Performability Analysis of Railway Systems

Train operations on railway networks rely on the availability and successful interaction of different subsystems, each consisting of multiple technical components. The heavy wear, long lifetime and adverse conditions the infrastructure is subject to gives rise to malfunctions and failures of components. Failure processes of individual components are reasonably well understood and have been analyzed using formal methods including fault trees and failure mode and effect analysis. Still, the understanding of the effects of infrastructure failures on the capacity and the performance of the entire network in operations remains limited to special cases or scenario-based approaches. In the present paper, we elaborate how formal methods used in risk assessment of individual components can be transferred to the analysis of the performance of the network. By identifying trains with their train paths we introduce a coupling between train operations and infrastructure asset management. We also discuss how a pre-defined level of service can be used to classify failure states or degraded modes of the system. Apart from a redundancy-based resilience analysis of network design the method can be used to identify critical elements and relevant failure scenarios, which can be used in timetabling applications.