Speciated OVOC and VOC emission inventories and their implications for reactivity-based ozone control strategy in the Pearl River Delta region, China

Science of the Total Environment 530–531 (2015) 393–402

Contents lists available at ScienceDirect

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

Speciated OVOC and VOC emission inventories and their implications for reactivity-based ozone control strategy in the Pearl River Delta region, China Jiamin Ou a, Junyu Zheng b,⁎, Rongrong Li a, Xiaobo Huang b, Zhuangmin Zhong b, Liuju Zhong c, Hui Lin a,d a

Institute of Space and Earth Information Science, The Chinese University of Hong Kong, Hong Kong, China College of Environment and Energy, South China University of Technology, University Town, Guangzhou 510006, PR China c Guangdong Provincial Environmental Monitoring Center, Guangzhou 510045, PR China d Department of Geography and Resource Management, The Chinese University of Hong Kong, Hong Kong, China b

H I G H L I G H T S • • • •

Speciated OVOC and VOC emissions and OFPs for the PRD region were estimated Emission- and OFP-based speciated and source characteristics were different Industrial sources, gasoline vehicles and motorcycles were major OFP-based sources OFP cap was proposed to regulate VOC control policies

a r t i c l e

i n f o

Article history: Received 20 March 2015 Accepted 16 May 2015 Available online 5 June 2015 Editor: D. Barcelo Keywords: VOC OVOC Speciated emission inventory OFP Reactivity-based control

a b s t r a c t The increasing ground-ozone (O3) levels, accompanied by decreasing SO2, NO2, PM10 and PM2.5 concentrations benefited from air pollution control measures implemented in recent years, initiated a serious challenge to control Volatile Organic Compound (VOC) emissions in the Pearl River Delta (PRD) region, China. Speciated VOC emission inventory is fundamental for estimating Ozone Formation Potentials (OFPs) to identify key reactive VOC species and sources in order to formulate efficient O3 control strategies. With the use of the latest bulk VOC emission inventory and local source profiles, this study developed the PRD regional speciated Oxygenated Volatile Organic Compound (OVOC) and VOC emission inventories to identify the key emission-based and OFP-based VOC sources and species. Results showed that: (1) Methyl alcohol, acetone and ethyl acetate were the major constituents in the OVOC emissions from industrial solvents, household solvents, architectural paints and biogenic sources; (2) from the emission-based perspective, aromatics, alkanes, OVOCs and alkenes made up 39.2%, 28.2%, 15.9% and 10.9% of anthropogenic VOCs; (3) from the OFP-based perspective, aromatics and alkenes become predominant with contributions of 59.4% and 25.8% respectively; (4) ethene, m/p-xylene, toluene, 1,2,4trimethyl benzene and other 24 high OFP-contributing species were the key reactive species that contributed to 52% of anthropogenic emissions and up to 80% of OFPs; and (5) industrial solvents, industrial process, gasoline vehicles and motorcycles were major emission sources of these key reactive species. Policy implications for O3 control strategy were discussed. The OFP cap was proposed to regulate VOC control policies in the PRD region due to its flexibility in reducing the overall OFP of VOC emission sources in practice. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Rapid economic development in recent decades led to serious air pollution issues characterized by high concentrations of PM2.5 and frequent photochemical smog events in the city clusters of China (Li

⁎ Corresponding author at: B4-514, College of Environment and Energy, South China University of Technology, South Campus, University Town, Guangzhou 510006, PR China. E-mail address: [email protected] (J. Zheng).

http://dx.doi.org/10.1016/j.scitotenv.2015.05.062 0048-9697/© 2015 Elsevier B.V. All rights reserved.

et al., 2013; Xue et al., 2014; Huang et al., 2012). In order to combat severe air pollution, Chinese central government issued new National Ambient Air Quality Standard (MEP, 2012) in 2012 and National Air Pollution Prevention Action Plan in 2013 (MEP, 2013), the outcomes from these policies and unprecedented control measures are encouraging: annual PM2.5 concentrations in 2014 were observed to nation-wide decrease, and the decreasing rates were even larger than 10% in some regions (CEMS, 2014). In the Pearl River Delta (PRD) region, the data from the regional monitoring network showed that annual averaged concentrations of SO2, NO2, PM10 and PM2.5 in 2014 decreased by 66%,

394

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

17%, 20% and 13% compared to the levels in 2006. However, ground O3 levels, instead, presented a slightly increasing trend from 0.048 mg/m3 in 2006 to 0.057 mg/m3 in 2014 in the PRD (GDEMC, 2015). The slightly increasing trend in ozone levels is not surprising. Studies showed that the O3 formation was generally under VOC-limited regime in the PRD region, especially in urban areas (Zhang et al., 2008; Guo et al., 2009). Compared to the decreasing trends in the emissions of NOx or other pollutants, VOC emission appeared to be an increasing one in the past decade, which unexpectedly elevated ground O3 levels in a VOC-limited regime area (Lu et al., 2013). This implies that the control of VOC emissions is a key to alleviate O3 pollution as well as organic compositions in the PM2.5 concentration in the PRD region. Recognizing the importance of reducing VOC emissions, the PRD local governments have put the control of VOC sources on the top agenda in the next years. However, due to the diversity, complexity, and un-organized characteristics of VOC emission sources, the characterization and control of VOC sources have been challenging. Different from other primary pollutants like SO2 and NOx, VOCs as a group include many hundreds of species; each one can react at different rates with a different reaction mechanism and has different O3 formation potentials. This means that the control of VOC emission sources needs to consider not only their emission amounts but also their chemical reactivity (e.g., MIR) (Avery, 2006). In other words, a reactivitybased control approach might be more efficient in alleviating O3 pollution than a traditional mass-based emission control approach. The principle of reactivity-based control approach prioritizes the control of highly reactive VOCs that contribute more to the O3 formation compared to those VOCs with relatively low reactivity (Derwent et al., 2007). The reactivity-based approach has been recommended by US EPA to encourage states to consider the reactivity of VOC in the development of Ozone State Implementation Plans (US EPA, 2005). It has also been applied to regulate aerosol coating (Aerosol Coating Rule) by Air Resource Board (ARB) of California in order to reduce O3 formation from aerosol coating product emissions (CARB, 2000), and to control highly reactive VOCs in the Houston/Galveston Area, one of O3 episodes frequently occurring areas (Stoeckenius and Russell, 2005). In Europe, reactivity-based strategies for photochemical O3 control have also been investigated and discussed (Derwent et al., 2007). Currently, mass-based VOC control measures were generally adopted in the PRD or other regions of China. In order to take reactivity-based control strategies to alleviate O3 pollution, the knowledge of speciated VOC emissions is essential to identify major VOC emission sources with high reactivity and key reactive VOC species. Zheng et al. (2009a) utilized a 2006-based PRD regional emission inventory (Zheng et al., 2009b) to characterize speciated VOC emissions and attempted to touch the reactivity-based control idea for the PRD region. However, due to lack of representative local source profiles and limitation in source coverage in the 2006-based PRD regional emission inventory (Zheng et al., 2009a,b), along with changes in VOC emission source characteristics arising from industrial structure adjustment and implementation of VOC source control, there is a need to update the PRD regional speciated VOC emission inventory and to conduct further study on reactivity-based O3 control strategies for the PRD region. In recent years, great efforts have been made in compiling PRD regional bulk VOC emission inventory including a more comprehensive coverage of VOC sources and the use of local emission factors (Yin et al., 2015). Also, local source profiles for major VOC sources (Zheng et al., 2013; Ou et al., 2014; Liu et al., 2008; HKPU and SCUT, 2014) have been established by field sampling on the PRD local VOC emission sources. Specially, Oxygenated Volatile Organic Compound (OVOC) source profiles for some important industrial VOC sources were also developed. These efforts made it possible to conduct an in-depth study on reactivity-based O3 control strategies for the PRD region, including the development of OVOC emission inventory. The important role of OVOC in O3 formation in the PRD region has been demonstrated by a recent ambient VOC grid measurement study (Louie et al., 2013),

