-ethyltoluene and m-ethyltoluene, should be given the priority when developing regulatory measures to control PM pollution due to the operations of petroleum refineries. In this study, all the samples were collected in a ten-day period, during which the VOC emissions may not be fully representative of

yearlong emissions from the petroleum refinery. However, during the study period the petroleum refinery operated according to the normal operation procedures for refineries, and therefore the samples collected in this study should reflect the general characteristics of typical source emissions of such refineries.

Table 2 Potential secondary organic aerosol attributed to VOCs (%). Location

Method

Aromatics

Alkenes

Alkanes

References

Refining area

FAC SOAP SOA yield FAC SOAP SOA yield FAC SOAP SOA yield FAC FAC

71.1 97.8 98.7 92.1 97.5 99.4 87.2 96.3 99.1 89.5 60.8

0.2 1.1 – – 0.1 – 0.1 0.7 – 5.1 12.1

28.7 1.0 1.3 7.9 2.4 0.6 12.8 3.0 0.9 5.7 7.3

This study

Chemical industry area

Wastewater treatment area

British Columbia Los Angeles

This study

This study

(Barthelmie and Pryor, 1997) (Grosjean, 1992)

References: Barthelmie, R. J., Pryor, S. C. 1997. Secondary organic aerosols: formation potential and ambient data. Sci. Total Environ. 205, 167–178. Grosjean, D. 1992. In situ organic aerosol formation during a smog episode: estimated production and chemical functionality. Atmos. Environ. 26A, 953–963.

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

12

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx

Table 3 Total SOA concentration formed from petroleum refinery VOC emissions and the top five contributors. Parameters

Refining area

Chemical industry area

Wastewater treatment area

FAC SOA (μg approach m−3) Top five contributors SOAP SOA (μg approach m−3)

183

3904

386

Toluene, benzene, methylcyclohexane, ethylbenzene, m,p-xylene 222

p-Diethylbenzene, m-diethylbenzene, o-ethyltoluene, m-ethyltoluene, toluene 2245

p-Diethylbenzene, o-ethyltoluene, toluene, m-ethyltoluene, m-diethylbenzene 233

Top five contributors SOA (μg

Toluene, ethylbenzene, benzene, m,p-xylene, o-xylene 30

o-Ethyltoluene, propylbenzene, m-ethyltoluene, toluene, ethylbenzene 3835

o-Ethyltoluene, propylbenzene, toluene, m-ethyltoluene, o-xylene 137

m−3) Top five contributors

C6–C8 aromatics

C9–C10 aromatics

C9–C10 aromatics

SOA yield

Acknowledgements The research is supported by National Natural Science Foundation of China (U1201232), the National Youth Natural Science Foundation of China (21406086), and the Fundamental Research Funds for the Central Universities (21614108). PT is funded by a QUT VC Fellowship. The authors are grateful to the Australia-China Centre for Air Quality Science and Management (ACC-AQSM) for helpful discussions in preparing the manuscript. We also thanks Mr. Yu Wang for his help with the drawing of the abstract graph.

Appendix A. Supplementary information Supplementary information to this article can be found online at http://dx.doi.org/10.1016/j.scitotenv.2017.01.179.

