Environmental Monitoring and Assessment (2005) 109: 227–242 DOI: 10.1007/s10661-005-6292-z

c Springer 2005


EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA (DELHI AND MUMBAI) ANJALI SRIVASTAVA∗ , A. E. JOSEPH, AJIT MORE and SUNIL PATIL National Environmental Engineering Research Institute (NEERI), Mumbai Zonal Centre, 89/B, Dr. A.B. Road, Worli, Mumbai, India (∗ author for correspondence, e-mail: [email protected])

(Received 14 May 2004; accepted 12 November 2004)

Abstract. Air pollution has assumed gigantic proportion killing almost half a million Asians every year. Urban pollution mainly comprises of emissions from buses, trucks, motorcycle other forms of motorized transport and its supporting activities. As Asia’s cities continue to expand the number of vehicles have risen resulting in greater pollution. Fugitive emissions from retail distribution center in urban area constitute a major source. Petrol vapours escape during refueling adding pollutants like benzene, toluene, ethylbenzene and xylene to ambient air. This paper discusses a study on fugitive emissions of Volatile Organic Compounds (VOC) at some refueling station in two metropolitan cities of India, i.e., Mumbai and Delhi. Concentration of VOCs in ambient air at petrol retail distribution center is estimated by using TO-17 method. Concentration of benzene in ambient air in Delhi clearly shows the effect of intervention in use of petroleum and diesel fuel and shift to CNG. Chemical Mass Balance (CMB) model is used to estimate source contributions. At Delhi besides diesel combustion engines, refueling emissions are also major sources. At Mumbai evaporative emissions are found to contribute maximum to Total VOC (TVOC) concentration in ambient air. Keywords: CMB, Delhi, Mumbai, petrol pump, VOC

1. Introduction Volatile Organic Compounds (VOCs) are important air pollutants in urban atmosphere. Some of the VOCs are toxic, potentially carcinogenic and mutagenic at concentrations levels found in urban environment (Edgerton et al., 1989). Exposure to VOCs is of concern as it may result in significant risk to human health. Atmospheric reactions of VOCs lead to secondary pollutants, which is turn cause the deterioration of air quality and damage to Corps and vegetation. VOCs are emitted from number of sources, which include urban and industrial activities, and natural sources. Previous studies have reported vehicular emissions as the major source of VOCs in urban air. Pjeffer (1994) reported that 90% of benzene in ambient air is from traffic. Vega et al. (2000) estimated that motor vehicle exhaust contributes 55% of non-methane hydrocarbons. Baldasano et al. (1998) have shown that 62% of VOC pollution in Martorell is Spain is road traffic. Besides vehicular exhaust, hot and cold soak emissions are also one the urban sources of VOCs. Evaporative emissions of VOCs from gasoline distribution network are also one of the important sources of VOCs in ambient air.

228

ANJALI SRIVASTAVA ET AL.

Transportation, storage and marketing petroleum liquids represents potential source of evaporation losses. Emissions occur due to temperature fluctuations, permeation of fuel through hoses and fitting etc High ambient temperatures cause large evaporative losses. Motor vehicle refueling is a potential source of evaporative emissions of VOCs. Vehicle refueling emission comes from vapours displaced from the automobile tank by dispensed gasoline and from spillage. Refuelling without control measures contributes to the gasoline emissions. In Mumbai and Delhi spillages are frequent due to manual refueling units. Atkinson has shown that atmospheric behaviour of VOCs is governed to a large extent by their life time. In the process of long range transport, the primary pollutants such as VOCs and NOx will react with atmospheric to produce secondary pollutants such as ozone and PAH etc. with different reaction rates. Highly reactive species will react near the vicinity of the sources, while slow reacting species may be transported to large distances. Toluene has much shorter life time than benzene (Singh and Zimmer Mon, 1992). So a higher Benzene to Toluene (B/T) ratio will be found in aged air via a long range transport. B/T ratio can be thus used as a tracer to predict long range transport. However, if sources of benzene or toluene other than vehicular exhaust are present B/T ratio cannot be considered as a tracer. High levels of VOCs have been observed in Asian Countries (Hussam et al., 2002; Iran et al., 1998). The present study focuses on estimating VOCs at urban petrol retail distribution centers in Mumbai and Delhi.

