Analysing power system cascading failures and service disruptions in a data-scarce environment: A case study of Vietnam

• Transparent, reproducible workflow to build overhead power grids from public data. • Map electricity end users by integrating Google Places API and OpenStreetMap. • Novel distance–capacity method links end users to substations. • Cascading failures modelled via power flow analysis, beyond topological methods. • Estimate service disruptions via post-disruption electricity-served ratio. Cascading failures in power grids can escalate localized disruptions into extensive service outages. This study presents a transparent, reproducible framework to build power grids from publicly available data, integrating commissioned power plants, substations, transmission lines, and hypothetical transformers, for large-scale power system modelling. To assess how grid disruptions lead to service interruptions for industrial users, we combine Google Places API industrial points of interest with OpenStreetMap industrial zones, and link them to substations based on capacity and proximity. We model initial disruptions under two scenarios: 1) random disruptions, implemented through percolation analysis to assess network robustness under stochastic component removals; and 2) targeted disruptions, generated by Monte Carlo sampling of wind-induced fragility curves derived from the 2024 Typhoon Yagi wind field to produce probabilistic failures. The analysis reveals distinct spatial variation in grid vulnerability and traces the propagation of failures through the network, quantifying downstream service disruptions for industrial users. System-wide simulations further identify critical transmission corridors with higher failure and overload risks. These insights provide a basis for targeted investment, resilience assessment, and future research on multi-hazard and adaptive protection strategies.