Citation

Mitchell JW. Fire Safety J. 2023; 140: e103879.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.firesaf.2023.103879

PMID

unavailable

Abstract

Utility-related ignitions have historically been correlated with catastrophic wildfire. First noted in Australia and Southern California, as climate-change related drought has increased catastrophic wildfires from utility ignitions now plague the western US. Regulatory changes now require California utilities to analyze and mitigate wildfire risk. This analysis reviews recent data and risk assessments from Pacific Gas and Electric Company (PG&E), Southern California Edison Company (SCE), and San Diego Gas and Electric Company (SDG&E), California's largest electrical utilities. While utilities have adopted data science methodology, covariates capturing extreme weather effects are missing and tools lack mechanisms to incorporate causal linkage between likelihood and consequence models. Consequently, risk models incorrectly prioritize risk drivers, such as vegetation, equipment, or external agents. Additionally, match-drop wildfire spread calculations fail to represent large fires due to limited run time. Risk models also fail to incorporate the health effects of wildfire smoke. Power shutoff is an effective mitigation during extreme weather events, but causes significant public harm. "Hardening" programs, especially undergrounding lines, are effective but their expense threatens public health for the poorest. Accurate balancing of wildfire risks, risk of power loss, and financial impacts on vulnerable populations, in conjunction with application of new technology is required to address the utility wildfire problem.

Language: en

Keywords

Electric utilities; Extreme weather; Fire growth; Ignition; Machine learning; Modeling; Risk assessment; Wildfire smoke; Wildfires