5. Results
   5.1.   Diffusive Sampling
   The results of the diffusive sampling measurements for each site are shown summarised in Table 7. For each service station, the following site descriptions have been used:
   
   

   Site 1:	Rear of forecourt
   Site 2:	Side of forecourt
   Site 3:	Side of forecourt
   Site 4:	Pumps
   Site 5:	Roadside (100 metres distance)
   Site 6:	Roadside (100 metres distance)

   
   Sites 1, 2 and 3 were located at varying distances from the petrol pumps along the forecourt perimeter, reflecting differences in forecourt size. Approximate distances from pumps for each of the three forecourt sites are given in Table 7.
   
   There was considerable variation in the concentrations measured at each location from one period to the next. This is consistent with other monitoring studies³ and is likely to reflect the influence of prevailing meteorological conditions. The highest mean concentration measured (19.1 ppb) was at the pumps (Site 4) at Station J; the lowest mean concentration (1.2 ppb) was measured at a roadside site (Site 5) at Station L.
   The individual two-weekly average concentrations for each site are provided in Appendix 3.
   
   5.1.1   Comparison with the Air Quality Standard
   A comparison between mean benzene concentrations recorded at each site with the Air Quality Standard of 5 ppb recommended in the Consultation Draft of the UK NAQS can be made by extrapolating the twelve weeks monitoring data to an annual average. Extrapolation in this manner does not allow a strict comparison with the AQS but does allow estimates of exceedances to be made. Such a comparison indicates that exceedances of the standard of 5 ppb were estimated to occur at nine of the 12 service stations, at a total of 15 monitoring locations. Of these nine stations, exceedances were only estimated to occur at locations remote from the fuel pumps at seven sites (Site 3, Station A; Site 1, Station D, Site 2, Station E; Sites 1,2 and 3 at Station K and Site 3 at Station J). There were no estimated exceedances of the AQS at any of the roadside monitoring sites.
   
   5.1.2.   Comparison of Site Categories
   A comparison of concentrations measured at the roadside, forecourt and fuel pumps sites at each service station is shown in Figure 1. For the purpose of this assessment the concentration at each sampling location has been determined from the mean of Sites 5 and 6 (roadside), Sites 1, 2 and 3 (forecourt) and Site 4 (pumps). Figure 1 clearly shows that mean concentrations were generally lowest at the roadside sites and highest at the sites located at the pumps, with intermediate concentrations recorded at the forecourt sites. At the roadside sites mean values ranged from 1.6 ppb (Station B, non-busy road with no controls) to 4.2 ppb (Station E, non-busy road with Stage I controls); at the forecourt sites values ranged from 1.8 ppb (Station L, a superstore site with Stage 2 controls) to 6.4 ppb (Station K, a superstore site with Stage I controls), and at the pumps values ranged from 2.7 ppb (Station B, non-busy road with no controls) to 19.1 ppb (Station J, busy road with Stage I controls).
   
   5.1.3   Spatial Variation at Individual Stations
   Table 7 indicates that at a number of the petrol stations there was considerable spatial variation in mean levels recorded at the forecourt sites, and no consistent pattern was identified. For example at Station A, mean levels at the forecourt ranged from 4.8 ppb (Site 1) to 8.0 ppb (Site 3); similarly at Station E, mean levels ranged from 2.5 ppb (Site 1) to 6.0 ppb (Site 2). At the four stations (A to D) where no Stage I vapour recovery systems were installed, the highest forecourt values were not necessarily measured at the sites located closest to the delivery tanks. Furthermore, at stations fitted with Stage I controls, the highest forecourt values were sometimes recorded at sites located next to the delivery tanks for example at Station J.
   
   At a number of stations there was also considerable variation in mean values recorded at the two roadside sites. For example at Station A mean values ranged from 2.2 to 4.3 ppb, whilst at other sites the roadside values were relatively similar, for example at Station H where mean values differed by only 0.3 ppb. The variation between the measured values at the two roadside sites at some petrol stations was thought to be mainly attributable to local site conditions such as proximity of sampler to road junctions (Station C) and car parking areas (Station A). In one instance, at Station L, differences between the two roadside locations reflected differences in traffic flows due to the configuration of the station; a mean of 1.2 ppb was recorded at the site located along the slip road to the station whilst a mean of 2.3 ppb was recorded at the second site on a major through road, where traffic levels were much higher.
   
