environmental and health impacts of the change in nmhcs ... · issn: 1680-8584 print / 2071-1409...

Special Session on Better Air Quality in Asia (I)

Aerosol and Air Quality Research, 20: 930–943, 2020 Copyright © Taiwan Association for Aerosol Research ISSN: 1680-8584 print / 2071-1409 online doi: 10.4209/aaqr.2019.09.0459

Environmental and Health Impacts of the Change in NMHCs Caused by the Usage of Clean Alternative Fuels for Vehicles Dalei Wu1, Leilei Fei2,3,4*, Zhisheng Zhang2,3, Yiqiang Zhang2,3, Youping Li5, Chuenyu Chan3, Xinming Wang4, Chaoping Cen2, Pu Li3,6, Lingwei Yu3,6 1 Institute of Environment and Development, Guangdong Academy of Social Sciences, Guangzhou 510635, China 2 Key Laboratory of Water and Air Pollution Control of Guangdong Province, South China Institute of Environmental Sciences, Ministry of Ecology and Environment, Guangzhou 510655, China

3 Sun Yat-sen University, Guangzhou 510275, China 4 State Key Laboratory of Organic Geochemistry and Guangdong Key Laboratory of Environmental Resources Utilization and Protection, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou 510640, China 5 College of Environmental Science and Engineering, China West Normal University, Nanchong 637009, China 6 Guangzhou Scie-Work Environmental Protection Technology Co., Ltd., Guangzhou 510635, China ABSTRACT

Vehicle exhaust and roadside air samples were collected at Guangzhou, Shenzhen and Chengdu in China, where liquefied

petroleum gas (LPG), liquefied natural gas (LNG) and compressed natural gas (CNG)—common fuel alternatives for vehicles worldwide—are used, respectively. The emission characteristics, ozone formation potential (OFP) and health risks of nonmethane hydrocarbons (NMHCs) in conventional and clean alternative fuel exhaust emitted by vehicles were assessed to explore the environmental and health effects from the change in NMHCs due to using clean alternative fuels. Our results indicate that the fuel type significantly impacted the composition of the roadside air. The OFP values for the total NMHCs were much lower in the CNG (2.5 ppmv), LNG (4.7 ppmv) and diesel (4.5 ppmv) exhaust than in the gasoline (94.3 ppmv) and LPG (23.1 ppmv) exhaust, indicating that using CNG and LNG may effectively reduce O3 formation due to vehicle exhaust. Additionally, the hazard quotient (HQ), hazard index (HI) and cancer risk (Risk) of NMHC species in the vehicle exhaust were calculated. Both the HI and Risk of these species in the exhaust from gasoline-powered vehicles greatly exceeded those from the other four types of vehicles, suggesting that using clean alternative fuels instead of gasoline benefits human health.

Keywords: VOCs; LPG; LNG; CNG; Ozone formation potential; Health risk assessment. INTRODUCTION

Nonmethane hydrocarbons (NMHCs) are important components of volatile organic compounds (VOCs), which are precursors of tropospheric ozone and secondary organic aerosols (Carter, 1994; Atkinson and Arey, 2003; Zhang et al., 2018). The abundance and speciation of NMHCs in the atmosphere significantly affect the atmospheric chemistry, environmental quality and human health (Guo et al., 2011). Vehicle exhaust emissions, known as important anthropogenic sources of VOCs worldwide, make a significant contribution to air pollution in China because of the increasing number of automobiles, especially in urban areas (Kelly and Zhu, 2016; Guo et al., 2017; Zhang et al., 2018). A multi-year study from * Corresponding author. Tel.: (+86-20)-85648556; Fax: (+86-20)-85648560 E-mail address: [email protected]

2005 to 2013 at a suburban site (Tung Chung) in Hong Kong showed that vehicular exhaust, gasoline evaporation and liquefied petroleum gas (LPG) usage accounted for 20.2 ± 6.2% and 25.4 ± 6.3% of the total NMHCs, respectively (Ou et al., 2015). A speciated VOC emission inventory in the Pearl River Delta (PRD) region established by Zheng et al. (2009) showed that gasoline vehicles, diesel vehicles and motorcycles were the largest contributors to VOCs emissions, accounting for 45.5–51.5%. In Beijing, a long-term measurement of NMHCs from 2004 to 2012 showed that transportation-related sources (gasoline evaporation and vehicle exhausts) accounted for 32–64% of the total NMHCs (Wang et al., 2015).

The degree of NMHC emissions from vehicles is affected by multiple factors, including fuel type, vehicle technology, emission control technology, traffic condition, driving mode, driving behavior and ambient meteorological condition, etc. (Guo et al., 2011). Various measures have been taken to reduce vehicle emissions, among which development and promotion of the usage of clean alternative fuels is one of the most commonly used strategy. LPG, liquefied natural gas (LNG) and

Wu et al., Aerosol and Air Quality Research, 20: 930–943, 2020 931

compressed natural gas (CNG) are widely used as alternative fuels for vehicles worldwide (Chang et al., 2001; Ristovski et al., 2005; Zhang et al., 2009; Li et al., 2014; Lounici et al., 2014; Agarwal et al., 2015). LPG consists of propane, isobutane and n-butane in major; LNG and CNG are mainly composed of methane and ethane (Chen et al., 2001; Kado et al., 2005; Lai et al., 2009). LPG, LNG and CNG are known as clean alternative fuels because of their higher efficiency and lower emission (Aslam et al., 2006; Lai et al., 2009; Osorio-Tejada et al., 2017). Previous studies showed that the exhaust from LPG, LNG and CNG vehicles had lower emission levels of carbon oxide (CO), hydrocarbons (HCs) and particles, compared to the exhaust from conventional vehicles (diesel and gasoline) (Chang et al., 2001; Aslam et al., 2006; Tsai et al., 2006; Wayne et al., 2009; Guo et al., 2011). Large-scale application of LPG, LNG and CNG could have an important impact on the composition of atmospheric NMHCs (Chen et al., 2001; Kado et al., 2005; Zielinska et al., 2014). Fuel types for vehicles can further affect human health as well, since some species of the NMHCs (e.g., benzene, toluene, ethylbenzene and xylene) in the exhaust are classified as hazardous air pollutants with potential carcinogenic effects (Edokpolo et al., 2014; Moolla et al., 2015).

With the vigorous promotion of energy structure adjustment and environmental pollution control in China in recent years, consumption of natural gas and LPG has increased to a certain extent. The share of natural gas in energy consumption has increased significantly from 4.2% in 2010 to 7.5% in 2017, and the consumption of LPG increased from 146.99 million tons in 2010 to 169.65 million tons in 2017 (Energy Statistics Division of National Bureau of Statistics, 2019). Meanwhile, the development of gas vehicles is vigorously promoted to deal with the increasingly serious air pollution problem in China. In 1999, Ministry of Science and Technology of the People’s Republic of China started the implementation of Air Purification Project: Clean Cars Action in 12 demonstration cities including Guangzhou. The number of gas vehicles increased rapidly from less than 10,000 in 1999 to 215,000 in 2004 (including 114,000 LPG vehicles and 101,000 CNG vehicles) (Zhang, 2008). By the end of 2004, there were 712 gas-filling stations in key cities (or regions), among which 357 were CNG-filling stations and 355 were LPG-filling stations (Wang, 2005). By the end of 2017, there were 6.08 million natural gas vehicles (5.73 million CNG vehicles and 0.35 million LNG vehicles) and 8,400 gas-filling stations in China, ranking the first in the world (www.sohu.com/a/2590 16455_463997). In this study, we conducted experiments at Guangzhou, Shenzhen and Chengdu in China, where LPG, LNG and CNG were used as alternative fuels for vehicles, respectively. Emission characteristics, ozone formation potential (OFP) and health risk assessment of NMHCs from five different fuel types of vehicles were discussed to explore the environmental and health impacts of the change of NMHCs caused by the usage of clean alternative fuels for vehicles. METHODS Site Description and Sampling Methods

Guangzhou and Shenzhen in the PRD region of South

China and Chengdu in the Chengdu-Chongqing Region (CCR) of Southwest China were selected as sampling sites for this study. Guangzhou is the capital city of Guangdong Province, which is also the economic and cultural center of the PRD region (Situ et al., 2013). Guangzhou began to promote the usage of LPG as bus and taxi fuel in 2003 (Tang et al., 2007). By 2009, Guangzhou owned the largest number of LPG buses in the world. By 2012, there have been 19,943 taxis and 11,911 buses in Guangzhou (Guangzhou Statistics Bureau, 2013), with LPG-powered taxis and buses accounting for 100% and over 80% of the total numbers of taxis and buses, respectively. As for the private motor vehicles, the total number was 1,646,930 (including 1,520,210 passenger vehicles, 121,870 trucks and 4,850 other vehicles) in 2012 in Guangzhou (Guangzhou Statistics Bureau, 2013). However, alternative fuels have not been widely promoted for private motor vehicles yet in Guangzhou. Shenzhen is another core city of the PRD region, adjacent to Hong Kong. Different from Guangzhou, Shenzhen has promoted the usage of LNG as an alternative fuel for bus instead of LPG since 2009. As one of the 13 pilot cities for demonstration and promotion of energy-saving and new-energy vehicles, Shenzhen was the city owning the most LNG vehicles in China by 2012, although there were only around 500 LNG buses in Shenzhen at that time. Chengdu is the capital city of Sichuan Province, which is a representative city where CNG vehicles were used at a relatively large scale in China. There were over 100,000 CNG vehicles in Chengdu by 2012, with CNG-powered taxis and buses accounting for over 90% of the total number of taxis and buses (Liao, 2013).

