Building Climate Resilient Utilities 2025

Infrastructure hardening and intelligent operations systems Effective early warnings are only valuable if the infrastructure can withstand the initial shock and operations can be intelligently managed during a crisis. Recognizing this, Chinese utilities are pursuing a dual strategy of physical fortification and operational digitalization. This involves deploying more resilient materials and hardware, while wrapping this hardened infrastructure in a layer of intelligent software, creating a “sentient” grid capable of self-assessment and rapid, semi- autonomous response. For example, advanced sensors (IoT devices) on transformers, switches and power lines constantly measure voltage, current and temperature. This data is fed into a central system, providing a live “health check” of the entire grid, far beyond a simple on/off status. Real-time data from monitoring terminals feeds into AI-driven models that can predict conditions such as icing up to three days in advance, enabling precise, proactive de-icing. The key to a semi-autonomous response is that the system intelligently takes pre-approved actions on its own but keeps a human operator in the loop for confirmation or high-level direction. China Southern Power Grid (CSG), one of China’s two power transmission giants, supplies electricity to more than 250 million people across five southern provinces. Managing a vast and climate-vulnerable service area, CSG has become a national leader in grid modernization and disaster resilience, fusing digital intelligence with robust physical infrastructure. Its systems have transformed disaster management – delivering faster recovery, higher accuracy and minimal damage even under extreme icing and typhoon events (see Case Study 5). CASE STUDY 5 China Southern Power Grid China Southern Power Grid (CSG) has pioneered a comprehensive defence system – spanning perception, prediction, response and recovery – by tightly integrating hardware upgrades with intelligent operations software. Icing disaster response: hardware-software synergy In regions prone to cold waves and freezing rain, such as Guizhou, CSG has established a multi-layered anti-icing system. The grid has deployed 108 sets of DC ice-melting devices and 620 online icing monitoring terminals, forming a “three-tier, four-protection” system: 1) DC ice-melting for main grids; 2) flexible ice-melting for county-level grids; 3) AC ice-melting for township grids; and 4) manual de-icing for village lines. This system can melt 10 centimetres of ice on 350 kilometres of 500kV lines within one hour, achieving “zero tripping, zero damage” for main grid equipment. Real-time data from monitoring terminals feeds into AI-driven models that predict icing conditions up to three days in advance, enabling precise, proactive de-icing. Digital twin technology simulates the de-icing process, optimizing strategies and ensuring near-100% success rates.26 Typhoon response: air-ground integrated operations During typhoons, CSG uses unmanned aerial vehicles (UAVs) and intelligent systems for rapid reconnaissance and restoration. For example, during Typhoon Wipha, “Wing Loong” drones penetrated cloud walls to relay real-time imagery to the emergency command centre and deployed airborne base stations to restore communications. Ground- based smart repair systems, informed by UAV data, automatically generate optimal repair routes, enabling a fully digitalized “reconnaissance-location-repair” workflow. The Guangdong grid’s “one map for emergency” system integrates wind, rain and disaster data, supporting rapid restoration: during Typhoon Butterfly, 60% of the 15,000 affected users had power restored within two hours. This swift response was enabled by AI-powered dispatchers such as the “Mingyue” system, whose rapid analysis and forecasting enabled crews to be dispatched almost immediately to the exact fault locations, streamlining the entire repair. Without such a system, simply diagnosing issues and coordinating a response could take hours. AI’s second-level risk analysis has reduced this process to just five minutes, with load forecasting accuracy reaching 98.3%. Meanwhile, CSG’s digital repair platforms, such as the “elink” system in Zhanjiang, enable real-time sharing of damage information and boost work order dispatch efficiency by 70%. In Xuwen, the integration of UAVs and manual teams enabled the restoration of power to 1,758 households within 45 minutes during a typhoon event. Source: China Southern Grid.27 Building Climate-Resilient Utilities: Lessons from China and Future Pathways 14