A comprehensive classification method for VOC emission sources to tackle air pollution based on VOC species reactivity and emission amounts

Guohao Li1，3，Wei Wei2，3，Xia Shao1，3，Lei Nie1，3，Hailin Wang1，3，Xiao Yan1，3，Rui Zhang1，3

1. Municipal Research Institute of Environmental Protection，Beijing 100037，China.

2. Key Laboratory of Beijing on Regional Air Pollution Control，Beijing University of Technology，Beijing 100124，China.

3. Key Laboratory of Beijing on VOC pollution control technology and application of urban atmosphere，Beijing 100037，China.

Abstract：In China，volatile organic compound（VOC）control directives have been continuously released and implemented for important sources and regions to tackleair pollution.The corresponding control requirements were based on VOC emission amounts（EA），but never considered the significant differentiation of VOC species in terms of atmospheric chemical reactivity.This will adversely influence the effect of VOC reduction on air quality improvement.Therefore，this study attempted to develop a comprehensive classification method for typical VOC sources in the Beijing-Tianjin-Hebei region（BTH），by combining the VOC emission amounts with the chemical reactivities of VOC species.Firstly，we obtained the VOC chemical profiles by measuring 5 key sources in the BTH region and referencing another 10 key sources，and estimated the ozone formation potential（OFP）per ton VOC emission for these sources by using the maximum incremental reactivity（MIR）index as the characteristic of source reactivity（SR）.Then，we applied the data normalization method to respectively convert EA and SR to normalized EA（NEA）and normalized SR（NSR）for various sources in the BTH region.Finally，the control index（CI）was calculated，and these sources were further classified into four grades based on the normalized CI（NCI）.The study results showed that in the BTH region，furniture coating，automobile coating，and road vehicles are characterized by high NCI and needto be given more attention；however，the petro-chemical industry，which was designated as an important control source by air quality managers，has a lower NCI.

Key words：VOC；Source chemical profile；Source classification；Beijing-Tianjin-Hebei（BTH）region

1　Introduction

The Beijing-Tianjin-Hebei（BTH）region is situated at 36°05'～42°40'N north latitude and 113°27'～119°50'E longitude，and includes the two municipalities of Beijing and Tianjin as well as 11 prefecture-level cities in Hebei province（Fig.1）.Over the past decades，the BTH region has become one of the most severely polluted regions in China，due tothe dramatic increase in economic level，industrial production and vehicle population（Ma et al.，2012）.In particular，ozone（O3）pollution represents a significant environmental issue in this region，followed by PM2.5.High concentrations of ground-level O3 occur frequently at many monitoring stationsin the BTH region in summer（Liu et al.，2005）.The hourly average O3 concentration has frequently exceeded the second grade of China’s National Ambient Air Quality Standard（200μg/m3），while an hourly O3 concentration of 412μg/m3 has been observed in Beijing（Ji et al.，2016）.It is well known that sunlight-driven photochemical reactionsresult in high ground-level O3 with the involvement of various precursors，including volatile organic compounds（VOCs）and nitrogen oxides（NOx）（Sillman，1999；Carter，1994）.However，the formation mechanisms for the high O3 vary greatly in different areas，mainly because the photochemical O3 production rateresponds nonlinearly to its VOC and NOx precursors.Recent studies have indicated that O3 formation was controlled byambient VOCs in urban areas of the BTH region，and the control of photochemical O3 can be largely accomplished by a concerted reduction of VOC and NOx emissions（Shao et al.2009）.

In recent years，scientists have developed highly species-resolved VOC emission inventoriesin the BTH region.In terms of the VOC emission inventory，Wei et al.（2008，2011）developed anemission inventory from anthropogenic sourcesin China for the year 2005，and obtainedthe VOC emission amounts in each province，including the BTH region.Li（2014）developed a VOC emission inventory in the BTH region for the year 2013，reporting anthropogenic VOC emission amounts of 285，243，and 1520kt/year respectively，in Beijing，Tianjin，and Hebei province.In terms of VOC source chemical profiles，local VOC profiles for majore mission sources have been measured in the BTH region，such as vehicular exhaust，coal burning，biomass burning，petrochemical industry，and solventuse forcoating and printing（Yuan et al.，2010；Wei et al.，2014；Yao et al.，2015a，2015b）.Yuan et al.（2010）reported onVOCs from solvent use in Beijing（e.g.，auto painting，furniture painting，architectural painting，and printing），and observed that toluene and C8 aromatics were the most abundant species in emissions from paint applications，while high-carbon alkanes and some aromatics were the dominant species in emissions from printing.Yao et al.（2015a，2015b）investigated VOC emissions from diesel vehicles and motor vehicles in Beijing，and found that aromatics and alkanes were the dominant VOC species in these emissions.Wei et al.（2014）observed that VOCs emitted from a refinery were characterized by C4～C6，propene，1-butene，benzene，and toluene.