References Bo, Y., Cai, H., Xie, S.D., 2008. Spatial and temporal variation of historical anthropogenic NMVOCs emission inventories in China. Atmos. Chem. Phys. 8, 7297–7316. Cetin, E., Odabasi, M., Seyfioglu, R., 2003. Ambient volatile organic compound (VOC) concentrations around a petrochemical complex and a petroleum refinery. Sci. Total Environ. 312, 103–112. Chen, C.-L., Shu, C.-M., Fang, H.-Y., 2006. Location and characterization of emission sources for airborne volatile organic compounds inside a refinery in Taiwan. Environ. Monit. Assess. 120, 487–498. Cheng, S., Li, W., Wei, W., Li, G., Wang, H., Jiang, C., et al., 2013. Refinery VOCs seasonal composition and analysis of ozone formation potential. J. Beijing Univ. Technol. 39, 438–444. Cohan, A., Eiguren-Fernandez, A., Miguel, A.H., Dabdub, D., 2013. Secondary organic aerosol formation from naphthalene roadway emissions in the South Coast Air Basin of California. Int. J. Environ. Pollut. 52, 206–224. de Gouw, J.A., Brock, C.A., Atlas, E.L., Bates, T.S., Fehsenfeld, F.C., Goldan, P.D., et al., 2008. Sources of particulate matter in the northeastern United States in summer: 1. Direct emissions and secondary formation of organic matter in urban plumes. J. Geophys. Res.-Atmos. 113, D08301. Dechapanya, W., Russell, M., Allen, D.T., 2004. Estimates of anthropogenic secondary organic aerosol formation in Houston, Texas special issue of aerosol science and technology on findings from the Fine Particulate Matter Supersites Program. Aerosol Sci. Technol. 38, 156–166. Derwent, R.G., Jenkin, M.E., Saunders, S.M., Pilling, M.J., 1998. Photochemical ozone creation potentials for organic compounds in northwest Europe calculated with a master chemical mechanism. Atmos. Environ. 32, 2429–2441. Derwent, R.G., Jenkin, M.E., Utembe, S.R., Shallcross, D.E., Murrells, T.P., Passant, N.R., 2010. Secondary organic aerosol formation from a large number of reactive man-made organic compounds. Sci. Total Environ. 408, 3374–3381. Durmusoglu, E., Taspinar, F., Karademir, A., 2010. Health risk assessment of BTEX emissions in the landfill environment. J. Hazard. Mater. 176, 870–877. Dusek, U., ten Brink, H.M., Meijer, H.A.J., Kos, G., Mrozek, D., Röckmann, T., et al., 2013. The contribution of fossil sources to the organic aerosol in the Netherlands. Atmos. Environ. 74, 169–176. Emanuelsson, E.U., Hallquist, M., Kristensen, K., Glasius, M., Bohn, B., Fuchs, H., et al., 2013. Formation of anthropogenic secondary organic aerosol (SOA) and its influence on biogenic SOA properties. Atmos. Chem. Phys. 13, 2837–2855. Ensberg, J.J., Hayes, P.L., Jimenez, J.L., Gilman, J.B., Kuster, W.C., de Gouw, J.A., et al., 2014. Emission factor ratios, SOA mass yields, and the impact of vehicular emissions on SOA formation. Atmos. Chem. Phys. 14, 2383–2397.