however, the OVOC emission characteristics have not been investigated in this region before. The main objectives of this manuscript are: (1) to develop PRD regional OVOC and VOC speciated emission inventories with the use of the latest VOC bulk emission inventory and recently established local source profiles; (2) to identify important OVOC and VOC sources and key VOC species in terms of their emission-based and reactivity-based contributions; (3) to provide recommendations on implementing reactivity-based O3 control strategies for the PRD region. The results and implications from this study can help the PRD local governments to take efficient O3 control strategies, which are of urgent need for their current VOC control practice. 2. Data and methods 2.1. Study domain The geographical areas for the PRD region in this study included the provincial capital Guangzhou, two special economic zones of Shenzhen and Zhuhai, Dongguan, Jiangmen, Foshan, Zhongshan, and parts of Huizhou and Zhaoqing. The location and geographic areas covered were illustrated in Fig. 1. The VOC sources covered in this study included: stationary combustion sources, on-road mobile sources, non-road mobile sources, industrial solvent use sources, industrial process sources, fuel storage and transportation source, biomass burning sources and biogenic sources. The detailed source classification can be referred to Yin et al. (2015). 2.2. Bulk emission inventory and source profiles As mentioned above, a 2010-based PRD bulk regional VOC emission inventory was used in this study (Yin et al., 2015). Compared to the previous PRD 2006-based VOC emission inventory used in Zheng et al. (2009a), there were a few improvements in the 2010-based PRD regional VOC emission inventory including: (1) more comprehensive source coverage. The missing sources such as some of industrial process and industrial solvent use and fuel storage and transportation sources were incorporated; (2) the overlapping sources in the previous study were eliminated; and (3) the use of the latest local emission factors for some important industrial VOC sources. For example, the VOC emissions for printing, furniture, shoe-making and other industrial sources were estimated based upon field investigation of VOC sources, and local emission factors developed for these sources. The details for these improvements can be found in Yin et al. (2015). The source profiles used in this study are almost from local source sampling in the PRD regions conducted in recent years, including VOC profiles of industrial solvent usage by Zheng et al. (2013) and HKPU & SCUT (2014), on-road mobile sources by HKPU & SCUT (2014) and Ou et al. (2014), and industrial process and other sources by Liu et al. (2008). Establishment of these source profiles followed strict QA/QC procedures and most of the source profiles have been peer-reviewed and publicly published. The source profiles used in this study to develop speciated VOC emissions were summarized in Appendix A. 2.3. Speciation of OVOC and VOC emission inventories Speciated OVOC and VOC emissions from anthropogenic sources were estimated on the basis of PRD regional 2010-based bulk emission inventory using sector-based source profiles for VOC speciation listed in Appendix A. The general method for bulk VOC emissions resolved into speciated VOC emissions was shown in Eq. (2.1) (Zheng et al., 2009a): n X ! ! Ei ¼ TE j  P j j¼1

ð2:1Þ

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

395

Fig. 1. Location and scope of the PRD region.

! in which, Ei is a vector representing the total emissions from all source sectors for the ith chemical specie (e.g., benzene, propane); TEj is the ! total emissions at the jth source sector; and P j is a vector containing source profile information for the jth source sector; and n is the number of source sectors conducted in this study. It must be pointed out that the speciated emissions for the OVOCs species were estimated only for industrial solvent usage, household solvent usage and architectural paints. The speciated OVOC emissions from mobile sources and other VOC sources were not estimated because the emission factors used in Hong Kong and the PRD region did not include OVOC emissions, and also OVOC species were not available in the current source profiles for these sources. 2.4. Estimation of biogenic speciated VOC emissions Speciated VOC emissions from biogenic sources were estimated by the Model of Emissions of Gases and Aerosols from Nature (MEGAN) (Guenther et al., 2006) for Hong Kong and the PRD region. The general method for estimating the emissions of isoprene and other biogenic VOCs (including some OVOCs species) can be summarized as follows:

2.5. Estimation of Ozone Formation Potential In this study, the Maximum Increment Reactivity (MIR) values were adopted to estimate the ozone formation potentials (OFPs) of different VOC species. MIR is a widely used concept to reflect the chemical reactivity of different VOC species, and the OFP is especially used in assessment of the roles of different VOC species in the O3 formation. In this study, on the basis of speciated VOCs emission inventory, the OFPs of individual VOC species were calculated by multiplying the emission by the corresponding MIR value (Zheng et al., 2009a), n

