Bull Environ Contam Toxicol (2015) 94:653–660 DOI 10.1007/s00128-015-1506-8
Spatial Distribution and Temporal Trends of VOCs in a Highly Industrialized Town in Turkey ¨ ztu¨rk • Pınar Ergenekon • Gaye O ¨ zdemir Sec¸kin Naciye O Su¨meyra Bayır
•
Received: 11 June 2014 / Accepted: 25 February 2015 / Published online: 5 March 2015 Ó Springer Science+Business Media New York 2015
Abstract An extensive monitoring study of volatile organic compounds (VOCs) was conducted at sites across the highly industrialized town of Dilovasi, northern Turkey to determine temporal and spatial trends in pollutant concentrations and relate to the effects of source locations, meteorology, and topography. Two-week passive samplers (Tenax tubes) were deployed at twelve sites from February to December 2012 and analysed using thermal desorption and gas chromatography with mass spectrometric detection (TD–GC-MS). Sampled total VOC (TVOC) levels were highest in the July through October period and were highest at low-altitude sites near industry facilities and vehicle traffic sources (148.3 lg/m3 at site 11, 154.1 lg/m3 at site 10) and lowest at high-altitude sites located furthest upwind from industry and traffic sources (78.4 lg/m3 at site 5 and 78.5 lg/m3 at site 6). Analysis of ‘‘T/B’’ ratios suggested that contributions to ambient VOC in Dilovasi are dominated by the town’s industrial sources. Meteorological conditions and the town’s basin topography were also found to significantly influence the city’s air quality, with strong winds from the NE observed to correlate with periods of higher sampled TVOC. Compared with other industrialized urban centers, the study revealed that there is significant toluene pollution in Dilovasi and recommended enhanced continuous monitoring at the city’s industrial and residential zones.
¨ ztu¨rk (&) S. Bayır N. O Department of Chemistry, Gebze Technical University, Kocaeli, Turkey e-mail: [email protected]; [email protected] ¨ . Sec¸kin P. Ergenekon G. O Department of Environmental Engineering, Gebze Technical University, Kocaeli, Turkey
Keywords Monitoring Air pollution Volatile organic compounds Toluene Basin topography Intraurban sampling Industry sources
Volatile organic compounds (VOCs) constitute an important group of air pollutants. Some of them can induce cancer, many of them play an active role in the formation of tropospheric ozone, cause photochemical smog formation and contribute to global warming (Atkinson 2005; Filella and Pen˜uelas 2006). The most abundant VOC species are benzene, toluene, and ethyl benzene, xylene (BTEX) (Parra et al. 2009).VOCs are released into atmosphere from both anthropogenic and natural sources. Anthropogenic sources are mainly fossil fuel use and industrial activities. Various household and personal care products also contribute significantly to VOC releases (Pearson 1982). Sampling onto solid adsorbents by active or passive sampling and then subsequent thermal desorption of trapped compounds into a capillary gas chromatography column is a widely used method for VOC monitoring. Combined thermal desorption gas chromatography mass spectrometry /flame ionization detection (TD–GC-MS/ FID) is a very sensitive and cost effective analytical system. Passive sampling of ambient VOCs is a preferred option for monitoring ambient or indoor air quality since samplers are easy for technicians to deploy in the field and samplers can be analysed for a range of compounds at low cost. Field monitoring studies have successfully used longterm diffusive sampling of one to 4 week periods to accurately determine time-weighted average (TWA) VOC concentrations (Thammakhet et al 2006; Martin et al 2010; Monn and Hangartner 1996).
123
654
It is well known that VOC concentrations are significantly higher in areas with industrial facilities and in large cities with high traffic densities. Dilovası, Turkey is a heavily industrialized town that is associated with high cancer incidence rates (i.e., three times the national average; Ministry of Health 2004). A recent study revealed that the risk of death due to cancer was 4.4 times higher for individuals residing in Dilovasi for more than 10 years compared to individuals with residence less than 10 years (Hamzaoglu et al. 2011). The city also has a basin topography that increases the severity of the exposure to air pollution by reducing dispersion, amplified during thermal inversion events. The town is also highly industrialized, with many facilities associated with significant VOC releases (e.g., metal, solvent storage, paint manufacturing) concentrated in the lowest part of the valley and highly populated residential areas distributed between the industrialized zones. For all these reasons, determination of the intraurban distribution of ambient VOC across the town via a comprehensive field sampling campaign was of vital importance. There have been few sampling studies for VOC in the region. Passive sampling and TD/GC–FID analysis were used to measure ambient concentrations of VOCs in Kocaeli and found that concentrations of benzene, toluene, ethylene, and xylene (i.e., BTEX compounds) were high along major roads, city centers and near industrial plants (Pekey and Yılmaz 2011). Statistical factor analysis revealed the vehicle exhaust and industrial activity as the potential pollution sources affecting the VOC concentrations (Pekey and Yılmaz 2011; Pekey and Arslanbas¸ 2008). However, preceding studies have used limited spatial sampling networks and sampled a limited time series and so have provided little information about air quality in Dilovasi’s residential zones relative to adjacent industrial zones. In this study, detailed VOC speciation and quantification was performed to better understand the overall profile and distribution of VOCs in Dilovası ambient air. Two-week integrated air samples were collected over a 10-month period from 12 sampling sites and analysed for speciated VOCs. Samples were collected on Tenax TA adsorbent by passive sampling and analyzed using TD/GC-MS. The study’s major objective was to identify the VOC profiles and levels across the study region, with emphasis on highly populated residential areas as well as to reveal any temporal or seasonal variations. The correlation between the most abundant species, meterological factors and VOC’s temporal and special variations are utilized to describe the general situation of exposure to VOC pollution in the town.
Materials and Methods Dilovasi is an industrial coastal town situated in northern Turkey by the Marmara sea. The climate in the region is
123
Bull Environ Contam Toxicol (2015) 94:653–660
temperate and highest temperatures are generally observed in July and August (Ergenekon et al 2009). The city covers a total area of 2000 ha and has a population of 50,000 residents. There are more than 150 industrial facilities operating in the city, most of which can be classified as chemical or metal industry. In the monitoring area of Dilovası district, 12 sampling points were selected at different distances to industrial plants and major highways (Fig. 1). Meteorological data (temperature, wind speed, direction etc.) were obtained from the air monitoring station of Dilovasi within the monitoring network of Environment and Urbanization Ministry at a distance of 150 m south of the site 10. Two-week integrated air samples were collected using passive stainless steel sampling tubes (Perkin Elmer) packed with Tenax TA media and capped with a diffusion end cap at the inlet and a Swagelok end cap at the other. Sampling tubes were conditioned prior to sampling by passing a stream of ultra-pure dry air at 90 mL/min at 300°C for 20 min. Sampling tubes were stored in the refrigerator at ?4°C prior to sampling and transferred to sampling point in protecting cooled storage container. The sampling period spanned 10 months (February to December 2012). Sampling tubes were collected in pairs from each site, capped with Swagelok end caps, stored in sealed glass jars and transported in cooled containers back to the laboratory for analysis. Measurements were made using a thermal desorption unit (Perkin Elmer Turbo Matrix) coupled to a GC (Perkin Elmer Clarus 680 GC) and a MS detector (Perkin Elmer Clarus 600 T MS). The VOCs were thermally desorbed for 10 min at 300°C with a flow of 30 mL/min of ultra-pure helium passing through and carrying the desorbed VOCs to a preconcentration trap (-10°C). After the focusing step, the analytes were desorbed from the trap by rapid heating to 250°C into the injector of the GC. The separation of VOCs was performed on a 30 m 9 0.32 mm 9 1.28 lm capillary column (DB-624) with ultra-pure helium as carrier gas. After sample injection, the column oven was maintained at 35°C for 2 min and then the temperature was first increased to 220°C at a rate of 5 °C/min (totaly 41 min oven program). The MS scanned from 35 to 400 a.m.u. in the electron impact mode. Four-point calibration (0.1, 1, 5, 10 ppm) was performed using liquid standards in methanol solutions. These liquid standards (AccuStandart M-502, 60 compounds) were injected into adsorbent tubes using a calibration rig (Markes) by passing a pure nitrogen stream for 2 min. through the adsorbent tubes ensuring that compounds to be adsorbed while methanol evaporated. Calibration standard tubes were desorbed and analyzed by using the same method parameters used in the analysis of sample tubes.
