Abstract
Air concentrations were obtained for 55 individual chemical components of gasoline measured by eight petroleum companies for service station attendants (n = 49), transport drivers (n = 49), and outside operators (n = 56) during the summer of 1984. Approximate 95 percent confidence intervals were computed for the mean exposures to each of the 55 components in the three job groups. Exposures of service station attendants were significantly lower when vapor recovery systems were present; exposures in the other two job groups were apparently unaffected by vapor recovery systems. The ratios of air concentrations of the 16 most abundant vapor components to the air concentrations of total hydrocarbons (THC) comprised 80-84 percent. It was observed that the vapor ratios calculated in this investigation were similar to those of the same 16 components reported in a previous investigation by McDermott and Killiany in 1978. This apparent stability of vapor ratios suggests the possibility that exposures of individual gasoline components might be estimated solely on the basis of THC measurements. This hypothesis was tested by comparing predicted air levels for 23 hydrocarbons with those actually observed in a recent investigation by Tironi et al (1986) and relatively good agreement was observed.
