Abstract: To address power supply pressure and safety hazards in high-density residential communities caused by dense population and surging electricity demand, this paper proposes an early warning method for electricity safety hazards based on electric power big data. It integrates multidimensional indicators, including average monthly electricity consumption per household, average capacity per household, and the ratio of electricity consumption to capacity, and applies the K-means clustering algorithm to analyze electricity consumption data of communities within the service area. The communities are divided into five categories, including high-density residential community, potential risk, high-load risk, high household count risk, and low risk. Differentiated risk early warning rules are established, and targeted governance measures are proposed.