Emission factor, relative ozone formation potential and relative carcinogenic risk assessment of VOCs emitted from manufacturing industries

Manufacturing industries are one of the important emission sectors of anthropogenic volatile organic compounds (VOCs). In this study, VOC emission factors, relative ozone formation potential (ROFP) and relative carcinogenic risk (RCR) were estimated for manufacturing industries (n = 13) located in central Taiwan. Emission samples were collected in stainless steel canisters and were analyzed with a system of gas chromatography-mass spectroscopy. Higher emission factors of total VOCs (∑VOCs) were observed for stencil printing (423 mg-VOC kg− 1) compared to other emission industries. Alkanes constituted the most prominent group of VOCs for steel foundry (42%), aluminum foundry (25%) and synthetic resin industries (25%). Oxygenated VOCs were the most abundant group in the organic solvent (80%), polyester resin (80%) and polyurethane (75%) industries. Moreover, emissions from acrylic resin manufacturing had a major contribution from aromatic compounds (> 95%). Toluene was the topmost compound in terms of its contribution to ∑VOCs in plastic tape manufacturing (44%), aluminum foundry (40%), steel foundry (12%), plastic coating (64%) and stencil printing (35%). Analysis of ozone formation potentials showed that the metal product and machinery acrylic resin manufacturing and stencil printing had a higher normalized relative ozone formation potential (ROFP) index and belonged to Level-I emission sources. However, in terms of the relative carcinogenic risk (RCR), integrated iron and steel manufacturing had the highest normalized RCR index that belonged to level-I emission sources. Level-I represents the most important VOC emission sources. This study provides a reactivity- and carcinogenicity-based approach to identify high-priority VOC emission sources. The results of this study would help formulate emission reduction policies and strategies for manufacturing industries.


Introduction
Manufacturing industries are important stationary emission sources of air pollutants worldwide. These emission sources have become one of the major contributors to air pollution due to a substantial increase in industrial production [1][2][3]. The pollutants such as particulate matter, oxides of carbon, oxides of nitrogen, polycyclic aromatic hydrocarbons and volatile organic compounds (VOCs) pose threats to human health and the environment [4][5][6][7]. VOCs are of interest in part because they participate in atmospheric photochemical reactions that contribute to ozone formation [8]. Tropospheric ozone is formed by chemical reactions involving airborne VOCs, airborne nitrogen oxides, and sunlight [9]. The oxidation products of VOCs may also get absorbed by the atmospheric aerosols [10]. VOCs contain several hundreds of species; each species can react with different rates and reaction mechanisms and have different potential for ozone formation [11]. Ozone formation potential (OFP) estimates the contribution of individual VOC species to ozone formation. Developing control strategies for VOC emission sources requires considering not only the emission amount of VOCs but also the chemical reactivity (i.e., maximum incremental reactivity, MIR) of VOC species [12]. The MIR based approach has been recommended by U.S. EPA [13]. Reactivity-based control strategies can be applied to various industrial emission sources to identify the highly reactive VOC emission sources that contribute more to the ozone formation. For example, in Europe, reactivity-based strategies for stationary emission sources resulted in the reduction of tropospheric ozone formation compared to simple massbased strategies [14].
VOC lifespan in the atmosphere could range from a few minutes to several months. So, some of the VOCs may travel over large distances and enter the human body mainly through inhalation or skin absorption causing varieties of health effects [15]. Human health effects can vary among VOC species, ranging from irritants to cancerous illnesses [16][17][18]. For example, exposure to BTEX (i.e., benzene, toluene, ethylbenzene and xylene) has been associated with toxicological effects such as depression, respiratory diseases, damage to the circulatory system and cancer [19][20][21][22]. Benzene, and 1, 3-butadiene have been identified as carcinogens [23], and wellknown risk factors for various types of cancers [24]. The main carcinogenic risks are associated with the blood (leukemia and non-Hodgkin lymphoma), lung, liver and biliary tract cancer [25][26][27]. With regard to these risks, the International Agency for Research on Cancer (IARC) has classified benzene as a Group-1 human carcinogen [24]. Moreover, other VOCs such as trichloroethylene and 1,2-dibromoethane are the probable human carcinogens (Group 2A, IARC), ethylbenzene and chloroform are possibly carcinogenic to human (Group 2B, IARC).
Several studies have been conducted on VOC emissions from sources such as road traffic, petrochemical industries, coal burning, biomass burning and solvent use [28][29][30]. These studies have reported significant differences in the source compositions between different regions. However, studies on VOC emissions from manufacturing industries are still limited [29,[31][32][33]. The major VOC emission sources include fossil fuel combustion, petrochemical, printing and solvent usage [29,31,34,35]. Wang et al. [35] studied the distribution of VOCs emitted from solvent usage in furniture paint, auto paint and printing ink. Their study indicated that the largest contributing groups among the measured VOCs were aromatics (52%) followed by alkanes (32%) and alkenes (16%). Aromatic compounds such as benzene has been found to be a common byproduct of the chemical manufacturing, petrochemical industries, production of xylene, toluene and other aromatic compounds, industrial solvent and printing sectors [23,24]. Yuan et al. [29] identified alkanes and aromatic compounds as the most contributing groups in the printing sector. Tsai et al. [36] reported toluene, 1,2,4-trimethyl benzene, m/p-xylene, 1-butene, ethylbenzene, and benzene predominantly emitted from an integrated iron and steel plant located in Southern Taiwan.
VOC emissions from the manufacturing industries have a significant impact on air quality and human health. Biological evidence supports the causal link between VOC species and certain cancer. For instance, exposure to benzene and 1,3-butadiene increases the risk of leukemia [21,[37][38][39]. Toluene causes neurological disorders [40]. Long-term exposure to xylene can cause headaches, tremors and impaired concentration [41]. Occupational epidemiological studies of petrochemical industry workers indicate that benzene may cause lung cancer, multiple myeloma and acute myelogenous leukemia [42,43]. Increased rates of leukemia risk have been shown in workers of synthetic rubber industries [37].
Stack emissions from different industries might differ greatly in the VOC species compositions. Thus, the emissions from different sources might play roles in atmospheric chemistry and human health hazards in different ways. It is essential to characterize VOC emission sectors, not only in terms of VOC emission concentrations but also in their chemical reactivity and associated health hazards. However, such data are still not sufficient for industrial emission sources. In this study, a few important industrial emission sources were investigated on the basis of emission factors, ROFP and RCR VOCs.

