National power transmission grids are increasingly exposed to diverse hazards, challenging their reliable and secure operation. Understanding the exposure of transmission assets to a comprehensive set of hazards is essential to assessing their vulnerability and ensuring system resilience. Probabilistic risk assessment (PRA) tools in power transmission aim to provide a stochastic description of asset exposure to hazards, such as random failures, windstorms, lightning strikes or ice. The current state of the art lacks comprehensive approaches to power transmission risk assessment that integrate asset-level analyses into a full-scale system assessment that embeds an exhaustive set of hazards.
In this study, we propose a Bayesian approach to estimate asset-specific fragility curves that quantify the conditional failure probability of assets under specific hazards. Hazard-specific finite-element fragility curves and a continental failure database are used to build physically informed prior knowledge. Regional historical failures and weather records are used to perform the Bayesian update. Finally, we develop a data-driven approach to estimate the common-cause exposure of multiple assets in close geographic proximity using historical observations.
The proposed PRA framework is applied to extra-high-voltage transmission assets in Switzerland using extensive historical and meteorological data from Switzerland and Northern Italy. The framework enables integrating asset-level risks into system-level reliability and resilience assessments. Once validated, the approach supports both short-term operational risk evaluation under hazard forecasts and long-term infrastructure planning under evolving climate conditions.