A Generalized Accelerated Failure Time Model to Predict Restoration Time from Power Outages

TL;DR

This paper introduces a spatially-aware generalized accelerated failure time model to predict power restoration times after hurricanes, considering hazard, environmental, and socio-demographic factors, with improved accuracy.

Contribution

It develops a novel GAFT model incorporating spatial dependence to better predict power outage restoration times during hurricanes.

Findings

Spatial dependence improves model fit by 12%.

Maximum wind speed and customer outage percentage strongly influence restoration time.

Socioeconomic factors like median income are significant predictors.

Abstract

Major disasters such as wildfire, tornado, hurricane, tropical storm, flooding cause disruptions in infrastructure systems such as power outage, disruption to water supply system, wastewater management, telecommunication failures, and transportation facilities. Disruptions in electricity infrastructures has a negative impact on every sector of a region, such as education, medical services, financial, recreation. In this study, we introduce a novel approach to investigate the factors which can be associated with longer restoration time of power service after a hurricane. We consider three types of factors (hazard characteristics, built-environment characteristics, and socio-demographic factors) that might be associated with longer restoration times of power outages during a hurricane. Considering restoration time as the dependent variable and utilizing a comprehensive set of county-level data, we have estimated a Generalized Accelerated Failure Time (GAFT) that accounts for spatial dependence among observations for time to event data. Considering spatial correlation in time to event data has improved the model fit by 12%. Using GAFT model and Hurricane Irma as a case study, we examined: (1) differences in electric power outages and restoration rates among different types of power companies: investor-owned power companies, rural and municipal cooperatives; (2) the relationship between the duration of power outage and power system variables, and socioeconomic attributes. We have found that factors such as maximum sustained wind speed, percentage of customers facing power outage, percentage of customers served by investor-owned power company, median household income, and number of power plants are strongly associated with restoration time. This paper identifies the key factors in predicting the restoration time of hurricane-induced power outages.