Description of emission sources
In this study, the following thirteen types of VOC emission sources were selected: organic solvent manufacturing, synthetic resin manufacturing, acrylic resin manufacturing, polyester resin manufacturing, plastic tape manufacturing, aluminum foundry, steel foundry, metal products and machinery industries, integrated iron and steel manufacturing, non-ferrous metal-based manufacturing, polyurethane (PU) leather manufacturing, plastic coating and stencil printing. All of the selected emission sources were located within Taichung city, Taiwan. The sample industries were selected to represent the major industrial emission sources located within Taichung city. The selected emission sources contributed to 73% of total VOC emissions within the city in the year 2016 [44]. The selected emission sources were grouped into the following five sectors on the basis The chemical manufacturing sector incorporates the chemical transformation of organic and inorganic raw materials and the formation of products. In the present study, this sector was represented by the industries that handled organic solvents and various types of resins. The plastic product manufacturing sector incorporated the industry that manufactured adhesive plastic tapes. The metal (primary and fabricated) manufacturing sector included iron and steel industries, metal casting industries, and nonferrous metal industries. In this study, only the sintering process of the integrated iron and steel industry was investigated. The leather/textile manufacturing sector consisted of the facility that handled synthetic leather, spinning natural and manmade fibers into yarns and threads. The printing sector included plastic coating, stencil printing, plate-making and bookbinding.
Detailed information about selected emission sources can be found in Table 1.

