Quantitative Analysis of the Derailment Characteristics of Loaded and Empty Unit Trains

Operation of unit trains has grown substantially over the past half century owing to their ability to provide economic and efficient transportation of bulk products. Although various aspects of train safety have been studied, there has been limited research examining the effect of train loading conditions on derailment occurrence, cause and severity. An algorithm was developed to identify derailments of loaded and empty unit trains on mainlines and sidings recorded in the Federal Railroad Administration database. A dataset of these accidents for the 15-year period of 2001–2015 was developed and analyzed. The frequency of derailments for both loaded and empty unit trains declined by more than 50%. The average number of cars derailed per accident fluctuated for both loading conditions, but showed no particular trend. Approximately five times more loaded unit train derailments were recorded in the database than empty unit trains, but in the absence of specific unit train traffic data, inferences about rates are not possible. Loaded unit trains were more than four times heavier than empty unit trains and loaded train derailments tended to involve more cars than empty trains. The distribution of derailment causes differed for loaded and empty unit trains. Loaded trains most frequently derailed because of broken rails and welds, while the leading cause of empty train derailments was obstructions, in particular severe weather. Over 90% of the derailments of loaded and empty unit trains considered in this study occurred on mainline tracks, and the distribution of causes differed between mainline and siding tracks.