Composition and Reactivity of Fuel Vapor Emissions from Gasoline-Oxygenate Blends 912429

Laboratory bench tests were conducted to simulate both the diurnal and the hot-soak (carburetor only) parts of an evaporative emissions test with gasolines containing various alcohols and ethers. The mass of vapor generated during each test and the detailed composition of the vapor were determined for each fuel. Using published atmospheric reactivity scales, the ozone-forming potential of the vapor was estimated. Based on the scale of maximum incremental reactivities, which the California Air Resources Board has proposed for future emissions standards, the diurnal test results showed that the addition of methyl tertiary-butyl ether (MTBE) or ethyl tertiary-butyl ether (ETBE) to gasoline resulted in about the same or lower total vapor reactivity compared to the gasoline alone. A blend of gasoline with ethanol or with a methanol-ethanol mixture also produced diurnal vapor with lower total reactivity, but only when the vapor pressures of the gasoline-oxygenate blends did not exceed that of the gasoline. At equal oxygenate concentrations in the fuel, MTBE was the most effective in reducing the reactivity per gram of vapor.

In the carburetor hot-soak tests, the total reactivity of the vapor was higher for all of the gasoline-oxygenate blends than for gasoline alone because the blends generated more vapor. However, the reactivity per gram of gasoline-oxygenate blend vapor was similar to or lower than that for gasoline.

CERTAIN OXYGENATES (alcohols and ethers) are required in gasoline in six local areas of the U. S. in the winter to help reduce vehicle carbon monoxide (CO) emissions [1]*. Although those requirements have been imposed by state and local governments, the federal Clean Air Act amendments of 1990 will require oxygenated fuels in 41 CO nonattainment areas (based on 1988-1989 data) beginning in 1992 and in at least the nine most severe ozone nonattainment areas beginning in 1995. Oxygenates help reduce vehicle CO and hydrocarbon (HC) emissions by effectively leaning the air-fuel mixture. This effect is very pronounced in older cars without closed-loop fuel metering systems, but it is also observed in late-model cars as well [2].

Ethanol is the most widely used alcohol in the U.S., with blends of approximately 10 percent ethanol in gasoline accounting for about 6 to 7 percent of the total gasoline market. Methyl tertiary-butyl ether (MTBE) is present in an estimated 15 to 20 percent of U.S. gasolines, at concentrations between 2 and 15 percent [3]. Other alcohols and ethers are not used to any significant extent at this time. However, Clean Air Act waivers granted by the U.S. Environmental Protection Agency (EPA) allow the use of up to 5 percent methanol (along with required cosolvents and corrosion inhibitors) or up to 16 percent tertiary-butyl alcohol. Other alcohols, and aliphatic ethers such as ethyl tertiary-butyl ether (ETBE), may be added to gasoline under the “substantially similar” definition [4] to provide up to 2.7 mass-percent oxygen in the final fuel. (This is equivalent to about 17 volume-percent ETBE.) With the much greater demand for oxygenates anticipated over the next five years, oxygenates other than ethanol and MTBE may find a place in the gasoline market.

In a recent study [5], we compared the effects of various oxygenates on the volatility characteristics (vapor pressure, distillation, and vapor-liquid ratio) of gasoline, with particular emphasis on ETBE. The addition of ETBE to gasoline reduced the vapor pressure of the fuel, making ETBE a valuable blending component where reduced fuel volatility is required. (Other low-volatility ethers such as methyl tertiary-amyl ether or ethyl tertiary-amyl ether would also significantly reduce gasoline vapor pressure.) When MTBE was added to a 9-psi vapor pressure gasoline, it had little effect on vapor pressure. Ten volume-percent ethanol increased fuel vapor pressure by about 1 psi.

That study [5] also included the measurement of the mass of fuel vapor which was generated from each gasoline-oxygenate blend during both the diurnal and the hot soak parts of the current evaporative emissions test. (The relationship of diurnal and hot-soak emissions to total vapor emissions will be discussed later.) In the diurnal test, the fuel in a fuel tank is heated from 15.6° to 28.g°C (60° to 84°F) in one hour. Because diurnal vapor generation is a function of fuel vapor pressure at the test temperature, gasoline-ETBE blends produced less vapor than the gasoline alone. When the vapor pressures (specifically, the dry vapor pressure equivalent, DVPE, which will be defined and discussed later) of the gasoline-ETBE blends were adjusted to match that of the gasoline, the mass of diurnal vapor produced was the same for the blend and the gasoline. Gasoline-MTBE blends produced about the same amount of vapor as the gasoline, while a gasoline-ethanol blend without vapor pressure adjustment produced more vapor.

In our one-hour hot-soak test [5], only the vaporization of fuel in a carburetor fuel bowl was considered. The maximum fuel temperature during the hot soak was 75.6°C, and the mass of vapor generated was a linear function of the percent of fuel distilled at that temperature in the standard ASTM D 86 distillation test.

The California Air Resources Board (CARB) has proposed reactivity-based exhaust emissions standards beginning in the mid-1990s [6]. With this system, the individual organic species in vehicle emissions are identified and quantified chromatographically, and a reactivity factor is applied to each species based on its reactivity in the atmosphere to form ozone (smog). The results are summed to obtain a total reactivity for the emissions. Thus, emissions from various alternative fuels and various vehicle technologies can be compared on a common basis. Although CARB's proposal does not presently include evaporative emissions, this requirement could be added in the future.

With this interest in emissions reactivity, we have extended our previous study [5] to compare the reactivity of fuel vapor generated from various gasoline-oxygenate blends during laboratory bench-test simulations of an evaporative emissions test. This report describes the results of those tests.