It isrecognized that different emission sources are characterized by different VOC chemical compositions，which then take part in atmospheric chemical processes indifferent ways.It is essential to scientifically understand the chemical composition of VOC emissions，not only in terms of a VOC emission inventory but also in VOC control policy-making（Watson et al.，2001；Zhang et al.，2008；Zheng et al.，2009）.However，research on the speciation of VOC emissions is limited in China，especially in the BTH region.The current VOC abatement policy is mainly based on the VOC emission amounts，never considering their chemical speciation.This will adversely influence the effect of VOC reductions on air quality improvement.

Therefore，the objective of this study wasto develop a comprehensive classification method for VOC emission sources based on VOC species reactivity and emission amount，through a case study in the BTH region.In order to improve the knowledge of the VOC chemical composition of various sources，VOC measurements were additionally carried out for 5 key sources.Through application of this classification method in the BTH region，an integrative control index could be estimated for individual VOC sources，and the rank of VOC sources as control targets could be determined.We believe that this study cangreatly help policy-makers to make effective decisions for VOC emission reductions，to tackle the regional air pollution in BTH.

2　Methods

2.1　Overview of VOCsources

In this study，15 anthropogenic VOC sources were considered for classification in the BTH region，including power plant coal burning，industrial coal burning，residential coal burning，biomass burning，coking industry，petrochemical industry，iron and steel industry，diesel vehicles，gasoline vehicles，motorcycles，automobile coating，furniture coating，architectural coating，printing and fuel oil evaporation.The VOC emission amounts of these 15 anthropogenic VOC sources accounted for more than 90% of the total anthropogenic VOC emission amounts in the BTH region.Our comprehensive literature research found that the chemical profile of VOCs for most sources had been measured and obtained inprevious studies，and the literatures were list in Table 1.For example，the VOC profiles from the emission sources of industrial coal burning，residential coal burning，biomass burning（wheat，corn，and wood），gasoline and diesel fuel evaporation，vehicle exhaust（gasoline vehicles，diesel vehicles，motorcycles），and the petrochemical industry were obtained from Liu et al.（2008）.The VOC profile from the architectural painting source was obtained from Yuan et al.（2010），and the VOC profile from the iron and steel facility sources（sintering，hot forming，cold forming）was found in Tsai et al.（2008）.

The major VOC species reported in the literature are as follows.The VOC species from biomassburning and coal burning were mainly acetylene，C2～C3 alkenes such as ethylene and propylene，alkanes like ethane and propane，and aromatics like benzene and toluene.The major VOC species from gasoline and diesel fuel evaporation were C4～C7 alkanes and alkenes，and aromatics，namely isobutene，n-butane，trans-2-butene，n-pentane，benzene，and toluene.The chemical compositions of VOCs from gasoline-powered vehicles were very similar to thosefrom diesel vehicles for compounds containing <10 carbonatoms；ethylene，is opentane，benzene，and toluene were the main species in exhaust emissions.Alkanes were the dominant VOC group from the petrochemical industry，and hexanewas the most abundant species，followed sequentially by cyclohexane，pentane，butane，isopentane，n-heptane，methylcyclohexane，2-methylpentane，and propane.The major VOC species in the emissions of water-based architectural paints were mainly C3～C6 alkanes and<C9 aromatics，namely toluene，benzene，n-butane，propane，and n-pentane；where as the principal emission species in solvent-based architectural paintswere C7～C8 aromatics，m/p-xylene，toluene，o-xylene，ethylbenzene，and styrene.The major VOC species in the exhaust from the sintering process included toluene，isopentane，benzene，1-butene，m/p-xylene，ethylbenzene，n-pentane，1，2，4-trimethylbenzene，n-butane，and 2-methylpentane；toluene，m/p-xylene，1，2，4-trimethylbenzene，benzene，1-butene，ethylbenzene，styrene，o-xylene，and octane were themajor VOCs from the hot forming process；toluene，benzene，isopentane，o-xylene，m/p-xylene，ethylbenzene，styrene，1-butene，and n-pentane were the major VOC species from the cold forming processes.

