Developing a safety risk model for railway network considering tunnel fall using Bayesian network

Evaluating the safety of the railway network as the fundamental part of the economy development reflects the safety of the entire transportation network of a country. One of the challenges threatening the railway safety is the railway tunnel conditions in terms of safe operating during operation period and analyzing the consequences of its hazards. Today, Bayesian network are the most identified methods to graphically model the causes of accidents and their conditional probabilities. In order to develop a model that best matches the real condition of the railway tunnels risks, modeling process divided in to two sections; in the first section, the compromised causes that identified and concluded by the specialists scored according to their frequency of happening and the related network constructed. Second section joints the cause and consequence analysis of different scenarios offering remedial actions and concluding the final model based on severity degrees defined. Outcome of the model shows the high probability for water penetration in tunnels without permanent lining as the most important cause and the first degree of the accident severity according to the guideline as the most frequent consequence of the risks identified.