Digital city information models to reduce power outages and enhance energy resilience for megacity Shenzhen

Extreme weather events are increasing the frequency and duration of urban power outages, threatening critical infrastructure and essential services in cities. This study presents an integrated approach to mitigate outage impacts through coordinated electric vehicle (EV) dispatch and strategic charging station planning, supported by a city information model (CIM). We develop an agent-based CIM for megacity Shenzhen, incorporating 115,544 buildings, 200,000 EVs, and a road network with 22,474 nodes and 36,195 edges, to mitigate power outages under extreme weather conditions. The baseline simulation shows significant spatial variation in outage counts and hours across urban building energy hubs under predefined grid supply thresholds. Two resilience strategies have been proposed, i.e., a vehicle-to-grid (V2G) emergency power dispatch framework, and a graph-based optimization model. The former is to redistribute electricity from available EVs to affected buildings, while the latter is to deploy bidirectional charging stations in high-risk areas to enhance coverage in critical zones. Results indicate that, compared to the baseline scenario, the V2G emergency power dispatch and a graph-based optimization model with additional bidirectional charging stations reduce total power outage counts by 11.1% and 11.3%, and power outage hours by 15.0% and 15.3%, respectively. Supplementary analyses on energy flows and economic costs including government subsidies for V2G participation demonstrate the broad feasibility of the proposed approach. This work provides a data-driven and urban-level strategy with optimal infrastructure deployment and flexible dispatch to improve urban energy resilience under climate-related disruptions.