Bull Environ Contam Toxicol (2015) 94:653–660
655
Fig. 1 The satellite image of study region (Dilovasi, Turkey) showing locations of sampling sites (Google, DigitalGlobe 2013)
Method detection limits (MDL) were calculated as three standard deviation of the blank measurement and ranged between 0.02 ng (n-dodecane) and 7.20 ng (m,p-xylenes); MDL are presented for all analysed species in Table 1. The data quality was controlled by checking relative standard deviation (RSD) of the replicates; replicate samples with RSD values higher than 20 % of the mean measured value were eliminated from the dataset. Blank data for conditioned sampling tubes were checked for any residues before use. In addition, toluened8 was injected to all tubes prior to GC/MS analysis for ensuring the accuracy of the analytical method. Field blanks were deployed at the sites at 2 month intervals and analysed for possible artifact formation. Field data and lab-analysed sample concentrations data was entered into Microsoft Excel, 2010 and SPSS 19.0 for calculation of time-weighted average concentrations, statistical correlation and temporal trend analysis. Wind rose analysis was completed using WRPLOT View software from Environmental Lakes Inc. Spatial distribution of sampled concentrations were mapped across the study area using ArcView GIS 10.1 (ESRI 2012) and a pollutant surface was constructed by interpolation (i.e., Spline, regulated type, weight parameter 0.1 and number of points parameter of 12).
Results and Discussion A total of 35 VOCs were identified in the sample data, classified as aromatic (Toluene, Ethylbenzene, m-p xylene, o-Xylene, Benzene, n-Propylbenzene, Isopropylbenzene,
1.3.5-Trimethylbenzene, tert-Butylbenzene, sec-Butylbenzene, Limonene, p-Isopropyltoluene, Styrene, Phenol, Acetophenone, Naphthalene,) halogenated (Dichloromethane, 1.2-Dichloroethene, Chloroform, 1,2-dichloroethane, Trichloroethene, Tetrachloroethene, Dibromochloromethane, 2-Chlorotoluene,1,2-Dichlorobenzene, trichlorofluoro-Methane, 1-chloro-2-methylbutane) and the others (Octane, Nonane, n-dodecane, n-tridecane, Pentane, 3-methylpentane, Methylcyclopentane, Ethanol). At most sites, all 35 VOC species were present at concentrations greater than their respective method detection limits. Summary statistics across all sites and samples are presented for all 35 analysed species in Table 1. TVOC values at 12 sampling points during the sampling period were shown in Fig. 2. Date in the x axis represents the starting day of the 2-week passive sampling period. Total volatile organic compound (TVOC) values vary from site to site, generally sites 9, 10, and 11 were the most polluted. It is clear that there is an increase in TVOC amounts at most of the sampling points during the sampling periods from 4th of July and August 15th to 7th of November. This might be caused by the emission variation in the sources or due to meteorological conditions prevalent in the region. The ‘‘high TVOC’’ period also generally aligns with the time of year where highest temperatures are seen (Rubin et al. 2006; Cetin et al. 2003). The hourly ambient temperature values for the sampling period ranged from -1 to 34°C. Temperatures over 20°C were observed between samplings of 23rd of May to 26th of September (warm period). However TVOC concentrations were not highly correlated with the temperature. The correlation coefficient (r) between TVOC and temperature is
123
656
Bull Environ Contam Toxicol (2015) 94:653–660
Table 1 Summary statistics of analysed VOC species Compounds
Species MDL as ng
Trichlorofluoro-Methane
0.59
% of site samples [MDL 63
Mean (SD) as lg/m3 1.00 (0.91)
Median (first quartile - third quartile) as lg/m3 1.00 (0.94–1.05)
Pentane
1.09
90
4.01 (2.86)
3.31 (2.83–5.20)
Ethanol
6.63
75
3.56 (4.26)
3.24 (2.61–4.59)
Dichloromethane
3.98
91
14.61 (31.37)
15.57 (12.07–16.48)
1-chloro-2-methylbutane
2.14
86
3.88 (3.34)
3.71 (3.48–3.90)
3-methylpentane
2.24
90
3.11 (2.83)
2.83 (2.76-3.27)
Methylcyclopentane 1.2-Dichloroethene
2.15 1.44
80 13
5.45 (9.53) 0.40 (0.28)
5.30 (4.61–6.48) 0.42 (0.33–0.48)
Chloroform
1.02
59
0.70 (0.84)
0.59 (0.53–0.86)
Benzene
1.2
99
2.83 (1.81)
2.92 (2.26–3.48)
1.2-dichloroethane
0.69
1
0.24 (0.08)
0.24 (0.24–0.24)
Trichloroethene
0.76
17
0.32 (0.18)
0.29 (0.25–0.39)
Toluene
6.38
100
41.91 (52.72)
42.96 (32.20–51.77)
Octane
0.13
100
0.68 (0.42)
0.60 (0.52–0.74)
Tetrachloroethene
0.65
53
0.67 (0.44)
0.65 (0.45–0.76)
Dibromochloromethane
0.74
7
0.79 (0.64)
0.52 (0.48–0.88)
Ethylbenzene
0.69
91
1.38 (0.80)
1.20 (1.04–1.92)
m-p xylene
7.2
91
4.83 (2.91)
4.16 (3.82–6.25)
o-Xylene
1.66
90
1.07 (0.70)
0.94 (0.76–1.50)
Nonane
0.34
90
3.62 (5.88)
3.76 (2.81–4.37)
Styrene
1.7
94
2.90 (2.07)
2.33 (1.65–3.89)
Isopropylbenzene n-Propylbenzene
0.41 0.4
10 46
0.13 (0.08) 0.24 (0.15)
0.14 (0.08–0.16) 0.23 (0.20–0.30)
2-Chlorotoluene
0.31
54
0.19 (0.15)
0.18 (0.16–0.21)
1.3.5-Trimethylbenzene
0.61
66
0.46 (0.31)
0.37 (0.32–0.67)
tert-Butylbenzene
0.96
62
4.55 (4.07)
3.88 (3.11–5.29)
sec-Butylbenzene
0.39
3
0.23 (0.23)
0.23 (0.23–0.23)
Limonene
0.67
97
1.92 (1.63)
1.81 (1.23–2.60)
p-Isopropyltoluene
0.46
23
0.32 (0.20)
0.19 (0.16–0.36)
1.2-Dichlorobenzene
0.41
78
0.28 (0.09)
0.27 (0.20–0.35)
Phenol
5.1
82
1.98 (1.20)
2.06 (1.67–2.21)
Acetophenone
0.81
87
0.99 (0.55)
1.01 (0.92–1.07)
n-dodecane
0.02
99
0.53 (0.31)
0.52 (0.46–0.60)
Naphthalene
0.62
64
0.66 (0.51)
0.58 (0.48–0.90)
n-tridecane
0.5
95
6.97 (5.22)
6.05 (2.30–8.65)
calculated as 0.36 but individual species displayed different degree of correlation. While r for toluene is 0.48, r for benzene is -0.89 suggesting warmer conditions cause the back diffusion of benzene from the adsorbent. In a study (Brandshaw and Ballantine 1995) back diffusion loss for benzene during a 14 day-sampling was given as 68 % while it is only 43 % for toluene. In addition since Benzene is not strongly adsorbed on Tenax TA, obviously back diffusion rates is going to increase with increasing temperature. Although tenax is not very strong adsorbent, once conditioned its low background concentration and high
123
temperature stability make it a preffered sorbent for monitoring a large group of VOC in ambient air (Harper 2000). 14 days of passive sampling for VOCs by using Tenax TA as in this study might be useful for preevaluation of the VOC pollution in a region if the limitations are carefully taken into consideration. The wind rose for the entire sampling period is presented in Fig. 3a; winds were predominantly from the NW and N and almost half of the winds were calm (wind speed \2.1 m/s). However, during periods with significantly higher sampled TVOC concentrations (i.e., 4 July and 15 August through 24
Bull Environ Contam Toxicol (2015) 94:653–660
657
Fig. 2 Temporal variation in TVOC (lg/m3) at all sampling sites
