Info

FIGURE 16.27 Measured ratio of VOC to NOt in the Los Angeles area during the 1987 Southern California Air Quality Study (SCAQS) to that calculated from emission inventories. (Data from Fujita et al., 1992.)

Figure 16.27 shows the results of one such comparison for the Los Angeles area during a comprehensive field campaign in 1987 when extensive chemical, meteorological, and emissions measurements were made (Fujita et al., 1992). The VOC/NOx ratio measured in both the summer and fall was approximately a factor of two larger than predicted from the emissions inventories for this region. A similar discrepancy was observed for the CO/NOj ratio. These discrepancies are not unique to Los Angeles. For example, Fig. 16.28 shows a similar comparison between measured values of this ratio and that calculated from the emissions inventories for a number of urban areas in the United States. Again, the observed VOC/NOx ratio was much higher than expected.

Clearly, such results suggest that either VOC emissions are larger than the estimates, NOx emissions are lower, or both. It is now known that it is primarily increased VOC emissions that are responsible for this discrepancy and that it is to a large extent the mobile source emissions that had been underestimated. This conclusion comes primarily from tunnel studies in which measurements of VOC, CO, and NOx concentrations are made in highway tunnels and compared to the calculated concentrations (or ratios of concentrations) predicted by mobile source emissions models. The latter have been developed at both the federal and state levels in the United States using emissions measurements on cars made under specified combinations of operating conditions and loads designed to represent an "average" driving cycle.

Figure 16.29 shows one such comparison using the 7C version of the California Air Resources Board mobile source emissions model known as EMFAC, which was in use at the time the ambient air measurements in Fig. 16.27 were made. The measured VOC/NOx ratios in the tunnel were a factor of 4.0 ± f.8 times larger than was predicted, and the CO/NOx a factor of 2.7 + 0.8 times larger. (Ratios are more accurately measured in such studies than absolute emissions, due to the need to correct for aerodynamic effects; e.g., see Rogak et al., 1998a.) As discussed by Pierson et al. (1990) and the National Research Council report (1991), the same sort of discrepancy was also observed in other tunnel studies as well as in measurements made by the side of the road that have been used to estimate emission rates. It was also shown in such tunnel studies that the NOx levels varied from 60 to f40% of those predicted by the emissions models. This suggests that it is primarily the VOC and CO that are being underestimated.

Since then, mobile source emissions models have been revised to take these increased emissions into account. As expected, the agreement between the observed and predicted concentration ratios is improved in most cases (e.g., see Gertler and Pierson, 1994; Gertler et al., 1996; Robinson et al., 1996; Pierson et al., 1996; Kirchstetter et al., 1996; McLaren et al., 1996b; and Pollack et al., 1998).

Similarly, increasing the on-road exhaust vehicle emissions by a factor of three in a Eulerian model was found to improve the agreement between predicted and observed concentrations of 03 and organics in the

St. Louis

Philadelphia

New York

Atlanta

Santa Barbara / Ventura

Dallas / Ft. Worth

St. Louis

Philadelphia

New York

Atlanta

Santa Barbara / Ventura

Dallas / Ft. Worth

Los Angeles

VOC/NOx | | Emissions inventory Ambient measurements

FIGURE 16.28 Comparison of VOC/NO, from measurements in ambient air to those calculated from emission inventories for some cities in the United States (adapted from National Research Council, 1991, based on data in Morris, 1990).

Los Angeles

VOC/NOx | | Emissions inventory Ambient measurements

FIGURE 16.28 Comparison of VOC/NO, from measurements in ambient air to those calculated from emission inventories for some cities in the United States (adapted from National Research Council, 1991, based on data in Morris, 1990).

Los Angeles area (Harley et al., 1993a, 1993b; Harley and Cass, 1995). This is not surprising in this case since emissions associated with automobile use are estimated to be a large fraction of the total organic emissions. Figure f6.30, for example, shows the relative contribu-

| Engine exhaust R^ Whole gasoline Gasoline vapors I I Waste & natural gas 0 Degreasing solvents

Ambient average

Inventory

FIGURE 16.30 Relative source contributions to VOC from a chemical mass balance model and from an emissions inventory (adapted from Harley et al., 1992). Note that this does not include all sources of VOC in this area.

| Engine exhaust R^ Whole gasoline Gasoline vapors I I Waste & natural gas 0 Degreasing solvents

Ambient average

Inventory

FIGURE 16.30 Relative source contributions to VOC from a chemical mass balance model and from an emissions inventory (adapted from Harley et al., 1992). Note that this does not include all sources of VOC in this area.

tions of engine exhaust, whole gasoline, gasoline vapors, waste and natural gas, and degreasing solvents to the total VOC estimated using a chemical mass balance model (Harley et al., 1992), compared to that in the basinwide emissions inventory; it should be noted that not included in these classifications are contributions from biogenic organics (which may be quite substantial in this area; e.g., see Harley and Cass, 1995), domestic solvent use, or surface-coating activities. However, it is clear from Fig. f6.30 that for the five particular source categories shown, the combination of organics from engine exhaust, whole gasoline, and gasoline vapors associated with the use of motor vehicles forms a very large portion, about 80%, with exhaust alone accounting for approximately a third of these five categories. Hence it is not surprising that increasing the vehicle emissions brings the models into better agreement with observations.

There are a variety of potential sources of this historical underestimation of VOC and CO mobile

VOC NOv

Was this article helpful?

0 0

Post a comment