2. Study Area Delhi is capital of India and hub of commercial, political, industrial, educational activities. It has tropical steppe climate. The general prevalence of continental air leads to relatively dry condition with extremely hot summers. Monthly mean temperature range from 14.3 ◦ C in January (minimum 8 ◦ C) to 34.5 ◦ C in June (maximum 47 ◦ C). The main seasonal climate influence is the monsoon, typically from June to October. The mean annual rainfall is 71.5 mm (maximum 211 mm July). Northwesterly winds usually prevail, however, in June and July south-easterly wind predominate. Winter witnesses long periods of calm. Mumbai is located on west coast of India. It is India’s one of the main industrial and commercial city situated on the coast of Arabian sea. It witnesses considerable sea port and airport activity. Mumbai has tropical savanna climate, with monthly relative humidity ranging between 57–87%. Annual average temperature is 25.3 ◦ C rising to maximum of 34.5 ◦ C in June and a minimum of 14.3 ◦ C in January. Average annual precipitation is 2078 mm prevailing in west and north-west with some southern component during June to August. VOCs in ambient air at three petrol pumps in Delhi and Mumbai each namely at IIT Crossing (A), Cannaught Place (B) and Race Course (C) in Dehli and Worli (D), Maheshwari Udyan (E) and Mahim (F) in Mumbai have been monitored once every

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

229

month during peak hours 8 am to 12 pm–5 pm to 9 pm. Sampling has been carried out at a height of approx. 5 to 6 feet in the middle of petrol pump where vehicles halt for filling petrol. All the petrol pumps selected in this study had heavy traffic inflow. Figure 1 shows the locations the sampling stations at Delhi and Mumbai, respectively.

Figure 1. Sampling locations at Delhi and Mumbai.

230

ANJALI SRIVASTAVA ET AL.

At Delhi, petrol pump A is close to traffic intersection, B is situated in a commercial area and C is on a busy road with trees lined on both sides of the road. Petrol pumps D and F at Mumbai are close to seashore approximately 300 m to 500 m away from the coast. Both the petrol pumps have the sewage outfall and sewage treatment plant close to it. The petrol pump E is close to traffic intersection.

3. Methodology USEPA (1985) study shows uncontrolled displacement losses from vehicles refueling for a particular set of conditions can be estimated using following equation. E R = 264.2[(−5.909) − 0.0949(
T ) + 0.0884(TD ) + 0.485(RVP)]

(1)

Where ER = Refueling emissions, mg/l;
T = Difference between temperature of fuel in vehicle tank and temperature of dispensed fuel in ◦ F; TD = Temperature of dispensed fuel ◦ F; RVP = Reid vapour pressure, psia. The quantity of displaced vapours depends on gasoline temperature, auto tank temperature, gasoline RVP and quantity of fuel dispensed. TVOC concentrations were determined using USEPA TO-17 method. The adsorption cartridges containing activated charcoal were fabricated from stainless pipe having length 19 cm and 4 mm ID with caps on both sides. Analytical grade activated charcoal was heated at 200 ◦ C for 12 h and cooled in vacuum desiccators. About 800 mg of preheated cooled activated charcoal was filled in the cartridge with glass wool plugs on either side of the adsorption cartridge. Desorption was carried out by heating at 220 ◦ C for 10 min. Varian make GCMS (Model Saturn 3) with injection mode of sample introduction with DB-624 capillary column of 30 m length. 0.32 mm ID and 1.8 micron film was used. Helium as carrier gas with flow rate of 1 ml/min with split ratio 1:25 was used GC oven was programmed for 35 ◦ C held for 2 min and ramped to 210 ◦ C with the rate of 5 ◦ C/min Ion trap temperature was maintained at 125 ◦ C while acquisition mass range was from 35 Dalton to 260 Dalton in EI Mode. Calibration curves were drawn for VOCs present in VOC Mix 15 of Dr. Erhenstrofer and these were quantified in this study. NIST 98 library was used to identify the VOCs. Matching of first three abundances were considered for identification.

