This paper investigates the influence of rail age, annual traffic density, and inspection frequency on broken-rail-caused train derailment risk. First, we estimate the probability of a broken-rail-caused train derailment based on a sequence of stochastic processes including rail defect formation, growth, detection and the likelihood that a broken rail causes a derailment. In addition to derailment frequency, we also estimate derailment severity, which is measured by the average number of railcars derailed per train derailment, based on FRA-reportable train derailment data. The preliminary risk analysis model provides a quantitative approach to understand broken rail risk, and potentially aid in development of effective ways to mitigate derailment risk.

This content is only available via PDF.
You do not currently have access to this content.