Gas samples were collected with evacuated 2 L stainless steel canisters (Entech Instruments Inc., USA). Both vehicle exhausts and roadside air were sampled. Table 1 presents a summary of sampling information in this study. For vehicle exhausts, each sample was collected directly from the tailpipes of 5 random vehicles using same fuels at idling state to reduce the influence of different vehicle and driving conditions. The sampling time of the vehicle exhaust gas samples was around 10 min for each vehicle (50 min for each mixed sample). Diesel, gasoline, LPG, LNG and CNG vehicles were involved. 1-hour roadside air samples were collected using canisters attached with a flow restrictor (Entech Instruments Inc., USA) during rush hours (08:00–10:00 and 17:00–20:00) at Guangzhou, Shenzhen and Chengdu. Representative busy roads without any significant stationary emission sources of NMHCs were selected in each city. The roadside air sampling location of Guangzhou was selected at Xin’gang West Road in Haizhu District (near the south gate of Sun Yat-sen University), which is a main traffic road in Guangzhou urban area. Roadside air samples in Shenzhen and Chengdu were collected by the side of Heping Road and Shennan East Road in Luohu District (two of the roads with the most intensive LNG buses in Shenzhen), and Renmin South Road (near Jinjiang Hotel) and 1st Ring Road (near Sichuan University) in Jinjiang District, respectively. A video camera was used to determine the vehicle counts. Average traffic flow (unidirectional traffic flow) of these selected roads during the sampling period was 2126, 1794 and 2002 vehicles per hour in Guangzhou, Shenzhen and Chengdu, respectively.


Table 1. Summary of sampling information.

Sample Type Sampling Site Sampling Period Sample Number Vehicle Exhaust Diesel Shenzhen 2011.09 6a

Gasoline Shenzhen 2011.09 6a LPG Guangzhou 2012.09 11a,b LNG Shenzhen 2012.10 7a CNG Chengdu 2012.09 12a,c

Roadside Air Shenzhen 2011.09 6 Guangzhou 2012.09 6 Guangzhou 2008.05 6 Chengdu 2012.09 6

a Each vehicle exhaust sample was collected from 5 random vehicles using same fuels. b 5 samples from LPG bus exhausts; 6 samples from LPG taxi exhausts. c 6 samples from CNG bus exhausts; 6 samples from CNG taxi exhausts.

The sampling locations were 2 meters from the roadside and 1.5 meters above ground. Weather conditions (including temperature, relative humidity, wind speed, air pressure, rainfall and cloud amount) during the sampling periods are presented in Table S1. It is worthy of notice that the differences of meteorological conditions in different sampling times and places could affect the sampling results, including the concentration of NMHCs and related photochemical reactions in the atmosphere. Wang et al. (2016) concluded that strong sunlight and low wind speed enhanced ozone production and the accumulation of ozone and its precursors. Tong et al. (2017) concluded that solar radiation and air temperature were positively correlated with ozone concentrations, while relative humidity and heavy precipitation were negatively correlated with ozone concentrations. Samples were collected under similar meteorological conditions in 3 cities to reduce the influence from meteorological factors, although the differences were inevitable. Laboratory Analysis

Gas samples were analyzed by State Key Laboratory of Organic Geochemistry (SKLOG), Guangzhou Institute of Geochemistry (GIG), Chinese Academy of Sciences (CAS). NMHCs were analyzed using a pre-concentrator (Model 7100; Entech Instruments Inc., USA) combined with a gas chromatography-mass selective detector/flame ionization detector (GC-MSD/FID; Model 5973N; Agilent Technologies, USA). In total 67 hydrocarbons were quantified by this system, including 29 alkanes, 19 alkenes, 1 alkyne and 18 aromatics. The method detection limits (MDLs) for each NMHC species are presented in Table S2. Details including analytical steps and parameters are described elsewhere (Zhang et al., 2012, 2015, 2018). In brief, NMHCs inside the canisters were firstly concentrated utilizing a liquid-nitrogen cryogenic trap at –160°C. The trapped NMHCs were then transferred by pure helium to a second-stage trap at –40°C with Tenax-TA as adsorbent. Most of H2O and CO2 were removed during these two steps. The second-stage trap was then heated to get NMHCs transferred by helium to a third-stage cryo-focus trap at –170°C. After the focusing step, the trap was rapidly heated and the NMHCs were transferred to the GC-MSD/FID system for quantification. The mixture was firstly separated by a HP-1 capillary column (60 m ×

0.32 mm × 1.0 µm; Agilent Technologies, USA) with helium as carrier gas, and then split into two: One is to a PLOT-Q column (30 m × 0.32 mm × 2.0 µm; Agilent Technologies, USA) followed by FID detection; the other one is to a stainless steel line (65 cm × 0.10 mm I.D.) followed by MSD detection. The GC oven temperature was initially programmed at –50°C, holding for 3 min; increasing to 10°C at 15°C min–1, then to 120°C at 5°C min–1, and then to 250°C at 10°C min–1 and holding for 10 min. The MSD was operated in electron impact ionization mode using selected ion monitoring (SIM) mode. Quality assurance and quality control procedures are available in the supplemental material. Ozone Formation Potential (OFP)

The OFP proposed by Carter (1994) is widely used to estimate the contribution of VOC to O3 formation (e.g., Barletta et al., 2008; Czader et al., 2008; Suthawaree et al., 2012; Li et al., 2017; Tsai et al., 2018). The OFP is calculated as the product of the concentration of an individual VOCs species and its corresponding maximum incremental reactivity (MIR) (Carter, 1994). MIR is defined as the maximum increment of O3 per additional individual VOCs added with the assumption of sufficient NOx and light intensity (Suthawaree et al., 2012). Updated MIR scales can be found in the study of Carter (2010). Source-oriented reactivities of most organic compounds in urban and industrial areas were in good agreement with the MIR scales in the study of Czader et al. (2008). In this study, the concentration data from conventional and clean alternative fuel vehicles were used to compare the OFP values for NMHC species in the exhaust from different types of vehicles.

Human Health Risk Assessment

Non-cancer and cancer risks were determined using non-carcinogenic hazard quotient (HQ) and cancer risk (Risk) guided by a United States Environment Protection Agency (U.S. EPA) recommended procedure for assessing inhalation risks (U.S. EPA, 2009). Hazard index (HI) is the sum of all HQs for multiple chemicals assessed via a hazard-based approach. In order to assess both non-cancer and cancer risks, the exposure concentration (EC) for each receptor exposed to contaminants via inhalation should be estimated first. The equations for calculating EC, HQ, HI and Risk are


presented below:

CA ET EF EDEC

AT

(1)

where, EC (µg m–3) = exposure concentration; CA (µg m–3) = contaminant concentration in air; ET (hours day–1) = exposure time; EF (days year–1) = exposure frequency; ED (years) = exposure duration; and AT (lifetime in years × 365 days year–1 × 24 hours day–1) = averaging time. HQ = EC/RfC (2) where, HQ (unitless) = hazard quotient; EC (µg m–3) = exposure concentration; and RfC (µg m–3) = reference concentration.

ii

HI HQ (3)

where, HI (unitless) = hazard index; and HQi (unitless) = hazard quotient of chemical i. Risk = EC × IUR (4) where, EC (µg m–3) = exposure concentration; and IUR (µg m–3)–1 = inhalation unit risk.

RESULTS AND DISCUSSION Emission Characteristics of NMHCs from Conventional and Clean Alternative Fuel Vehicle Exhausts

Fig. 1 presents major species contributions (% by volume of the total NMHCs) to the exhausts from diesel-, gasoline-, LPG-, LNG- and CNG-powered vehicles. Table 2 presents the top ten most abundant NMHC species in the exhaust from each type of vehicles. C2–C4 alkenes and aromatics were the major NMHCs emitted from diesel-powered vehicles, with ethene, acetylene and propane the top three most abundant NMHC species. The concentration of ethene, acetylene and propane in diesel vehicle exhaust was 0.2, 0.15 and 0.15 ppmv, respectively. These three species accounted for 43.5% of the total NMHCs. Benzene and toluene were important NMHCs components for diesel vehicle exhaust as well, accounting for 6.3% and 2.6%, respectively. In addition, the concentration of long-chain alkanes including n-nonane, n-decane, n-dodecane and n-undecane ranged from 0.05 ppmv to 0.1 ppmv, accounting for 4.2–8.8%. For gasoline vehicle exhaust, C5–C8 alkanes, C2–C4 alkenes and aromatics were the major NMHCs components, with BTEX accounting for 33.2% of the total NMHCs. The concentration of toluene, m/p-xylene, ethylbenzene and benzene in gasoline vehicle exhaust was 2.61, 1.37, 1.35 and 1.32 ppmv, accounting for 13.1%, 6.8%, 6.7% and 6.5% of the total NMHCs, respectively. The concentration of total NMHCs in gasoline vehicle exhaust (19.95 ppmv) was much larger than that in diesel vehicle exhaust (1.28 ppmv). The major species in diesel or gasoline vehicle emissions obtained in this study are similar to the results in other studies (Liu et al., 2008; Guo et al., 2011).

