Functional reliability analysis on metro network considering spatial variation of seismic ground motions and population distribution

The metro system is a critical component of urban lifeline infrastructure. The catastrophic failure of Daikai Station during the 1995 Kobe earthquake has drawn global attention to the seismic performance of metro systems. Seismic research at both the structural and network levels of metro systems is accordingly of paramount engineering and academic significance. Based on seismic hazard analysis and seismic fragility models of metro structures, this study proposes a method for calculating the connectivity and functional reliability of metro systems subject to earthquakes using the Monte Carlo simulation method. The proposed method is applied to the metro system in the main urban area of Beijing to verify its practicality and compare the differences between the connectivity reliability and the functional reliability, thereby demonstrating the necessity of functional reliability analysis. To support functional analysis, this study proposes a method for dynamically deriving origin-destination (OD) matrix data between stations using population heat maps. Its validity is verified through benchmarking against historical OD data, providing a pragmatic solution to the long-standing challenge of OD data acquisition in metro system seismic studies. For seismic ground motions with a 50-year exceedance probability 10%, 75.0% of stations in Beijing metro system can maintain effective connectivity. The closed-loop structure of the operating lines significantly enhances system reliability, whereas the reliability of branch-line stations remains comparatively weak. Functional and connectivity reliability values across stations are approximately proportional; however, functional reliability is consistently higher across most stations, indicating that connectivity-only assessments may underestimate the actual seismic performance of metro stations.