OFP i ¼ ∑ j¼1 Ei; j  MIRi

ð2:3Þ

In which, OFPi is the OFP of the ith VOC chemical species; Ei,j is the speciated emission of the ith chemical species at the jth source sector; and MIRi is the MIR value of the ith VOC chemical species. The updated MIR values from Cater (2008) were adopted in this study. 3. Results and discussion 3.1. Speciated OVOC emissions

E ¼ ε  γCE  γAge  γSM  ρ

ð2:2Þ

in which, ε is the emission factor; γ represents the activity factor, which is calculated by γCE (the average canopy influence of leaf PPFD and temperature), γAge (the canopy-weighted average factor), γSM (the emission factor influenced by soil moisture); and ρ is canopy loss and production. Meteorological and land cover data are fundamental information to drive the MEGAN to estimate biogenic emissions. In this study, the Weather Research and Forecast (WRF) model was coupled with Meteorology–Chemistry Interface Processor (MCIP) to provide the meteorological inputs, such as temperature and solar radiation. The land cover data, including Leaf Area Index (LAI) and percentage of each Plant Functional Type (PFT) were acquired from default land values inherent in the MEGAN.

In the PRD region, the OVOC emissions in 2010 for industrial solvents, household solvents, architectural paints and biogenic sources were estimated to be 297.4 kt, accounting for 19.9% of the total VOC emissions. It should be noted that OVOC emissions from other sources (e.g., mobile sources) were not estimated in this study, therefore, emissions of OVOCs and their contributions in VOC emission inventory were underestimated in this study. As shown in Table 1, methyl alcohol, acetone and ethyl acetate contributed to 24.3%, 18.2% and 11.1% of OVOC emissions. Anthropogenic emission of methyl alcohol was not estimated and its high emission was contributed by biogenic sources. Acetone was mainly emitted from industrial solvents and biogenic sources. In the shoe-making industry in the PRD region, acetone was widely used as constituents of adhesives (Zheng et al., 2013; HKPU & SCUT, 2014). Over 95% of ethyl acetate emission was emitted from industrial solvents since it was

396

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

Table 1 OVOCs emission for the PRD region in 2010 (unit: kt/yr). OVOCs species Methyl alcohol Acetone Ethyl acetate 2-Butanone Ethanol Acetic acid butyl ester Acetaldehyde Isopropyl alcohol Methyl isobutyl ketone 1-Methoxy-2-propanol 1-Butanol 2-Methoxy-1-propanol Aceticacid, methylester 1-Methoxy-2-propylacetate Dimethoxymethane Acetic acid 1-methylpropyl ester Diethyl ether Formaldehyde cis-3-Hexenal Cyclohexanone Formic acid Acetic acid 2-Ethyl-1-hexanol Carbonicacid, dimethylester α,α-Dimethylbenzenemethanol 2-Methyl-2-propanol trans-2-Hexenal cis-3-Hexenyl acetate cis-3-Hexenol Linalool 4-Methyl-benzaldehyde 1-Phenyl-1-butene Homosalate Others a a

Industrial solvents

Household solvents

Architectural paints

22.4 32.0 23.3 1.4 12.7

1.1 0.0 0.0 5.7 0.0

8.8 0.9 0.0 0.7 1.1

6.8 4.5 4.5 4.4 4.4 4.2 3.7 2.7 2.6 0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.4

0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.8 0.4 0.0

Biogenic sources

Total

72.2 21.7

72.2 54.0 32.9 23.5 18.1 13.7 10.3 6.9 4.5 4.5 4.4 4.4 4.2 3.7 3.5 3.0 2.4 2.3 2.3 2.2 1.8 1.8 1.6 1.6 1.5 1.4 1.4 1.4 1.4 1.3 1.2 1.0 1.0 6.0

0.2 10.3 10.3

2.3 2.3 2.2

0.0

0.0 1.8 1.8

1.6 1.6 0.8 1.4

0.0 0.0 0.5 0.0

0.0 0.0 0.2 0.0 1.4 1.4 1.4 1.3

1.1 1.0

0.0 0.0

5.7

0.3

0.0 0.0 1.0

Emissions of other OVOC species can be referred to Appendix B.

typically used in gravure printing, furniture-making and shoe-making industries in the PRD region. Besides, 2-butanone, ethanol, acetic acid butyl ester, acetaldehyde and isopropyl alcohol (IPA) made up 7.9%, 6.1%, 4.6%, 3.5% and 2.3% of OVOC emissions, respectively. More than 90% of 2-butanol, acetic acid butyl ester and IPA were emitted from industrial solvents, e.g., shoe-making, printing and metal-surface coating sectors (Zheng et al., 2013; HKPU & SCUT, 2014). For most areas of the world, OVOC emissions were not regularly reported. The Air Toxic emission inventories in San Diego County (SDC) of California (Air Pollution Control District, San Diego County, 2014) and Greater Vancouver Regional District (GVRD) of Canada (Environment

Emission contribution

75%

PRD US-SDC Canada-GVRD

50%

25%

0%

Canada, 2014) included a few OVOC species, they were used for comparison with estimations in this study. Since OVOC emissions from mobile source were not estimated in this study, OVOC emissions from mobile sources were excluded in the following comparison. Reported OVOC emissions were normalized to their total emissions in each inventory. Due to differences in estimated species, strictly speaking, emission contributions cannot be compared directly. But the estimated emission contributions were still useful to show the dominant OVOC species in different areas. As shown in Fig. 2, as an important constituent of biogenic emissions, methyl alcohol was consistently the major OVOC species in the RPD region as well as SDC and GVRD. Acetone was not quantified in the inventory of SDC, but it was consistently the major OVOC species in the PRD and GVRD in Canada. Acetone was mainly emitted from industrial sectors and biogenic sources in the PRD region and GVRD. In addition, 2-butanone, acetaldehyde and IPA were also notable OVOC constituents in the PRD, SDC and GVRD, but the contribution of IPA in SDC was much higher, which might be attributed to the high contribution of printing industry to the local OVOC emissions. Two important OVOC species in the PRD, i.e., ethyl acetate and ethanol, were not estimated in SDC and GVRD. In brief, methyl alcohol and acetone were common high-contributing species in these areas since they were typical biogenic species or raw materials which were widely used in industries. For other species that were less important, some discrepancies were observed given the differences in the quantified species and local source characteristics. 3.2. Speciated VOC emissions and their source characterization

Fig. 2. Comparison of major OVOC species in the PRD and other areas.