Emission characteristics of NMHCs were quite different between exhausts from conventional fuel vehicles and clean alternative fuel vehicles. Propane, n-butane and isobutane were the top three most abundant species of NMHCs from LPG vehicle exhaust, with the concentration of 16.65, 2.67 and 1.87 ppmv, respectively. Major components of LPG vehicle exhaust were similar with those of LPG fuel, indicating that high concentration and contribution of propane, n-butane and isobutane in the exhaust mainly came from the leaked or unburned LPG fuel. This result is similar to that by Lai et al. (2009) and Ho et al. (2013). It is noted that ethene, propene and acetylene, which were the top three most abundant species of NMHCs from diesel vehicle exhaust, were major components of LPG vehicle exhaust as well. LNG and CNG were two existing forms of natural gas, resulting in similar compositions of NMHCs from LNG and CNG vehicle exhausts. Ethane and propane contributed more than 75% to the total NMHCs for both LNG and CNG vehicle exhausts. High ethane and propane in the air were also observed in Cairo, Karachi and Dhaka where CNG is widely used (Abu-Allaban et al., 2002; Barletta et al., 2002; Suthawaree et al., 2012). Besides, n-butane and isobutane were also important components of LNG and CNG vehicle exhausts.

Apart from the differences among the components of different types of fuels, whether an original or modified engine was used by the alternative fuel vehicles affected the composition and abundance of NMHCs as well. Previous studies showed that the incomplete combustion and leakage of LPG associated with modified engines played an important role on the air quality in Guangzhou (Tsai et al., 2006; Tang et al., 2008; Lai et al., 2009). Similarly, although the chemical compositions of LNG and CNG are quite close, LNG vehicles in Shenzhou emitted higher concentration of NMHCs than CNG vehicles in Chengdu as a result of the engines. The LNG vehicles we sampled in 2012 in Shenzhou were equipped with engines modified from diesel internal combustion engines. However, the CNG vehicles in Chengdu were equipped with original engines specifically developed for CNG. Thus, it is recommended to promote alternative fuel vehicles with original engines, because of the higher emission levels of air pollutants in the exhaust from vehicles with modified engines compared with those from vehicles with original engines. The Change of NMHCs in the Roadside Air Caused by the Usage of Clean Alternative Fuels for Vehicles

Fig. 2 presents major species contributions (% by volume of the total NMHCs) to the roadside air of Guangzhou, Shenzhen and Chengdu together with 1-hour average traffic composition during the sampling period. Propane, n-butane and isobutane were the top three most abundant species of NMHCs in the roadside air of Guangzhou, accounting for 40.4%, 17.8% and 10.2%, respectively. They were the major components of both LPG fuel and LPG vehicle exhaust (accounting for 57.7%, 15.1% and 11.0%, respectively), as shown in Table 2. Proportion of these three species indicate that LPG vehicle exhaust was the main source of propane, n-butane and isobutane in the roadside air at Guangzhou.


Fig. 1. Major species contributions (% by volume of the total NMHCs) to the diesel, gasoline, LPG, LNG and CNG vehicle exhausts.

Similar results were reported by Tang et al. (2007) and Lai et al. (2009) as well. Guangzhou began to promote the usage of LPG as bus and taxi fuel in 2003 (Tang et al., 2007). The concentration of propane in roadside air was 16.55 ppbv in 2000 in Guangzhou (Tang, 2007; Tsai, 2007) as shown in Table 3, and it increased significantly after the promotion of LPG vehicles to 102.31 ppbv and 63.87 ppbv in 2008 and 2012, respectively. Meanwhile, the concentrations of other common pollutants from gasoline and diesel vehicles (e.g.,

toluene, benzene, ethene and acetylene) decreased significantly in 2008 and 2012 compared to those in 2000, indicating the improvement of the roadside air quality because of the upgrade of fuel quality and vehicle emission standards in China during this period. Generally, the total concentration of the first twelve NMHC species in Table 3 decreased from 512.12 ppbv in 2000 to 193.98 ppbv and 144.36 ppbv in 2008 and 2012, suggesting the decrease of total NMHCs concentration after the promotion of LPG vehicles.


As for Chengdu, ethane was the most abundant species of NMHCs in the roadside air, accounting for 20.8%. The concentration of ethane in the roadside air in Chengdu (11.42 ppbv) in 2012 was significantly higher than that in Guangzhou (4.14 ppbv in 2012) and Shenzhen (1.66 ppbv in 2011). No obvious correlation was found between the concentration of ethane and the tracers of gasoline and diesel vehicles (benzene and acetylene). Ethane was not a major component of gasoline and diesel vapors, suggesting that gasoline and diesel vehicles were not the main contributors to the high concentration of ethane in the roadside air of Chengdu. It is well known that household natural gas emission was also an important source of ethane in urban atmosphere (Barletta et al., 2002; Yuan et al., 2013). Fig. S1 presents the comparison of major NMHC species of roadside and urban ambient air in Chengdu. However, it can be seen that the concentration of ethane in the roadside air (11.4 ppbv) was far higher than that in the urban ambient air (5.7 ppbv), suggesting that ethane from the vehicles instead of household natural gas emissions dominated the high concentration of ethane in the roadside air of Chengdu. Considering that CNG buses and taxis accounted for 19.1% of the total traffic during the sampling period, and ethane was the main component of the exhaust from CNG buses and taxis, it can be inferred that the relatively high concentration of ethane in the roadside air of Chengdu was mainly contributed by the exhaust emission of CNG buses and taxis. High concentration of ethane was also observed in the roadside air of Karachi (Barletta et al., 2002) and Dhaka (Suthawaree et al., 2012), where CNG is widely used as an alternative fuel.

Different from Guangzhou and Chengdu, LNG vehicle exhaust showed little effect on the roadside air of Shenzhen because of relatively small number of LNG vehicles in Shenzhen. In 2011, there were only about 500 LNG vehicles in Shenzhen. LNG vehicles accounted for only 1.5% of the total number of vehicles during the sampling period. LNG vehicle exhaust tracer (ethane) accounted for only 6.9% of NMHCs in the roadside air of Shenzhen, at the same level as the proportion of ethane in other cities without using clean alternative fuels for vehicles (Tsai et al., 2006). Table 3 presents the mixing ratios for the selected hydrocarbons measured in roadside air of different cities. It can be seen that the concentration of most species decreased significantly from 2000 to 2008 in Guangzhou except propane, which is a tracer of the usage of LPG. Different from Guangzhou, concentrations of all the hydrocarbons listed in Table 3 decreased significantly from 2000 to 2011 in Shenzhen, with the total concentration of the first twelve NMHC species in Table 3 decreasing from 123.8 ppbv to 19.5 ppbv. The concentration of LNG tracer (ethane) decreased from 3.1 ppbv to 1.66 ppbv instead of increasing after the application of LNG buses, indicating that the decrease of NMHCs in roadside air of Shenzhen mainly resulting from the upgrade of fuel quality and vehicle emission standards instead of the promotion of LNG vehicles, because of the limited number of LNG vehicles in Shenzhen (3.4% of the total number of buses). Although we have already chosen the city (Shenzhen) and roads (Heping Road and Shennan East Road) with the

Tab

le 2

. Top

ten

mos

t abu

ndan

t NM

HC

s sp

ecie

s in

the

exha

usts

fro

m d

iese

l-, g

asol

ine-

, LP

G-,

LN

G-

and

CN

G-p

ower

ed v

ehic

les.

D

iese

l G

asol

ine

LP

G

LN

G

CN

G

Spe

cies

C

a %

b S

peci

es

Ca

%b

Spe

cies

C

a %

b S

peci

es

Ca

%b

Spe

cies

C

a %

b 1

Eth

ene

0.2

15.4

T

olue

ne

2.61

13

.1

Prop

ane

16.6

5 57

.7

Eth

ane

5.25

54

.5

Eth

ane

1.29

74

.1

2 A

cety

lene

0.

16

14.1

m

/p-X

ylen

e 1.

37

6.8

n-B

utan

e 2.

67

15.1

Pr

opan

e 3.

02

21.8

Pr

opan

e 0.

14

8.1

3 P

rope

ne

0.15

13

.9

Eth

ylbe

nzen

e 1.

35

6.7

Isob

utan

e 1.

87

11.0

n-

But

ane

1.19

10

.5

Eth

ene

0.07

4.

5 4

n-U

ndec

ane

0.1

8.8

Ben

zene

1.

32

6.5

Prop

ene

1.46

7.

1 Is

obut

ane

0.85

6.

9 n-

But

ane

0.05

2.

7 5

n-D

odec

ane

0.09

8.

3 Is

open

tane

1.

32

6.5

Eth

ene

0.5

4.0

Eth

ene

0.24

3.

3 Is

obut

ane

0.04

2.

3 6

Ben

zene

0.

07

6.3

Eth

ene

1.01

6.

