Statistical Modeling of Freight Train Derailments at Highway-Rail Level Crossings

Implementation of highway-rail level crossing warning systems, educational programs and research on crossings have all contributed to a steady reduction in the risk of level crossings to North American highway users over the past several decades. Much less attention has been given to understanding the effect of level crossings on train safety and risk. Incidents at highway-rail level crossings can have serious consequences for the public and the railroads alike, especially if an incident results in a train derailment involving passengers or hazardous materials. The goal of past research has been to identify and understand the physical factors leading to these derailments. These physical factors point to a subset of level crossings with the highest derailment risk including train speed, highway vehicle speed, highway vehicle type, and incident type. The results of this paper can be used to provide railway decision makers with a metric for prioritizing level crossing upgrades and closures.