Then，this study carried outa measurement campaign for the weak sources in the BTH region，including automobile coating，furniture coating，print，power plants，and coking plants.

2.2　Measurement campaign of VOC chemical profiles for weak sources

2.2.1　Schedule of field sampling

In this study，VOC samples were collected from a variety of emission sources，including automobile coating，furniture coating，printing，power plants，and coking plants.Table 2 lists the sampling information.The emissions from the painting，printing and coking were classified as “fugitive emission”，while the emission from the stacks of coal burning was considered as “stack emission”.Different methods were used for fugitive and stack emission sampling.In terms of fugitive emission，ambient air samples around the industrial units were directly collected into a 3.2 L stainless steel canister with a flow-controller（Silonite，EntechInstrument，USA）within 60min（Tang et al.，2007）.An ozone scrubber with anhydrous sodium sulfate in a glass tube was connected to the inlet ofeach sampling canister to remove high levels of ozone for fugitive emission sampling.In terms of stack emission sampling，a glass tube filled with anhydrous sodium sulfate and glass wool in series were used to remove the water vapor，ozone，and particulate matters，and a stainless-steel probe with a temperature-controlled thermocouple was linked to a condensate trap for stack gas cooling，and finally linked to a 3.2 L stainless steel canister with a flow-controller（Silonite，Entech Instrument，USA）.The stack emission sampling generally lasted 10min.In total，45 VOC samples were collected and sent to the laboratory for analysis.

Before sampling，each canister was cleaned with high-purity nitrogen（99.999%）for three cycles of filling（～30psia）and evacuating（～100mtorr），and finall yevacuated to a vacuum state（～30mtorr）.

（1）Automobile coating and furniture coating

A typical automobile 4S（Sale-Sparepart-Service-Survey）shopand a typical furniture factory in Heibei province were selected for sampling the VOC emissions.Samples collected in the three different workshops corresponded to the three most important processes in coating，namely primer coating，finish-coat coating，and paint-drying.The paints used in the auto garage were all nitrocellulose-based resin paints，while the paints used in the furniture factory were polyurethane-based resin paints（PU paints）and nitrocellulose-based resin paints（NC paints）.The VOC samples for painting were collected in the spray booth during the primer paint-spraying period and the finish coat paint-spraying period.The samples for the paint-drying process were collected during the drying process after painting.Three samples were collected for each coating process，with a total of 9 samples collected in the automobile 4S shop，and 18 samples were collected in the furniture factory.

（2）Print

A typical printing factory in Heibei province，which mainly prints books，magazines，newspapers，and packaging materials，was selected for sampling the VOC emissions.The VOC samples were collected around the operating offset lithographic printing press equipment.A total of 6 samples were collected in the printing factory.

（3）Coking plant

Two coking plants in Heibei province were selected for sampling the VOC emissions，and two coke oven types（i.e.，top charging coke oven and tamping coke oven）in each plant were investigated.The VOC samples were collected usinga 6 L stainless steel canister with a flow-controller on the top of the two coke ovens，over a period of 6 hoursforeach sample.Three samples were collected for each coke oven，for a total of 6 samples collected in each of the twocoking plants.

（4）Power plant

Two power plants in Heibei province were selected for sampling the VOC emissions；two boiler types（i.e.，circulating fluidized bed boiler and pulverized coal fired boiler）in each plant corresponding to different coal types were investigated.The VOC samples were collected in the stack after the desulfurization and dust removal device.Three samples were collected for each boiler，for a total of 6 samples collected in each ofthe two power plants.

