| Summary: | Light-duty vehicles produce exhaust air toxic species in addition to regulated pollutants. Although numerous studies have evaluated the impacts of fuel composition on air toxics, there has been no comprehensive modeling study that considered data from all available sources. To examine the effects of real-world fuel composition on air toxics, black carbon (BC), and particle number (PN) emissions, we conducted a thorough evaluation of US emission studies, pooled data from the Federal Test Procedure (FTP) and Unified Cycle (LA92) driving schedules, and developed models based on fuel and vehicle properties. To address nonlinear blend property responses to ethanol concentration, we built separate linear models (split models) for low (up to 10% ethanol by volume) and mid blends. We then estimated differences in emissions from expected market fuel compositions. We observed differences between the operation of port fuel injection (PFI) and gasoline direct injection (GDI) engines and between test cycles and phases. Results showed that aromatics raised 1,3-butadiene, benzene, toluene, ethylbenzene, and xylene emissions, while ethanol raised acetaldehyde and lowered 1,3-butadiene. Ethanol raises octane number and enables reduction of reformate (and hence aromatics) in a blend. For aldehydes, 1,3-butadiene, and BC the effects differed between PFI and GDI vehicles. For ethanol and acetaldehyde production by PFI and for BTEX, BC, and PN from GDI vehicles, there were differences between comprehensive (i.e., single linear fit) and split models, confirming nonlinearity of the actual ethanol relationships and supporting use of separate models for ethanol blends above and below 10%. To the extent that PFI and GDI models differ, emissions inventory calculations should treat them separately. Estimated market fuel blend compositions for summer regular gasoline yielded quantitative projections of ethanol blend levels on air toxics, BC, and PN emissions. Projections were consistent for all projected fuel grades (summer regular, winter regular, and premium).
|
|---|