RISK ANALYSIS AT HIGHWAY-RAILWAY GRADE CROSSINGS USING COUNT-DATA STATISTICAL MODELING: A CASE STUDY OF THE TAIWAN RAILWAYS ADMINISTRATION SYSTEM

Abstract

A highway-railway grade crossing (HRGX) is a spatial location where highway and rail users share the same right-of-way (ROW). Despite the number of HRGXs managed by Taiwan Railways Administration (TRA) is relatively small, a significant number of traffic collisions and severe consequences at the HRGXs in Taiwan have signaled an urgent need for applying appropriate models to investigate the key factors that are associated with the casualty risk level of an HRGX. This study applies Poisson-based count-data statistical model to analyze crash risk levels in terms of number of fatalities or injuries of the investigated HRGXs in a given time period. In the empirical study, using the recent HRGX crash and inventory data provided by the TRA, the causal relationships between crash risk and a set of influence factors are analyzed and policy implications are discussed. Finally, elasticity analysis based on the calibrated model is also conducted. The empirical study results indicate that traffic exposure related variables such as daily trains and/or average highway daily vehicular traffic are the key factors that are responsible for the occurrence of a traffic collision(s) at HRGXs. Besides, highway and/or railway geometric characteristics and/or sight distance also play different roles on the risk levels of an HRGX.