VOCs sampling and analysis
VOC emissions from manufacturing industries can be categorized into fugitive emissions and stack emissions. The current study was conducted only for stack emissions. Emission samples were collected using 6 L fused silica stainless steel canisters (Entech Instruments, Cata-log# 29-10,622) that had been pre-cleaned with high purity-nitrogen and evacuated with an automated canister cleaner. A flow controller and Teflon tubing were used to extract the exhaust gas from emission stacks to the evacuated canisters. The volume of flue gas was measured using a gas flowmeter. The selection of the sampling location in the stack is important to obtain representative samples. To minimize the effects of process variables, monitoring was performed when there was constant flow through stacks. Stack-ports were set by the local environmental protection body for routine monitoring.
The collected samples were analyzed with the Gas Chromatography (7890) Mass Spectroscopy (5977B) system (GC/MS, Agilent Technologies). The details of the GC/MS analysis is described elsewhere [45]. The internal calibration method was applied for the quantification of the VOCs. The standard mixture of gases (A715.15B) from Taiwan National Institute for Environmental Analysis was used as the external standard. Bromochloromethane, 1,4-difluorobenzene, chlorobenzene-d5 and 4bromofluorobenzene were used as internal standards. A total of 72 VOC species were quantified in this study. Based on the functional groups, these VOCs were classified as alkanes (24 species), alkenes (8 species), aromatics (18 species), oxygenated VOCs (8 species), halocarbons (13 species) and others (1 species) ( Table 2). Emission factors of the detected VOC species were calculated for each of the emission sources by using Eq. (1).
The calculated emission factor of each VOC species was normalized (x i, j ) to the ∑VOCs emission factors in the source sample for each emission source. The detailed concentrations and emission factors can be found in Additional file 1 (Table S1 and Table S2 of Supplemental Materials).

Quality assurance and quality control (QA/QC)
The canisters were pre-cleaned before sampling with ultra-pure nitrogen (99.999%) to remove water vapor and contaminants. Canisters were cleaned for 12 cycles of filling and evacuation using a canister cleaning system (3100A, Entech). Replicates were performed for each sample to minimize analytical errors. The blank analysis was run before each sample analysis.

ROFP
The absolute OFP of VOC species is usually calculated by multiplying its concentration by its MIR value [46,47]. However, it is usually the relative importance of each VOC species in comparison with the other emission sources that is of more practical significance. Knowing the relative importance of different VOCs allows for targeting more reactive VOCs sources, hence more efficient and flexible for sources comparing strategies. Therefore, ROFP was calculated on the basis of the VOC emission factor (normalized to 1) and the MIR value of each VOC (Eq. (2)).
where, ROFP j is for source j (g-O 3 g-VOCs − 1 ), x i, j is the ratio of EF of VOC species i to ∑VOCs for source j, and MIR i is the value of species i as proposed by Carter [11].

RCR
The absolute carcinogenic risk is usually estimated by multiplying the actual VOC concentration and its carcinogenic risk factor [48]. However, to ensure comparability among different emission sources, carcinogenicity was calculated for each emission source on a relative basis (Eq. (3)).
where, RCR j is the source j mg − 1 m 3 , x i, j is the ratio of emission factor of VOC species i to ∑VOCs for source j, UR i is the carcinogenic risk factor for VOC species i (μg m − 3 ) − 1 , and f i is the unit conversion factor. The U.S. EPA developed Integrated Risk Information System to provide carcinogenic risk factors for VOCs (Additional file 1: Table S3).

Normalized ROFP and RCR
After obtaining the ROFP and RCR values for different emission sources, they were sorted from high to low.
where, NROFP j is the normalized ROFP index of source j, ROFP min is the minimum ROFP j among VOC sources, ROFP max is the maximum ROFP j among VOC sources, NRCR j is the normalized RCR index of sources j. RCR min is the minimum RCR j among VOC sources, RCR max is the maximum RCR j among VOC sources.

Emission factors
The ∑VOCs emission factors of the manufacturing sectors are shown in Fig. 1. The sequence of ∑VOC emission factors for the five groups of manufacturing sectors was printing > plastic > metal > chemical > leather. The ∑VOCs emission factors of manufacturing sectors were observed to be ranging from 0.00267 to 423 mg-VOC kg − 1 . Among all sectors, the maximum emission factors of ∑VOC were observed for stencil printing (423 mg-VOC kg − 1 ) followed by plastic tape (24.9 mg-VOC kg − 1 ), metal product and machinery (10.3 mg-VOC kg − 1 ) and organic solvent (4.4 mg-VOC kg − 1 ). For synthetic resin, acrylic resin, polyester resin, integrated iron and steel, aluminum foundry, steel foundry, non-ferrous metal, and polyurethane (PU) leather and plastic coating, ∑VOCs emission factors were found ≤1 mg-VOC kg − 1 .
Stencil printing had the highest emission factor which could be due to the presence of a large number of organic compounds in printing inks. Toluene was the most abundant VOC species measured in emissions from a Chinese printing industry [49]. In general, toluene, ethylbenzene, xylenes and ethyl acetate were common components in ink solvents [34,49]. Some other organic compounds such as acetate, glycolic acid butyl ester, butyl glycol were also used in printing inks [50], but their percentage varied largely due to the heavy use of various solvent-based inks and paint solvents. Printing inks are made up of pigments, dyes, additives and carrier solvents. Different types of printing sectors have variable ink flow properties, which range from extremely thin watery through highly viscous to dry powder.