0 A

cety

lene

0.

4 1.

7 Is

open

tane

0.

21

1.1

n-P

enta

ne

0.04

2.

3 7

n-D

ecan

e 0.

05

4.2

1,2,

4-T

rim

ethy

lben

zene

1.

13

5.5

1-B

uten

e 0.

21

1.0

tran

s-2-

But

ene

0.08

0.

2 Is

open

tane

0.

03

1.8

8 n-

Non

ane

0.05

4.

2 M

ethy

l-to

luen

e 0.

99

4.9

Isop

enta

ne

0.14

0.

9 n-

Pen

tane

0.

08

0.4

Ben

zene

0.

02

1.1

9 T

olue

ne

0.03

2.

6 n-

Pen

tane

0.

91

4.4

Eth

ane

0.09

0.

5 Pr

open

e 0.

04

0.1

2-M

ethy

lpen

tane

0.

02

1.1

10

Isob

utan

e 0.

03

2.3

2-M

ethy

lpen

tane

0.

85

4.1

tran

s-2-

But

ene

0.09

0.

5 1-

But

ene

0.02

0.

1 3-

Met

hylp

enta

ne

0.01

0.

8 T

otal

NM

HC

s 1.

28

100

Tot

al N

MH

Cs

19.9

5 10

0 T

otal

NM

HC

s 28

.85

100

Tot

al N

MH

Cs

9.64

10

0 T

otal

NM

HC

s 1.

74

100

a Con

cent

ratio

n (p

pmv)

. b

Vol

ume

frac

tion.


Fig. 2. Major species contributions (% by volume of the total NMHCs) to the roadside air of (a) Guangzhou, (b) Shenzhen and (c) Chengdu; a1–c1 were 1-hour average traffic composition during the sampling period.

largest number of LNG buses at that time, it was still difficult to observe significant changes of NMHCs in the roadside air caused by the promotion of LNG vehicles in Shenzhen. In the case of Chengdu, the concentration of ethane was about 6 times higher than that in Shenzhen, due to a larger-scale promotion of CNG vehicles (over 90% of the total number of taxis and buses).

Impact of Clean Alternative Fuel Usage on the OFP

The OFP values for total NMHCs in the exhaust from each type of vehicles were shown in Table 4. In general, the OFP value for total NMHCs in the gasoline vehicle exhaust was the highest (94.3 ppmv), followed by that in the LPG vehicle exhaust (23.1 ppmv). The OFP values for total NMHCs in the CNG, LNG and diesel vehicle exhausts were much lower, with the values of 2.5, 4.5 and 4.7 ppmv, respectively.

For diesel vehicles, the OFP value for total NMHCs in the

exhaust was much lower than that in the gasoline and LPG vehicle exhausts as a result of relatively low concentration of total NMHCs and lower MIR of diesel fuel markers such as long-chain alkanes (e.g., n-undecane and n-dodecane) in the diesel vehicle exhaust (Liu et al., 2008). Table 4 presents the top ten largest OFP values for major NMHC species in the exhaust from each type of vehicles. It can be seen that alkenes were the largest contributors to the total OFP for diesel vehicle exhaust. Among alkenes, high contributions of ethene, propene and 1-butene to the total OFP were observed, accounting for 32.8%, 32.0% and 4.9%, respectively. Apart from alkenes, aromatics and acetylene were important contributors to the total OFP for diesel vehicle exhaust as well. As for gasoline vehicles, aromatics and alkenes were main contributors to the total OFP, with m/p-xylene and 1,2,4-trimethylbenzene the largest two contributors (accounting for 10.7% and 10.6%, respectively). The contribution ratios of m/p-xylene and 1,2,4-trimethylbenzene


Tab

le 3

. Mix

ing

ratio

s fo

r se

lect

ed N

MH

Cs

mea

sure

d in

roa

dsid

e ai

r of

dif

fere

nt c

itie

s (p

pbv)

.

NM

HC

s

Gua

ngzh

ou

She

nzhe

n C

heng

du

Hon

g K

ong

(MK

AQ

MS

stat

ion)

Hon

g K

ong

(MK

AQ

MS

stat

ion)

K

arac

hi

Dha

ka

Thi

s st

udy

Thi

s st

udy

Tan

g, 2

007;

T

sai,

2007

T

his

stud

y T

ang,

200

7;

Tsa

i, 20

07

Thi

s st

udy

Ho

et a

l., 2

013

Hua

ng

et a

l., 2

015

Bar

letta

et

al.,

200

2 S

utha

war

ee

et a

l., 2

012

2012

20

08

2000

20

11

2000

20

12

2003

20

11

1988

–199

9 20

11

LP

G

LP

G

– L

NG

–

CN

G

LP

G

LP

G

CN

G

CN

G

Eth

ane

4.14

6.

16

13.8

9 1.

66

3.1

11.4

2 3.

60

2.68

83

11

5.28

P

ropa

ne

63.8

7 10

2.31

16

.55

3.66

5.

9 3.

56

6.63

8.

14

40

29.9

4 n-

But

ane

27.7

6 17

.41

24.5

1.

78

8.8

2.18

7.

14

11.4

2 73

7.

10

Isob

utan

e 15

.72

14.7

8 16

.81

2.32

7.

7 2.

49

4.69

6.

68

24.9

4.

36

n-P

enta

ne

2.07

2.

47

15.0

9 0.

73

5 1.

31

0.92

0.

60

102

1.94

Is

open

tane

4.

28

7.7

67.6

3 1.

75

29.2

3.

22

2.99

1.

09

74

2.10

E

then

e 6.

43

16.4

8 11

9.6

1.52

20

.8

4.61

8.

94

4.08

12

9 6.

12

Pro

pene

3.

01

5.1

30.7

7 0.

75

6.4

1.28

2.

27

1.57

39

0.

63

Isop

rene

3.

37

1.07

12

.29

1.02

4.

1 0.

25

0.25

0.

26

1.2

– A

cety

lene

5.

09

11.4

14

0.83

1.

75

21.3

5.

76

7.71

4.

34

109

6.23

B

enze

ne

1.68

2.

93

21.6

4 0.

76

4.1

2.81

1.

41

0.91

19

.7

2.57

T

olue

ne

6.94

6.

17

32.5

2 1.

82

7.4

4.11

7.

30

2.74

44

9.

03

Eth

ylbe

nzen

e 0.

60

1.53

–

0.59

–

0.74

0.

69

0.52

–

0.41

o-

Xyl

ene

0.63

1.

03

– 0.

18

– 0.

78

0.60

0.

26

12

0.41

m

/p-X

ylen

e 1.

44

2.61

–

0.34

–

2.1

1.45

0.

78

34

0.80

Tab

le 4

. Top

ten

larg

est O

FP

val

ues

of m

ajor

NM

HC

s sp

ecie

s in

the

exha

usts

fro

m d

iese

l-, g

asol

ine-

, LP

G-,

LN

G-

and

CN

G-p

ower

ed v

ehic

les.

D

iese

l G

asol

ine

LP

G

LN

G

CN

G

Spe

cies

O

FP

a %

b S

peci

es

OF

Pa

%b

Spe

cies

O

FP

a %

b S

peci

es

OF

Pa %

b S

peci

es

OF

Pa %

b 1

Eth

ene

1462

32

.8%

m/p

-Xyl

ene

1013

9 10

.7%

Pro

pene

87

88

38.1

% E

then

e 15

48 3

2.7%

Eth

ene

1104

43

.5%

2

Pro

pene

14

29

32.0

% 1

,2,4

- T

rim

ethy

lben

zene

99

70

10.6

% P

ropa

ne

5553

24

.1%

Eth

ane

866

18.3

% E

than

e 73

9 29

.1%

3

1-B

uten

e 21

9.1

4.9%

E

then

e 74

55

7.9%

E

then

e 23

40

10.1

% n

-But

ane

679

14.4

% P

ropa

ne

133

5.2%

4

1,2,

4- T

rim

ethy

lben

zene

17

5.8

3.9%

T

olue

ne

7041

7.

5%

n-B

utan

e 18

98

8.2%

P

ropa

ne

664

14.1

% P

rope

ne

82.4

3.

2%

5 m

/p-X

ylen

e 12

8.7

2.9%

1-

But

ene

6831

7.

2%

Isob

utan

e 14

30

6.2%

Is

obut

ane

532

11.3

% t

rans

-2-B

uten

e 74

.6

2.9%

6

1-P

ente

ne

104.

9 2.

4%

tran

s-2-

But

ene

5463

5.

8%

1-B

uten

e 11

53

5.0%

tr

ans-

2-B

uten

e 11

6 2.

5%

n-B

utan

e 51

.9

2.0%

7

Ace

tyle

ne

81.5

1.

8%

o-X

ylen

e 44

54

4.7%

tr

ans-

2-B

uten

e 82

0.8

3.6%

Is

open

tane

95

.1

2.0%

Is

open

tane

50

.1

2.0%

8

1,2,

3- T

rim

ethy

lben

zene

78

.8

1.8%

ci

s-2-

But

ene

4431

4.

7%

cis-

2-B

uten

e 44

8.6

1.9%

1-

But

ene

58.5

1.

