A multi-hazard approach to assess severe weather-induced major power outage risks in the US

作者：Mukherjee Sayanti^*; Nateghi Roshanak; Hastak Makarand

来源：Reliability Engineering & System Safety, 2018, 175: 283-305.

DOI：10.1016/j.ress.2018.03.015

• Major power outages
• Multi-hazard risk analytics
• Extreme event risk
• State-level risk analytics
• Support vector machines
• Random forest
• Hybrid statistical learning model

Severe weather-induced power outages affect millions of people and cost billions of dollars of economic losses each year. The National Association of Regulatory Utility Commissioners have recently highlighted the importance of building electricity sector's resilience, and thereby enhancing service-security and long-term economic benefits. In this paper, we propose a multi-hazard approach to characterize the key predictors of severe weather induced sustained power outages. We developed a two-stage hybrid risk estimation model, leveraging algorithmic data-mining techniques. We trained our risk models using publicly available information on historical major power outages, socio-economic data, state-level climatological observations, electricity consumption patterns and land-use data. Our results suggest that power outage risk is a function of various factors such as the type of natural hazard, expanse of overhead T&D systems, the extent of state-level rural versus urban areas, and potentially the levels of investments in operations/maintenance activities (e.g., tree-trimming, replacing old equipment, etc.). The proposed framework can help state regulatory commissions make risk-informed resilience investment decisions.

• 出版日期2018-7