Large-scale data analytics for resilient recovery services from power failures

Massive power failures are induced frequently by natural disasters. A fundamental challenge is how recovery can be resilient to the increasing severity of disruptions in a changing climate. We conduct a large-scale study on recovery from 169 failure events at two operational distribution grids in the states of New York and Massachusetts. Guided by unsupervised learning from non-stationary data, our analysis finds that under the widely adopted prioritization policy favoring large failures, recovery exhibits a scaling property where a majority (similar to 90%) of customers recovers in a small fraction (similar to 10%) of total downtime. However, recovery degrades with the severity of disruptions: large failures that cannot recover rapidly increase by similar to 30% from the moderate to extreme events. Prolonged small failures dominate entire recovery processes. Further, our analysis demonstrates the promise of mitigating the degradation by enhancing recovery of a small fraction of large failures through distributed generation and storage.

Automated summary

The study reveals that prioritizing large failures can help the majority of customers to recover quickly from massive power outages induced by natural disasters. However, as the severity of disruptions increases, recovery capability degrades and prolonged small failures dominate the entire recovery processes. Enhancing recovery of a small fraction of large failures through distributed generation and storage shows promise in mitigating the degradation.