Quantification of gasoline-ethanol blend emissions effects

Abstract

ABSTRACT Emissions levels from current gasoline spark-ignited engines are low, and emissions changes associated with the blending of ethanol into gasoline are small and difficult to quantify. Addition of ethanol, with a high blending octane number, allows a reduction in aromatics in market gasoline. Blending behavior of ethanol is nonlinear, altering the distillation curve, including the 50% temperature point, T50. Increase in gasoline direct injection (GDI) engine technology in the fleet challenges ability of older models based on port fuel injection (PFI) results to predict the overall air quality impact of ethanol blending. Five different models derived from data collected through U.S. Environmental Protection Agency Energy Policy Act (EPAct) programs were used to predict LA92 Phase 1 particulate matter (PM) emissions for summer regular (SR) E0 (gasoline with 0% ethanol by volume), E10 (gasoline with 10% ethanol) and E15 (gasoline with 15% ethanol). Substantial reductions of PM for E10 and E15 relative to E0 were predicted when aromatics were displaced by ethanol to maintain octane rating. SR E0 and E10 were also matched to linear combinations of EPAct fuels and results showed a 35% PM reduction for SR E10 relative to SR E0. For GDI vehicles the Coordinating Research Council (CRC) E-94-3 study found that E10 had 23% or 29% PM increase. However, CRC E-129 found an E10 PM reduction of 10% when one E0 fuel and its splash blended (SB) E10 were compared. Both CRC project E-129 SB data and fuel triplets selected from the EPAct study showed variation for E15 emissions, although E-129 suggests that E15 in GDI offers about a 25% reduction of PM with respect to E0. Overall, data suggest that ethanol blending offers a modest to a substantial reduction of cold-start PM mass if aromatic levels of the finished products are reduced in response to ethanol addition. Implications: Studies of exhaust emissions effects of ethanol blending with gasoline vary in conclusions. Blending properties are nonlinear. Modeling of real-world emissions effects must consider all fuel composition adjustments and property changes associated with ethanol addition. Aromatics are reduced in E10 or E15, compared with E0, and distillation changes. PFI-derived models show reductions in cold-start PM for expected average E10 versus E0 pump fuel, due to reduced aromatic content. Relative emissions effects from older technology (PFI) engines do not predict newer engine (GDI) results reliably, but recent GDI data show reduced cold-start PM when ethanol displaces aromatics.