October sampling period), prevailing wind directions were predominantly from the NE and wind speeds were higher, as presented in Fig. 3b. This observation suggests that in addition to source emission variations, when the strong winds blow from NE direction residential areas downwind of the industrial zones, such as the southern coast of the Marmara Sea, may also be exposed to high VOC levels and monitoring at these residential areas is recommended. Averaged concentrations of 13 most abundant compounds at all sites can be seen in Fig. 4 . Those compounds are namely benzene, styrene, 3-methylpentane, nonane, 1-cloro-2-methylbutane, pentane,ethanol,
m-p-xylenes, tert-butylbenzene, methylcyclopentane, ntridecane, dichloromethane and toluene. Toluene is the most abundant compound during entire monitoring period at all sampling. Toluene and total of selected 13 VOC concentrations are lowest at the sites with the highest altitutudes (Sites 5, 6 and 7). In addition to their high altitude, these sites are also farther from Dilovasi sources and located upwind of major source regions (i.e., based on prevailing winds, see Fig. 3a). Site 1 also shows a similar clean profile due to its location where no major sources exist in the vicinity and surrounding topography allows greater dispersion of pollutants (see Fig. 1). The
Fig. 3 Wind rose for the sampling period: a 10-month sampling period; b period corresponding to high sampled TVOC (4 July and 15 August through 24 October)
123
658
Bull Environ Contam Toxicol (2015) 94:653–660
Fig. 4 Concentrations (lg/m3) of selected compounds at sampling sites; carcinogenics is sum of Group1 and Group 2B compounds (IARC)
basin topography of the town reduces dispersion and increases build up of ambient pollutant levels and consequently the most polluted sites (sites 10, 11) were those located near major emission sources and also found at the bottom of the valley (lowest altitudes). Sites 3, 9, and 11 seem to be the most polluted site in terms of toluene. The figure also depicts the sum of the concentrations of carcinogenic species in Group1 (carcinogen to human), Group 2A (probably carcinogenic to human), and Group 2B (possibly carcinogenic to human) classified by International Agency for Research on Cancer (IARC). In terms of carcinogenic species measured in Dilovasi only benzene is in Group 1, dichloromethane is in Group 2A, and other four compounds of chloroform, ethylbenzene, styrene, and naphthalene are from Group 2B. Total concentration of these six compounds was highest at the Site 11. Measuring the other Group 1 species such as 1,3-butadiene, dioxins, and benz(a)-pyrene could give complete spatial distribution of the carcinogenic species in the area. Locations of major chemical, dye and metal industry facilities (known to have VOC emissions due to their process involving solvent use) and average sampled TVOC concentrations were mapped across the study area in Fig. 5; the interpolated pollutant surface shows the spatial distribution of ambient TVOC levels. Highest levels of TVOC are concentrated in the southern coast of the town surrounding sites 9, 10 and 11 while areas N and E of this region have much lower ambient TVOC concentrations. The locations of the metal, dye and chemical industries with possible VOC emission were shown in the figure. Overall, the zones with most dense industry facility siting coincide with the most polluted zones of Dilovasi. This also confirms the pollution was carried to the southern direction under the prevailing winds from N, NE, and NW. Correlation coefficients between VOC species were also calculated and showed: high correlation between 1,2-dichloroethene and toluene (0.73), dicholoromethane
123
and nonane (0.81), and m,p-xylene and ethyl-benzene (0.91); very low negative correlation between toluene and benzene (-0.13). The ambient levels of toluene and several additional sampled compounds (i.e., inclusive of carcinogenic species) sampled in Dilovasi is compared with sampling results from other urban and industrial urban centers in Table 2. Data given for this work in Table 2 are mean values averaged over 10-month sampling period at the stated sites. This showed that Dilovasi is generally highly polluted in toluene relative to other cities, particularly at the industrial zone sampling sites. Toluene to benzene ratios (T/B) of 1.5–2 (i.e., depending on proximity to emission source) have been reported for sites influenced primarily by traffic sources (Gelencser et al. 1997) while sites influenced primarily by industrial emissions have been reported T/B ratios greater than 5 (Barletta et al 2008). In this study, due to the fact that very high negative correlation of benzene with temperature caused by the back diffusion, T/B ratios were calculated for only cold period (i.e. before May 23 and after September 26 as a total of 11 samplings) T/B ratios ranged between 2.86 and 13.1 suggesting that contributions to ambient toluene in Dilovasi are dominated by industrial sources rather than traffic. Sum of carcinogenics which is the total of IARC Group 1, 2A and 2B carcinogenics (6 compounds) is very high at the site 11. The VOC monitoring study conducted at 12 sites across the Dilovasi region revealed meaningful spatial and temporal trends in the region’s ambient air quality. Sites located furthest from industry and traffic sources (site 5 and 6) had the lowest sampled TVOC levels. Conversely, sites located near both industry facilities and vehicle traffic (sites 10 and 11) had very high sampled TVOC levels. Analysis of T/B ratios suggested that contributions to ambient VOC in Dilovasi are dominated by the city’s industrial sources. Meteorological conditions and the city’s
Bull Environ Contam Toxicol (2015) 94:653–660
659
Fig. 5 Spatial distribution of average-sampled TVOC concentrations (lg/m3) in Dilovası
Table 2 Average sampled VOC levels (lg/m3) in Dilovasi ambient air compared with other urban and industrial centers (no data denoted by ’nd’) Compounds
This work
USAa
Chinab
Turkeyc
Japand
Nigeriae
Chinaf
Site 5
Site 11
Michigan
Industrial
Ankara
Yokohama city
Apapa
Guangzhou
Dongguan
Industrial background
Industrial
Urban
Urban
Urban
Industrial
Urban
Industrial
Commercial
Industrial
Toluene (T)
32.76
56.00
3.30
15.70
7.89
12.43
7.91
19.70
22.10
23.07
DCM*
10.57
14.60
nd
nd
nd
nd
nd
14.12
nd
nd
Chloroform*
0.54
0.60
nd
1.20
nd
nd
nd
18.38
nd
nd
n-tridecane
1.24
12.70
nd
nd
nd
1.21
0.75
nd
nd
nd
m-p xylene
2.52
6.80
3.00
2.80
2.21
4.91
2.43
53.22
6.67
6.38
Benzene (B)*
2.49
3.90
1.20
6.70
2.18
3.39
2.11
22.26
6.54
4.02
Ethylbenzene*
0.68
2.10
0.70
2.00
0.85
6.34
2.74
14.58
5.38
4.60
Naphthalene*
0.50
1.52
0.30
nd
nd
1.04
1.04
14.88
nd
nd
Styrene*
0.97
3.32
nd
nd
0.41
nd
nd
nd
nd
nd
15.75
26.04
2.2
9.9
3.44
10.77
5.89
84.22
11.92
8.62
Sum of Cargc. a
Jiaa et al. (2012),
b
Pan et al.(2011),
c
Yurdakul et al. (2013),
d
e
f
Tiwari et al. (2010), Chekwube et al.(2012), Barletta et al. (2008)
