Quantifying the Social Costs of Power Outages and Restoration Disparities Across Four U.S. Hurricanes

The multifaceted nature of disaster impact reveals that areas with concentrated populations may contribute more to overall economic burden, while sparsely populated but highly impacted regions suffer disproportionately at the individual level. This study presents a framework that quantifies the societal costs of power outages by converting customer-weighted outage exposure into monetized social losses. This framework integrates welfare metrics with three recovery measures—average outage-days per customer, restore duration, and relative restoration rate—computed from sequential EAGLE-I observations and linked to Zip Code Tabulation Area-level demographics. Through a standardized pipeline applied to four U.S. hurricanes—Beryl (2024, Texas), Helene (2024, Florida), Milton (2024, Florida), Ida (2021, Louisiana), we produce the first cross-event, fine-scale accounting of societal outage costs and their underlying drivers. The analysis reveals substantial heterogeneity in impacts: Ida generated $1.50 billion in total deprivation costs ($1,757 per capital) Milton, $1.26 billion ($387 per capita); Beryl, $629 million, ($674 per capita); and Helene, $411 million ($285 per capita). Across all events, the results document consistently regressive burden distributions, with deprivation-cost shares declining systematically with income. Mechanistic analysis demonstrates that deprivation costs increase with restoration duration and decrease with relative restoration rates. Explainable modeling identifies restore duration as the dominant predictor of costs, while unsupervised clustering reveals distinct recovery typologies invisible to conventional reliability metrics. By shifting from simple outage counts and subjective assessments to welfare-based, exposure-weighted measurement, this study delivers three key contributions: (I) a transferable framework for quantifying outage impacts and equity, (ii) comparable cross-event evidence linking restoration dynamics to social losses, and (iii) actionable spatial analyses and community typologies that enable equity-informed restoration prioritization and resilience investment decisions.