According to the 2010 VOC emission inventory in the PRD region, anthropogenic sources made up 70.8% of the total emissions while biogenic emission contributed to the remaining 29.2%. Table 2 listed the

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

397

Table 2 Speciated anthropogenic VOC emissions and OFPs for the PRD region in 2010 (unit: kt/yr). Species

Emissions

Species

MIR

kt/yr Alkanes n-Hexane Isopentane Cyclohexane 2-Methylpentane 3-Methylpentane n-Pentane n-Heptane n-Butane 3-Methylhexane 2-Methylhexane Ethane Propane Methylcyclopentane Isobutane Methylcyclohexane n-Octane n-Nonane n-Undecane 2,3-Dimethylbutane n-Decane Other alkanes a Alkenes Ethene Acetylene Propene 1-Butene Other alkenes a Aromatics Toluene Benzene m/p-Xylene Ethylbenzene Styrene 1,2,4-Trimethylbenzene m-Ethyltoluene o-Xylene 1,2,3,5-Tetramethylbenzene 1,2,3-Trimethylbenzene p-Ethyltoluene 1,3,5-Trimethylbenzene o-Ethyltoluene 4-Ethyl-1,2-dimethyl-benzene 1,3-Dimethylbenzene o-Cymene n-Propylbenzene p-Diethylbenzene Isopropylbenzene Other aromatics a Halocarbons 1,2-Dichloroethene Other halocarbons a OVOCs Ethylacetate Acetone 2-Butanone Acetic acid butyl ester Ethanol Isopropyl alcohol Other OVOCs Others a

298.2 27.2 26.0 24.1 22.6 18.1 16.4 13.7 13.5 12.9 12.3 12.1 11.4 10.7 8.9 8.5 7.2 6.6 5.5 4.9 4.8 31.0 115.7 36.9 22.0 19.9 8.0 28.8 414.5 70.7 54.2 37.4 34.9 32.5 29.0 20.7 17.8 14.9 11.9 9.0 8.1 7.9 7.8 7.8 7.5 5.9 5.9 5.5 25.1 16.2 8.4 7.9 168.0 32.9 32.3 23.3 13.7 7.8 6.9 51.1 44.3

OFPs kt/yr

Alkanes Isopentane 2-Methylpentane n-Hexane 3-Methylpentane Cyclohexane Methylcyclopentane n-Pentane 3-Methylhexane Other alkanes a Alkenes Ethene Propene 1-Butene trans-2-Butene cis-2-Butene 1,3-butadiene 2-Methyl-2-butene trans-2-Pentene Acetylene 1-Pentene cis-2-Pentene Other alkenes a Aromatics m/p-Xylene Toluene 1,2,4-Trimethylbenzene m-Ethyltoluene 1,2,3-Trimethylbenzene 1,2,3,5-Tetramethylbenzene o-Xylene Ethylbenzene 1,3,5-Trimethylbenzene 1,3-Dimethylbenzene 4-Ethyl-1,2-dimethyl-benzene Styrene o-Ethyltoluene o-Cymene p-Ethyltoluene 1-Ethyl-2,3-dimethylbenzene Benzene 2-Ethyl-1,4-dimethyl-benzene p-Diethylbenzene 1-Methyl-3-propylbenzene p-Xylene Other aromatics a OVOCs 2-Butanone Ethylacetate MethylIsobutylketone Other OVOCs a

1.36 1.4 1.14 1.69 1.14 2.05 1.22 1.5

8.88 11.6 9.57 15.2 14.3 12.5 14.2 10.5 0.95 7.07 10.3

7.56 3.93 8.83 7.39 11.9 9.26 7.58 2.96 11.8 9.73 7.54 1.66 5.54 5.43 4.39 10.2 0.69 7.54 4.39 7.08 5.78

1.45 0.59 3.78

347.9 35.4 31.7 31.0 30.5 27.4 21.9 20.0 19.3 130.7 933.4 328.1 230.7 76.7 45.2 34.9 32 31.1 30.4 20.9 17.8 17.4 68.0 2146.6 282.8 277.9 256 153.1 141.9 137.9 134.6 103.3 95.8 75.9 59.1 53.9 44.0 40.5 39.4 38.9 37.4 30.4 26.0 19.6 16.4 82.7 185.0 33.8 19.4 17.1 114.7

Emissions of others species can be referred to Appendix C.

speciated VOC emissions and their OFPs from anthropogenic sources. Fig.3a presented the source contributions to anthropogenic emissions. For source contribution characteristics, as shown in Fig. 3a, industrial-related sources and on-road mobile sources were two important source categories of anthropogenic VOC emissions. Industrial solvents and industrial process were responsible for 38.8% and 17.9% of

anthropogenic VOC emissions, while gasoline vehicles, motorcycles and diesel vehicles made up 14.5%, 10.5% and 3.2%. For speciated characteristics, as shown in Fig. 4, 39.2% of anthropogenic VOCs were aromatics. Alkanes, OVOCs and alkenes accounted for 28.2%, 15.9% and 10.9% of anthropogenic VOCs respectively. The top 10 emission contributing species were showed in Fig. 5. Aromatic

398

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

(a) Source contributions to anthropogenic VOC emissions (%) Biomass burning, 3.7 Others, 1.2

Fuel distribution Architectural &transport, paints, 3.5 3.0 Household solvents, 2.3

Stationary combustion, 1.4

Gasoline vehicles, 14.5

Diesel vehicles, 3.2 Motorcycles , 10.5

Industrial solvents, 38.8

Industrial process, 17.9

(b) Source contributions to anthropogenic OFPs (%)

3.3. Source- and species-based Ozone Formation Potentials

Fuel distribution & transport, 4.5

Biomass burning, 5.1

Architectural paints, 3.1 Household solvents, 1.6

Others, 1.5 Stationary combustion, 1.8

Gasoline vehicles, 18.4 Industrial solvents 33.4

Diesel vehicles, 4.2 Motorcycles, 13.7 Industrial process, 12.8

Fig. 3. Source contributions of (a) anthropogenic emissions and (b) anthropogenic OFPs in the PRD region for the year of 2010.

species took up 6 places in the 10 top anthropogenic VOC species. Toluene and benzene were the largest and second largest emission contributing species with contributions of 6.7% and 5.1% respectively, while m/ p-xylene (3.5%), ethyl benzene (3.3%), styrene (3.1%) and 1,2,4trimethyl benzene (2.7%) were also high contributing aromatic species in anthropogenic emissions. In addition, ethene and n-hexane made up 3.5% and 2.6% of anthropogenic emissions, which was the highestcontributing species in alkanes and alkenes, respectively. Ethyl acetate and acetone were the highest and second highest contributing OVOC species, with contributions of 3.3% and 3.1% to anthropogenic emissions.