2.2.2　VOC analysis

The VOC samples were concentrated in apre-concentrator（Model 7100，Entech Instrument，USA），and analyzed by agas chromatography（GC）system（Model7890A，Agilent，USA）equipped with a quadrupole mass spectrometer（MS）system（Model5975C，Agilent，USA）and a flame ionization detector（FID）.The C2～C3 hydrocarbons were separated on an Al2O3 porous layer open tubular（PLOT）column（30m×0.32mm×3.0μm，J&W Scientific，USA）and quantified by FID.The C4～C12 hydrocarbons were separated on a DB-624 column（60m×0.32mm×1.8μm，J&W Scientific，USA）and quantified by MS.The system is described elsewhere in detail（Liu et al.，2008）.The pre-concentrator had a three-stage trapping system.In the first stage，the VOCs were adsorbed on glass beads at -150℃ and desorbed at 15℃.In the second stage，they were trapped on Tenax at -30℃ and desorbed at 180℃.Finally，in the third stage，the VOCs were condensed on a transfer line at -160℃ and heated rapidly to 65℃.Then，the concentrated VOCs were injected into the gas chromatograph.Helium was used as the carrier gas for GC at a flow rate of 1.3mL/min.The GC oven temperature was initially programmed at -50℃，holding for 2 min，and increasing to 220℃ at 6℃/min.The mass spectrometer was operated in SCAN mode and scanned from 20 to 300.The ionization method was electron impact（EI，70 eV），and the source temperature was 220℃.

The photochemical assessment monitoring system（PAMS）standard gas（57 species，Spectra Gases，USA）was used to confirm compounds’ retention times and identify compounds.In total，55 species were identified and measured in this study.The target species were quantified by using the multipoint external calibration method.Calibration was performed at 5 different concentrations from 1 to 20 ppbv by the 55 PAMS gas standard.1，4-Difluorobenzene and chlorobenzene-d5 were chosen as internal standards in the samples.The method detection limits（MDLs）of the various VOC species ranged from 0.009 to 0.035 ppbv（Liu et al.，2008）.

2.3　VOC emission source classification

Fig.2 presents an overview of the VOC emission source classification process.The VOC emission amount（EA）from a source was chosen as one important factor，and this information can be found from the emission inventory.In this study，the VOC emission inventory of the BTH region was obtained from Li（2014）.The VOC emission source reactivity（SR）was chosen as another important factor for classification，it is defined as the ozone formation potential（OFP）of a unit VOC emission mass from a source，and is calculated using the following equation（Na and Kim，2007）：

　　（1）

where SRj is the emission source reactivity of VOC source j（ton O3/ton VOCs），and is the estimated ozone formation amount when 1 ton of VOCs is emitted from source j；fi， j is the mass fraction of VOC species i in source j，and MIRi is the maximum incremental reactivity（MIR）value of species i（g O3/g VOCs）as proposed by Carter（2008）.

After obtaining the EA and SR values for different emission sources，they were sorted from high to low.The range method was then applied to obtain the normalized emission amount index（NEA）and normalized source reactivity index（NSR）using the following equations：

　　（2）

　　（3）

where EAj is the VOC emission amount of source j（ton/year），EAMin is the minimum EAj among VOC sources，EAMax is the maximum EAj among VOC sources，NEAj is the normalized VOC emission index of source j.SRMin is the minimum SRj among VOC sources，SRMax is the maximum SRj among VOC sources，NSRj is the normalized index of source j.

After normalization，relative weights were arbitrarily assigned to NEAj and NSRj，and then a control index（CI）was calculated using the following equation：

CIj=k1×NEAj+k2×NSRj　　（4）

wherek is the weight，k1+k2=1.0；in this study，NEAj and NSRj have the same weight，so k1=k2=0.5；CIj is the control index of source j.

The obtained CI value was then normalized as follows：

　　（5）

where CIMin is the minimum of CIj，CIMax is the maximum of CIj，NCIj is the normalized control index（NCI）of source j.

The obtained NCI value was then used for VOC emission source classification.In this study four levels were identified for VOC emission sources，in order to distinguish the VOC emission sources needing priority control by environmental protection departments.Levels Ⅰ，Ⅱ，Ⅲ and Ⅳ VOC emission sources had NCI between 0.75～1.00，0.50～0.75，0.25～0.50，and 0～0.25，respectively.Level Irepresent the most important VOC emission sources，which should require priority control.

3　Results and discussion

3.1　VOC characteristics of emission sources

In this study，55 VOC species（C2～C12）were detected in the collected samples，including 29 alkanes，10 alkenes，and 16 aromatic hydrocarbons.Naphthalenes，oxygenated hydrocarbons and halogenated hydrocarbons were not detected.

