Regarding risk management in railway systems, a series of methodologies in the European standard EN 50126 are well defined and applied on many railway systems. The purpose of EN 50126 is to manage hazards throughout the entire V-typed lifecycle, including the design, construction, operation, and all the way to decommissioning. Hazard’s causes, mitigations, and residual risk class are updated accordingly during the entire life cycle. In common railway practices, risk management starts at the beginning of the project; thus the hazard database is often built up by system providers and then transferred to operators. Since the providers’ duty is to meet the quantitative requirements of safety and availability, and it is difficult to implement quantitative analysis of natural disasters, especially when the provider is not usually held liable, most providers only focus on equipment failure and neglect management of hazards that are beyond our control, such as flooding, earthquake, mudslide, etc. In many cases, these disasters will become the main threats to the system after the handover. It is necessary to recheck the hazard database, reconsider the measurements, and then mitigate the impact of disasters.

The proposed case study was implemented on the Alishan Forest Railway, which is a traditional railway system and has operated for more than 100 years since its establishment. The system is famous for its beautiful high mountain railway and Z-shaped switchback lines. Unfortunately, because parts of the railroads were damaged by typhoons or earthquakes in recent years, these disasters had led to significant rockslides and collapsed tunnel. In order to understand the risk of disasters and required mitigations, this study followed EN 50126 to reconsider the impact of natural disasters and planned to improve the existing hazard database. However, the hazard database does not have data on the Alishan Forest Railway from the beginning, and we needed to take other railway systems into consideration and select for the applicable ones. The discussed hazards and corresponding mitigations will also be entered into the database. The result demonstrates that we can clarify the impact of natural disasters and understand how many possible mitigations the system has. It helps operators to find out the weakness of the system and then prioritizes the action plans. By adopting plan-do-check-act cycle in ISO 31000, the operators could trace the performance of the mitigations during the practical operation, and then modify the action plans to improve the performance. In general, this study achieved in showing how to analyze the risk of disasters systematically, and this methodology could be applicable in both railway systems and other domains.