Alkanes Anthropogenic Emission

28.2%

10.9%

39.2%

15.9%

Alkenes Aromatics

Anthropogenic OFP 9.6%

25.8%

59.4%

5.1%

OVOCs Halocarbons

0%

20%

The source characteristics for the top 10 species were summarized below. Toluene, benzene and m/p-xylene were mainly emitted from industrial sources and on-road mobile sources. Industrial sources and onroad mobile sources made up comparable contributions to these three aromatics. For example, industrial sources contributed 51.8% of toluene emissions, while on-road mobile sources made up 34.7%. Over 95% of styrene was emitted from industrial sources. In particular, woodfurniture manufacturing and metal surface coating were major industrial sectors contributing to styrene emissions. 1,2,4-Trimethyl benzene was mainly generated from on-road mobile sources, e.g., gasoline vehicles, motorcycles and diesel vehicles. Ethene was a combustion product. It was emitted by fuel combustion in mobile sources, petrochemical industry and biomass burning in the PRD region. Acetone was widely used in the shoe-making industrial sector and architectural paints in the PRD region, which were two major sources of acetone emissions. More than 95% of ethyl acetate was attributed to industrial sources, including wood-furniture manufacturing, printing, shoe-making and metal surface coating. For n-hexane, industrial sources were major emission sources and to a less extent, on-road mobile sources. Industrial sources and on-road mobile sources were major sources of these key emission contributing species.

40%

60%

80%

100%

Others

Fig. 4. Speciated characteristics of anthropogenic emissions and OFPs in the PRD region for the year of 2010.

On the basis of speciated VOC emissions, species- and source-based OFP were estimated. Anthropogenic sources accounted for 64.7% of the total OFPs in the PRD region, while biogenic source was responsible for 35.3%. Compared with the emission-based contribution, the contribution of anthropogenic sources in OFPs decreased while biogenic source contribution increased. For individual anthropogenic sources, their OFP-based contributions were also different from the emissionbased contributions. As shown in Fig. 3b, industrial-related sources and on-road mobile sources were still the two major source categories contributing to the anthropogenic OFPs. But compared to their emission-based contributions, the OFP-based contribution of industrial-related sources decreased while the contribution of on-road mobile source increased. Specifically, industrial solvents and industrial process contributed to 33.4% and 12.8% of the anthropogenic OFPs, which were lower than their contributions to anthropogenic emissions (38.8% and 17.9%). On the other hand, the OFP-based contributions of gasoline vehicles and motorcycles were 18.4% and 13.7%, higher than their emission-based contributions of 14.5% and 10.5%. Changes in emission-based and OFP-based contributions were attributed to the speciated characteristics of emissions sources. Compared with the source profiles of industrial-related sources, there were generally higher proportions of reactive alkenes and aromatics such as ethene, propene and 1,2,4-trimethyl benzene in source profiles of on-road mobile sources, leading to the increase of on-road mobile source contribution in anthropogenic OFPs. Regarding the speciated characteristics of anthropogenic OFPs, as shown in Fig. 4, OFP-based contributions of VOC groups were different from their emission-based contributions. Aromatic and alkenes made up 59.4% and 25.8% of anthropogenic OFPs, which were higher than their corresponding emission-based contributions of 39.2% and 10.9%. On the contrary, alkanes and OVOCs each contributed to 28.2% and 15.9% of anthropogenic emissions, but they only made up 9.6% and 5.1% of the anthropogenic OFPs. It was observed that the speciated characteristics of anthropogenic OFPs were quite different from anthropogenic emissions. The top 10 high contributing species to anthropogenic OFPs were showed in Fig. 5, along with their emission-based contributions and MIR values. Ethene was the largest contributing species to anthropogenic OFPs with contribution of 9.0%. From the perspective of emission, toluene was the largest contributor and its emission was about 2 times of ethene. However, the MIR value of ethene (8.88) is more than 2 times of toluene (3.93). Combined with emission magnitude and reactivity,

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

(a) Top 10 VOC species contributing to anthropgenic emissions Toluene

(b) Top 10 VOC species contributing to anthropgenic OFPs, their emission-based contributions and MIR value

6.7

Benzene

9.1

Ethene

5.1

7.8

m/p-Xylene

m/p-Xylene

3.5

Toluene

Ethene

3.5

1,2,4Trimethylbenzene

Ethylbenzene

3.3

7.7

7.1

6.4

Propene 4.2

Ethylacetate

3.1

m-Ethyltoluene

Styrene

3.1

1,2,3Trimethylbenzene

3.9

Acetone

3.1

1,2,3,5Tetramethylbenzene

3.8

1,2,4Trimethylbenzene

2.7

o-Xylene

n-Hexane

2.6

Ethylbenzene

0

2

4

6

399

8

3.7

2.9

0

Emission contribution (%)

OFP contribution Emission contribution MIR

5

10

15

OFP contribution (%) / Emission contribution (%) /MIR value

Fig. 5. Key contributing species to (a) anthropogenic VOC emissions and (b) anthropogenic OFPs in the PRD region for the year of 2010.