* Carcinogenic by IARC
basin topography also significantly influence the city’s air quality, with strong winds from the NE observed to correlate with periods of higher sampled TVOC. Therefore
emissions from the intensely located metal, chemical, dye production and solvent storage facilities in the middle of the town, will possibly be carried to the southern end of the
123
660
town and areas across the Marmara sea. From these findings, continuous monitoring for selected VOCs at Dilovasi and southern vicinity can be recommended. The study revealed that there is significant toluene pollution in Dilovası, compared with other industrialized urban centers. Passive sampling can be used as a cost-effective method of monitoring ambient air quality across the Dilovası urban region, as was demonstrated in this study. Finally, studies of chronic health risk associated with exposure to ambient VOC can also be recommended to further reveal relationships between ambient air quality and public health. ¨ BI˙TAK, The Acknowledgments This work is supported by TU Scientific and Technological Research Council of Turkey under Research project 110Y138. We also thank Meltem Celen, Ph.D, for her help with ArcView use.
References Atkinson R (2005) Atmospheric chemistry of VOCs and NOx. Atmos Environ 34:2063–2101 Barletta B, Meinardi S, Simpson IJ, Zoub S, Rowland FS, Blake DR (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 Bradshaw NM, Ballantine JA (1995) Confirming the limitations of diffusive sampling using Tenax TA during long term monitoring of the environment. Environ Technol 16:433–444 Cetin E, Odabasi M, Seyfioglu R (2003) Ambient volatile organic compound (VOC) concentrations arounda petrochemical complex and a petroleum refinery. Sci Total Environ 312:103–112 Chekwube OC, Majebi OJ, Sanni LM (2012) Baseline levels of volatile organic compounds (VOCs) Pollution in Apapa Industrial Areas of Lagos State, Southwestern-Nigeria. Arch Appl Sci Res 4(6):2564–2571 ¨ ztu¨rk N, Tavs¸ an S (2009) Environmental air levels of Ergenekon P, O VOCs by thermal desorption-gas chromatography in an Industrial Town. Fresenius Environ Bull 18:1999–2003 Filella I, Pen˜uelas J (2006) Daily, weekly, and seasonal time courses of VOC concentrations in a semi-urban area near Barcelona. Atmos Environ 40:7752–7769 Gelencser A, Siszle K, Hlavay J (1997) Toluene-Benzene concentration ratio as a tool for characterizing the distance from vehicular emission sources. Environ Sci Technol 31:2869–2872
123
Bull Environ Contam Toxicol (2015) 94:653–660 Hamzaog˘lu O, Etiler N, Yavuz CI, Caglayan C (2011) The causes of deaths in an industry-densed area: example of Dilovasi (Kocaeli). Turk J Med Sci 41:369–375 Harper M (2000) Sorbent trapping of volatile organic compounds from air. J Chromatogr A 885:129–151 Jiaa C, Battermanb SA, Relyeaa GE (2012) Variability of indoor and outdoor VOC measurements: an analysis using variance components. Environ Pollut 169:152–159 Martin NA, Leming EJ, Henderson MH, Lipscombe RP, Black JK, Jarvis SD (2010) Verification of diffusive and pumped samplers for volatile organic compounds using a controlled atmosphere test facility. Atmos Environ 44:3378–3385 Monn C, Hangartner M (1996) Passive sampling of aromatic volatile organic compounds (VOC) in ambient air Switzerland. Environ Technol 17:301–307 Pan Y, Liu Q, Liu FF, Qian GR, Xu ZP (2011) Regional assessment of ambient volatile organic compounds from biopharmaceutical R&D complex. Sci Total Environ 409:4289–4296 Parra MA, Elustondo D, Bermejo R, Santamarı´a JM (2009) Ambient air levels of volatile organic compounds (VOC) and nitrogen dioxide (NO2) in a medium size city in Northern Spain. Sci Total Environ 407:999–1009 Pearson CR (1982) C1 and C2 halocarbons, anthropogenic compounds. In: Hutzinger O (ed) The handbook of environmental chemistry, vol 3, Part B. Springer, New York, pp 69–88 Pekey H, Arslanbas¸ D (2008) The relationship between indoor, outdoor and personal VOC concentrations in homes, offices and schools in the 120 metropolitan region of Kocaeli, Turkey. Water Air Soil Pollut 191:113–129 Pekey B, Yılmaz H (2011) The use of passive sampling to monitor spatial trends of volatile organic compounds (VOCs) at an industrial city of Turkey. Microchem J 97:213–219 Rubin JI, Kean AJ, Harley RA, Millet DB, Goldstein AH (2006) Temperature dependence of volatile organic compound evaporative emissions from motor vehicles. J Geophys Res Atmos 111:D03305. doi:10.1029/2005JD006458 Thammakhet C, Muneesawang V, Thavarungkul P, Kanatharana P (2006) Cost effective passive sampling device for volatile organic compounds monitoring. Atmos Environ 40:4589–4596 Tiwari V, Hanai Y, Masunaga S (2010) Ambient levels of volatile organic compounds in the vicinity of petrochemical industrial area of Yokohama, Japan. Air Qual Atmos Health 3:65–75 Turkish Ministry of Health (2004) National burden of disease and cost-effectiveness study report. Turkish Ministry of Health, Ankara Yurdakul S, Civan M, Tuncel G (2013) Volatile organic compounds in suburban Ankara atmosphere, Turkey: sources and variability. Atmos Res 120:298–311
Spatial Distribution and Temporal Trends of VOCs in a Highly Industrialized Town in Turkey ¨ ztu¨rk • Pınar Ergenekon • Gaye O ¨ zdemir Sec¸kin Naciye O Su¨meyra Bayır
•
Received: 11 June 2014 / Accepted: 25 February 2015 / Published online: 5 March 2015 Ó Springer Science+Business Media New York 2015
Abstract An extensive monitoring study of volatile organic compounds (VOCs) was conducted at sites across the highly industrialized town of Dilovasi, northern Turkey to determine temporal and spatial trends in pollutant concentrations and relate to the effects of source locations, meteorology, and topography. Two-week passive samplers (Tenax tubes) were deployed at twelve sites from February to December 2012 and analysed using thermal desorption and gas chromatography with mass spectrometric detection (TD–GC-MS). Sampled total VOC (TVOC) levels were highest in the July through October period and were highest at low-altitude sites near industry facilities and vehicle traffic sources (148.3 lg/m3 at site 11, 154.1 lg/m3 at site 10) and lowest at high-altitude sites located furthest upwind from industry and traffic sources (78.4 lg/m3 at site 5 and 78.5 lg/m3 at site 6). Analysis of ‘‘T/B’’ ratios suggested that contributions to ambient VOC in Dilovasi are dominated by the town’s industrial sources. Meteorological conditions and the town’s basin topography were also found to significantly influence the city’s air quality, with strong winds from the NE observed to correlate with periods of higher sampled TVOC. Compared with other industrialized urban centers, the study revealed that there is significant toluene pollution in Dilovasi and recommended enhanced continuous monitoring at the city’s industrial and residential zones.
