Forecasting Class I Railroad Fuel Consumption by Train Type

Abstract

Fuel efficiency is an important consideration in evaluating public-sector investments in multimodal corridors. Two approaches are typically used in corridor studies to forecast railroad diesel fuel consumption: (1) system-average efficiency factors, and (2) detailed analytical estimates derived from train performance calculators. The former method is easy to apply but may not be reflective of the actual mix of trains used. The second method is data- and time-intensive and can only be effectively implemented with specialized software that is not publicly available. An intermediate method is introduced in this paper which allows distinctions among the types of trains that might be utilized in a corridor (e.g., unit, way, and through). A model is estimated from publicly available data that has excellent statistical properties and quantifies the absolute and relative fuel efficiencies of train options. The analysis demonstrates that using system-average fuel consumption factors may significantly understate railroad fuel economy when traffic moves in unit trains instead of mixed train service. The new method offers greater accuracy than system-average comparisons yet is much less data-intensive than train performance calculators or analytical methods.