   

   5.1.4   Influence of Fuel Sales Details of fuel sales at each service station during the twelve week monitoring period are shown in Table 8. There was considerable variation in total fuel sales at each station, ranging from 230,000 litres at Station C to 2,281,000 litres at Station K, a superstore outlet. Generally the total sales reflect the number of pumps at each station, with six pumps operational at Station C, rising to 36 at Station K.	Table 8 Fuel Sales (litres) for the Period 2.4.96 to 3.7.96
   	Station Grand ID Total A 338,000 B 489,000 C 230,000 D 1,727,145 E 425,000 F 1,189,261	Station Grand ID Total G 1,750,000 H 935,860 I 830,032 J 1,402,688 K 2,281,000 L 1,916,000

   
   
   5.1.5   Normalised Benzene Concentrations
   It is difficult to directly compare measured concentrations between service stations, due to the wide range of station types included in the survey (ranging from 6 to 36 pumps). If it is assumed that emissions from the fuel pumps are the principal source of benzene emissions in the forecourt area, then the source strength will be proportional to the volume of petrol sales, although it is recognised that this is a simplified approach. In order to more fully compare benzene concentrations measured within the forecourt at each service station, the data have been normalised with respect to total sales of leaded and unleaded petrol, and expressed in terms of ppt/klitres. These normalised values are illustrated in Figure 2 for the forecourt and pump sites, and shows a marked difference in the distribution with respect to Figure 1
   
   5.1.6 Effects of Stage I and II Controls
   These normalised data can be used to indicate the effects of Stage I and Stage II controls on benzene levels measured at the forecourt and pump sites. The normalised mean values for the forecourt and pump sites are described in Table 9, according to the type (if any) of controls fitted.
   
   Table 9 Normalised Mean Benzene Levels Recorded at Forecourt and Pump Locations (ppt/litre)
   

   Controls Fitted	Forecourt	Pump
   None	13.4	17
   Stage I	4.8	9.9
   Stage II	1.0	2.6

   
   The measured benzene levels at forecourt sites fitted with Stage I controls were less than half of those recorded at sites with no controls. The influence of Stage I controls in reducing forecourt concentrations can be further seen in Figure 2, which shows that the highest concentrations were recorded at Stations A and C, which have no controls fitted. Measured benzene levels at the pump sites at stations fitted with Stage I controls were approximately 40% lower than at stations where no controls were fitted.
   The effects of Stage II controls are more dramatic, resulting in a 93% reduction in forecourt benzene concentrations, and an 85% reduction in concentrations at the pumps. However, it must be emphasised that these results are based upon data collected from a single service station.
   
   5.1.7   Influence of Road Characteristics
   Figure 3 illustrates the mean concentrations for the two roadside sites adjacent to each station. There is no clear difference in concentrations between the 'non-busy' and 'busy' categories with an overall mean value of 2.7 ppb recorded for the 'non-busy' category compared to 2.9 ppb for the 'busy' roadside category.
   
   5.2   Continuous Sampling
   Concentrations of benzene monitored at Station H are described in Figure 4 as 30-minute averages over the period of sampling. The highest 30-minute concentration was 13.2 ppb, and the maximum hourly mean recorded was 9.4 ppb. An average concentration of 2.8 ppb was recorded over the four week monitoring period. Daily means ranged from 1.7 ppb to 4.5 ppb.
   
   The inlet to the benzene analyser was located approximately 2 metres from the delivery pumps. Petrol delivery dates are indicated on Figure 4 which shows no consistent relationship between delivery dates and peak concentrations.
   
   Full details of the daily, hourly and 30-minute averages are provided in Appendix 4.
   
   5.2.1   Comparison with Diffusion Tube Data
   At Station H, benzene diffusion tubes were exposed at a location approximately 2 m from the inlet to the BTX analyser (Site 1). Unfortunately concurrent data are not available, and as a result only an approximate comparison can be made between the data. The data collected using both methods is summarised in Table 10 which indicates a reasonably good correlation between the two methods.
   
   Table 10 Comparison between BTX Analyser and Diffusion Tube
   

   Method	Range (ppb)	Mean (ppb)
   Diffusion Tube (2/4/96 - 3/7/96)	1.3 - 2.2	1.7
   BTX (13/3/96 - 12/4/96)	-	2.8

---

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