¨ ztu¨rk (&) S. Bayır N. O Department of Chemistry, Gebze Technical University, Kocaeli, Turkey e-mail: [email protected]; [email protected] ¨ . Sec¸kin P. Ergenekon G. O Department of Environmental Engineering, Gebze Technical University, Kocaeli, Turkey
Keywords Monitoring Air pollution Volatile organic compounds Toluene Basin topography Intraurban sampling Industry sources
Volatile organic compounds (VOCs) constitute an important group of air pollutants. Some of them can induce cancer, many of them play an active role in the formation of tropospheric ozone, cause photochemical smog formation and contribute to global warming (Atkinson 2005; Filella and Pen˜uelas 2006). The most abundant VOC species are benzene, toluene, and ethyl benzene, xylene (BTEX) (Parra et al. 2009).VOCs are released into atmosphere from both anthropogenic and natural sources. Anthropogenic sources are mainly fossil fuel use and industrial activities. Various household and personal care products also contribute significantly to VOC releases (Pearson 1982). Sampling onto solid adsorbents by active or passive sampling and then subsequent thermal desorption of trapped compounds into a capillary gas chromatography column is a widely used method for VOC monitoring. Combined thermal desorption gas chromatography mass spectrometry /flame ionization detection (TD–GC-MS/ FID) is a very sensitive and cost effective analytical system. Passive sampling of ambient VOCs is a preferred option for monitoring ambient or indoor air quality since samplers are easy for technicians to deploy in the field and samplers can be analysed for a range of compounds at low cost. Field monitoring studies have successfully used longterm diffusive sampling of one to 4 week periods to accurately determine time-weighted average (TWA) VOC concentrations (Thammakhet et al 2006; Martin et al 2010; Monn and Hangartner 1996).
123
654
It is well known that VOC concentrations are significantly higher in areas with industrial facilities and in large cities with high traffic densities. Dilovası, Turkey is a heavily industrialized town that is associated with high cancer incidence rates (i.e., three times the national average; Ministry of Health 2004). A recent study revealed that the risk of death due to cancer was 4.4 times higher for individuals residing in Dilovasi for more than 10 years compared to individuals with residence less than 10 years (Hamzaoglu et al. 2011). The city also has a basin topography that increases the severity of the exposure to air pollution by reducing dispersion, amplified during thermal inversion events. The town is also highly industrialized, with many facilities associated with significant VOC releases (e.g., metal, solvent storage, paint manufacturing) concentrated in the lowest part of the valley and highly populated residential areas distributed between the industrialized zones. For all these reasons, determination of the intraurban distribution of ambient VOC across the town via a comprehensive field sampling campaign was of vital importance. There have been few sampling studies for VOC in the region. Passive sampling and TD/GC–FID analysis were used to measure ambient concentrations of VOCs in Kocaeli and found that concentrations of benzene, toluene, ethylene, and xylene (i.e., BTEX compounds) were high along major roads, city centers and near industrial plants (Pekey and Yılmaz 2011). Statistical factor analysis revealed the vehicle exhaust and industrial activity as the potential pollution sources affecting the VOC concentrations (Pekey and Yılmaz 2011; Pekey and Arslanbas¸ 2008). However, preceding studies have used limited spatial sampling networks and sampled a limited time series and so have provided little information about air quality in Dilovasi’s residential zones relative to adjacent industrial zones. In this study, detailed VOC speciation and quantification was performed to better understand the overall profile and distribution of VOCs in Dilovası ambient air. Two-week integrated air samples were collected over a 10-month period from 12 sampling sites and analysed for speciated VOCs. Samples were collected on Tenax TA adsorbent by passive sampling and analyzed using TD/GC-MS. The study’s major objective was to identify the VOC profiles and levels across the study region, with emphasis on highly populated residential areas as well as to reveal any temporal or seasonal variations. The correlation between the most abundant species, meterological factors and VOC’s temporal and special variations are utilized to describe the general situation of exposure to VOC pollution in the town.
Materials and Methods Dilovasi is an industrial coastal town situated in northern Turkey by the Marmara sea. The climate in the region is
123
Bull Environ Contam Toxicol (2015) 94:653–660
temperate and highest temperatures are generally observed in July and August (Ergenekon et al 2009). The city covers a total area of 2000 ha and has a population of 50,000 residents. There are more than 150 industrial facilities operating in the city, most of which can be classified as chemical or metal industry. In the monitoring area of Dilovası district, 12 sampling points were selected at different distances to industrial plants and major highways (Fig. 1). Meteorological data (temperature, wind speed, direction etc.) were obtained from the air monitoring station of Dilovasi within the monitoring network of Environment and Urbanization Ministry at a distance of 150 m south of the site 10. Two-week integrated air samples were collected using passive stainless steel sampling tubes (Perkin Elmer) packed with Tenax TA media and capped with a diffusion end cap at the inlet and a Swagelok end cap at the other. Sampling tubes were conditioned prior to sampling by passing a stream of ultra-pure dry air at 90 mL/min at 300°C for 20 min. Sampling tubes were stored in the refrigerator at ?4°C prior to sampling and transferred to sampling point in protecting cooled storage container. The sampling period spanned 10 months (February to December 2012). Sampling tubes were collected in pairs from each site, capped with Swagelok end caps, stored in sealed glass jars and transported in cooled containers back to the laboratory for analysis. Measurements were made using a thermal desorption unit (Perkin Elmer Turbo Matrix) coupled to a GC (Perkin Elmer Clarus 680 GC) and a MS detector (Perkin Elmer Clarus 600 T MS). The VOCs were thermally desorbed for 10 min at 300°C with a flow of 30 mL/min of ultra-pure helium passing through and carrying the desorbed VOCs to a preconcentration trap (-10°C). After the focusing step, the analytes were desorbed from the trap by rapid heating to 250°C into the injector of the GC. The separation of VOCs was performed on a 30 m 9 0.32 mm 9 1.28 lm capillary column (DB-624) with ultra-pure helium as carrier gas. After sample injection, the column oven was maintained at 35°C for 2 min and then the temperature was first increased to 220°C at a rate of 5 °C/min (totaly 41 min oven program). The MS scanned from 35 to 400 a.m.u. in the electron impact mode. Four-point calibration (0.1, 1, 5, 10 ppm) was performed using liquid standards in methanol solutions. These liquid standards (AccuStandart M-502, 60 compounds) were injected into adsorbent tubes using a calibration rig (Markes) by passing a pure nitrogen stream for 2 min. through the adsorbent tubes ensuring that compounds to be adsorbed while methanol evaporated. Calibration standard tubes were desorbed and analyzed by using the same method parameters used in the analysis of sample tubes.