2%

Isob

utan

e 44

.5

1.8%

9

Tol

uene

76

.2

1.7%

1,

2,3-

Tri

met

hylb

enze

ne

4274

4.

5%

Ace

tyle

ne

118.

2 0.

5%

Pro

pene

55

.4

1.2%

ci

s-2-

But

ene

40.1

1.

6%

10 t

rans

-2-B

uten

e 61

.2

1.4%

P

rope

ne

4265

4.

5%

1-H

exen

e 11

4.6

0.5%

n-

Hex

ane

54.3

1.

1%

n-P

enta

ne

33.1

1.

3%

T

otal

NM

HC

s 44

59

100%

T

otal

NM

HC

s 94

321

100%

T

otal

NM

HC

s 23

087

100%

T

otal

NM

HC

s 47

28 1

00%

T

otal

NM

HC

s 25

38 1

00%

a O

zone

for

mat

ion

pote

ntia

l (pp

bv).

b C

ontr

ibut

ion

ratio

.


to the total OFP were obviously higher than the volume fractions of these two species (6.8% and 5.5%, respectively). On the contrary, toluene contributed only 7.5% to the total OFP, though it was the most abundant NMHCs specie in the gasoline vehicle exhaust. For LPG vehicles, alkenes and alkanes accounted for the largest fraction to the total OFP, with propene, propane and ethene the top three contributors (accounting for 38.1%, 24.1% and 10.1%, respectively). Although propane accounted for 57.7% of total NMHCs in the LPG vehicle exhaust in this study, it contributed only 24.1% to OFP because of its relatively low MIR. This finding is similar to that by Chen et al. (2001), which found that LPG species accounted for 69% of the NMHCs but only contributed 15% to the production of O3 in Santiago, Chile. Thus, using LPG as an alternative fuel for gasoline has been reported as an effective measure to reduce O3 formation (Chang et al., 2001; Luis et al., 2003; Guo et al., 2011). In the case of LNG and CNG vehicle exhausts, ethene was the largest contributor to the total OFP of NMHCs, followed by ethane. Ethene and ethane contributed over 50% to the total OFP of NMHCs in both LNG and CNG vehicle exhausts. Ethene accounted for 32.7% and 43.5% of the total OFP of NMHCs in LNG and CNG vehicle exhausts, respectively. However, the concentration of ethene accounted for only 3.3% and 4.5% of the total NMHCs in the exhaust of LNG and CNG vehicles, respectively. It seems that the MIR of each species was a more important factor for the calculation of OFP values for NMHCs from vehicles powered by different fuels, especially for the vehicles with clean alternative fuels. Results indicate that fuel type had a significant impact on the OFP for NMHCs in the vehicle exhausts. Compared with gasoline, the usage of clean alternative fuels (especially CNG and LNG) for vehicles could lead to effective reduction of O3 formation from vehicle exhausts.

Top ten largest OFP values for major NMHC species in roadside air of different cities were shown in Table 5. The OFP value for total NMHCs in roadside air of Guangzhou was the highest (454.8 ppbv in 2008 and 267.7 ppbv in 2012), and the OFP values in Shenzhen and Chengdu were relatively low, with the values of 68.8 and 134.8 ppbv, respectively. Ethene was the largest contributor to the total OFP of NMHCs in roadside air of all these 3 cities. The significant drop of OFP values in Guangzhou from 2008 to 2012 could be explained by the upgrade of both fuel quality and vehicle emission standards in China during this period. Similar conclusions were also proposed by Zhang et al. (2018). In addition, as shown in both Tables 4 and 5, LNG and CNG seemed to be better than LPG considering the aim of reducing O3 formation from vehicle exhausts. Health Risk Assessment of NMHCs in the Roadside Air

Table 6 presents the HQ, HI and Risk of NMHC species in the roadside air samples at Guangzhou, Shenzhen and Chengdu (assuming ET = 24 hours day–1, EF = 365 days year–1, ED = 70 years (average human life span) and AT = 70 years × 365 days year–1 × 24 hours day–1). HQs of eleven NMHC species in total in the roadside air samples at Guangzhou and Chengdu were calculated for non-cancer risk assessment,

Tab

le 5

. Top

ten

larg

est O

FP

val

ues

of m

ajor

NM

HC

s sp

ecie

s in

roa

dsid

e ai

r of

dif

fere

nt c

itie

s.

G

uang

zhou

(20

08)

Gua

ngzh

ou (

2012

) S

henz

hen

(201

1)

Che

ngdu

(20

12)

S

peci

es

OF

Pa

%b

Spe

cies

O

FP

a %

b S

peci

es

OF

Pa

%b

Spe

cies

O

FP

a %

b 1

Eth

ene

121.

9

26.8

%

Eth

ene

47.6

17

.8%

E

then

e 11

.2

16.3

%

Eth

ene

34.1

25

.3%

2

Pro

pane

49

.1

10.8

%

Isop

rene

30

.7

11.5

%

Isop

rene

9.

3 13

.5%

m

/p-X

ylen

e 15

.5

11.5

%

3 P

rope

ne

47.9

10

.5%

P

ropa

ne

30.7

11

.5%

P

rope

ne

7.0

10.2

%

Pro

pene

12

.1

9.0%

4

tran

s-2-

But

ene

26.9

5.

9%

n-B

utan

e 28

.3

10.6

%

Tol

uene

4.

9 7.

2%

Tol

uene

11

.1

8.2%

5

m/p

-2-B

uten

e 21

.7

4.8%

P

rope

ne

28.3

10

.6%

1-

But

ene

4.9

7.1%

1-

But

ene

7.1

5.2%

6

1-B

uten

e 21

.3

4.7%

Is

obut

ane

19.0

7.

1%

tran

s-2-

But

ene

2.9

4.1%

o-

Xyl

ene

5.1

3.8%

7

m/p

-Xyl

ene

19.3

4.

2%

Tol

uene

18

.7

7.0%

Is

obut

ane

2.8

4.1%

Is

open

tane

4.

4 3.

3%

8 Is

obut

ane

17.9

3.

9%

m/p

-Xyl

ene

10.7

4.

0%

m/p

-Xyl

ene

2.6

3.7%

1,

2,4-

TM

B

3.5

2.6%

9

n-B

utan

e 17

.8

3.9%

1-

But

ene

9.1

3.4%

Is

open

tane

2.

4 3.

5%

tran

s-2-

But

ene

3.3

2.4%

10

T

olue

ne

16.7

3.

7%

Isop

enta

ne

5.9

2.2%

m

/p-2

-But

ene

2.2

3.3%

m

/p-2

-But

ene

3.1

2.3%

Tot

al N

MH

Cs

454.

8 10

0%

Tot

al N

MH

Cs

267.

7 10

0%

Tot

al N

MH

Cs

68.8

10

0%

Tot

al N

MH

Cs

134.

8 10

0%

a Ozo

ne f

orm

atio

n po

tent

ial (

ppbv

). b

Con

trib

utio

n ra

tio.


including BTEX (benzene, toluene, ethylbenzene, m/p-xylene and o-xylene), styrene, 1,2,3-trimethylbenzene, 1,2,4-trimethylbenzene, 1,3,5-trimethylbenzene, n-hexane and cyclohexane. HQ of n-hexane was not calculated at Shenzhen, since n-hexane was not detected in the roadside air samples. Risks of benzene were calculated for cancer risk assessment at all these 3 cities. RfC and IUR of each chemical used in this study came from the U.S. EPA Integrated Risk Information System (IRIS) (www.epa.gov/iris).

As suggested by U.S. EPA, a value of HQ or HI < 1 implies no significant non-cancer risk, which is considered to be below the level of concern; a value ≥ 1 implies a significant non-cancer risk, which increases with the increasing of HQ or HI value (Bamuwamye et al., 2017; Widiana et al., 2019). In this study, all the HQs and HIs were < 1. The HI of NMHCs in the roadside air samples at Guangzhou, Shenzhen and Chengdu was 0.29, 0.11 and 0.48, respectively, suggesting that non-cancer risks of NMHCs in the roadside air at all these 3 cities were within the safety limit. Li et al. (2013) measured thirty-one kinds of NMHCs in the ambient air at a downtown site of Guangzhou in 2009. They reported that the HI of NMHCs in the air samples was 0.54, which was larger than that at Guangzhou in this study but still below the reference level (HI = 1). Gao et al. (2012) evaluated the health risk of atmospheric VOCs at Tianjin in 2011, with the HI of 0.42, which was also higher than those at Guangzhou and Shenzhen, but slightly lower than that of Chengdu. As for cancer risk assessment, U.S. EPA proposed 1.0E-06 as a management goal, though Risks within the range of 1.0E-06 to 1.0E-04 were considered acceptable for regulatory purposes, depending on the situation and circumstances of exposure (Qu et al., 2015). When assuming that residents exposed all 24 hours in the roadside air every day as an extreme upper limit (assuming ET = 24 hours day–1

as shown in Table 6), the Risk of NMHCs in the roadside air samples at Guangzhou, Shenzhen and Chengdu was 4.2E-05, 1.9E-05 and 7.0E-05, respectively. To control the Risk below 1.0E-06, the exposure time in the roadside air should be controlled below 0.6, 1.3 and 0.3 hours day–1 at Guangzhou, Shenzhen and Chengdu when other factors remain unchanged, respectively. It is worthy of notice that both HI and Risk values of NMHCs in the roadside air of Chengdu were higher than those in Guangzhou and Shenzhen, because of the higher concentrations of hazardous species in Chengdu as shown in Table 6. However, as shown in Fig. S1, the concentrations of most NMHC species (ethane not included) in roadside air were quite close to that in urban ambient air (sampled on a 60 m high building) in Chengdu, indicating that most of the hazardous NMHCs were dominated by other emission sources instead of vehicles.

