Mixed-Model Statistical Analysis of Fuel, Equipment, Mileage, and Driving Schedule Effects on Particulate Matter Emissions from Heavy Diesel-Powered Vehicles

Abstract

An extensive experimental program has been conducted to evaluate the comparative effects of California Air Resources Board diesel fuel and an ultra-low-sulfur (S) diesel (with and without aftermarket passive filtering devices) on mass emissions of particulate matter (PM) in heavy vehicles. Tests have been performed on 20 Class 8 trucks at two high-mileage levels using two different driving schedules. The design of the test program facilitates the use of mixed-model statistical analysis, which allows more appropriate treatment of the explanatory variables than normally achieved. The analysis suggests that the ultra-low-S diesel fuel yields extremely low mean PM emissions when used in conjunction with a particulate filter, even at high mileage, but that the estimates are highly variable. The high degree of uncertainty, caused at least in part by large vehicle-to-vehicle variation, may obscure the true PM response and adversely impact attainment of increasingly more stringent diesel PM emissions standards in the United States.