3.4 Analysis of Daily Roadside Enhancement of concentration for Benzene and 1,3-Butadiene
As NO_x has been well characterised in the past, and is measured in numerous locations, the "roadside enhancement" of benzene and 1,3-butadiene were compared with similar NO_x measurements.

The daily mean roadside enhancement of the NO_x, benzene and 1,3-butadiene concentrations have been calculated for the Marylebone Road monitoring site for the period January to July 1998. Daily mean background NO_x concentrations from the London Bloomsbury monitoring site and daily mean background benzene and 1,3-butadiene concentrations from the London UCL monitoring site were subtracted to calculate this roadside enhancement. The scatter plots (Figures 3.2 and 3.3) show the excellent correlation between the enhanced roadside concentration of these three pollutants.

Benzene to NO_x and 1,3-butadiene to NO_x ratios from both measurements and road link emissions estimates for Marylebone Road were then calculated as follows (with concentrations expressed in µg m-3):-

Roadside enhancement of Benzene	= cBz/NOx	= 0.041
Roadside enhancement of NOx

Benzene emissions from road link	= qBz/NOx	= 0.020
NOx emissions from road link

Comparison of the benzene:NO_x emissions (qBz/NOx) and benzene:NO_x roadside enhancement (cBz/NOx) indicates that for the same estimated emission, the benzene roadside enhancement is approximately twice that of NO_x (i.e. c_Bz/NOx » 2.q_Bz/NOx). So, the coefficient for benzene roadside enhancement kBz will be approximately twice that for NO_x (i.e. k_Bz » 2.k_NOx), which is as observed: k_NOx = 5.73 and k_Bz = 11.5 (from section 3.1 and section 3.3).

For 1,3-butadiene the situation is somewhat different. Due to the limited number of roadside 1,3-butadiene measurement sites it has not been possible to determine k₁₃ directly. The alternative method (Method 2) using only the Marylebone Road site has been used. k₁₃ is determined by using the relationship between the roadside enhancement of 1,3-butadiene and NO_x derived from the Marylebone Road data. From previous definitions it can be shown that:

k13 =	c13/NOx	.kNOx = 7.5
	q13/NOx

Evidently there is better agreement between the concentration and emissions ratios for 1,3-butadiene and NO_x than for benzene and NO_x.


3.4.1 Factors Affecting the Benzene:NO_x Road Link Emissions Ratio and Benzene:NO_x Roadside Enhancement
There are several possible explanations for the observed differences between c_Bz/NOx and qBz/NOx -i.e. the observed difference between the benzene to NO_x ratio of roadside enhancement, and the ratio of benzene to NO_x road link emissions. These originate from several assumptions made in the methodology for estimating the road link emissions which may not be entirely appropriate for the Marylebone Road site.

First, and thought to be the most important, the benzene to NO_x ratio of emissions from road vehicles (q_Bz/NOx) is strongly speed dependent. Currently, the NAEI methodology assumes an average speed of 55 km h-1 (34 miles h-1), which is the default value for a dual carriageway with a speed limit of 40 miles h-1. It is thought that this is an overestimate of the average vehicle speed. Figure 3.4 illustrates the speed dependence of the benzene to NO_x emissions (q_Bz/NOx), and the values of q_Bz/NOx and c_Bz/NOx are indicated on this plot.

The value of c_Bz/NOx suggests an average vehicle speed of 15 to 20 km h-1, which is more consistent with the value of 11 miles h-1 (17.7 km h-1) for average traffic speed in the central London area, given in Road Transport Statistics for London (1996).

Similar comparisons for 1,3-butadiene between c_13/NOx and q_13/NOx suggest a higher average vehicle speed. The estimated vehicle speed from the value of c_13/NOx is approximately 45 kmh-1, as shown in Figure 3.5. While there is twice as much roadside enhancement of benzene concentrations than is predicted by the emissions estimates, 1,3-butadiene concentrations are closer to, though not exactly equal to, the expected values for this road link.

It is recommended that the estimates for road vehicle speeds in the London area be addressed by the NAEI in the near future. Examination of likely vehicle speeds at the monitoring sites shown in Figures 3.4 and 3.5 may help in this analysis.

In addition to the above, it should be noted that evaporative losses of benzene are not included in the road link emissions data. The evaporative losses of benzene contribute 7% to the total vehicle emissions, when expressed as a national average. Also, it has recently been suggested that prolonged engine idling can reduce the efficiency of some types of catalytic converters. This would increase the emissions of benzene relative to NO_x, but at present it is not possible to determine whether this impact on the benzene to NO_x emission ratio would be significant.