Bull Environ Contam Toxicol (2015) 94:653–660
655
Fig. 1 The satellite image of study region (Dilovasi, Turkey) showing locations of sampling sites (Google, DigitalGlobe 2013)
Method detection limits (MDL) were calculated as three standard deviation of the blank measurement and ranged between 0.02 ng (n-dodecane) and 7.20 ng (m,p-xylenes); MDL are presented for all analysed species in Table 1. The data quality was controlled by checking relative standard deviation (RSD) of the replicates; replicate samples with RSD values higher than 20 % of the mean measured value were eliminated from the dataset. Blank data for conditioned sampling tubes were checked for any residues before use. In addition, toluened8 was injected to all tubes prior to GC/MS analysis for ensuring the accuracy of the analytical method. Field blanks were deployed at the sites at 2 month intervals and analysed for possible artifact formation. Field data and lab-analysed sample concentrations data was entered into Microsoft Excel, 2010 and SPSS 19.0 for calculation of time-weighted average concentrations, statistical correlation and temporal trend analysis. Wind rose analysis was completed using WRPLOT View software from Environmental Lakes Inc. Spatial distribution of sampled concentrations were mapped across the study area using ArcView GIS 10.1 (ESRI 2012) and a pollutant surface was constructed by interpolation (i.e., Spline, regulated type, weight parameter 0.1 and number of points parameter of 12).
Results and Discussion A total of 35 VOCs were identified in the sample data, classified as aromatic (Toluene, Ethylbenzene, m-p xylene, o-Xylene, Benzene, n-Propylbenzene, Isopropylbenzene,
1.3.5-Trimethylbenzene, tert-Butylbenzene, sec-Butylbenzene, Limonene, p-Isopropyltoluene, Styrene, Phenol, Acetophenone, Naphthalene,) halogenated (Dichloromethane, 1.2-Dichloroethene, Chloroform, 1,2-dichloroethane, Trichloroethene, Tetrachloroethene, Dibromochloromethane, 2-Chlorotoluene,1,2-Dichlorobenzene, trichlorofluoro-Methane, 1-chloro-2-methylbutane) and the others (Octane, Nonane, n-dodecane, n-tridecane, Pentane, 3-methylpentane, Methylcyclopentane, Ethanol). At most sites, all 35 VOC species were present at concentrations greater than their respective method detection limits. Summary statistics across all sites and samples are presented for all 35 analysed species in Table 1. TVOC values at 12 sampling points during the sampling period were shown in Fig. 2. Date in the x axis represents the starting day of the 2-week passive sampling period. Total volatile organic compound (TVOC) values vary from site to site, generally sites 9, 10, and 11 were the most polluted. It is clear that there is an increase in TVOC amounts at most of the sampling points during the sampling periods from 4th of July and August 15th to 7th of November. This might be caused by the emission variation in the sources or due to meteorological conditions prevalent in the region. The ‘‘high TVOC’’ period also generally aligns with the time of year where highest temperatures are seen (Rubin et al. 2006; Cetin et al. 2003). The hourly ambient temperature values for the sampling period ranged from -1 to 34°C. Temperatures over 20°C were observed between samplings of 23rd of May to 26th of September (warm period). However TVOC concentrations were not highly correlated with the temperature. The correlation coefficient (r) between TVOC and temperature is
123
656
Bull Environ Contam Toxicol (2015) 94:653–660
Table 1 Summary statistics of analysed VOC species Compounds
Species MDL as ng
Trichlorofluoro-Methane
0.59
% of site samples [MDL 63
Mean (SD) as lg/m3 1.00 (0.91)
Median (first quartile - third quartile) as lg/m3 1.00 (0.94–1.05)
Pentane
1.09
90
4.01 (2.86)
3.31 (2.83–5.20)
Ethanol
6.63
75
3.56 (4.26)
3.24 (2.61–4.59)
Dichloromethane
3.98
91
14.61 (31.37)
15.57 (12.07–16.48)
1-chloro-2-methylbutane
2.14
86
3.88 (3.34)
3.71 (3.48–3.90)
3-methylpentane
2.24
90
3.11 (2.83)
2.83 (2.76-3.27)
Methylcyclopentane 1.2-Dichloroethene
2.15 1.44
80 13
5.45 (9.53) 0.40 (0.28)
5.30 (4.61–6.48) 0.42 (0.33–0.48)
Chloroform
1.02
59
0.70 (0.84)
0.59 (0.53–0.86)
Benzene
1.2
99
2.83 (1.81)
2.92 (2.26–3.48)
1.2-dichloroethane
0.69
1
0.24 (0.08)
0.24 (0.24–0.24)
Trichloroethene
0.76
17
0.32 (0.18)
0.29 (0.25–0.39)
Toluene
6.38
100
41.91 (52.72)
42.96 (32.20–51.77)
Octane
0.13
100
0.68 (0.42)
0.60 (0.52–0.74)
Tetrachloroethene
0.65
53
0.67 (0.44)
0.65 (0.45–0.76)
Dibromochloromethane
0.74
7
0.79 (0.64)
0.52 (0.48–0.88)
Ethylbenzene
0.69
91
1.38 (0.80)
1.20 (1.04–1.92)
m-p xylene
7.2
91
4.83 (2.91)
4.16 (3.82–6.25)
o-Xylene
1.66
90
1.07 (0.70)
0.94 (0.76–1.50)
Nonane
0.34
90
3.62 (5.88)
3.76 (2.81–4.37)
Styrene
1.7
94
2.90 (2.07)
2.33 (1.65–3.89)
Isopropylbenzene n-Propylbenzene
0.41 0.4
10 46
0.13 (0.08) 0.24 (0.15)
0.14 (0.08–0.16) 0.23 (0.20–0.30)
2-Chlorotoluene
0.31
54
0.19 (0.15)
0.18 (0.16–0.21)
1.3.5-Trimethylbenzene
0.61
66
0.46 (0.31)
0.37 (0.32–0.67)
tert-Butylbenzene
0.96
62
4.55 (4.07)
3.88 (3.11–5.29)
sec-Butylbenzene
0.39
3
0.23 (0.23)
0.23 (0.23–0.23)
Limonene
0.67
97
1.92 (1.63)
1.81 (1.23–2.60)
p-Isopropyltoluene
0.46
23
0.32 (0.20)
0.19 (0.16–0.36)
1.2-Dichlorobenzene
0.41
78
0.28 (0.09)
0.27 (0.20–0.35)
Phenol
5.1
82
1.98 (1.20)
2.06 (1.67–2.21)
Acetophenone
0.81
87
0.99 (0.55)
1.01 (0.92–1.07)
n-dodecane
0.02
99
0.53 (0.31)
0.52 (0.46–0.60)
Naphthalene
0.62
64
0.66 (0.51)
0.58 (0.48–0.90)
n-tridecane
0.5
95
6.97 (5.22)
6.05 (2.30–8.65)
calculated as 0.36 but individual species displayed different degree of correlation. While r for toluene is 0.48, r for benzene is -0.89 suggesting warmer conditions cause the back diffusion of benzene from the adsorbent. In a study (Brandshaw and Ballantine 1995) back diffusion loss for benzene during a 14 day-sampling was given as 68 % while it is only 43 % for toluene. In addition since Benzene is not strongly adsorbed on Tenax TA, obviously back diffusion rates is going to increase with increasing temperature. Although tenax is not very strong adsorbent, once conditioned its low background concentration and high
123
