Estiamtion of Local Fleet Characteristics Data for Improved Emission Inventory Development

摘要

Considerable effort in recent years has been focused on the improvement of on-road mobile source emission factors with much less attention paid to the refinement of activity and fleet characteristics estiamtes. Current emissiosn modeling practices commonly use emission factor model defaults or statewide averages for fleet and activity data. As part of the U.S. EPA's Emission Inventory Improvement Program (EIIP), ENVIRON develoepd methodologies to derive locality-specific fleet characteristics data from existing data sources in order to improve local emission inventory estimates. Data sources examined included remote sensing studies and inspection and maintenance (I/M) program data. IN this paper, we focus on two specific examples: (1) the calculation of mileage accumulation rates from Arizona I/M program data, and (2) the calculation of registration distribution from a Sacramento remote sensing database. In both examples, differences exist between the calcualted distributions and those currently used for air quality modeling, resulting in significant impacts on the estiamted mobile source emissions inventory. For example, use of the automobile registration distriubtion data derived from the Sacramento Pilot I/M Program remote sensing database results in an increase in estimated automobile TOG, CO and NO_x of 15, 24 and 17 percent, respectively, when used in place of the default registration distribution in the current California Air Resources Board MVE17G emissions model.