VOC species compositions
The VOC emission factors have been expressed as the percentage of each species relative to ∑VOCs emission factors. The identified VOCs were classified into the following five categories: alkanes, alkenes, aromatics, halocarbons and oxygenated VOCs. As presented in Fig. 2, alkanes formed the dominant VOC group in steel foundry (42%), aluminum foundry (25%) and synthetic resin (25%). A higher proportion of alkenes were observed for metal products and machinery (60%) followed by other emission sectors. Zhao et al. [33] observed a high percentage of alkanes (26%) followed by alkynes (16%), aromatics (14%) and alkenes (11%) in the emissions from iron and steel sectors. In the integrated iron and steel sector, aromatic compounds were the dominant (84%) VOC group followed by oxygenated VOC and alkenes. Similar results were reported by Tsai et al. [36] that showed a high contribution (45-70%) of aromatic compounds from integrated iron and steel industry. The higher proportion of alkanes and alkenes emissions indicates insufficient oxidation of volatile components released from the fuels [46]. Chemical sectors such as organic solvent and polyester resin manufacturing were found to have more than 75% ∑VOCs contributed by oxygenated VOCs. Acrylic resin, plastic tape, integrated iron and steel, plastic coating and stencil printing sectors showed the major fraction of ∑VOCs contributed by aromatics (65-99%). Tsai et al. [36] reported the high abundance of aromatic species, including, toluene, 1,2,4 trimethylbenzene, xylene and benzene in the emissions from an integrated iron and steel industry. Yuan et al. [29] reported that the aromatics were the most abundant VOC group in the emissions from solvent usage in auto-painting, architectural painting and printing. The results provide only the functional groups of VOCs for different manufacturing sectors showed the VOCs emission characteristics from different product processes in three plants. Detailed profiles of the VOCs (top 5 contributors) for each manufacturing sector are presented in Table 3 and discussed in the following subsections.

Chemical manufacturing
The top five VOC species emitted from four chemical manufacturing sectors, such as organic solvent, synthetic resin, polyester resin and acrylic resin industries are shown in Table 3. The VOCs emitted from organic solvent manufacturing were mainly oxygenated VOCs, such as ethyl acetate and acetone, accounting for > 65% of ∑VOCs. Methyl isobutyl ketone, toluene and isopropyl alcohol accounted for a minor fraction of ∑VOCs. Synthetic and polyester resin manufacturing sectors predominantly emitted oxygenated VOCs like acetone (56%, 0.023 mg-VOC kg − 1 ) and ethyl acetate (80%, 0.026 mg-VOC kg − 1 ). However, the acrylic resin industry had higher contributions from aromatics (> 95% of ∑VOCs) such as m/p/o-xylene, ethylbenzene, and toluene. Most organic resins are solvent-borne, which contain a mixture of organic solvents, many of which are VOCs. The common solvents used in resins contain toluene, xylene, benzene, ethyl ketone, and methyl isobutyl ketone. Acrylic resin is widely used in paint formulations, and the paints emit predominantly aromatics compounds, such as benzene, toluene, ethylbenzene and xylene [28,51]. However, the polyester resin was found to have the emission contributions mainly from oxygenated VOCs, as they are formed by the reaction of dibasic organic acids and polyhydric alcohols [52]. The difference in VOC emissions was probably related to the raw materials and products originating from the chemical manufacturing sectors.