temperature stability make it a preffered sorbent for monitoring a large group of VOC in ambient air (Harper 2000). 14 days of passive sampling for VOCs by using Tenax TA as in this study might be useful for preevaluation of the VOC pollution in a region if the limitations are carefully taken into consideration. The wind rose for the entire sampling period is presented in Fig. 3a; winds were predominantly from the NW and N and almost half of the winds were calm (wind speed \2.1 m/s). However, during periods with significantly higher sampled TVOC concentrations (i.e., 4 July and 15 August through 24
Bull Environ Contam Toxicol (2015) 94:653–660
657
Fig. 2 Temporal variation in TVOC (lg/m3) at all sampling sites
October sampling period), prevailing wind directions were predominantly from the NE and wind speeds were higher, as presented in Fig. 3b. This observation suggests that in addition to source emission variations, when the strong winds blow from NE direction residential areas downwind of the industrial zones, such as the southern coast of the Marmara Sea, may also be exposed to high VOC levels and monitoring at these residential areas is recommended. Averaged concentrations of 13 most abundant compounds at all sites can be seen in Fig. 4 . Those compounds are namely benzene, styrene, 3-methylpentane, nonane, 1-cloro-2-methylbutane, pentane,ethanol,
m-p-xylenes, tert-butylbenzene, methylcyclopentane, ntridecane, dichloromethane and toluene. Toluene is the most abundant compound during entire monitoring period at all sampling. Toluene and total of selected 13 VOC concentrations are lowest at the sites with the highest altitutudes (Sites 5, 6 and 7). In addition to their high altitude, these sites are also farther from Dilovasi sources and located upwind of major source regions (i.e., based on prevailing winds, see Fig. 3a). Site 1 also shows a similar clean profile due to its location where no major sources exist in the vicinity and surrounding topography allows greater dispersion of pollutants (see Fig. 1). The
Fig. 3 Wind rose for the sampling period: a 10-month sampling period; b period corresponding to high sampled TVOC (4 July and 15 August through 24 October)
123
658
Bull Environ Contam Toxicol (2015) 94:653–660
Fig. 4 Concentrations (lg/m3) of selected compounds at sampling sites; carcinogenics is sum of Group1 and Group 2B compounds (IARC)
basin topography of the town reduces dispersion and increases build up of ambient pollutant levels and consequently the most polluted sites (sites 10, 11) were those located near major emission sources and also found at the bottom of the valley (lowest altitudes). Sites 3, 9, and 11 seem to be the most polluted site in terms of toluene. The figure also depicts the sum of the concentrations of carcinogenic species in Group1 (carcinogen to human), Group 2A (probably carcinogenic to human), and Group 2B (possibly carcinogenic to human) classified by International Agency for Research on Cancer (IARC). In terms of carcinogenic species measured in Dilovasi only benzene is in Group 1, dichloromethane is in Group 2A, and other four compounds of chloroform, ethylbenzene, styrene, and naphthalene are from Group 2B. Total concentration of these six compounds was highest at the Site 11. Measuring the other Group 1 species such as 1,3-butadiene, dioxins, and benz(a)-pyrene could give complete spatial distribution of the carcinogenic species in the area. Locations of major chemical, dye and metal industry facilities (known to have VOC emissions due to their process involving solvent use) and average sampled TVOC concentrations were mapped across the study area in Fig. 5; the interpolated pollutant surface shows the spatial distribution of ambient TVOC levels. Highest levels of TVOC are concentrated in the southern coast of the town surrounding sites 9, 10 and 11 while areas N and E of this region have much lower ambient TVOC concentrations. The locations of the metal, dye and chemical industries with possible VOC emission were shown in the figure. Overall, the zones with most dense industry facility siting coincide with the most polluted zones of Dilovasi. This also confirms the pollution was carried to the southern direction under the prevailing winds from N, NE, and NW. Correlation coefficients between VOC species were also calculated and showed: high correlation between 1,2-dichloroethene and toluene (0.73), dicholoromethane
123
and nonane (0.81), and m,p-xylene and ethyl-benzene (0.91); very low negative correlation between toluene and benzene (-0.13). The ambient levels of toluene and several additional sampled compounds (i.e., inclusive of carcinogenic species) sampled in Dilovasi is compared with sampling results from other urban and industrial urban centers in Table 2. Data given for this work in Table 2 are mean values averaged over 10-month sampling period at the stated sites. This showed that Dilovasi is generally highly polluted in toluene relative to other cities, particularly at the industrial zone sampling sites. Toluene to benzene ratios (T/B) of 1.5–2 (i.e., depending on proximity to emission source) have been reported for sites influenced primarily by traffic sources (Gelencser et al. 1997) while sites influenced primarily by industrial emissions have been reported T/B ratios greater than 5 (Barletta et al 2008). In this study, due to the fact that very high negative correlation of benzene with temperature caused by the back diffusion, T/B ratios were calculated for only cold period (i.e. before May 23 and after September 26 as a total of 11 samplings) T/B ratios ranged between 2.86 and 13.1 suggesting that contributions to ambient toluene in Dilovasi are dominated by industrial sources rather than traffic. Sum of carcinogenics which is the total of IARC Group 1, 2A and 2B carcinogenics (6 compounds) is very high at the site 11. The VOC monitoring study conducted at 12 sites across the Dilovasi region revealed meaningful spatial and temporal trends in the region’s ambient air quality. Sites located furthest from industry and traffic sources (site 5 and 6) had the lowest sampled TVOC levels. Conversely, sites located near both industry facilities and vehicle traffic (sites 10 and 11) had very high sampled TVOC levels. Analysis of T/B ratios suggested that contributions to ambient VOC in Dilovasi are dominated by the city’s industrial sources. Meteorological conditions and the city’s
Bull Environ Contam Toxicol (2015) 94:653–660
659