3.1.1　Furniture coating

Nearly all of the paints used in the furniture manufacturing industry（>96%）were solvent-based.Table 3 lists the 10 most abundant species and their percentages in the VOCs from two different furniture coating emissions（PU and NC paints）.The number listed in this table represents the average VOC profile for the three different painting processes，as mentioned above.The polyurethane paints and NC paints are associated with different top 10 VOC species due to different paint formulations.For example，styrene was included in the top 10 species from PU paints but not from NC paints.In general，the top 10 species were mainly<C9 aromatics.

PU paint：polyurethane-based resin paints；NC paint：nitrocellulose-based resin paints.

3.1.2　Power plant

The VOC species emitted from thepower plant coal burning process depend largely on the coal type.Table 4 lists the 10 most abundant species and their percentages in VOCs from the emissions of the two boiler types.It was found that benzene had a higher percentage in VOCs from the circulating fluidized bed boiler than the pulverized coal fired boiler，but n-decane showed opposite variation.This difference was caused by the different coal types used，because low quality coal（e.g.，raw coal，coal gangue）was used as fuel by the circulating fluidized bed boiler，while the pulverized coal-fired boiler usually burned high-quality thermal coal.

3.1.3　Coking plant

Coking plant emissionsarean important anthropogenic VOC source in the BTH region.Table 5 lists the 5 most abundant species and their percentages in VOCs from the emissions of the two coke oven types.It can be found that five hydrocarbons，namely acetylene，benzene，ethane，toluene，and propene，accounted for>86% of VOCs from the coking process emission.The content of benzene in VOCs was higher from the top charging coke oven than from the tamping coke oven，but the acetylenecontent showed opposite variation.This difference was due tothe different coal types used for coking，with high-quality coal（e.g.，coking coal）being used as material by the top charging coke oven，and low-quality coal（e.g.，coking coal mixed with gas coal and lean coal）by the tamping coke oven.

3.1.4　Overview of the source chemical profiles

Table 6 lists the detailed VOC profiles measured for different emission sources，while the profile from furnitureand auto painting was obtained by averaging the measured VOC profiles from the three different painting processes，and the VOC profile for furniture painting was calculated by averaging from two different solvents.The VOC profile for printing was obtained by averaging the measured VOC profiles from the printing process.The VOC profile for power plant coal burning emission was obtained using the average value measured from the two different boiler types，and the profile for the coking process was calculated by averaging the measured values from the two different coke ovens.

Number in bracket represents standard deviation.nd：not detected.

As shown in Table 6，o-xylene was the most abundant VOC species from furniture painting emissions，accounting for 11.6% of its total VOCs.Other major species from furniture painting emissions were 1，2，4-trimethylbenzene，ethylbenzene，m-ethyltoluene，toluene，m/p-xylene，p-ethyltoluene，and 1，3，5-trimethylbenzene，accounting for 11.5%，11.4%，8.5%，8.1%，11.4%，7.6%，6.5%，and 6.3% of its total VOCs，respectively.In total，aromatic species accounted for 93.9% of the furniture painting VOC emissions.

It can be found fromTable 6 that o-xylene was the most abundant VOC species from auto painting emission，accounting for 21.3% of its total VOCs，followed by ethylbenzene，toluene，m/p-xylene，and 1，2，4-trim ethylbenzene，accounting for 12.6%，10.4%，8.2%，and 7.1%，respectively.In total，aromatic species accounted for 83.2% of VOCs from auto factory painting emissions.

VOC emissions from printing factories are mainly from ink evaporation and detergents used to clean press machinery.The VOCs from printing factory emissions mainly comprised C10～C12 alkanes and aromatics，and n-undecane，n-decane，n-nonaneaccounted for 6.0%，3.9%，and 2.7% of the total VOCs；1，2，4-trimethylbenzene，ethylbenzene，m/p-xylene m-ethyltoluene，o-xylene，and p-ethyltoluene accounted for 6.8%，6.0%，5.4%，5.2%，4.7%，and 4.2% of the total VOCs.In total，the alkanes and aromatic compounds accounted for 44.8% and 47.4% of the total VOCs from printing emissions.

