The Occurrence Regulation And Security Impact Study Of Liquefied Gases Railway Tank Car Release Accidents

With the rapid development of national economy,China has entered an era of heavy and chemical industries.While the annual volumes of railway hazardous goods are increasing,various types and commodities of hazardous goods have complicated properties,among which the most frequent accidents are liquefied gases railway tank car release accidents.During the transit,there exits lots of potential risks that would threat personal and property safety.Therefore,a study of the occurrence regulation and security impact study of liquefied gases tank car release accidents is needed to offer support to railway transportation safety.First,by collecting the release data of PHMSA and Chinese liquefied gases tank car,a statistical characteristic analysis and summary of the accident regulation are studied.Second,using SPSS Modeler 14.0 software to analyze the correlations among accident related factors and mining for the internal association rules of release accidents with Apriori.Third,transportation models of liquefied tank car release accident are built to calculate the system reliability parameters.Finally,a case-based reasoning to search for the most likable case is built based on PHMSA release accident data.The studies show that Jan,Jun and Oct are the high-incidence seasons while Sep is the lowest.What's more,most of the liquefied railway tank car releases happened in transit and resulted in low leakage.Chlorine,Ammonia anhydrous and LPG are key commodities.Valves are the most frequent accident settings with piping and fittings occupied the second place.Usually tank car settings failed with loose components in a release.The release accidents can be classified into NAR and AR accidents,and the safety factors of the former are not related to velocity while the latter'^ main safety factors are related to velocity.Based on PHMSA's data,American liquefied tank car comprehensive system's steady state availability is 0.01 year and the mean time to fail is 0.02 year.