Fig. 5 Spatial distribution of average-sampled TVOC concentrations (lg/m3) in Dilovası
Table 2 Average sampled VOC levels (lg/m3) in Dilovasi ambient air compared with other urban and industrial centers (no data denoted by ’nd’) Compounds
This work
USAa
Chinab
Turkeyc
Japand
Nigeriae
Chinaf
Site 5
Site 11
Michigan
Industrial
Ankara
Yokohama city
Apapa
Guangzhou
Dongguan
Industrial background
Industrial
Urban
Urban
Urban
Industrial
Urban
Industrial
Commercial
Industrial
Toluene (T)
32.76
56.00
3.30
15.70
7.89
12.43
7.91
19.70
22.10
23.07
DCM*
10.57
14.60
nd
nd
nd
nd
nd
14.12
nd
nd
Chloroform*
0.54
0.60
nd
1.20
nd
nd
nd
18.38
nd
nd
n-tridecane
1.24
12.70
nd
nd
nd
1.21
0.75
nd
nd
nd
m-p xylene
2.52
6.80
3.00
2.80
2.21
4.91
2.43
53.22
6.67
6.38
Benzene (B)*
2.49
3.90
1.20
6.70
2.18
3.39
2.11
22.26
6.54
4.02
Ethylbenzene*
0.68
2.10
0.70
2.00
0.85
6.34
2.74
14.58
5.38
4.60
Naphthalene*
0.50
1.52
0.30
nd
nd
1.04
1.04
14.88
nd
nd
Styrene*
0.97
3.32
nd
nd
0.41
nd
nd
nd
nd
nd
15.75
26.04
2.2
9.9
3.44
10.77
5.89
84.22
11.92
8.62
Sum of Cargc. a
Jiaa et al. (2012),
b
Pan et al.(2011),
c
Yurdakul et al. (2013),
d
e
f
Tiwari et al. (2010), Chekwube et al.(2012), Barletta et al. (2008)
* Carcinogenic by IARC
basin topography also significantly influence the city’s air quality, with strong winds from the NE observed to correlate with periods of higher sampled TVOC. Therefore
emissions from the intensely located metal, chemical, dye production and solvent storage facilities in the middle of the town, will possibly be carried to the southern end of the
123
660
town and areas across the Marmara sea. From these findings, continuous monitoring for selected VOCs at Dilovasi and southern vicinity can be recommended. The study revealed that there is significant toluene pollution in Dilovası, compared with other industrialized urban centers. Passive sampling can be used as a cost-effective method of monitoring ambient air quality across the Dilovası urban region, as was demonstrated in this study. Finally, studies of chronic health risk associated with exposure to ambient VOC can also be recommended to further reveal relationships between ambient air quality and public health. ¨ BI˙TAK, The Acknowledgments This work is supported by TU Scientific and Technological Research Council of Turkey under Research project 110Y138. We also thank Meltem Celen, Ph.D, for her help with ArcView use.
References Atkinson R (2005) Atmospheric chemistry of VOCs and NOx. Atmos Environ 34:2063–2101 Barletta B, Meinardi S, Simpson IJ, Zoub S, Rowland FS, Blake DR (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 Bradshaw NM, Ballantine JA (1995) Confirming the limitations of diffusive sampling using Tenax TA during long term monitoring of the environment. Environ Technol 16:433–444 Cetin E, Odabasi M, Seyfioglu R (2003) Ambient volatile organic compound (VOC) concentrations arounda petrochemical complex and a petroleum refinery. Sci Total Environ 312:103–112 Chekwube OC, Majebi OJ, Sanni LM (2012) Baseline levels of volatile organic compounds (VOCs) Pollution in Apapa Industrial Areas of Lagos State, Southwestern-Nigeria. Arch Appl Sci Res 4(6):2564–2571 ¨ ztu¨rk N, Tavs¸ an S (2009) Environmental air levels of Ergenekon P, O VOCs by thermal desorption-gas chromatography in an Industrial Town. Fresenius Environ Bull 18:1999–2003 Filella I, Pen˜uelas J (2006) Daily, weekly, and seasonal time courses of VOC concentrations in a semi-urban area near Barcelona. Atmos Environ 40:7752–7769 Gelencser A, Siszle K, Hlavay J (1997) Toluene-Benzene concentration ratio as a tool for characterizing the distance from vehicular emission sources. Environ Sci Technol 31:2869–2872
123
Bull Environ Contam Toxicol (2015) 94:653–660 Hamzaog˘lu O, Etiler N, Yavuz CI, Caglayan C (2011) The causes of deaths in an industry-densed area: example of Dilovasi (Kocaeli). Turk J Med Sci 41:369–375 Harper M (2000) Sorbent trapping of volatile organic compounds from air. J Chromatogr A 885:129–151 Jiaa C, Battermanb SA, Relyeaa GE (2012) Variability of indoor and outdoor VOC measurements: an analysis using variance components. Environ Pollut 169:152–159 Martin NA, Leming EJ, Henderson MH, Lipscombe RP, Black JK, Jarvis SD (2010) Verification of diffusive and pumped samplers for volatile organic compounds using a controlled atmosphere test facility. Atmos Environ 44:3378–3385 Monn C, Hangartner M (1996) Passive sampling of aromatic volatile organic compounds (VOC) in ambient air Switzerland. Environ Technol 17:301–307 Pan Y, Liu Q, Liu FF, Qian GR, Xu ZP (2011) Regional assessment of ambient volatile organic compounds from biopharmaceutical R&D complex. Sci Total Environ 409:4289–4296 Parra MA, Elustondo D, Bermejo R, Santamarı´a JM (2009) Ambient air levels of volatile organic compounds (VOC) and nitrogen dioxide (NO2) in a medium size city in Northern Spain. Sci Total Environ 407:999–1009 Pearson CR (1982) C1 and C2 halocarbons, anthropogenic compounds. In: Hutzinger O (ed) The handbook of environmental chemistry, vol 3, Part B. Springer, New York, pp 69–88 Pekey H, Arslanbas¸ D (2008) The relationship between indoor, outdoor and personal VOC concentrations in homes, offices and schools in the 120 metropolitan region of Kocaeli, Turkey. Water Air Soil Pollut 191:113–129 Pekey B, Yılmaz H (2011) The use of passive sampling to monitor spatial trends of volatile organic compounds (VOCs) at an industrial city of Turkey. Microchem J 97:213–219 Rubin JI, Kean AJ, Harley RA, Millet DB, Goldstein AH (2006) Temperature dependence of volatile organic compound evaporative emissions from motor vehicles. J Geophys Res Atmos 111:D03305. doi:10.1029/2005JD006458 Thammakhet C, Muneesawang V, Thavarungkul P, Kanatharana P (2006) Cost effective passive sampling device for volatile organic compounds monitoring. Atmos Environ 40:4589–4596 Tiwari V, Hanai Y, Masunaga S (2010) Ambient levels of volatile organic compounds in the vicinity of petrochemical industrial area of Yokohama, Japan. Air Qual Atmos Health 3:65–75 Turkish Ministry of Health (2004) National burden of disease and cost-effectiveness study report. Turkish Ministry of Health, Ankara Yurdakul S, Civan M, Tuncel G (2013) Volatile organic compounds in suburban Ankara atmosphere, Turkey: sources and variability. Atmos Res 120:298–311
