Methods used to detect additional sources using Toluene/Benzene ratios in Windsor, Ontario (2004-2006)

Volatile organic compounds were measured using 3M #3500 organic vapor samples (Guillevan, Montreal) in the city of Windsor, Ontario, Canada. This monitoring study was carried out over a three year period (2004-2006) during four seasons (winter, spring, summer, fall), for a total of 12 two-week sampling periods. The sites were not in identical locations from year to year, but remained the same for each season within a given year (approximately 50 sites). Due to missing data in the summer of 2005, only 2004 and 2006 data was used. Benzene is predominantly emitted from traffic whereas toluene is generated from both traffic and solvents. The ratio between toluene and benzene (T/B) can act as an indicator of traffic emissions when this ratio is within the range of 1.5-4.3, as reported by previous studies (Hoque et al., 2008; Liu et al., 2009). For ratios >4.3, solvent source impacts are likely. The highest observed T/B ratios were in summer and the lowest in winter. In summer more than 50% of the sites had ratios above 4.3, whereas in spring and fall the number of sites decreased and in winter there were zero sites. Low winter temperatures affect the photochemistry of these species; therefore, the T/B ratio alone cannot consistently identify the solvent source influenced sites. Three alternative methods were explored. In the first method the sites were split into three groups; high, medium, and low, according to the values of the Toluene/(M+P)-Xylene (T/(M+P)) ratio. Within in each group, the T/B ratios were calculated and sites with a T/B value ≥25% higher than the perspective group T/B mean were identified as having a solvent source. In winter 2004 and 2006, 3 and 2 sites were identified respectively. The same sites were discovered in at least 2 other seasons. The second method considered the 10 sites with the highest T/B values in each season. The majority of the top ten sites reappeared in at least two seasons of a year. In the third method the T/B ratios of winter, spring, and fall, were scaled to have the lower limit match that of summer. The number of sites having scaled ratios above 4.3 for the three seasons in 2004 was within 8% of summer, whereas for 2006 they were within 50% of summer. Each of the three methods worked to a certain degree. It is suggested that seasonal variability be considered when using T/B ratios to identify traffic and solvent sources.