Evaluating the seismic resilience of urban power distribution networks under spatially variable ground motions

To accurately assess the seismic resilience of urban power distribution networks and achieve dynamic seismic risk zoning and classification, this study proposes a comprehensive framework for seismic resilience and risk assessment that incorporates the spatially variable ground motions. First, a spatially correlated peak ground acceleration is generated based on historical ground motion records and attenuation relationships. Subsequently, the proposed Space-G network modelling method is adopted to construct the power distribution network topology and extract component ground motion parameters. Damage scenarios are then simulated using Monte Carlo Simulation methods using fragility models of substations and lines. The network functionality and connectivity are evaluated from damage samples, while the temporal evolution of functionality during both the damage and recovery phases is quantified through multiple performance indicators. This allows for comprehensive system resilience and dynamic seismic risk assessment. The proposed framework is verified through a case study of a power distribution network. Results indicate that the proposed Space-G method effectively captures the network topology; after considering spatial ground motion variability, the connectivity rate of the same node varies by 7.4%–15.7%, and the overall resilience decreases by approximately 20%. Furthermore, the dynamic risk analysis illustrates the spatial–temporal evolution of regional seismic risk.