Anthropogenic emission/OFP reduction (%)

the OFP of ethene was higher than that of toluene and became the largest contributing species to anthropogenic OFPs. Among the top 10 VOC species contributing to anthropogenic emissions, 5 species were not in the list of top 10 OFP-based contributing species. The 5 species were benzene, ethyl benzene, ethyl acetate, styrene, acetone and n-hexane, of which MIR values were lower than 2. For example, benzene was the second largest species in anthropogenic emissions with percentage of 5.1%. But the MIR of benzene was only 0.69 and therefore its contribution to anthropogenic OFPs was just 1.0% and ranked as the 23th place in OFP contributions. On the contrary, the other 5 species with lower emissions but higher reactivity were ranked in the top 10 OFPcontributing species. The 5 species were propene, m-ethyl toluene, 1,2,3-trimethyl benzene, 1,2,3,5-tetramethyl benzene and o-xylene, of which MIR values were larger than 7. For instance, propene was ranked as the 19th highest contributing species to anthropogenic emissions but it was the 5th largest OFP-based contributor. Since OFP was an 100 90 80 70 60 50 40 30 20 10 0

80

83

80

52 Emission OFP

Emission-based control

Reactivity-based control

Fig. 6. Reduction rates of anthropogenic emissions and OFPs from emission-based and OFP-based controls.

integration of both emission magnitude and reactivity, there were obvious variations between the emission-based and OFP-based key species. High emission contributing species does not necessarily have the same importance in OFP contributions.

3.4. Implications for reactivity-based ozone control strategy in the PRD region As shown in Section 3.2 and 3.3, differences were observed in the emission-based and OFP-based speciated characteristics and source contributions. These differences may lead to different policy implications for O3 control strategies. In this study, emissions of more than 130 VOC species were estimated. Toluene, benzene, ethene, m/p-xylene and the other 40 VOC species together made up 80% of the VOC emissions, in a decreasing order. From the perspective of emission-based control, the above 44 high emission-based contributing species should be targeted. Consequently, as shown in Fig. 6, control of the 44 species could lead to 80% of VOC emission reduction, which can benefit 83% reduction of anthropogenic OFPs. On the other hand, from the perspective of reactivity-based control, high OFP-contributing species would be targeted with priority instead of high emission-contributing species. Species-based OFPs revealed that ethene, m/p-xylene, toluene, 1,2,4trimethyl benzene and the other 24 species were the high OFPcontributing species that contributed to 80% of the anthropogenic OFPs. If emissions of the 28 high OFP-contributing species were abated completely, VOC emission could be reduced by 52% but the reduction of anthropogenic OFPs could be up to 80%. In other words, as shown in Fig. 6, the reactivity-based control is more efficient than the emission-based approach by achieving similar OFP reduction via less emission reduction.

400

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

Ethene m/p-Xylene Toluene 1,2,4-Trimethylbenzene Propene m-Ethyltoluene 1,2,3-Trimethylbenzene 1,2,3,5-Tetramethylbenzene o-Xylene Ethylbenzene 1,3,5-Trimethylbenzene 1,3-Dimethylbenzene 1-Butene 4-Ethyl-1,2-dimethyl-benzene Styrene trans-2-Butene o-Ethyltoluene 1-Ethyl-2,3-dimethylbenzene p-Ethyltoluene o-Cymene Benzene cis-2-Butene Isopentane 2-Butanone 2-Methyl-2-butene 1,3-butadiene n-Hexane 2-Methylpentane 3-Methylpentane

0%

20%

Stationary combustion On-road mobile sources Household solvents

40%

60%

Industrial process Gasoline evaporation Architectural paints

80%

100%

Industrial solvents Non-road mobile sources Biomass burning

Fig. 7. Key OFP-contributing species and their emission sources.

The effectiveness of reactivity-based control lies in the targeting on high OFP-contributing species instead of high emission-contributing species. The 28 high OFP-contributing species and their emission source contributions were listed in Fig. 7. The source contribution patterns for the 28 key species could be summarized as: (1) for ethene, propene and 1-butene, over 80% of their emissions were attributed to on-road mobile sources, industrial process and biomass burning; (2) for m/p-xylene, toluene, m-ethyl toluene and o-xylene, the major contributing sources were on-road mobile sources and industrial solvents; (3) for 1,2,4-trimethyl benzene, 1,2,3-trimethyl benzene, o-ethyl toluene, pethyl toluene and isopentane, on-road mobile sources were dominant, and to a lesser extent, industrial solvents; (4) for 1,2,3,5-tetramethyl benzene, ethyl benzene, 4-ethyl-1,2-dimethyl benzene, 1-ethyl-2,3-dimethyl benzene, o-cymene and 2-butanone, industrial solvents was the major emission source; (5) 1,3-diemthyl benzene was emitted from architectural paints, industrial solvents and on-road mobile sources; (6) for trans-2-butene, cis-2-butene and 2-methyl-2-butene, gasoline evaporation, on-road mobile sources and biomass burning were major contributors; (7) for benzene, 1,3-butadiene and nhexane, industrial process was the most important contributor, followed by on-road mobile sources, industrial solvents or biomass burning; (8) for 2-methyl pentane and 3-methyl pentane, on-road mobile sources, industrial solvent and industrial process were the major contributing sources. Generally speaking, on-road mobile sources (mainly gasoline vehicles and motorcycles) and industrial solvents were the two most important sectors responsible for emissions of the 28 high OFP-contributing species, but speciated source characteristics were also notable suggesting that specific control on other sources was also needed. In other words, the reactivity-based control strategy in the PRD region should target on the 28 key reactivity species from the

above emission sources in order to more efficiently alleviate O3 pollution. In recent years, the PRD region issued a series of air pollution control measures which has been proved efficient in reducing primary pollutant concentrations and even PM2.5 levels (GDEMC, 2015). Nevertheless, the slightly increase trend of ground-level O3 concentrations initiated a serious challenge in reducing VOC emissions in this region. Facing this challenge, the PRD local government has taken the control of VOC emission sources as a priority task in the next five year plan, and issued a series of control policies, measures, or emission standards, such as tightening standards on vehicular emissions and fuels, restrictions on motorcycles, encouraging usage of water-borne or low VOC-content paints and solvents, gasoline recovery and others (GDEPD, 2013; Government of Guangdong Province, 2014). By reviewing these control policies or measures, it was found that most of these policies or measures focus on the reduction of total VOC emissions. The chemical reactivity in different VOC species and reactivity-based source controls were not taken into account at all. Specifically, realizing the fact that a large amount of VOC emissions are emitted from solvent-borne varnishes or finishes for printing, furniture and other industries, ambitious goals on substitution with water-borne varnishes have been proposed: for any new-registered industrial enterprises of metal surface coating, the proportion of water-borne or other low-emission solvent should be more than 50% and even 80% for auto-mobile manufacturing (Government of Guangdong Province, 2013). In the adjacent Hong Kong, the VOC Regulation has been issued since 2007 to regulate the total amount of VOCs in the contents of certain products (HKEPD, 2007). However, experiences from some developed countries have proved that such emission-based control might lead to unexpected increase of OFPs though the total VOC emissions