Feng, T., Li, G., Cao, J., Bei, N., Shen, Z., Zhou, W., et al., 2016. Simulations of organic aerosol concentrations during springtime in the Guanzhong Basin, China. Atmos. Chem. Phys. 16, 10045–10061. Gariazzo, C., Pelliccioni, A., Di Filippo, P., Sallusti, F., Cecinato, A., 2005. Monitoring and analysis of volatile organic compounds around an oil refinery. Water Air Soil Pollut. 167, 17–38. Gentner, D.R., Isaacman, G., Worton, D.R., Chan, A.W.H., Dallmann, T.R., Davis, L., et al., 2012. Elucidating secondary organic aerosol from diesel and gasoline vehicles through detailed characterization of organic carbon emissions. Proc. Natl. Acad. Sci. 109, 18318–18323. Goodman-Rendall, K.A.S., Zhuang, Y.R., Amirav, A., Chan, A.W.H., 2016. Resolving detailed molecular structures in complex organic mixtures and modeling their secondary organic aerosol formation. Atmos. Environ. 128, 276–285. Gordon, T.D., Presto, A.A., May, A.A., Nguyen, N.T., Lipsky, E.M., Donahue, N.M., et al., 2014. Secondary organic aerosol formation exceeds primary particulate matter emissions for light-duty gasoline vehicles. Atmos. Chem. Phys. 14, 4661–4678. Grosjean, D., 1992. In situ organic aerosol formation during a smog episode: Estimated production and chemical functionality. Atmos. Environ. Part A 26, 953–963. Grosjean, D., Seinfeld, J.H., 1989. Parameterization of the formation potential of secondary organic aerosols. Atmos. Environ. (1967) 23, 1733–1747. Guo, H., Ling, Z.H., Cheng, H.R., Simpson, I.J., Lyu, X.P., Wang, X.M., et al., 2017. Tropospheric volatile organic compounds in China. Sci. Total Environ. 574, 1021–1043. Hallquist, M., Wenger, J.C., Baltensperger, U., Rudich, Y., Simpson, D., Claeys, M., et al., 2009. The formation, properties and impact of secondary organic aerosol: current and emerging issues. Atmos. Chem. Phys. 9, 5155–5236. Henze, D.K., Seinfeld, J.H., Ng, N.L., Kroll, J.H., Fu, T.M., Jacob, D.J., et al., 2008. Global modeling of secondary organic aerosol formation from aromatic hydrocarbons: high- vs. low-yield pathways. Atmos. Chem. Phys. 8, 2405–2420. Hodzic, A., Kasibhatla, P.S., Jo, D.S., Cappa, C.D., Jimenez, J.L., Madronich, S., et al., 2016. Rethinking the global secondary organic aerosol (SOA) budget: stronger production, faster removal, shorter lifetime. Atmos. Chem. Phys. 16, 7917–7941. Hoffmann, T., Odum, J., Bowman, F., Collins, D., Klockow, D., Flagan, R.C., et al., 1996. Aerosol formation potential of biogenic hydrocarbons. J. Aerosol Sci. 27 (Supplement 1), S233–S234. Hoyt, D., Raun, L.H., 2015. Measured and estimated benzene and volatile organic carbon (VOC) emissions at a major U.S. refinery/chemical plant: comparison and prioritization. J. Air Waste Manage. Assoc. 65, 1020–1031. Hu, D., Bian, Q., Li, T.W.Y., Lau, A.K.H., Yu, J.Z., 2008. Contributions of isoprene, monoterpenes, β-caryophyllene, and toluene to secondary organic aerosols in Hong Kong during the summer of 2006. J. Geophys. Res.-Atmos. 113 (n/a-n/a). Huang, X.H.H., Ip, H.S.S., Yu, J.Z., 2011. Secondary organic aerosol formation from ethylene in the urban atmosphere of Hong Kong: a multiphase chemical modeling study. J. Geophys. Res.-Atmos. 116 (n/a-n/a). Huang, C., Wang, H.L., Li, L., Wang, Q., Lu, Q., de Gouw, J.A., et al., 2015. VOC species and emission inventory from vehicles and their SOA formation potentials estimation in Shanghai, China. Atmos. Chem. Phys. 15, 11081–11096. Johnson, D., Utembe, S.R., Jenkin, M.E., 2006. Simulating the detailed chemical composition of secondary organic aerosol formed on a regional scale during the TORCH 2003 campaign in the southern UK. Atmos. Chem. Phys. 6, 419–431. Kalabokas, P.D., Hatzianestis, J., Bartzis, J.G., Papagiannakopoulos, P., 2001. Atmospheric concentrations of saturated and aromatic hydrocarbons around a Greek oil refinery. Atmos. Environ. 35, 2545–2555. Kleindienst, T.E., Corse, E.W., Li, W., McLver, C.D., Conver, T.S., Edney, E.O., et al., 2002. Secondary organic aerosol formation from the irradiation of simulated automobile exhaust. J. Air Waste Manage. Assoc. 52, 259–272. Kleindienst, T.E., Jaoui, M., Lewandowski, M., Offenberg, J.H., Lewis, C.W., Bhave, P.V., et al., 2007. Estimates of the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at a southeastern US location. Atmos. Environ. 41, 8288–8300. Kleinman, L.I., Springston, S.R., Daum, P.H., Lee, Y.N., Nunnermacker, L.J., Senum, G.I., et al., 2008. The time evolution of aerosol composition over the Mexico City plateau. Atmos. Chem. Phys. 8, 1559–1575. La, Y.S., Camredon, M., Ziemann, P.J., Valorso, R., Matsunaga, A., Lannuque, V., et al., 2016. Impact of chamber wall loss of gaseous organic compounds on secondary organic aerosol formation: explicit modeling of SOA formation from alkane and alkene oxidation. Atmos. Chem. Phys. 16, 1417–1431.