HQ, HI and Risk of NMHC species in the exhausts from diesel-, gasoline-, LPG-, LNG- and CNG-powered vehicles were calculated to estimate the health effect of different fuels for vehicles, as shown in Table 7 (assuming ET = 24 hours day–1, EF = 365 days year–1, ED = 70 years (average human life span) and AT = 70 years × 365 days year–1 × 24 hours day–1). Since people were scarcely possible to expose to the extreme circumstances (pure vehicle exhaust) for a whole day (ET = 24 hours day–1) as assumed above,

Tab

le 6

. HQ

, HI

and

Ris

k of

NM

HC

s sp

ecie

s in

the

road

side

air

sam

ples

at G

uang

zhou

, She

nzhe

n an

d C

heng

dua .

Spe

cies

R

fCb

(µg

m–3

) IU

Rb

(m3 µ

g–1)

EC

(µ

g m

–3)

HQ

R

isk

Gua

ngzh

ou

She

nzhe

n C

heng

du

Gua

ngzh

ou

She

nzhe

n C

heng

du

Gua

ngzh

ou

She

nzhe

n C

heng

du

Ben

zene

3.

0E+

01

7.80

E-0

6 5.

4 2.

4 9.

0 1.

8E-0

1 8.

1E-0

2 3.

0E-0

1 4.

2E-0

5 1.

9E-0

5 7.

0E-0

5 T

olue

ne

5.0E

+03

–

26.1

6.

9 15

.5

5.2E

-03

1.4E

-03

3.1E

-03

– –

– E

thyl

benz

ene

1.0E

+03

–

2.6

2.6

3.2

2.6E

-03

2.6E

-03

3.2E

-03

– –

– m

/p-X

ylen

e 1.

0E+

02

– 6.

3 1.

5 9.

1 6.

3E-0

2 1.

5E-0

2 9.

1E-0

2 –

– –

o-X

ylen

e 1.

0E+

02

– 2.

7 0.

8 3.

4 2.

7E-0

2 7.

6E-0

3 3.

4E-0

2 –

– –

Sty

rene

1.

0E+

03

– 0.

5 0.

6 4.

4 4.

8E-0

4 5.

7E-0

4 4.

4E-0

3 –

– –

1,2,

3-T

rim

ethy

lben

zene

2.

0E+

02

– 0.

8 0.

6 5.

4 3.

9E-0

3 3.

1E-0

3 2.

7E-0

2 –

– –

1,2,

4-T

rim

ethy

lben

zene

4.

0E+

03

– 2.

1 0.

9 6.

4 5.

4E-0

4 2.

3E-0

4 1.

6E-0

3 –

– –

1,3,

5-T

rim

ethy

lben

zene

4.

0E+

03

– 0.

4 0.

3 7.

4 9.

7E-0

5 6.

5E-0

5 1.

8E-0

3 –

– –

n-H

exan

e 7.

0E+

02

– 8.

7 –

8.4

1.2E

-02

– 1.

2E-0

2 –

– –

Cyc

lohe

xane

6.

0E+

03

– 0.

7 0.

4 9.

4 1.

1E-0

4 5.

9E-0

5 1.

6E-0

3 –

– –

HI

2.

9E-0

1 1.

1E-0

1 4.

8E-0

1

a Ass

umin

g E

T =

24

hour

s da

y–1, E

F =

365

day

s ye

ar–1

, ED

= 7

0 ye

ars

and

AT

= 7

0 ye

ars

× 3

65 d

ays y

ear–1

× 2

4 ho

urs

day–1

. b

RfC

and

IU

R o

f ea

ch c

hem

ical

use

d in

this

stu

dy c

ame

from

the

US

EPA

Int

egra

ted

Ris

k In

form

atio

n S

yste

m (

IRIS

).


relative magnitude of the values of HQ, HI and Risk were considered here, instead of absolute values. Results show that both HI and Risk of NMHC species in the exhausts from gasoline-powered vehicles were much larger than those from the other four types of vehicles, indicating a larger non-cancer and cancer risk of gasoline vehicle exhaust. The ratios of HI (or Risk) of gasoline vehicular exhausts to another type of vehicular exhausts were calculated to compare the health impact of different fuels on human beings. A ratio > 1 implies a smaller negative health impact of this fuel compared with gasoline, and the negative health impact decreases with the increase of the ratio. The ratio of HIGasoline to HILPG, HILNG and HICNG was 139, 658 and 13; the ratio of RiskGasoline to RiskLPG, RiskLNG and RiskCNG was 114, 1577 and 15, respectively. Generally, the usage of clean alternative fuels for vehicles had a positive health impact on human beings compared with gasoline. Additionally, LPG seems to be not that ideal as an alternative fuel for vehicles, considering that the concentration of total NMHCs in the exhaust from LPG vehicles was higher than that from gasoline and diesel vehicles as shown in Table 2. However, both HI and Risk of NMHC species in the exhaust from LPG vehicles was lower than those from vehicles using traditional fuels, indicating that less hazardous NMHCs were produced by LPG vehicles. Besides, LPG vehicles produced much less particles than diesel vehicles (Chan et al., 2007), which is an important reason for the promotion of LPG vehicles in Guangzhou.

CONCLUSIONS AND INSPIRATIONS

In this study, the emission characteristics, OFP and health risks of NMHCs in conventional and clean alternative fuel exhaust emitted by vehicles in Guangzhou, Shenzhen and Chengdu were assessed to explore the environmental and health effects from the change in NMHCs due to using clean alternative fuels. Our results show that the fuel type significantly impacted the composition of the roadside air. The promotion of LPG-powered buses and taxis in Guangzhou after 2003 significantly increased the amount of propane in the roadside air, whereas the wide usage of CNG vehicles produced a high concentration of ethane in the roadside air of Chengdu. The total OFP value for the major NMHC species in the exhaust from each type of vehicle was also calculated in this study. The OFPs for the total NMHCs were much lower in the exhaust of vehicles powered by CNG (2.5 ppmv), LNG (4.7 ppmv) or diesel (4.5 ppmv) than those powered by gasoline (94.3 ppmv) or LPG (23.1 ppmv). Hence, substituting gasoline with CNG and LNG may effectively reduce O3 formation from vehicle exhaust. Furthermore, the HQ, HI and Risk of NMHC species in the exhaust were calculated to estimate the health effects of different vehicle fuels, revealing that both the HI and Risk of the species in the exhaust from gasoline-powered vehicles greatly exceeded those from the other four types of vehicles. The ratios of HIGasoline to HILPG, HILNG and HICNG were 139, 658 and 13, and the ratios of RiskGasoline to RiskLPG, RiskLNG and RiskCNG were 114, 1577 and 15, respectively, suggesting that substituting gasoline with clean alternative fuels benefits human health.

Tab

le 7

. HQ

, HI

and

Ris

k of

NM

HC

s sp

ecie

s in

the

exha

usts

fro

m d

iese

l-, g

asol

ine-

, LP

G-,

LN

G-

and

CN

G-p

ower

ed v

ehic

lesa .

Spe

cies

R

fCb

(µg

m–3

) IU

Rb

(m3

µg–1

) E

C (

µg

m–3

) H

Q

Ris

k D

iese

l G

asol

ine

LP

G L

NG

C

NG

D

iese

l G

asol

ine

LP

G

LN

G

CN

G

Die

sel

Gas

olin

e L

PG

L

NG

C

NG

B

enze

ne

3.0E

+01

7.8

0E-0

6 22

7.1

5243

.8

46.5

3.

4 35

1.3

7.6E

+00

1.7

E+

02

1.5E

+00

1.1

E-0

1 1.

2E+

01 1

.8E

-03

4.1E

-02

3.6E

-04

2.6E

-05

2.7E

-03

Tol

uene

5.

0E+

03 –

12

2.9

7243

.7

65.6

3.

2 45

6.9

2.5E

-02

1.4E

+00

1.

3E-0

2 6.

5E-0

4 9.

1E-0

2 –

– –

– –

Eth

ylbe

nzen

e 1.

0E+

03 –

56

.6

3712

.8

4.6

1.9

160.

7 5.

7E-0

2 3.

7E+

00

4.6E

-03

1.9E

-03

1.6E

-01

– –

– –

– m

/p-X

ylen

e 1.

0E+

02 –

69

.1

3775

.0

7.3

1.4

426.

5 6.

9E-0

1 3.

8E+

01

7.3E

-02

1.4E

-02

4.3E

+00

–

– –

– –

o-X

ylen

e 1.

0E+

02 –

36

.4

2035

.3

2.9

0.9

145.

9 3.

6E-0

1 2.

0E+

01

2.9E

-02

8.5E

-03

1.5E

+00

–

– –

– –

Sty

rene

1.