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

were reduced. For example, solvent producers could increase the content of higher-reactivity chemical species in the water-borne varnish formulations or resin systems to meet the standard on total VOC amount in the product, but consequently increased the overall OFPs of the final product, as occurred in California before (Wendoll, 2003; Avery, 2006). These experiences and research findings conducted in this study provide important policy implications for the PRD region to formulate O3 control policies. It is the time for the PRD local governments to shift the current emission-based O3 control strategy to the reactivitybased O3 control one. However, the PRD governments need to be cautious in taking a reactivity-based control strategy approach. Currently, in a reactivity-based O3 strategy, regulating the contents in VOC species with high reactivity in varnishes, solvents or fuels or substituting high reactivity VOC species with low reactivity ones is a typical approach for reducing the OFPs (Avery, 2006). However, such an approach may not be feasible in practice since some high reactivity species are necessary to maintain the performance of a product. In this study, we proposed that the PRD governments can establish the OFP caps in regulating VOC control policies in order to alleviate O3 pollution. In other words, the governments can use the OFP as an important metric in formulating O3 control policies and evaluating their effectiveness. Producers or emission contributors can take various measures flexibly to reach an overall decrease of OFPs under the regulated OFP caps. For example, for the solvent use source, one of the major VOC contributors to O3 formation, solvent producers can either reduce the contents of VOC species with high reactivity or remain to use some high reactivity VOC species, while solvent users can either reduce the total VOC emissions (including OVOC emissions) by installing various control measures or adopt the solvents with low reactivity. Whatever measures they take, an important standard is to evaluate whether or not these measures can meet the regulated OFPs. This might be a more flexible reactivity-based approach in practice to regulate O3 control policies. 4. Summary and conclusions Control of VOC emission sources has become a priority work in the PRD region in order to alleviate increasing regional O3 pollution and reduce the secondary organic aerosol compositions in PM2.5 pollution. Characterization of major VOC sources and identification of key VOC species are of great importance in formulating VOC control measures and policies. This study developed a 2010-based PRD regional speciated OVOC and VOC emission inventories with the use of the latest bulk VOC emission inventory and local source profile database established in recent years. Emission-based major VOC sources and key VOC species contributors were characterized and identified. Especially, major VOC sources with high reactivity and key reactive VOC species were identified in terms of OFPs, which provide important policy implications for implementing reactivity-based O3 control strategies in the PRD region. The OVOCs emissions from industrial solvents, household solvents, architectural paints and biogenic sources were estimated, and the methyl alcohol, acetone and ethyl acetate were their major species. In terms of anthropogenic VOC emissions, the top 10 VOC species were toluene, benzene, m/p-xylene, ethene, ethyl benzene, ethyl acetate, styrene, acetone, 1,2,4-trimethyl benzene and n-hexane. Aromatics, alkanes, OVOCs and alkenes made up 39.2%, 28.2%, 15.9% and 10.9% of anthropogenic VOCs emissions. Industrial solvents, industrial process, gasoline vehicles and motorcycles were major anthropogenic VOC emission contributors, responsible for 38.8%, 17.9%, 14.5%, and 10.5% of total anthropogenic VOC emissions, respectively. However, in terms of OFP-based contributions, the shares of industrial solvents and industrial process decreased to 33.4% and 12.8%, while gasoline vehicles and motorcycles increased to 18.4% and 13.7%. Aromatics and alkenes become predominant groups in OFP contributions with up to 59.4% and 25.8%, while alkanes and OVOCs only made up 9.6% and 5.1% of anthropogenic OFPs. The top 10 contributing species to anthropogenic OFPs ,

401

were ethene, m/p-xylene, toluene, 1,2,4-trimethyl benzene, propene, m-ethyl toluene, 1,2,3-trimethyl benzene, 1,2,3,5-tetramethyl benzene, o-xylene and ethyl benzene. Moreover, analysis showed that by targeting the high OFPcontributing species rather than the high emission-contributing species, reactivity-based control was more efficient than the emission-based approach by achieving similar OFP reductions via less emission reduction. Ethene, m/p-xylene, toluene, 1,2,4-trimethyl benzene and the other 24 high OFP-contributing species were the key reactive species that should be targeted in reactivity-based control strategy. The OFP cap was proposed to regulate VOC control policies, this might be a more flexible approach in reducing the overall OFP of VOC emission sources in practice. Acknowledgments This work was supported by the NSFC-National Distinguished Young Scholar Science Fund (No. 41325020), Chinese Academy of Science -The PRD gridded emission inventory study under the atmospheric haze key project (No. XDB05020303), the Innovation and Technology Fund of Hong Kong (No. ITS/042/12FP), and the National Key Basic Research Program of China (No. 2015CB954103). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2015.05.062. References Air Pollution Control District, San Diego County, 2014. California Air Toxic “Hot Spot” Information and Assessment Act (AB2588)—2013 Air Toxic “Hot Spots” Program Report for San Diego County. Available via:. http://www.sdapcd.org/homepage/ public_part/workshops/2013_THS_Program_Report.pdf. Avery, R.J., 2006. Reactivity-based VOC control for solvent products: more efficient ozone reduction strategies. Environ. Sci. Technol. 40 (16), 4845–4850. California Air Resources Board (CARB), 2000. Updated informative digest. Adoption of Amendments to the Regulation for Reducing Volatile Organic Compound Emissions From Aerosol Coating Products and Tables of Maximum Incremental Reactivity (MIR) Values, and Adoption of Amendments to ARB Test Methodp. 310. Carter, W., 2008. Reactivity Estimates for Selected Consumer Product Compounds. California: Center for Environmental Research and Technology, College of Engineering, University of California. February 19, 2008. Available via:. http://www.cert.ucr. edu/~carter/pubs/aminrep.pdf. China Environmental Monitoring Station (CEMS), 2014. Report on air quality in Jing-jin-ji. Yangtze River Delta and Pearl River Delta region (from 2014 January to December). Derwent, R.G., Jenkin, M.E., Passant, N.R., et al., 2007. Reactivity-based strategies for photochemical ozone control in Europe. Environ. Sci. Pol. 445–453. Environment Canada, 2014. Air and climate indicators—ambient levels of ozone. Available via https://www.ec.gc.ca/indicateurs-indicators/default.asp?lang=en& n=9EBBCA88-1. Government of Guangdong, 2013. Action Plan on Clear Air Act in Guangdong (2013 to 2015). Available via http://www.gd.gov.cn/govpub/bmguifan/201303/t20130326_ 176695.htm. Government of Guangdong, 2014. Action plan on prevention and control of air pollution in Guangdong Province. Available via http://zwgk.gd.gov.cn/006939748/201402/ t20140214_467051.html. Guang Dong Environmental Protection Department (GDEPD), 2014. The 2013 Action Plan for the 12th Air Pollution Control Plan of the Guangdong Province. Available via http://zwgk.gd.gov.cn/006940060/201303/t20130327_370666.html. Guangdong Environmental Monitoring Centre (GDEMC), 2015. Monitoring Results of the Pearl River Delta Regional Air Quality Monitoring Network. Guenther, A., Karl, T., Harley, P., et al., 2006. Estimates of global terrestrial isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature). Atmos. Chem. Phys. 6, 3181–3210. Guo, H., Jiang, F., Cheng, H.R., et al., 2009. Concurrent observations of air pollutants at two sites in the Pearl River Delta and the implication of regional transport. Atmos. Chem. Phys. 9, 7343–7360. Hong Kong Environmental Protection Department (HKEPD), 2007. Air Pollution Control-Volatile Organic Compounds Regulations (CAP 311 W). Available via: http://www.epd.gov.hk/epd/english/environmentinhk/air/prob_solutions/voc_reg. html#point_3. Hong Kong Polytechnic University and South China University of Technology (HKPU & SCUT), 2014. Characterization of VOC sources and integrated photochemical ozone analysis in Hong Kong and the Pearl River Delta region. Final Report Submitted to Hong Kong Environmental Protection Department.

