Rail is a safe and efficient mode of transporting hazardous materials (hazmat). In the past decade, the hazmat traffic transported by unit trains has significantly increased in the United States. As a result, a comprehensive understanding of the safety and risk of hazmat unit trains is important and can contribute to the identification, evaluation, and implementation of risk mitigation strategies. Limited prior research has focused on unit train derailment risk analysis. This paper develops a quantitative analysis of freight unit train derailment characteristics and compares those statistics to non-unit, manifest trains (mixed trains). Mainline freight train derailment data on Class I railroads between 1996 and 2018 were analyzed for hazmat unit trains, non-hazmat unit trains, and manifest trains. Derailment rates, measured by three traffic exposure metrics (train-miles, ton-miles, and car-miles) were estimated and compared. The analyses showed that a unit train has a 30% lower derailment rate in terms of ton-miles and car-miles than manifest trains, while the derailment rate per million train-miles of unit trains is slightly greater than that of manifest trains. Loaded unit trains have roughly four-fold higher derailment rate in terms of train-miles and car-miles than that of empty unit trains. Within loaded unit trains, hazmat unit trains have lower derailment rates than non-hazmat unit trains. Overall, heavier and shorter loaded unit trains tend to have greater derailment rates in terms of all three traffic exposure metrics. A causal analysis was also conducted for the three types of train. Infrastructure causes were the most frequent in all train types and length followed by equipment-related causes. These statistics provided important information for rational allocation of risk mitigation resources to improve rail hazmat transportation safety.

The need for shared freight and passenger rail corridors in the United States is increasing due to the growing demand for regional and intercity passenger transport. Several researches have been conducted on reducing the risk of freight train accidents, but little research has been done on the risk of passenger train accidents. The accident rates of passenger trains have declined in the past two decades; however, faster and more frequent passenger train services require even higher safety standards, and therefore further reduction to the risk of passenger train accidents is needed. The research presented in this paper analyzed the passenger train accidents in the United States using the Federal Railroad Administration train accident database to understand the trend of passenger train accident rates, the frequency and severity of different types of accidents, and to explore the major factors that cause them. Derailments and collisions were identified as the most significant types of passenger train accidents, and track failures and human factors, respectively, were the primary causes of those accidents. Accidents caused due to human factors and train operations such as train speed violations and failure to obey signals are often high-consequence accidents and therefore pose the greatest risk. Higher risk infrastructure-related factors include track geometry defects and broken rails or welds. This study on passenger train accidents provides a solid foundation for further research on improving the safety of passenger rail and shared-use rail corridors.

Adjacent-track accidents (ATAs) have been identified as an important hazard on shared-use rail corridors. In these train accidents derailed railroad equipment intrudes on (fouls) an adjacent track, disrupts operations, and potentially causes a collision with trains operating on the fouled tracks. Derailments without intrusion may cause equipment and infrastructure damage, passenger casualties, and disturbances to system operation; however, an intrusion may be even more severe because of the potential involvement of multiple trains. Opportunities for ATAs have increased in recent years because of expanded passenger and transit service on freight railroad trackage, right-of-way, and corridors and because of increased multiple tracks related to capacity expansion projects. This paper presents a probabilistic risk assessment methodology for analyzing ATA risk. An event tree is created to identify scenarios for ATAs, and a fault tree analysis is performed to identify basic events that contribute to such accidents. The quantitative probability of an ATA is derived by using Boolean algebra on the basis of the results of the fault tree analysis.

No abstract