Developing Vehicle Classification Inputs for Project-Level MOVES Analysis

Abstract

The motor vehicle emission simulator (MOVES) model is the primary regulatory model for estimating automobile emissions in the United States. The model requires refined input data; otherwise, internal model assumptions that are not necessarily representative of the project being modeled can dominate the outputs. For example, project-level on-road fleet composition is highly dependent on local vehicle use; hence, MOVES default inputs and regional distributions are not likely to apply (and MOVES estimates for project-level analyses are especially sensitive to vehicle source type distribution). Unfortunately, developing project-level source type distributions can be challenging for model users. This research proposes a procedure for developing MOVES vehicle source type distribution inputs that uses the FHWA vehicle classification scheme, Environmental Protection Agency certification data, state registration data, along with on-road license plate and video data. A case study of I-85 near Atlanta, Georgia, is presented to illustrate the importance of distinguishing within light-duty vehicle classes for hydrocarbon and carbon monoxide estimations, and between the single-unit heavy-duty truck (HDT) and combination HDT classes for nitrogen oxide and particulate matter estimation. The analysis suggests that the most important work is to generate on-road distributions of HDTs with respect to single-unit and combination trucks rather than to use regional defaults. The case study results show the need for locally derived vehicle class inputs for MOVES for project-level analysis and calls for an alternative MOVES vehicle class input option that uses regulatory class distributions because the default vehicle class distribution embedded in MOVES may sometimes be unrealistic.