It can be observed from Table 6 that n-decane was the most abundant VOC species from power plant coal burning emission，accounting for 18.3% of its total VOCs.Other major species were benzene，acetylene，and ethane，accounting for 16.5%，15.9%，and 11.3%，respectively.In total，alkanes，alkenes and aromatics species accounted for 42.9%，22.1% and 35.0% of VOCs from power plant coal burning emissions，respectively.

The major VOCs species from coking plants were acetylene，benzene，and ethane，accounting for 36.3%，20.5%，and 18.5% of VOCs，respectively.The alkanes，alkenes and aromatics species accounted for 24.1%，45.6% and 30.3% of the VOCs from the coking plants，respectively.

3.2　Classification of VOC emission sources in BTH region

3.2.1　VOC emission amounts in BTH region

The VOC emission amounts for the year 2013 were obtained from the inventory of the BTH region developed by Li（2014）.Fig.3 presents the VOC emission information from different sources and cities.It can be observed from Fig.3 that the VOC emissions were 2047 kton in the entire BTH region，and the VOC emissions from Beijing，Tianjin，and Heibei province were 285，242，and 1520 kton，respectively.It can be found that Beijing hadthe largest VOC emission（285 kton），followed by Tianjin（242 kton），Tangshan（213 kton），Shijiazhuang（205 kton）and Baoding（200 kton）.Gasoline vehicles represent the largest emission source in Beijing，whose emission amount was 94 kton.Gasoline vehicles and the petrochemical industry represent the largest emission sources in Tianjin，whoseemission amounts were 39 and 35 kton.However，in Tangshan，Shijiazhuang and Baoding，industrial coal burning，biomass burning，and motorcycles were the top three VOC emission sources.

3.2.2　VOC emission source reactivity

Fig.4 presents the calculated VOCs emission source reactivities（SRs）.SRswere calculated by using the VOC source profiles measured in Table 6 and other source profiles in the literature in Table 1.The SR of furniture painting emission was the highest（i.e.，5.89 ton O3/ton VOCs）among all the VOC sources，followed by emissions from auto painting（5.59 ton O3/ton VOCs），diesel vehicles（5.29 ton O3/ton VOCs），fuel evaporation（4.67 ton O3/ton VOCs），and gasoline vehicles（4.52 ton O3/ton VOCs）.However，the SR for power plant coal burning emissions was the lowest，because C10～C12 alkanes with low reactivities accounted for a major percentage in the total VOCs.As illustrated in Fig.4，contributions to SRs from solvent-using emissions（i.e.，furniture painting，auto painting，architectural painting，and printing）andiron and steel facility emissions were mainly from aromatic compounds.However，alkene compounds made a major contribution to the SRs of other anthropogenic sources.Given their higher SR values，controlling equal amounts of VOC emissions from furniture painting，auto painting，and diesel vehicle emissions would result in more reduction of ozone formation in the VOC-limited urban areas.

3.2.3　VOC emission source classification

Fig.5 presents the VOC emission source classification resultsfor various sources and cities ofthe BTH region.It was found that gasoline vehicle emission in Beijing had the highest NCI（1.00）.Level I VOC emission sources include gasoline vehicles，auto painting，and furniture painting in Beijing，as well as furniture painting in Tianjin and Langfang.Level Ⅱ VOC sources include gasoline vehiclesin Tianjin，biomass burning in Baoding，Cangzhou，Xingtai，Handan and Shijiazhuang，motorcycles in Tangshan，Baoding，Handan，and Shijiazhuang，industrial coal burning in Tangshan and Baoding，as well as the diesel vehicles，auto painting，and furniture painting in every city of BTH.Level Ⅲ VOC sources mainly include industrial coal burning，biomass burning，iron and steel facilities，printing，and fuel evaporation in every city of BTH.Level IV VOC sources were mainly power plant coal burning，residential coal burning，coking，petrochemical industry and architectural painting.

In terms of emission source classification for the entire BTH region，it can be illustrated from Fig.6 that furniture painting emission had the highest NCI.The painting applications（e.g.，auto painting and furniture painting）and vehicles（especially diesel vehicles）had higher NCI than other VOC emission sources，and these sources should be controlled with priority in order to reduce ambient O3 concentrations in the urban areas.In terms of emission source classification for every city in the BTH region，the NCI shows little variation between the cities in the BTH region.

