Correlation Between Speciated Hydrocarbon Emissions and Flame Ionization Detector Response for Gasoline/Alcohol Blends

Abstract

The U.S. renewable fuel standard has made it a requirement to increase the production of ethanol and advanced biofuels to 36 billion by 2022. Ethanol will be capped at 15 billion, which leaves 21 billion to come from other sources such as butanol. Butanol has a higher energy density and lower affinity for water than ethanol. Moreover, alcohol fueled engines in general have been shown to positively affect engine-out emissions of oxides of nitrogen and carbon monoxide compared with their gasoline fueled counterparts. In light of these developments, the variety and blend levels of oxygenated constituents is likely to increase in the foreseeable future. The effect on engine-out emissions for total hydrocarbons is less clear due to the relative insensitivity of the flame ionization detector (FID) toward alcohols and aldehydes. It is well documented that hydrocarbon (HC) measurement using a conventional FID in the presence of oxygenates in the engine exhaust stream can lead to a misinterpretation of HC emissions trends for alcohol fuel blends. Characterization of the exhaust stream for all expected hydrocarbon constituents is required to accurately determine the actual concentration of unburned fuel components in the exhaust. In addition to a conventional exhaust emissions bench, this characterization requires supplementary instrumentation capable of hydrocarbon speciation and response factor independent quantification. Although required for certification testing, this sort of instrumentation is not yet widely available in engine development facilities. Therefore, an attempt is made to empirically determine FID correction factors for oxygenate fuels. Exhaust emissions of an engine fueled with several blends of gasoline and ethanol, n-butanol and iso-Butanol were characterized using both a conventional FID and a Fourier transform infrared. Based on these results, a response factor predicting the actual hydrocarbon emissions based solely on FID results as a function of alcohol type and content is presented. Finally, the correlation derived from data presented in this study is compared with equations and results found in the literature.