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179

Z. Zhang et al. / Science of the Total Environment xxx (2017) xxx–xxx Li, J., Xie, S.D., Zeng, L.M., Li, L.Y., Li, Y.Q., Wu, R.R., 2015. Characterization of ambient volatile organic compounds and their sources in Beijing, before, during, and after AsiaPacific Economic Cooperation China 2014. Atmos. Chem. Phys. 15, 7945–7959. Li, L., Tang, P., Nakao, S., Kacarab, M., Cocker 3rd., D.R., 2016. Novel approach for evaluating secondary organic aerosol from aromatic hydrocarbons: unified method for predicting aerosol composition and formation. Environ. Sci. Technol. 50, 6249–6256. Lin, T.-Y., Sree, U., Tseng, S.-H., Chiu, K.H., Wu, C.-H., Lo, J.-G., 2004. Volatile organic compound concentrations in ambient air of Kaohsiung petroleum refinery in Taiwan. Atmos. Environ. 38, 4111–4122. Liu, Y., Shao, M., Fu, L., Lu, S., Zeng, L., Tang, D., 2008. Source profiles of volatile organic compounds (VOCs) measured in China: Part I. Atmos. Environ. 42, 6247–6260. Liu, T., Wang, X., Deng, W., Hu, Q., Ding, X., Zhang, Y., et al., 2015. Secondary organic aerosol formation from photochemical aging of light-duty gasoline vehicle exhausts in a smog chamber. Atmos. Chem. Phys. 15, 9049–9062. Martín-Reviejo, M., Wirtz, K., 2005. Is benzene a precursor for secondary organic aerosol? Environ. Sci. Technol. 39, 1045–1054. Mo, Z., Shao, M., Lu, S., Qu, H., Zhou, M., Sun, J., et al., 2015. Process-specific emission characteristics of volatile organic compounds (VOCs) from petrochemical facilities in the Yangtze River Delta, China. Sci. Total Environ. 533, 422–431. Na, K., Song, C., Cocker Iii, D.R., 2006. Formation of secondary organic aerosol from the reaction of styrene with ozone in the presence and absence of ammonia and water. Atmos. Environ. 40, 1889–1900. Nordin, E.Z., Eriksson, A.C., Roldin, P., Nilsson, P.T., Carlsson, J.E., Kajos, M.K., et al., 2013. Secondary organic aerosol formation from idling gasoline passenger vehicle emissions investigated in a smog chamber. Atmos. Chem. Phys. 13, 6101–6116. Odum, J.R., Jungkamp, T.P.W., Griffin, R.J., Flagan, R.C., Seinfeld, J.H., 1997. The atmospheric aerosol-forming potential of whole gasoline vapor. Science 276, 96–99. Ortega, A.M., Hayes, P.L., Peng, Z., Palm, B.B., Hu, W., Day, D.A., et al., 2016. Real-time measurements of secondary organic aerosol formation and aging from ambient air in an oxidation flow reactor in the Los Angeles area. Atmos. Chem. Phys. 16, 7411–7433. Ots, R., Young, D.E., Vieno, M., Xu, L., Dunmore, R.E., Allan, J.D., et al., 2016. Simulating secondary organic aerosol from missing diesel-related intermediate-volatility organic compound emissions during the Clean Air for London (ClearfLo) campaign. Atmos. Chem. Phys. 16, 6453–6473. Pandya, G.H., Gavane, A.G., Bhanarkar, A.D., Kondawar, V.K., 2006. Concentrations of volatile organic compounds (VOCs) at an oil refinery. Int. J. Environ. Stud. 63, 337–351. Pereira, K.L., Hamilton, J.F., Rickard, A.R., Bloss, W.J., Alam, M.S., Camredon, M., et al., 2015. Insights into the formation and evolution of individual compounds in the particulate phase during aromatic photo-oxidation. Environ. Sci. Technol. 49, 13168–13178. Rao, B.P., Ansari, F., Ankam, S., Kumar, A., Pandit, V.I., Nema, P., 2005. Estimating fugitive emission budget of volatile organic carbon (VOC) in a petroleum refinery. Bull. Environ. Contam. Toxicol. 75, 127–134. Sato, K., Nakao, S., Clark, C.H., Qi, L., Cocker Iii, D.R., 2011. Secondary organic aerosol formation from the photooxidation of isoprene, 1,3-butadiene, and 2,3-dimethyl-1,3butadiene under high NOx conditions. Atmos. Chem. Phys. 11, 7301–7317. Shrivastava, M.K., Donahue, N.M., Pandis, S.N., Robinson, A.L., 2008. Effects of gas particle partitioning and aging of primary emissions on urban and regional organic aerosol concentrations. J. Geophys. Res. 113.