3.3　Suggestions for VOC abatement policy based on the classification results

The National Action Plan on Air Pollution Control（2013～2017）that was released in 2012 explicitly required the reduction of anthropogenic VOCs，and focused on the sources of petrochemical industry，organic chemical industry，packaging printing and industrial coating.The requirements are as follows：implementation of leak detection and repair technology was required in the petrochemical industry；completion of oil and gas recovery and the control of gas stations，oil storage tanks and oil tankers within a limited time；improvement of the volatile organic limit standards for paint，adhesives and other products，and promotion of the use of water-based paint，to encourage the production，sale and use of low toxicity，low-VOC solvents.The national measures for pollution charges for VOCs were released in 2015，and pollution charges were levied for the petrochemical industry and packaging printing.However，the VOC emissions from industrial coating and vehicles have not received enough attention in the national policy.The corresponding control requirements were decided primarily based on VOC emission amounts（EA），but never considered the significant differentiation of VOC species in atmospheric chemical reactivity and significant differentiation of VOC sources in VOC species profiles.In this study，the VOC emission amount and source reactivity were used in the VOC source classification，and our classification results imply that the sources of furniture painting，auto painting and vehicles（especially diesel vehicles）have high NCI，and need priority control in the BTH region.The VOC source classification result has guiding significance for regional VOC priority emission abatement control.

Conclusions

In this study，the VOC source profiles were examined for furniture coating，auto coating，printing，power plants，and coking plants in the Beijing-Tianjin-Hebei（BTH）region.Analysis of source reactivities indicated that the ozone formation potential is highest for furniture coating，followed by auto coating emissions，diesel vehicle emissions，fuel evaporation，and gasoline vehicles.The VOC emission amount index（NEA）and the source reactivity index（NSR）were chosen to build the VOC source classification method，and this method was applied in the BTH region.The classification results showed that gasoline vehicle emission in Beijing was a Level I VOC source in the BTH region.Furniture coating，auto coating and vehicles（especially diesel vehicles）had high NCI，and need priority control in the BTH region.

Acknowledgements

This research was supported by the National Key Technology Support Programof China（Nos.2014 BAC23B05 & 2014 BAC23B02），the Natural Sciences Foundation of China（No.51478017），the Youth Individual Project of Beijing Talents Training Fund（No.2015000021733G170），as well as the Ministry of Environmental Protection Special Funds for Scientific Research on Public Causes（No.201409016）.The authors are grateful to the anonymous reviewers for their insightful comments.

References

[1]　Carter，W.P.L，1994.Development of ozone reactivity scales for volatile organiccompounds.J.Air Waste Manage.Assoc.，44：881-899.

[2]　Carter，W.P.L，2008.Development of the Saprc-07 Chemical Mechanism and Updated Ozone Reactivity Scales，Final Report to the California Air Resources Board Contract No.03-318.Air Pollution Research Center and College of Engineering，Center for Environmental Research and Technology，University of California，Riverside，California，USA.

[3]　Gao，Y.，Zhang，M.G，2010.Sensitivity analysis of surface ozone to emission controls in Beijing and its neighboring area during the 2008 Olympic Games.J.Environ.Sci-CHINA，24：50-61.

[4]　Guo，H.，Jiang，F.，Cheng，H.R.，Simpson，I.J.，Wang，X.M.，Ding，A.J.，et.al.，2009.Concurrent observations of air pollutants at two sites in the Pearl River Delta and the implication of regional transport.Atmos.Chem.and Phys.，9：7343-7360.

[5]　Ji，D.S.，Gao，W.K.，Zhang，J.K.，Yu，M.，Zhou，L.X.，Yu，P.F.，et.al.，2016.Investigating the evolution of summertime secondary atmosphericpollutants in urban Beijing.Sci.Total Environ.，572：289-300.

[6]　Li，G.H.，2014.Research on typical VOCs emission sources classification and optimized reduction strategy.Beijing University of Technology，Beijing，China.

[7]　Liu，Y.，Shao，M.，Lu，S.H.，Chang，C.C.，Wang，J.L.，Chen，G.，2008.Volatile Organic Compound（VOC）measurements in the PearlRiver Delta（PRD）region，China.Atmos.Chem.and Phys.，8：1531-1545.