Plastic tape manufacturing
The VOC species emitted from the plastic tape manufacturing industry were mainly aromatics such as toluene, ethyl acetate, m/p-xylene and ethylbenzene, accounting for 44% (11 mg-VOC kg − 1 ), 16% (4.0 mg-VOC kg − 1 ), 10% (2.5 mg-VOC kg − 1 ), 10% (2.5 mg-VOC kg − 1 ) and 5% (1.3 mg-VOC kg − 1 ) of ∑VOCs emission factors, respectively. The reason for the high emission of toluene may be related to the nature of raw material used in plastic tape manufacturing. Plastic tape products included petroleum, petroleum by-products, natural rubber, acrylic resins, silicone rubber, dispersions, polymers and solvents. Solvent-based adhesive tape applications require high-stress resistance. There is no solvent-free adhesive tape available that shows equivalent properties. Therefore, there is no alternative to solvent-based adhesives in the high-quality range. All stages of the tape manufacturing process included hot melts releasing VOCs into the atmosphere. Tsubaki et al. [53] found toluene to be the major VOC species emitted from the double-sided pressure-sensitive adhesive tape.

Metal manufacturing
Toluene, ethyl acetate, acetone, 1-butene, n-butane, cis-2-butene, benzene, and ethanol were the main VOC species emitted from the metal-based manufacturing industries. Toluene was the most abundant VOC species in the emissions from both aluminum and steel foundries, accounting for 40% (0.012 mg-VOC kg − 1 ) and 12% (0.00031 mg-VOC kg − 1 ) of the ∑VOCs emission factors, respectively. Benzene had the highest contribution (41%, 0.024 mg-VOC kg − 1 ) in the integrated iron and steel sector. Samples from metal product and machinery and non-ferrous metal showed the major fractions of 1butene and ethyl acetate emissions, accounting for 40% (4.2 mg-VOC kg − 1 and 38% (0.0041 mg-VOC kg − 1 ) of the ∑VOCs emission factors, respectively (Table 3). VOC compositions in the emissions from the integrated iron and steel manufacturing sector were similar to those reported by the previous work [36] for four processes of the integrated iron and steel industries. The study reported benzene, toluene, xylene, n-butane and 2methylpentane for the sintering process as the major VOCs. Toluene, 1,2,4-trimethyl benzene, isopentane, m/ p-xylene, 1-butene, ethylbenzene, benzene, trichloroethylene, n-hexane and n-pentane were the major VOC species in coke making exhausts [36].

Leather manufacturing
The major VOCs emitted from PU leather industry were methyl ethyl ketone, toluene and methyl methacrylate, accounting for 60% (0.40 mg-VOC kg − 1 ), 24% (0.16 mg-VOC kg − 1 ) and 13% (0.085 mg-VOC kg − 1 ) of ∑VOCs, respectively. The contributions of major VOC species detected in this study were similar to those reported in the earlier studies. Chang and Lin [54] obtained a higher amount of toluene and methyl ethyl ketone in emissions from coating, drying and surface treating processes of the PU leather industry. Wang et al. [55] reported a high      quantity of 2-butanone, toluene and ethyl acetate in different areas of the manufacturing facility such as manufacturing department, semi-finished raw material department, resin warehouses and outside vicinity of the industries. Organic solvents are widely used in the PU industries which could be a potential source of VOCs. The species of organic solvents include toluene, methyl ethyl ketone and dimethylformamide. Some solvents are added as thinners and additives to avoid excessive viscosity of polyurea-formaldehyde in PU industry [54]. These results showed that the VOC species compositions were attributable to the organic solvent and other raw materials used in PU industries.

Printing sectors
The printing sector in the present study included plastic coating and stencil printing (Table 3). Toluene was the most abundant species in the emissions of plastic coating and stencil printing, accounting for 64% (0.0099 mg-VOC kg − 1 ) and 35% (149 mg-VOC kg − 1 ) of ∑VOCs emission factors, respectively. However, methyl methacrylate, ethyl acetate, acetone, methyl ethyl ketone and 2-methyl hexane accounted for a minor fraction of ∑VOCs for both sectors. Organic solvents are widely used in plastic coating and ink printing [29] which could make an important source of aromatic compounds. Li et al. [34] obtained a higher percentage of alkanes (45%) and aromatics (47%) in printing ink. The study reported that VOC species emitted from the ink printing sectors were mainly, alkanes and aromatics. Some other studies found that toluene, benzene, and some oxygenated organics are the typical species emitted form printing sectors [28,34,56]. Zheng et al. [28] reported aromatics (e.g., benzene and toluene) as the major species of letterpress printing, while ethyl acetate and isopropyl alcohol were important VOCs in offset printing and gravure printing processes. Raw materials and derivatives might be the responsible factor for the observed differences in the emissions from different printing sectors.

