6. Discussion
   
   6.1   Benzene Levels at Forecourt and Pump Sites
   Comparison of benzene levels measured within the three site types at each station (roadside, forecourt and pumps) illustrates the strong influence of petrol station emissions to measured benzene levels within the forecourt and at the pumps, with mean values recorded at these sites generally elevated above levels recorded at the roadside sites. The much higher levels of benzene measured at the pumps reflects the significance of emissions during refuelling, specifically the displacement of headspace vapour from car tanks.
   
   Elevated concentrations measured within the forecourt areas could not be attributed to specific emission sources such as the delivery tanks, and there was a wide spatial variation in measured levels at individual stations, reflecting the influence of a variety of emission sources and orientation of the diffusion tubes in relation to the prevailing wind direction. Spatial variation in forecourt levels at individual stations also partly reflects differences in the location of the diffusion tube in relation to distances from the pumps, for example at Station A the highest forecourt mean of 8 ppb (Site 3) was recorded at a distance of approximately 3 m from the pumps, with a mean of 4.8 ppb recorded at a distance of approximately 5 m. However due to the influence of other factors, such as wind direction, a clear relationship between distance from emission source and forecourt levels was not observed at all stations, for example at Station E a mean of 6 ppb was recorded at Site 2, located approximately 15 m from the pumps, compared to a mean of 3 ppb recorded at Site 3 which was located approximately 5 m from the pumps.
   
   6.2   Benzene Levels at Roadside Sites
   Benzene levels measured at the roadside sites will be strongly influenced by traffic volumes. However there was no clear difference in measured values between the 'non-busy' and 'busy' roadside categories, and in the absence of specific traffic flow data, it is not possible to examine this relationship in detail. The lack of any clear relationship may be partly attributable to differences in local site factors, such as sampler height and local meteorological conditions.
   
   6.3   Influence of Road Traffic Emissions on Forecourt Benzene Levels
   Road traffic emissions will contribute to measured benzene levels at the front of the petrol station forecourt. This influence will decline with increasing distance from the kerbside, and data from previous monitoring studies suggests that there will be little effect beyond a distance of approximately 20 metres.
   
   Owing to the influence of a variety of factors on levels recorded at the roadside sites and the subsequent variability in the data, it is difficult to estimate the contribution of road traffic emissions to measured forecourt levels. Furthermore, in many cases, site layout dictated the positioning of the samplers with the result that the forecourt sites and roadside sites were not located at identical distances from the kerb. However a very crude estimate can be made of the contribution to levels recorded at Sites 2 and 3 (located at the sides of the forecourt) by comparing the mean for these two sites with the mean value for the two adjacent roadside sites. The complex configurations for Sites K and L meant that such that a simple comparison was not valid, and have not been included in this assessment.
   The data from Site 2 at stations C and H have been ignored as the traffic volumes past Sites 2 and 3 differed significantly owing to the presence of road junctions. Furthermore contributions from G and H are not included as the % contribution was a negative figure.
   
   Table 11: Estimated Contribution of Roadside Emissions to Forecourt Benzene Levels
   

   Station	Controls Fitted	Road Type	Roadside Mean (Sites 5 and 6)	Forecourt Mean (Sites 2 and 3)	% Contribution of Roadside Emissions
   A	None	Non-Busy	1.7	5	34%
   B	None	Non-Busy	0.1	1.3	8%
   C	None	Busy	0.8	1.5 *Site 3 only	27%
   D	None	Busy	1.7	3.2	53%
   E	Stage I	Non-Busy	2.7	2.8	96%
   F	Stage I	Non-Busy	0.4	1.3	31%
   G	Stage I	Busy	2.6	0.7	N/A
   H	Stage I	Busy	1.2	0.8 *Site 3 only	N/A
   I	Stage I	Busy	1.2	2.1	57%
   J	Stage I	Busy	1.8	4.3	42%

   Note: A value of 1.5 ppb has been subtracted from all the above figures to take account of the ambient background concentration
   
   Table 11 indicates that the estimated contribution of roadside emissions showed considerable variation between the stations, ranging from 8% at Station B (no controls, non-busy road) to 96% at Station E (Stage I, non-busy). Such variation may partly reflect differences in the diffusion tube locations relative to emission sources between the stations. Given this fact and the crude manner in which the calculations have been made it is difficult to examine the data set for any differences in the contribution of roadside emissions according to station type. Calculation of the mean contribution of roadside emissions at sites according to station type indicates that the estimated contribution was slightly higher at sites at which Stage I controls were fitted; the mean contribution at sites with no controls by non-busy roads and busy roads and at sites with Stage I fitted by non-busy and busy roads was 21%, 40%, 63% and 50% respectively. A higher contribution of roadside emissions to forecourt levels at sites at which Stage I controls are fitted may reflect observations that Stage I controls result in a reduction in benzene emissions to the forecourt area.
   Across all ten stations at which calculations were made roadside emissions contributed approximately 44% to benzene levels measured at the front of the petrol station forecourt. This is comparable with the results of the CONCAWE preliminary study which indicated that the contributions from traffic and forecourt activities to measured levels at the petrol station boundary were approximately equal. The CONCAWE year long study, which followed the preliminary study, indicated, unsurprisingly, that the influence of road traffic emissions on benzene levels at petrol stations was less in semi-rural locations, contributing only 15% to the overall mean concentration at the service station.
   
   6.4   Comparison with the Air Quality Standard
   A strict comparison of the data with the AQS recommended in the UK National Air Quality Strategy AQS cannot be made as the measurements recorded in this study only represent twelve weeks sampling. The mean values can thus only be used to provide an estimate of any exceedances of the AQS.
   Whilst the AQS of 5 ppb was exceeded at nine of the 12 stations, exceedances were only measured at locations remote from the immediate vicinity of the pumps at five stations. Personal exposure to concentrations at the pumps will only occur over a relatively short period during refuelling; exceedances of the AQS at these sites are not as critical as at those locations where longer term exposure is likely. Exceedances of the AQS were estimated to occur at seven forecourt sites. Two of these sites were located adjacent to residential housing (Stations A and E), thus exceedances at these sites are therefore of much greater significance as residents may be exposed to such concentrations for significant amounts of time.
   The forecourt sites at which exceedances of the AQS were recorded were located at varying distances from the pumps, at Station A levels exceeded 5 ppb at the site located approximately 3 m from the pump, whilst at Stations D, J and K exceedances were recorded at sites located at 12 m from the pumps.
   
   

   ---

   Report prepared by Stanger Science and Environment
   Site prepared by the National Environmental Technology Centre, part of AEA Technology, on behalf of the UK Department of the Environment, Transport and the Regions