4. Results and Discussions The minimum and maximum temperatures observed during the sampling period and average temperature during 7 am to 11 pm are given in Table I. On an average temperature of dispensed fuel was observed to be 2 ◦ C less than ambient temperature

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

231

TABLE I Ambient temperature during sampling in ◦ C Minimum

Maximum

Average∗

Delhi

∗

March April May June July August September October November December January February

11.6 17 21.2 24.8 27 25.2 22.6 16.6 9.4 6.2 5.6 7.2

March April May June July August September October November December January February

21 22.3 24.5 21.6 23.5 23.9 23.4 19.8 19.5 17.6 16.3 13.9

37.8 42.4 44.8 41.8 40.6 36.4 37.6 34.6 34.4 27.6 24 30.4 Mumbai 39.1 37.2 37.1 33.6 32 31.3 33.4 36 36.5 35.3 33.6 34.5

24.7 29.7 33 33.3 33.8 30.8 30.1 25.6 21.9 16.9 14.8 18.8 30.5 29.75 30.8 27.6 27.75 27.6 28.4 27.9 28 26.45 24.95 24.2

Average during 8 am to 11 pm.

during 7 am to 11 pm at Mumbai. At Delhi this temperature varied from 3 ◦ C during April, May, June and July to 2 ◦ C below ambient temperature during rest of the month from 7 am to 11 pm. The temperature of the fuel in the vehicle tank varied from vehicle to vehicle. At Mumbai average
T was observed to range from 1.8 ◦ C to 2 ◦ C during the year. During April, May, June and July average value of
T was observed to 0.5 ◦ C and during rest of the month’s 2 ◦ C at Delhi. Table II gives the data on petrol and diesel fuel consumption and registered vehicles during the year 2001 at Mumbai and Delhi. In India RVP of petrol should be in range 30–60 as per BIS specifications. However, a study carried out by Centre for Science and Environment (CSE) shows the RVP is in the range of 55–60.

232

ANJALI SRIVASTAVA ET AL.

TABLE II Fuel consumption and registered vehicle count at Delhi and Mumbai

Petrol Diesel Vehicle count

Delhi

Mumbai

568000 kl 1223000 kl 3456579

406344 kl 322464 kl 1029563

Source: Petroleum coordination committee and road transport office.

Considering average temperatures as given in Table I, the refueling emissions during different seasons at Mumbai and Delhi are estimated using Equation 1, and are given in Table III. Total annual evaporative emissions is estimated to be 3.685 Gg in Delhi and 2.664 Gg in Mumbai during the year 2001. 5. Ambient Concentrations USEPA has identified 188 gaseous pollutants as Hazardous Air Pollutants (HAPs) on the basis of there toxicity, affect on human health, ecology and climate (US, 1990, Clean Air Act Amendments, Section, 112). Amongst the VOCs identified at petrol pumps in Mumbai 14 fall under the category of HAPs and 11 VOCs identified at Delhi petrol pumps are classified as HAPs. Most of the HAPs identified are contents of exhaust emissions of cars and evaporative emissions of petrol. Table IV gives the list of VOCs identified at Delhi and Mumbai, respectively. Annual and seasonal means of VOCs at Delhi are presented in Figure 2. Besides VOCs originating from diesel and gasoline vehicle exhaust and evaporation, chlorinated VOCs have also been identified. These could be considered to be originating from sewage man holes, dry cleaning solvent use and biogenic emissions. Concentrations of TVOC and benzene during the months of April to July shows drastic reduction as compared to previous months. This can be attributed to implementation TABLE III Evaporative emissions during refueling (g/l) Season

Delhi

Mumbai

Summer Monsoon Winter Annual average Total annual emission during 2001

6.658 6.657 6.224 6.487 3.685 (Gg)

6.680 6.554 6.485 6.556 2.664 (Gg)

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

233

TABLE IV VOCs identified at petrol pumps in Mumbai and Delhi Mumbai CAS No. Compound name 56-23-5 67-66-3 71-43-2 71-55-6 74-86-2 74-87-3 74-97-5 74-98-6 75-07-0 75-09-2 75-15-0 75-27-4 75-34-3 75-69-4 75-83-2 76-13-1 78-78-4 79-01-6 91-20-3 95-47-6 95-63-6 96-14-0 96-37-7 98-06-6 98-82-8 99-87-6 100-41-4 103-65-1 104-51-8 105-05-5 106-42-3 106-42-3 106-46-7 107-06-2 108-38-3 108-38-3

Carbon tetrachloride Chloroform Benzene Ethane, 1,1,1-trichloroAcetylene Chloromethane Methane bromochloro Propane Acetaldehyde Methylene chloride Carbon disulfide Methane bromodichloro Ethene 1,1 dichloro Trichlorofluromethane 2-2 Dimethyl butane 1,1,2,Trichloro-1, 2,2Trifluoroethane iso-Pentane Trichloroethylene Naphthalene 1,2-Dimethyl benzene 1,2,4-Trimethyl benzene 3-Methyl pentane Methylcyclopentane t-butylbenzene Isopropylbenzene P Isopropyl toluene Ethylbenzene Propylbenzene n-butylbenzene 1,4-Di ethylbenzene 1,4-Dimethyl benzene o-Xylene 1,4 dichlorobenzene Ethane 1,2 dichloro 1,3-Dimethyl benzene m-Xylene

Delhi

Mahalaxmi Race Cannught Mahim Worli udyan Course Place IIT gate EPA M.S.