ROFP
The top five VOCs in terms of their contributions to ROFP in different emission sectors are shown in Fig. 3. The top five VOC species for organic solvent included ethyl acetate, toluene, 1,2,4-trimethylbenzene, m-xylene and m-ethyl toluene that together accounted for 75% of the total ROFP. In synthetic resin industries, the top five ROFP VOCs accounted for 72% of the total ROFP. They included aromatics (toluene and m-xylene), alkanes (methylcyclohexane), and oxygenated VOCs (acetone and methyl ethyl ketone). In the polyester resin industry, oxygenated and aromatic species were predominant contributors to the total ROFP, whereas in the acrylic resin industry, the aromatic VOCs (m/p/o-xylene, ethylbenzene and toluene) were the major contributors to ROFP. 1-butene and ethyl acetate were the dominant contributors to the total ROFP for metal product and machinery and non-ferrous metal sectors, respectively. Similarly, toluene was the highest contributor followed by m/p-xylene and ethyl acetate in the plastic tape manufacturing sector. Likewise, toluene was the dominant contributor of total ROFP for the aluminum foundry, steel foundry, integrated iron and steel, PU leather and plastic coating sectors. For the stencil printing sector, 1, 2,3-trimethyl benzene was the major contributor to the total ROFP. Moreover, several aromatic VOCs including 1,3,5-trimethyl benzene, 1,2,3-trimethylebenzene and methyltoluene were also obtained as the top five VOCs contributing to the total ROFP in the stencil printing sector.
The ROFPs of VOCs varied widely across different emission sectors (Additional file 1: Fig. S1). The total ROFP of metal product and machinery emission was the highest (7.6 g-O 3 g-VOCs − 1 ) among all emission sources, followed by acrylic resin manufacturing (7.4 g-O 3 g-VOCs − 1 ), stencil printing (6.5 g-O 3 g-VOCs − 1 ), and plastic tape manufacturing (4.6 g-O 3 g-VOCs − 1 ). In the metal product and machinery sector, alkenes were the main contributors, accounting for 88% of total ROFP. However, in acrylic resin manufacturing, printing and plastic tape sectors, aromatics contributed to 99.9, 99.8 and 85% of total ROFP, respectively. The major aromatic VOCs for these emission sources were 1-butene, m-xylene, 1,2,4-trimethylbenzene and toluene. The total ROFP for synthetic resin manufacturing emissions was the lowest because the oxygenated VOCs with low reactivity accounted for the major fraction of ∑VOCs. The ROFP contributions for acrylic resin, plastic tape, metal manufacturing (integrated iron and steel, aluminum and steel foundries) and printing (plastic coating and stencil printing) emissions were mainly from aromatic VOCs. However, oxygenated VOCs were the major contributing group to the ROFP for other manufacturing sectors (i.e., organic solvent, polyester resin, no-ferrous metal and PU leather). Tsai et al. [36] reported toluene, 1-butene, m/p-xylene, o-xylene, 1,2,4-trimethylbenzene, ethylbenzene, 1,3.5-trimethylbenzene, 1,2, 3-trimethylbenzene as the major VOCs in terms of OFP in the emissions from four process units of an integrated iron and steel plant. The sequence of OFP for four processes was as follows: cold forming ≈ sintering > hot forming > cakemaking. Furthermore, OFP of cold forming and sintering process is about 4 and 5 times higher than that of hot forming and cakemaking processes, respectively. Ou et al. [57] reported that industrial solvents and gasoline vehicles contributed 33 and 18% of anthropogenic OFP of VOC emissions, respectively, whereas, motorcycles and industrial processes accounted for 14 and 13%, respectively. Li et al. [34] also reported a similar VOC emission pattern for iron and steel and printing sectors. A detailed comparison of the results with literature can be found in the supplementary material (Additional file 1: Table S4). It should be noted that the ozone formation potential depends on the speciated VOC emission concentrations and their MIR values. Thus, VOCs with high emission factors may not always be the major contributors to ozone formation.