402

J. Ou et al. / Science of the Total Environment 530–531 (2015) 393–402

Huang, X.F., Zeng, L.W., Yu, G.H., et al., 2012. Black carbon aerosol characterization in a coastal city in South China using a single particle soot photometer. Atmos. Environ. 51 (5), 21–28. Li, Y., Lau, A.K.H., Fung, J.C.H., et al., 2013. Systematic evaluation of ozone control policies using an ozone source apportionment method. Atmos. Environ. 76, 136–146. Liu, Y., Shao, M., Fu, L.L., et al., 2008. Source profiles of volatile organic compounds (VOCs) measured in China: part I. Atmos. Environ. 42, 6247–6260. Louie, P.K.K., Ho, J.W.K., Tsan, R.C.W., et al., 2013. VOCs and OVOCs distribution and control policy implications in Pearl River Delta Region, China. Atmos. Environ. 76, 125–135. Lu, Q., Zheng, J.Y., Ye, S.Q., et al., 2013. Emission trends and source characteristics of SO2, NOx, PM10 and VOCs in the Pearl River Delta region from 2000 to 2009. Atmos. Environ. 76, 11–20. Ministry of Environmental Protection (MEP), P.R. of China, 2012. China National Ambient Air Quality Standard (GB 3095-2012). Available via http://kjs.mep.gov.cn/hjbhbz/ bzwb/dqhjbh/dqhjzlbz/201203/t20120302_224165.htm. Ministry of Environmental Protection (MEP), P.R. of China, 2013. The State Council issues action plan on prevention and control of air pollution introducing ten measures to improve air quality. Available via http://english.mep.gov.cn/News_service/infocus/ 201309/t20130924_260707.htm. Ou, J.M., Feng, X.Q., Liu, Y.C., et al., 2014. Source characteristics of VOCs emissions from vehicular exhaust in the Pearl River Delta region. Acta Sci. Circumst. 34 (4), 826–834 (in Chinese). Stoeckenius, T., Russell, J., 2005. Survey of HRVOC Regulations. Report Houston Advanced Research Center Project No. H-12-2004-EE-UT-TI(582-4-6587). Environ International Corp., Novato, CA.

U.S. EPA, 2005. Revisions to the California state implementation plan and revision to the definition of volatile organic compounds (VOC): removal of VOC exemptions for California's aerosol coating products reactivity-based regulation. Fed. Regist. 70, 53930–53935. Wendoll, R., 2003. Draft Environmental Assessment for Proposed Amended Rule 1113Architectural Coatings-Comment Letter to South Coast Air Quality Management District. Dunn-Edwards Corporation. Xue, L.K., Wang, T., Gao, J., et al., 2014. Ground-level ozone in four Chinese cities: precursors, regional transport and heterogeneous processes. Atmos. Chem. Phys. 14, 13175–13188. Yin, S.S., Zheng, J.Y., Lu, Q., et al., 2015. A refined 2010-based VOC emission inventory and its improvement on modeling regional ozone in the Pearl River Delta Region, China. Sci. Total Environ. 514, 426–438. Zhang, Y.H., Su, H., Zhong, L.J., et al., 2008. Regional ozone pollution and observationbased approach for analyzing ozone precursor relationship during the PRIDEPRD2004 campaign. Atmos. Environ. 42 (25), 6203–6218. Zheng, J., Shao, M., Che, W., et al., 2009a. Speciated VOC emission inventory and spatial patterns of ozone formation potential in the Pearl River Delta, China. Environ. Sci. Technol. 43, 8580–8586. Zheng, J.Y., Zhang, L.J., Che, W.W., et al., 2009b. A highly resolved temporal and spatial air pollutant emission inventory for the Pearl River Delta Region, China, and its uncertainty assessment. Atmos. Environ. 43, 5112–5122. Zheng, J.Y., Yu, Y.F., Mo, Z.W., et al., 2013. Industry sector-based volatile organic compound (VOC) source profiles measured from manufacturing workshop and exhaust chimney in the Pearl River Delta, China. Sci. Total Environ. 456–457, 127–136.