#
#
#


#
#
#










#
#
#
#



#
#
#







#






#


*




















#
#
#


#
#
#


#
#
#



#
#
#
#








#
#
#



#
#
#



#


#


#


#



#

#


#



#
#


#
#


#
#
#

#

+




#
#




* * * * *


#






#
#


#
#
#




#



#





#
#
#


#
#
#



#
#
#
#
#
#


#


#

* * *

+ +

*

+

*

+

*


#
#


#


#

* *

+

(Continued on next page.)

234

ANJALI SRIVASTAVA ET AL.

TABLE IV (Continued ) Mumbai CAS No. 108-67-8 108-87-2 108-88-3 115-07-1 115-11-7 115-11-7 124-18-5 124-48-1 135-98-8 142-28-9 513-35-9 541-05-9

Compound name

Benzene 1,3,5-trimethylMethylcyclohexane Toluene Propene iso-Butene 2- Methyl-1- propene n-Decane Dibromochloro methane Sec butyl benzene Propane 1,3 dichloro 2-Methyl –2- butene Hexamethylcyclotrisiloxane 556-67-2 Octamethylcyclotetrasiloxane 563-46-2 2-Methyl-1-butene 565-59-3 2,3-Dimethyl benzene 565-71-7 iso propyl benzene 584-94-1 2,3-Dimethyl hexane 591-76-4 2-Methyl hexane 591-76-4 2 –Dimethyl hexane 592-41-6 n-Hexane 594-20-7 Propane2,2 dichloro 611-14-3 1-Ethyl-2-Methylbenzene 616-09-1 Methylene chloride 616-12-3 (E)-3-Methyl-2-Pentene 622-96-8 1-Ethyl-4-Methyl benzene 674-76-0 (E)-4-Methyl-2-Pentene 763-29-1 2-Methyl-1-Pentene 1120-21-4 n-Undecane 1640-89-7 Ethylcyclopentane 2213-23-2 2,4-Dimethyl heptane 2216-34-4 4-Methyl octane 7642-09-3 (Z)-3-Hexene 7688-21-3 (Z)-2-Hexene 13269-52-8 (E) –3- Hexene ∗

Delhi

Mahalaxmi Race Cannught IIT Mahim Worli udyan Course Place gate EPA M.S.


#

#




#

#




#

#




#











#

#


#

#


#














#
#






























#




#














#




#













#















#



*









VOC appearing in the USEPA list of Hazardous Air Pollutants in CAAA Title III. VOC quantified using VOC Mix-15. + Mobile Source air toxic. #


# *




+

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

235

Figure 2. Seasonal and annual averages concentrations at Delhi petrol pumps.

of government policy of banning petrol and diesel in three wheelers, taxi and buses in Delhi. CNG was introduced as alternate fuel for these vehicles by Government of Delhi (Auto Fuel Policy, Govt. of India, 2001). Annual and seasonal means of VOCs at Mumbai are presented in Figure 3. It is observed that chlorinated VOCs which can be associated with oceanic emissions, sewage sludge, dry cleaning, degreasing and solvent use have been identified

236

ANJALI SRIVASTAVA ET AL.

Figure 3. Seasonal and annual average concentrations at petrol pumps of Mumbai.

in Mumbai. Beside these VOCs like acetylene, benzene, toluene, xylenes, 1,2,4 trimethyl benzene, ethylbenzene, ndecane, undecane, n-hexane etc which are considered to be constituent of petrol and diesel exhaust and petrol vapours (Fujita et al., 1997; Watson et al., Spiecer et al., 1992; Scheff et al., 1989; Battelle report, 1992) have as well been identified. Concentration of benzene dominates TVOCs concentrations both in Mumbai and Delhi. Benzene to Toluene ratio of 0.5 has been associated with vehicular exhaust emissions. (Sweet and Vertmettc, 1992, Scheff and Wadden, 1993). In the present

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

237

study benzene to toluene (B/T) ratio of 11.74 at Mumbai and 8.11 at Delhi has been observed. This indicates presence of additional sources of benzene besides vehicular exhaust. At petrol pumps it could be evaporative emissions from vehicles, storage tanks and spillages. The total spillages could not be estimated. Approximately 2–3 ml of gasoline is spilled per vehicle refueled.

