Abstract View
Improved Prediction of Near Roadside Vehicle Emissions from PEMS and Laboratory Measurements
AYLA MORETTI, David R. Cocker III, Matthew Barth, University of California, Riverside
Abstract Number: 176
Working Group: Urban Aerosols
Abstract
Vehicle emissions are measured using dynamometers (coupled with test-cell emission instruments) and/or portable emissions measurement systems (PEMS); however, these systems operate at temperature and dilution ratios not representative of the ambient atmosphere. Estimates of near-road particulate matter (PM) concentrations using these emission factors (EF) within emission models, such as the EPA’s Motor Vehicle Emission Simulator (MOVES), are not in agreement with measured near-road PM concentrations. A majority of differences between the near-road ambient studies and MOVES could be due gas-particle partitioning that occurs immediately after the emissions rapidly dilute and cool in the ambient atmosphere. Gas-particle partitioning suggests that we need a better way to predict roadside emissions by extrapolating from PEMS and dynamometer-based measurements.
This research uses published volatility basis set (VBS) data coupled with new experimental data-using a variable residence and dilution tunnel connected to an engine. Using the VBS approach, the gas-particle partitioning of OA from a gasoline vehicle was modeled using Python to create a correction factor that can work with the outputted MOVES EF to correct for primary PM2.5 from gasoline vehicles. This correction factor helps to bridge the gap between regulatory model estimations and what is measured near-roadways. Results indicate that, as suspected, the gas-particle partitioning plays a major role in final PM levels present in the atmosphere due to vehicle exhaust. This research explores sensitivity of dilution and temperature and shows that there is a bias in predicted roadside PM using the current transportation models.