0E+

03 –

–

227.

8 –

– 8.

1 –

2.3E

-01

– –

8.1E

-03

– –

– –

– 1,

2,3-

Tri

met

hylb

enze

ne

2.0E

+02

–

43.8

15

40.0

1.

7 3.

6 21

0.3

2.2E

-01

7.7E

+00

8.

6E-0

3 1.

8E-0

2 1.

1E+

00 –

–

– –

– 1,

2,4-

Tri

met

hylb

enze

ne

4.0E

+03

–

88.0

33

13.0

3.

0 3.

1 47

8.6

2.2E

-02

8.3E

-01

7.5E

-04

7.6E

-04

1.2E

-01

– –

– –

– 1,

3,5-

Tri

met

hylb

enze

ne

4.0E

+03

–

23.1

11

29.3

0.

8 0.

8 20

8.5

5.8E

-03

2.8E

-01

2.0E

-04

1.9E

-04

5.2E

-02

– –

– –

– n-

Hex

ane

7.0E

+02

–

14.6

25

36.2

76

.7

154.

0 12

2.1

2.1E

-02

3.6E

+00

1.

1E-0

1 2.

2E-0

1 1.

7E-0

1 –

– –

– –

Cyc

lohe

xane

6.

0E+

03 –

9.

2 28

8.7

10.7

0.

3 73

.3

1.5E

-03

4.8E

-02

1.8E

-03

5.7E

-05

1.2E

-02

– –

– –

– H

I

9.0E

+00

2.5

E+

02

1.8E

+00

3.8

E-0

1 1.

9E+

01

a Ass

umin

g E

T =

24

hour

s da

y–1, E

F =

365

day

s ye

ar–1

, ED

= 7

0 ye

ars

and

AT

= 7

0 ye

ars

× 3

65 d

ays y

ear–1

× 2

4 ho

urs

day–1

. b

RfC

and

IU

R o

f ea

ch c

hem

ical

use

d in

this

stu

dy c

ame

from

the

US

EP

A I

nteg

rate

d R

isk

Info

rmat

ion

Sys

tem

(IR

IS).


This study confirms that using alternative fuel for vehicles critically improves urban air quality. Although LPG, LNG and CNG, which produce less ozone and fewer adverse health effects than gasoline, are all suitable alternative fuels, LNG and CNG seem to be better choices due to generating NMHCs with lower total OFPs in the vehicle exhaust. Moreover, the following issues should be addressed. Firstly, since vehicles with modified engines exhaust higher levels of air pollutants than those with custom-built engines, we recommend promoting alternative fuel vehicles running on custom-built engines. Secondly, the engine must be compatible with and adjusted for the vehicle, and appropriate exhaust after-treatment devices should be deployed as well. Last but not least, alternative fuel vehicles should be repaired and maintained daily to optimize their performance. ACKNOWLEDGMENTS

This work is financially supported by the Fundamental Research Funds for Central Public Welfare Research Institutes (No. PM-ZX703-201803-067); State Key Laboratory of Organic Geochemistry, GIGCAS (No. SKLOG-201718); the Science and Technology Key Projects of Guangdong Province (2017A030223005) and the Science and Technology Program of Guangzhou City (201804010147). SUPPLEMENTARY MATERIAL

Supplementary data associated with this article can be found in the online version at http://www.aaqr.org. REFERENCES Abu-Allaban, M., Gertier, A.W. and Lowenthal, D.H. (2002).

A preliminary apportionment of the sources of ambient PM10, PM2.5, and VOCs in Cairo. Atmos. Environ. 36: 5549–5557.

Agarwal, A.K., Bothra, P. and Shukla, P.C. (2015). Particulate characterization of CNG fuelled public transport vehicles at traffic junctions. Aerosol Air Qual. Res. 15: 2168–2174.

Aslam, M., Masjuki, H., Kalam, M., Abdesselam, H., Mahlia, T. and Amalina, M. (2006). An experimental investigation of CNG as an alternative fuel for a retrofitted gasoline vehicle. Fuel 85: 717–724.

Atkinson, R. and Arey, J. (2003). Atmospheric degradation of volatile organic compounds. Chem. Rev. 103: 4605–4638.

Bamuwamye, M., Ogwok, P., Tumuhairwe, V., Eragu, R., Nakisozi, H. and Ogwang, P. (2017). Human Health Risk Assessment of Heavy Metals in Kampala (Uganda) Drinking Water. J. Food Res. 6: 6–16.

Barletta, B., Meinardi, S., Simpson, I.J., Khwaja, H.A., Blake, D.R. and Rowland, F.S. (2002). Mixing ratios of volatile organic compounds (VOCs) in the atmosphere of Karachi, Pakistan. Atmos. Environ. 36: 3429–3443.

Barletta, B., Meinardi, S., Simpson, I.J., Zou, S., Rowland, F.S. and Blake, D.R. (2008). Ambient mixing ratios of nonmethane hydrocarbons (NMHCs) in two major urban centers of the Pearl River Delta (PRD) region: Guangzhou

and Dongguan. Atmos. Environ. 42: 4393–4408. Carter, W.P.L. (1994). Development of Ozone Reactivity

Scales for Volatile Organic Compounds. J. Air Waste Manage. Assoc. 44: 881–899.

Carter, W.P.L. (2010). Development of the SAPRC-07 chemical mechanism. Atmos. Environ. 44: 5324–5335.

Chan, T., Ning, Z., Wang, J., Cheung C., Leung, C. and Hung, W. (2007). Gaseous and particle emission factors from the selected on-road petrol/gasoline, diesel, and liquefied petroleum gas vehicles. Energy Fuels 21: 2710–2718.

Chang, C., Lo, J. and Wang, J. (2001). Assessment of reducing ozone forming potential for vehicles using liquefied petroleum gas as an alternative fuel. Atmos. Environ. 35: 6201–6211.

Chen, T., Simpson, I.J., Blake, D.R. and Rowland, F.S. (2001). Impact of the leakage of liquefied petroleum gas (LPG) on Santiago air quality. Geophys. Res. Lett. 28: 2193–2196.

Czader, B.H., Byun, D.W., Kim, S.T. and Carter, W.P.L. (2008). A study of VOC reactivity in the Houston-Galveston air mixture utilizing an extended version of SAPRC-99 chemical mechanism. Atmos. Environ. 42: 5733–5742.

Edokpolo, B., Yu, Q. and Connell, D. (2014). Health risk assessment of ambient air concentrations of benzene, toluene and xylene (BTX) in service station environments. Int. J. Environ. Res. Public Health 11: 6354–6374.

Energy Statistics Division of National Bureau of Statistics (2019). China energy statistics yearbook 2018. China Statistics Press, Beijing.

Gao, S., Bai, Z., Wang, X., Han, B. and Wang, Z. (2012). Cancer risk assessment of volatile organic compounds in an ornament market in Tianjin, China. 2012 International Conference on Biomedical Engineering and Biotechnology, IEEE computer society, pp. 1247–1250.

Guangzhou Statistics Bureau (2013). Guangzhou statistical yearbook 2013. China Statistics Press, Beijing.

Guo, H., Ling, Z.H., Cheng, H.R., Simpson I.J., Lyu, X.P., Wang, X.M., Shao, M., Lu, H.X., Ayoko, G., Zhang, Y.L., Saunders, S.M., Lam, S.H.M., Wang, J.L. and Blake, D.R. (2017). Tropospheric volatile organic compounds in China. Sci. Total Environ. 574: 1021–1043.

Guo, H., Zou, S., Tsai, W., Chan, L. and Blake, D.R. (2011). Emission characteristics of nonmethane hydrocarbons from private cars and taxis at different driving speeds in Hong Kong. Atmos. Environ. 45: 2711–2721.

Ho, K., Ho, S., Lee, S., Louie, P., Cao, J. and Deng, W. (2013). Volatile organic compounds in roadside environment of Hong Kong. Aerosol Air Qual. Res. 13: 1331–1347.

Huang, Y., Ling Z., Lee, S., Ho, S., Cao, J., Blake, D., Cheng, Y., Lai, S., Ho, K., Gao, Y. and Louie, P. (2015). Characterization of volatile organic compounds at a roadside environment in Hong Kong: An investigation of influences after air pollution control strategies. Atmos. Environ. 122: 809–818.

Kado, N.Y., Okamoto, R.A., Kuzmicky, P.A., Kobayashi, R., Ayala, A., Gebel, M.E., Rieger, P.L., Maddox, C. and Zafonte, L. (2005). Emissions of toxic pollutants from


compressed natural gas and low sulfur diesel-fueled heavy-duty transit buses tested over multiple driving cycles. Environ. Sci. Technol. 39: 7638–7649.

Kelly, F.J. and Zhu, T. (2016). Transport solutions for cleaner air. Science 352: 934–936.

Lai, C., Chang, C., Wang, C., Shao, M., Zhang, Y. and Wang, J. (2009). Emissions of liquefied petroleum gas (LPG) from motor vehicles. Atmos. Environ. 43: 1456–1463.