6. Source Apportionment Chemical Mass Balance model of USEPA was used to study the contributions of different sources. Source emission estimates have been taken from USEPA Speciate 3.2 database. The sources considered are diesel internal combustion engines which comprise of trucks, off road equipment, stationery engines for pumps and generators. Composite vehicles emissions consist of heavy and light duty cars and small engines. Emission inventory do not usually contain breakdown by cold starts and visibly smoking vehicles. Such instances are quite common in Delhi and Mumbai. However, the present source profile considered does not include this source profile for want of data. Evaporative emissions comprises of refueling stations. Hot soak vehicle speciate profile was modified with reference to benzene, toluene, ethylbenzene, xylene, propylbenzene, 1,3,5 trimethyl benzene, 1,2,4 trimethyl benzene and naphthalene content. The modification done was based on analysis of petrol vapour samples collected from petrol pumps of four companies supplying petrol in Indian viz. IOC, IBP, HPCL and BPCL. Auto repair emissions include vehicle repair garages and shops. Degreasing and dry cleaning comprises of stripping and spraying additives. Dry cleaning activity was merged with degreasing because the profiles were collinear. Natural gas combustion emissions include natural gas combustion engines. Sludge emissions consist of emissions from sewage treatment plants and open defecation. A profile for emissions from oceans was built based on study of Dewulf and Langenhove (1997). Table V gives source profile estimates used in CMB model. Figures 4 and 5 gives the source contribution to total VOCs at Delhi and Mumbai. It shows that diesel internal combustion engine emission dominant in Delhi and evaporative emissions dominate in Mumbai. This results corroborates with the fact that large number of diesel generator sets are used in Delhi in residential, commercial and industrial areas as a backup to power supply. Emission from internal combustion diesel engine thus outweigh evaporative emissions in Delhi. At Mumbai generators sets are generally used at construction sites and industrial areas. Contribution from sea are observed at sites close to the sea at Mumbai. Table VI gives the correlation coefficient r2 values of the source contribution estimate along with range of ratio R/U, which is a residual, i.e., signed difference of measured and calculated concentration of species divided by the uncertainty of that residual. Acceptable values of R/U lie in the range of 0.5 to 2 (Watson et al., 1997). Positive value of R/U indicates that more than one source is contributing to that species

0.02155 0.091 0.0257 0.0146 0.1659 0.3248 0 0 0 0 0 0 0 0 0 0 0

0.095 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Asphalt road 0.002 0.00206 0.00042 0.00016 0.04054 0.02214 0.00066 0.0002 0.0009 0.0002 0.00022 0.0082 0.0493 0.04456 0.00008 0 0

Degreasing & dry cleaning 0.13655 0.00325 0.04195 0.06055 0.23315 0.1409 0.00105 0.0085 0.0265 0.0121 0.0051 0 0 0 0.0002 0 0

Composite vehicle 0.0011 0 0.0001 0.0004 0.0002 0 0 0.0001 0.0002 0.0001 0 0 0 0 0 0 0

Natural gas 0.0302 0 0.0078 0 0.0645 0 0 0 0 0 0 0.001 0 0.179 0.0028 0.0456 0.0973

Sludge 0.122 0 0.0184 0.0075 0.0499 0.1349 0.0031 0.007 0.0484 0.0144 0.0126 0 0 0 0 0 0

Diesel internal combustion engine 0 0 0 0 0 0 0 0 0 0 0 0.1 0 0.1753 0 0.1132 0.4384

Oceanic 0.3811 0.0003 0.068 0 0.3829 0.2313 0.0012 0.0139 0.0416 0.0194 0.0073 0 0 0 0 0 0

Evaporative emissions

BENZE – Benzene, BUTBN – Butylbenzene, ETBZ – Ethylbenzene, NAPH – Naphthalene, TOLUE – Toluene, MPOXY – M,P,O xylene, IPRBZ – iso-Propylbenzene, N PRBZ = n-Propylbenzene, BZ124M – 1,2,4 Trimethylbenzene, BZ135M – 1,3,5 Trimethylbenzene, BZ123M – 1,2,3 Trimethylbenzene, DICLME – Dichloromethane, DILBZ – Dichlorobenzene, TRCLET – Trichloroethane, CLBZ – Chlorobenzene, CCL4 – Carbon tetrachloride, CHLFR – Chloroform.