RCR
The RCR of each of the emission sources investigated in this study has been presented in Additional file 1: Fig.   S2. The RCR of emissions from the integrated iron and steel sector was the highest among all tested emission sources with the risk value of 41 × 10 − 8 mg − 1 m 3 , followed by steel foundry (4.0 × 10 − 8 mg − 1 m 3 ) and nonferrous metal industry (9.2 × 10 − 9 mg − 1 m 3 ). RCR values of other emission sectors such as organic solvent, synthetic resin, acrylic resin, plastic tape, aluminum foundry, PU leather, plastic coating and stencil printing ranged from 1.2 × 10 − 10 -9.1 × 10 − 9 mg − 1 m 3 . The RCRs were not calculated for polyester resin and metal product and machinery because of the unavailability of the carcinogenic risk factor values for the VOCs detected in the emissions of those sources. The highest relative RCRs were observed for the integrated iron and steel manufacturing sector and steel foundry because those sources emitted a high amount of benzene. Several VOC species, including benzene, chloroform, trichloroethylene, bromoform and 1,3-butadiene are classified as carcinogenic compounds according to the Integrated Risk Information System. Industrial manufacturing sectors are reported as the major VOC emission sources [28,36]. Several previous studies have assessed cancer risks for inhalation exposure to VOCs in industrial areas, but most of them focus on the ambient VOCs [58][59][60].

Specific potential emission sources
A comparison of potential VOC emission sources based on ROFP and RCR between different emission sectors is presented in Fig. 4. NROFP and NRCR were categorized into four levels: Level-I (0.75-1.00), Level-II (0.50-0.75), Level-III (0.25-0.50) and Level-IV (0-0.25) where Level-I represents the most important VOC emission sources and Level-IV represents the least important emission source in terms of NROFP and NRCR. Metal product and machinery, acrylic resin and stencil printing sectors belonged to the Level-I emission sources (NROFP index: > 0.75). However, in terms of the NRCR, it was found that the integrated iron and steel had the highest NRCR index that belonged to Level-I emission sources. Therefore, it can be concluded that not only the single factor such as OFP but also the cancer risk should be taken into consideration for the identification of potentially important VOC emission sources.

Conclusions
In this study, 13 types of stationary emission sources were investigated for emission factors of speciated VOCs and associated ROFP and RCR. ROFP and RCR were used for the identification of potentially important VOC emission sectors. The results showed that the maximum ∑VOCs emission factors were observed for stencil printing (423 mg-VOC kg − 1 ) among all emission sources. Alkanes formed the dominant VOC group in the steel foundry (42% of ∑VOCs emission factors), aluminum foundry (25% of ∑VOCs emission factors) and synthetic resin (25% of ∑VOCs emission factors). However, in the chemical sector (synthetic and polyester resin), oxygenated VOCs such as acetone (56% of ∑VOCs emission factors) and ethyl acetate (80% of ∑VOCs emission factors) were the major contributors to total VOCs. Xylene, ethylbenzene, and toluene formed > 95% of ∑VOCs in the emissions from acrylic resin manufacturing. Analysis of ROFP showed that metal product and machinery sector, acrylic resin manufacturing and stencil printing were potentially important emission sources in terms of OFP. In terms of the NRCR, integrated iron and steel sector was the potentially important source with the highest NRCR index. The results suggested that the reactivityand carcinogenicity-based approach is required for prioritizing the emission sources. Toluene, xylene, 1,2,4-trimethylbenzene and other high OFP-contributing species were the major reactive species that could be targeted while developing control strategies. In addition, carcinogenic VOCs such as benzene, chloroform, trichlorethylene, bromoform and 1, 3-butadiene should also be targeted for emission control.
Additional file 1 Table S1 Average concentrations of the VOC species identified in different manufacturing sectors. Comparison of ozone formation potential of VOC obtained in this study and with previous studies. Fig. S1 Relative ozone formation potential of VOCs for different emission sources. Fig. S2 Relative carcinogenic risk of VOCs for different emission sources.