13

Song, Y., Dai, W., Shao, M., Liu, Y., Lu, S., Kuster, W., et al., 2008. Comparison of receptor models for source apportionment of volatile organic compounds in Beijing, China. Environ. Pollut. 156, 174–183. Sonibare, J.A., Akeredolu, F.A., Obanijesu, E.O.O., Adebiyi, F.M., 2007. Contribution of volatile organic compounds to Nigeria's Airshed by petroleum refineries. Pet. Sci. Technol. 25, 503–516. Sun, J., Wu, F., Hu, B., Tang, G., Zhang, J., Wang, Y., 2016. VOC characteristics, emissions and contributions to SOA formation during hazy episodes. Atmos. Environ. 141, 560–570. Volkamer, R., Jimenez, J.L., San Martini, F., Dzepina, K., Zhang, Q., Salcedo, D., et al., 2006. Secondary organic aerosol formation from anthropogenic air pollution: rapid and higher than expected. Geophys. Res. Lett. 33, L17811. von Stackelberg, K., Buonocore, J., Bhave, P., Schwartz, J., 2013. Public health impacts of secondary particulate formation from aromatic hydrocarbons in gasoline. Environ. Health 12, 19. Watson, J.G., Chow, J.C., Fujita, E.M., 2001. Review of volatile organic compound source apportionment by chemical mass balance. Atmos. Environ. 35, 1567–1584. Wei, W., Wang, S., Chatani, S., Klimont, Z., Cofala, J., Hao, J., 2008. Emission and speciation of non-methane volatile organic compounds from anthropogenic sources in China. Atmos. Environ. 42, 4976–4988. Wei, W., Wang, S., Hao, J., Cheng, S., 2011. Projection of anthropogenic volatile organic compounds (VOCs) emissions in China for the period 2010–2020. Atmos. Environ. 45 (38), 6863. Wei, W., Cheng, S., Li, G., Wang, G., Wang, H., 2014a. Characteristics of ozone and ozone precursors (VOCs and NOx) around a petroleum refinery in Beijing, China. J. Environ. Sci. 26, 332–342. Wei, W., Cheng, S., Li, G., Wang, G., Wang, H., 2014b. Characteristics of volatile organic compounds (VOCs) emitted from a petroleum refinery in Beijing, China. Atmos. Environ. 89, 358–366. Wei, W., Lv, Z., Yang, G., Cheng, S., Li, Y., Wang, L., 2016. VOCs emission rate estimate for complicated industrial area source using an inverse-dispersion calculation method: a case study on a petroleum refinery in Northern China. Environ. Pollut. 218, 681–688. Yen, C.-H., Horng, J.-J., 2009. Volatile organic compounds (VOCs) emission characteristics and control strategies for a petrochemical industrial area in middle Taiwan. J. Environ. Sci. Health A 44, 1424–1429. Zhang, X., Cappa, C.D., Jathar, S.H., McVay, R.C., Ensberg, J.J., Kleeman, M.J., et al., 2014. Influence of vapor wall loss in laboratory chambers on yields of secondary organic aerosol. Proc. Natl. Acad. Sci. 111, 5802–5807. Zhang, Q.J., Beekmann, M., Freney, E., Sellegri, K., Pichon, J.M., Schwarzenboeck, A., et al., 2015. Formation of secondary organic aerosol in the Paris pollution plume and its impact on surrounding regions. Atmos. Chem. Phys. 15, 13973–13992. Zhao, Y., Hennigan, C.J., May, A.A., Tkacik, D.S., de Gouw, J.A., Gilman, J.B., et al., 2014. Intermediate-volatility organic compounds: a large source of secondary organic aerosol. Environ. Sci. Technol. 48, 13743–13750.

Please cite this article as: Zhang, Z., et al., Emission characteristics of volatile organic compounds and their secondary organic aerosol formation potentials from a petroleum r..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.179