BENZE BUTBN ETBZ NAPH TOLUE MPOXY IPRBZ N PRBZ BZ124M BZ135M BZ123M DICLME DICLBZ TRCLET CLBZ CCL4 CHLFR

Repair auto

TABLE V Source profile estimates used (fractions by weight of modelled species)

238 ANJALI SRIVASTAVA ET AL.

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

239

Figure 4. Source contribution to total VOC at petrol pumps of Delhi.

and negative values indicate insufficient contribution to that species, may be due to missing source. At Delhi R/U values were observed to lie in acceptable range. Improper representation of concentration of chlorinated species, iso-propyl benzene, butyl benzene, toluene, xylene and naphthalene in source profile is observed. Diesel internal exhaust engine emissions are found to contribute maximum followed by composite vehicular exhaust and evaporative emissions. Natural gas emissions also constitute a significant proportion of TVOC. Wide use of DG sets is reflected from large proportion of diesel internal combustion engine.

240

ANJALI SRIVASTAVA ET AL.

TABLE VI Source contribution estimates R/U Site

R2

IIT gate Cannught place Race course

0.87 0.82 0.87

Worli naka Maheshwari udyan Mahim

0.98 0.64 0.99

Range

TVOC

Delhi −4.1 to 3.2 0.2 −4.2 to 3.6 0.2 −2.2 to 2.7 0.5 Mumbai −1.2 to 2.1 1.3 −4.5 to 2.5 −1.7 −3.2 to 1.5 1.5

Extreme R/U species

Methylene chloride, Naphthalene Methylene chloride, Naphthalene Iso-Propylbenzene, Toluene 1,3,5 Trimethylbenzene, Chloroform Methylene chloride, Dichlorobenzene Benzene, 1,1,1 Trichloroethane

Figure 5. Source contribution to total VOC at petrol pumps of Mumbai.

EMISSIONS OF VOCs AT URBAN PETROL RETAIL DISTRIBUTION CENTRES IN INDIA

241

At Mumbai R/U values for TVOCs lie well within the acceptable range. However, R/U values for different species indicate improper representation of certain species concentration in source profile. Chlorinated compounds, benzene and 1,3,5 trimethyl benzene are the major species which shows improper representation. Besides evaporative emissions, vehicle exhaust, diesel internal combustion engines and degreasing activities form the major source of VOC emissions. Petrol pumps at Worli and Mahim are situated close to sewage outfalls. Emissions associated with sludge and ocean are observed at these locations. Degreasing activity is observed at all the three petrol pumps. Data on volume of petrol dispensed and spilled during sampling period was not available; hence, a correlation between quantity of petrol/diesel dispensed spilled and VOC levels observed could not be established. 7. Conclusions The results of the present study shows that concentrations of VOCs in ambient air at refueling facilities in Delhi and Mumbai are high. Refuelling losses also contribute to VOC concentrations in ambient air. Considering the adverse impacts of VOCs on human health, climate and ecosystem, need of vapour recovery system at refueling units is evident. The vapour recovery system will help in reducing the ambient concentrations of TVOCs in air and as well save the gasoline of the order of three Giga grams per year at two cities Delhi and Mumbai in India. To control VOCs in air the management strategy should thus focus on cost effective vapour recovery system at refueling stations and in vehicles. An effective inspection and maintenance programme and measure against tempering and misfuelling can reduce evaporative and exhaust VOC emissions. Besides fuel volatility control can also be an effective measure to reduce VOC emission from in use motor vehicles. The results indicate additional sources of benzene, one of them could be adulteration of fuels. The observed concentrations are much higher than those recommended by WHO and is a potential hazard to workers at refueling stations. Acknowledgments The authors are thankful to Dr. Sukumar Devotta, Director, NEERI, for his encouragement and Central Pollution Control Board for funding this study.

References Alkinson, R.: 2000, ‘Atmospheric chemistry of VOCs and NOx’, Atmos. Environ. 34, 2063– 2101. Bald Asano, J. M., Delgado, R. and Calbo, J.: 1998, Applying Receptor Models to Analyze Urban/ Suburban VOCs–Air Quality in Martorell (Spain).