[8]　Liu，Y.，Shao，M.，Zhang，J.，Fu，L.L.，Lu，S.H.，2005.Distribution and source apportionment of ambient volatileorganic compounds（VOCs）in Beijing city，China.J.Environ.Sci.Heal.，40：1843-1860.

[9]　Ma，J.，Xu，X.B.，Zho，C.S.，Yan，P.，2012.A review of atmospheric chemistry research in China：photochemical smog，haze pollution，and gas-aerosol interactions.Adv.Atmos.Sci.，29：1006-1026.

[10]　Sillman，S.，1999.The relation between ozone，NOx and hydrocarbons in urban and polluted rural environments.Atmos.Environ.，33：1821-1845.

[11]　Shao，M.，Zhang，Y.H.，Ceng，L.M.，Tang，X.Y.，Zhang，J.，Zhong，L.J.，et al.，2009.Ground-level ozone in the Pearl River Delta and the roles of VOC and NOx in its production.J.Environ.Manage.，90：512-518.

[12]　Streets，D.G.，Fu，J.H.S.，Jang，C.J.，Hao，J.M.，He，K.B.，Tang，X.Y.，et.al.，2007.Air quality during the 2008 Beijing Olympic Games.Atmos.Environ.，41：480-492.

[13]　Tsai，J.H.，Lin，K.H.，Chen，C.Y.，Lai，N.，Ma，S.Y.，Chiang，H.L.，2008.Volatile organic compound constituents from an integrated iron and steel facility.J.Hazard.Mater.，157：569-578.

[14]　Wang，T.，Ding，A.，Gao，J.，and Wu，W.S.，2006.Strong ozone production in urban plumes from Beijing，China.Geophys.Res.Lett.，33：320-337.

[15]　Wei，W.，Wang，S.X.，Chatani，S.，Klimont，Z.，Cofala，J.，Hao，J.M.，2008.Emission and speciation of nonmethane volatile organic compounds from anthropogenic sources in China.Atmos.Environ.，42：4976-4988.

[16]　Wei，W.，Wang，S.X.，Hao，J.M.and Chen，S.Y.，2011.Projection of Anthropogenic volatile Organic Compounds（VOCs）Emissions in China for the Period 2010-2020.Atmos.Environ.，45：6863-6871.

[17]　Wei，W.，Cheng，S.Y.，Li，G.H.，Wang，G.，Wang，H.Y.，2014.Characteristics of volatile organic compounds（VOCs）emitted from apetroleum refinery in Beijing，China.Atmos.Environ.，89：358-366.

[18]　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.

[19]　Yao，Z.L.，Shen，X.B.，Ye，Y.，Cao，X.Y.，Jiang，X.，Zhang，Y.Z.，et.al.，2015a.On-road emission characteristics of VOCs from diesel trucks inBeijing，China.Atmos.Environ.，103：87-93.

[20]　Yao，Z.L.，Wo，B.B.，Shen，X.B.，Cao，X.Y.，Jiang，X.，Ye，Y.，et.al.，2015b.On-road emission characteristics of VOCs from rural vehicles and their ozone formation potential in Beijing，China.Atmos.Environ.，105：91-96.

[21]　Yuan，B.，Shao，M.，Lu，S.H.，Wang，B.，2010.Source profiles of volatile organic compounds associated with solvent use in Beijing.Atmos.Environ.，44：1919-1926.

[22]　Zhang，Y.H.，Su，H.，Zhong，L.J.，et al.，2008.Regional ozone pollution and observation-based approach for analyzing ozone-precursor relationship during the PRIDE-PRD2004 campaign.Atmos.Environ.，42：6203-6218.

[23]　Zheng，J.Y.，Shao，M.，Che，W.W.，Zhang，L.J.，Zhong，L.J.，Zhang，Y.H.，et al.，2009.Speciated VOC emission inventory and spatial pattern of ozone formation potentials in the Pearl River Delta，China.Environ.Sci.Technol.，43：8580-8586.

[24]　Zheng，J.Y.，Yu，Y.F.，Mo，Z.W.，Zhang，Z.，Wang，X.M.，Yin，S.S.，et.al.，2013.Industrial sector-based volatile organic compound（VOC）source profiles measured in manufacturing facilities in the Pearl River Delta，China.Sci.Total Environ.，456-457：127-136.