Public Review Draft

The ARB staff has presented a methodology for developing speciation for both exhaust and evaporative organic gas emissions from motor vehicles for 3 MTBE free gasolines:

Unoxygenated 2.0% oxygen ethanol 3.5% oxygen ethanol

ARB staff relied on results from earlier test programs and ARB’s Predictive Model to estimate how the speciation of organic gas emissions will change relative to the baseline MTBE gasoline and comments received by Dr. Rob Harley of UC Berkeley (see ?). A basic assumption is that all 3 MTBE free gasolines will comply with ARB’s regulations.

In addition to the organic gas species profiles developed by ARB staff for 2.0% and 3.5% oxygen ethanol gasolines, Dr. Harley has suggested that the headspace evaporative organic gas emissions profiles (used to represent diurnal evaporative emissions), developed by ARB staff for the 2 ethanol gasolines, may be too high in ethanol emissions and as a result too low in emissions of other species. Also, he suggested that the liquid gasoline composition be used as an alternate representation for hot soak vapors.

Photochemical grid model simulations were performed for both the ARB evaporative profiles as well as those recommended by Dr. Harley. Summaries of several important characteristics of the emission profiles are shown in Tables 1 through 7. (Harley’s recommended profiles are designated with a trailing “H”.) Tables 1 through 6 compare the weight percent of six selected organic gas species for all categories and gasolines used in the airshed modeling. The six species are: ethanol, benzene, formaldehyde, acetaldehyde, 1,3-butadiene, and methane.

Table 1 shows the weight percent of ethanol in the motor vehicle emission categories. Note that ARB and Dr. Harley’s estimates for hot soak and headspace vapors are very different for the two ethanol gasolines. The use of the two estimates of evaporative emissions does lead to a large range in expected ethanol emissions.

Table 2 shows the estimated benzene weight percents for the emission categories. Since there is no difference expected in the benzene content in any of the gasolines, there is not much difference in the expected benzene in any of the emission categories. Dr. Harley’s headspace profiles contain twice the benzene (0.80% vs. 0.36%) content as ARB’s estimates, but since the weight percents are very low in the headspace vapors, the overall benzene inventory will not be very different.

Tables 3 through 6 show acetaldehyde, formaldehyde, 1,3-butadiene, and methane. These compounds are not found in the gasoline nor in the evaporative emissions so only the exhaust comparisons are shown. Since acetaldehyde is a product of ethanol combustion, it is expected to be higher as the ethanol content of gasoline increases. As seen in Table 3, acetaldehyde emissions are expected to be highest for the ethanol blends.

Exhaust emissions of 1,3-butadiene, formaldehyde, and methane are expected to be similar for all 4 gasolines.

Table 7 shows the specific reactivity (SR) for all emission categories. The maximum incremental reactivity (MIR) values used to calculate the specific reactivity for each category are the same as those adopted for use in ARB’s Low Emission Vehicle program.

Note that the unoxygenated gasoline SRs are highest for all source categories. This is due to the replacement of lower reactivity oxygenates with higher reactivity alkanes or aromatics.

Figures 1 through 11 show a more complete comparison of the species profiles for each emission category. There are about 180 organic species identified if motor vehicle emissions. These figures contain 7 categories of “lumped” species (butanes, pentanes, C6+ alkanes, etc) and 11 explicit species.

Figures 1 through 3 show the profiles for the liquid gasoline, hot soak, and headspace vapors. The unoxygenated gasoline has the highest alkane emissions; the evaporative emissions are also the highest in aromatic content.

Figures 4 through 7 show how ARB’s evaporative emission profiles compare to those suggested for use by Dr. Harley. Replacing the ARB hot soak emission profiles with the liquid profile results in large speciation differences for both the 2.0% and 3.5% oxygen ethanol gasolines. The liquid gasoline has much higher alkane content then ARB’s hot soak emissions. This also results in lower content of all other gasoline components including ethanol, especially toluene and ethanol. Dr. Harley’s headspace estimated are lower in ethanol content especially for the 3.5% oxygen ethanol gasoline. The biggest change in using Dr. Harley’s profile is to reduce the amount of ethanol emissions.