242

ANJALI SRIVASTAVA ET AL.

Centre for Science and Environment India: 2002, A Report on the Independent Inspection of Fuel Quality at Fuel Dispensing Stations, Oil Depots and Tank Lorries. Dewulf, J. and Langenhove, H. V.: 1997, ‘Analytical techniques for the determination of measurement data of 7 chlorinated C1 and C2 hydrocarbons and 6 monocylic aromatic hydrocarbons in remote air masses: An overview’, Atoms. Environ. 31(20), 3291–3307. Edgerton, S. A., Holdernm, M. W. and Smith, D. I.: 1989, ‘Inter urban comparison of ambient volatile organic compound concentration’, JAPCA 39, 729–732. Fujita, I. M., Lu, Z, Sheetz, L., Harshfield, G. and Zjekinska, B.: 1997, Determination of Mobile Source Emission Source Fraction Using Ambient Field Measurements, Prepared for Coordinating Research Council, Atlanta, G.A. Desert Research Institute, Reno, NV. Hussam, A. A., Alauddin, M., Khan, A. H., Choudhari, D., BiBi, H., Bhatacharjee, M. and Sultana, S.: 2002, ‘Solid phase micro-extraction: Measurement of volatile organic compounds in Dhaka city air pollution’, J. Env. Sc. and Health Port – A Toxic /Hazard. Subst. Environ. Engg. A37(7), 1223–1239. Ge, I. L. and Sollars, C. J.: 1998, ‘Ambient air levels of volatile organic compounds in Latin America and Asian cities’, Chemosphere 36, Xw 11, 24-97-250. Pfeffere, H. V.: 1994, ‘Ambient air concentrations of pollutants at traffic related sites in urban areas of north Rhire-West-Phalia, Germany’, Sci. Total Environ. 146/147, 263–273. Scheff, P. A., Waddon, R. A., Bate, B. A. and Aronian, P. F.: 1989, ‘Source finger prints for receptor modelling of volatile organics’, J. Air Poll Cont. Assess. 39, 469–478. Scheff, P. A. and Wadden, R. A.: 1993, ‘Reactor modelling of volatile organic compounds emission inventory and validation’, Environ. Sci. Technol. 27(41), 617–625. Singh, H. B. and Zimmermon, P. B.: 1992, Gaseous Pollutants Characterization and Cycling O. Nriaga (ed.) Vol., 24, pp. 177, Wiley Services on Advances in Environment Science and Technology. Swect, C. W. and Vertmette, S. J.: 1992, ‘Toxic volatile organic compounds in urban air in illinois’, Environ. Sci. Technol. 26(1), 165–173. Spicer, C. W., Buxton, B. E., Holdren, M. W., Kelley, T. J., Rust, S. W., Ramamurthi, Smith, D. L., Pate, AD., Sverdrup, G. M., Chuong, J. C. and Shah, J.: 1992, Variability and Source Attribution of Hazardous Urban Air Pollutants – The Columbus Field Study. Battelle Final Report, US. EPA Contract No. 68-D-80082, June 1992. Report on the Expert Committee on Auto Fuel Policy, Chapter 14, Government of India, 2001. US Environmental Protection Agency, Ann. Arbor, MI: 1985, Refueling Emissions from Uncontrolled Vehicles, EPA-AA-SDSB-85-6. USEPA, CMB 8: 2001, Office Air Quality Planning and Standards, Research Triangle Park, NC, 27711-EPA-45R/01-XXX. Vega, E., Mugica, V., Carmona, R. and Valencia, E.: 2000, ‘Hydrocarbon source apportionment in Mexio city using the chemical mass balance receptor model’, Atmos. Environ. 34, 4121–4129. Watson, J. G., Chow, J. C., Fujita, E. M.: 2001, ‘Review of volatile organic compound source apportionment by chemical mass balance’, Atmos. Environ. 35, 1567–1584. Watson, J. G., Robinson, N. F., Lewis, C. W., Coulter, C. T., Chow, J. C., Fujita, E. M., Lowenthal, D. H., Conner, T. L., Henry, R. C. and Wills, R. D.: 1997, Chemical and Mass Balance Receptor Model Version 8 (CMB) User manual, Prepared for USEPA Research Triangle Park, NC, Desert Research Institute, Reno, M.V.