Development of a Predictive Model for Gasoline Vehicle Particulate Matter Emissions

The relationship between gasoline properties and vehicle particulate matter emissions was investigated, for the purpose of constructing a predictive model. Various chemical species were individually blended with an indolene base fuel, and the solid particulate number (PN) emissions from each blend were measured over the New European Driving Cycle (NEDC). The results indicated that aromatics with a high boiling point and a high double bond equivalent (DBE) value tended to produce more PN emissions. However, high boiling point components with low DBE values, such as paraffins, displayed only a minor effect on PN. Upon further analysis of the test results, it was also confirmed that low vapor pressure components correlated with high PN emissions, as might be expected based on their combustion behavior. A predictive model, termed the PM Index, was constructed based on the weight fraction, vapor pressure, and DBE value of each component in the fuel. It was confirmed that the PM Index could accurately predict not only the total PN trend but also total particulate matter (PM) mass, regardless of engine type or test cycle. A large number of gasoline samples were collected in various countries, and submitted for detailed hydrocarbon analysis (DHA). Using the resulting hydrocarbon speciation data, a PM Index distribution was calculated for each country from which fuel samples were acquired. Based on the range of PM Indices encountered, it was estimated that the highest PM Index fuel would produce 10 times the PM emissions of that of the lowest PM Index fuel. Therefore, it was concluded that worldwide PM emissions can be reduced not only through improvements in engine hardware, but also through improvements in fuel quality.