Li, B., Ho, S.S.H., Xue. Y., Huang, Y., Wang, L., Cheng, Y., Dai, W., Zhong, H., Cao, J. and Lee, S. (2017). Characterizations of volatile organic compounds (VOCs) from vehicular emissions at roadside environment: The first comprehensive study in Northwestern China. Atmos. Environ. 161: 1–12.

Li, L., Li, H., Wang, X. Zhang, X. and Wen, C. (2013). Pollution characteristics and health risk assessment of atmospheric VOCs in the Downtown area of Guangzhou, China. Environ. Sci. 34: 4558–4564. (in Chinese)

Li, L., Chen, Y., Zeng, L., Shao, M., Xie, S., Chen, W., Lu, S., Wu, Y. and Cao, W. (2014). Biomass burning contribution to ambient volatile organic compounds (VOCs) in the Chengdu–Chongqing Region (CCR), China. Atmos. Environ. 99: 403–410.

Liao, Z. (2013). Development and prospect of Sichuan natural gas vehicle market. Technol. Market 20: 327–328. (in Chinese)

Liu, Y., Shao, M., Fu, L., Lu, S., Zeng, L. and Tang, D. (2008). Source profiles of volatile organic compounds (VOCs) measured in China: Part I. Atmos. Environ. 42: 6247–6260.

Lounici, M., Loubar, K., Tarabet, L., Balistrou, M., Niculescu, D.C. and Tazerout, M. (2014). Towards improvement of natural gas-diesel dual fuel mode: An experimental investigation on performance and exhaust emissions. Energy 64: 200–211.

Luis, J.L.J., Julio, S.F., Uriel, G.M. and Emmanuel, G.O. (2003). Liquefied petroleum gas effect on ozone formation in Mexico city. Atmos. Environ. 37: 2327–2335.

Moolla, R., Curtis, C. J. and Knight, J. (2015). Assessment of occupational exposure to BTEX compounds at a bus diesel-refueling bay: A case study in Johannesburg, South Africa. Sci. Total Environ. 537: 51–57.

Osorio-Tejada, J.L., Llera-Sastresa, E. and Scarpellini, S. (2017). Liquefied natural gas: Could it be a reliable option for road freight transport in the EU? Renew. Sust. Energ. Rev. 71: 785–795.

Ou, J., Guo, H., Zheng, J., Cheung, K., Louie, P.K.K., Ling, Z. and Wang, D. (2015). Concentrations and sources of non-methane hydrocarbons (NMHCs) from 2005 to 2013 in Hong Kong: A multi-year real-time analysis. Atmos. Environ. 103: 196–206.

Qu, C., Li, B., Wu, H., Wang, S. and Giesy, J. (2015). Multi-pathway assessment of human health risk posed by polycyclic aromatic hydrocarbons. Environ. Geochem. Health 37: 587–601.

Ristovski, Z.D., Jayaratne, E.R., Morawska, L., Ayoko, G.A. and Lim, M. (2005). Particle and carbon dioxide emissions from passenger vehicles operating on unleaded

petrol and LPG fuel. Sci. Total Environ. 345: 93–98. Situ, S., Guenther, A., Wang, X., Jiang, X., Turnipseed, A.,

Wu, Z., Bai, J. and Wang, X. (2013). Impacts of seasonal and regional variability in biogenic VOC emissions on surface ozone in the Pearl River Delta region, China. Atmos. Chem. Phys. 13: 11803–11817.

Suthawaree, J., Sikder, H., Jones, C., Kato, S., Kunimi, H., Kabir, A. and Kajiii, Y. (2012). Influence of extensive compressed natural gas (CNG) usage on air quality. Atmos. Environ. 54: 296–307.

Tang, J. (2007). Characterization and speciation of non-methane hydrocarbons (NMHCs) and their implications on the atmosphere of South China. Ph.D. thesis, The Hong Kong Polytechnic University, Hong Kong.

Tang, J., Chan, L., Chan, C., Li, Y., Chang, C., Liu, S., Wu, D. and Li, Y. (2007). Characteristics and diurnal variations of NMHCs at urban, suburban, and rural sites in the Pearl River Delta and a remote site in South China. Atmos. Environ. 41: 8620–8632.

Tang, J., Chan, L., Chan, C., Li, Y., Chang, C., Wang, X., Zou, S., Barletta, B., Blake, D. and Wu, D. (2008). Implications of changing urban and rural emissions on non-methane hydrocarbons in the Pearl River Delta region of China. Atmos. Environ. 42: 3780–3794.

Tong, L., Zhang, H., Yu, J., He, M., Xu, N., Zhang, J., Qian, F., Feng, J. and Xiao, H. (2017). Characteristics of surface ozone and nitrogen oxides at urban, suburban and rural sites in Ningbo, China. Atmos. Res. 187: 57–68.

Tsai, J., Yao, Y., Huang, P. and Chiang, H. (2018). Fuel economy and volatile organic compound exhaust emission for motorcycles with various running mileages. Aerosol Air Qual. Res. 18: 3056–3067.

Tsai, W., Chan, L., Blake, D.R. and Chu, K. (2006). Vehicular fuel composition and atmospheric emissions in South China: Hong Kong, Macau, Guangzhou, and Zhuhai. Atmos. Chem. Phys. 6: 3281–3288.

Tsai, W. (2007). Non-methane hydrocarbon characteristics of motor vehicular emissions in the Pearl River Delta region. The Hong Kong Polytechnic University, Hong Kong.

U.S. EPA (2009). Risk Assessment Guidance for Superfund, Volume I: Human Health Evaluation Manual (Part F, Supplemental Guidance for Inhalation Risk Assessment). United States Environmental Protection Agency , USA, https://www.epa.gov/sites/production/files/2015-09/doc uments/partf_200901_final.pdf.

Wang, B. (2005). Achievements and prospects of clean vehicle action in China. Automot. Eng. 27: 643–647. (in Chinese)

Wang, M., Shao, M., Chen, W., Lu, S., Liu, Y., Yuan, B. and Li, Y. (2015). Trends of non-methane hydrocarbons (NMHC) emissions in Beijing during 2002–2013. Atmos. Chem. Phys. 15: 1489–1502.

Wang, T., Xue, L., Brimblecombe, P., Yun, F., Li, L. and Zhang, L. (2016). Ozone pollution in China: A review of concentrations, meteorological influences, chemical precursors, and effects. Sci. Total Environ. 575: 1582–1596.

Wayne, W., Sandoval, J.A. and Clark, N.N. (2009).


Emissions benefits from alternative fuels and advanced technology in the U.S. transit bus fleet. Energy Environ. 20: 497–515.

Widiana, D.R., Wang, Y., You, S., Yang, H., Wang, L., Tsai, J. and Chen, H. (2019). Air pollution profiles and health risk assessment of ambient volatile organic compounds above a municipal wastewater treatment plant, Taiwan. Aerosol Air Qual. Res. 19: 375–382.

Yuan, Z., Zhong, L., Lau, A., Yu, J. and Louie, P. (2013). Volatile organic compounds in the Pearl River Delta: Identification of source regions and recommendations for emission-oriented monitoring strategies. Atmos. Environ. 76: 162–172.

Zhang, F. (2008). Statistical analysis and study on the failure of LPG buses in Guangzhou. South China Agricultural University, Guangzhou.

Zhang, J., Wang, Y., Wu, F., Lin, H. and Wang, W. (2009). Nonmethane hydrocarbon measurements at a suburban site in Changsha City. China. Sci. Total Environ. 408: 312–317.

Zhang, Y., Wang, X., Blake, D.R., Li, L., Zhang, Z., Wang, S., Guo, H., Lee, F.S.C., Gao, B., Chan, L., Wu, D. and Rowland, F.S. (2012). Aromatic hydrocarbons as ozone precursors before and after outbreak of the 2008 financial crisis in the Pearl River Delta region, south China. J. Geophys. Res. 117: D15306.

Zhang, Y., Wang, X., Zhang, Z., Lu, S., Huang, Z. and Li, L. (2015). Sources of C2-C4 alkenes, the most important ozone nonmethane hydrocarbon precursors in the Pearl River Delta region. Sci. Total Environ. 502: 236–245.

Zhang, Y., Yang, W., Simpson, I., Huang, X., Yu, J., Huang, Z., Zhang, Z., Liu, D., Huang, Z., Wang, Y., Pei, C., Shao, M., Blake, D.R., Zheng, J., Huang, Z. and Wang, X. (2018). Decadal changes in emissions of volatile organic compounds (VOCs) from on-road vehicles with intensified automobile pollution control: Case study in a busy urban tunnel in south China. Environ. Pollut. 233: 806–819.

Zheng, J., Shao, M., Che, W., Zhang, L., Zhong, L., Zhang, Y. and Streets, D. (2009). Speciated VOC emission inventory and spatial patterns of ozone formation potential in the Pearl River Delta, China. Environ. Sci. Technol. 43: 8580–8586.

Zielinska, B., Campbell, D. and Samburova, V. (2014). Impact of emissions from natural gas production facilities on ambient air quality in the Barnett Shale area: A pilot study. J. Air Waste Manage. Assoc. 64: 1369–1383.

Received for review, September 27, 2019 Revised, January 13, 2020

Accepted, January 16, 2020

environmental and health impacts of the change in nmhcs ... · issn: 1680-8584 print / 2071-1409...

Documents