Abstract: Power outages following hurricanes disrupt social welfare, business operations, and the functioning of public facilities in the affected areas. Extended outages can severely impact community well-being, particularly among vulnerable populations such as the elderly and individuals with medical conditions. While previous studies have investigated disparities in outage recovery, many of them focus on single events and rely on general social vulnerability indices, limiting their ability to identify consistent patterns or capture the specific factors that heighten vulnerability to power disruptions. Moreover, few studies have examined the influence of economic activity or the presence of critical facilities on recovery time. This study analyzes power outage recovery trajectories across 27 U.S. states affected by 21 hurricanes between 2017 and 2024. Using Spatial Durbin Model (SDM) regression analysis, we assess whether counties’ socioeconomic vulnerability, levels of economic activity, and the number of critical facilities consistently explain differences in post-hurricane power outage recovery time. A customized outage-focused Social Vulnerability Index (SVI) was used to more accurately capture population vulnerability to power disruptions. Results indicate that the SVI does not have a consistent effect on recovery time, suggesting no systematic disparities based solely on socioeconomic characteristics. Median income, representing household-level wealth, is negatively correlated with recovery time, while higher county-level economic capacity (measured by GDP per capita) is associated with longer recovery; although both effects vary with outage severity and the inclusion of other model control variables. The number of critical facilities does not show a consistent effect once other county characteristics are accounted for.  

Publication link: https://ieeexplore.ieee.org/document/11240106

Data and Codes: https://github.com/amsalman1987/NSF-CAREER-Research-Products

Abstract: Water distribution systems are critical to the stability and well-being of communities but remain highly vulnerable to disruptions from natural disasters and aging. These disruptions do not impact all populations equally, making it essential to incorporate differential vulnerability into water infrastructure management and planning. A common method for assessing such disparities is the use of a social vulnerability index (SVI). While various SVIs have been developed by researchers and government agencies, they are often generic—intended for general disaster preparedness or infrastructure challenges—and may omit key factors relevant to water service disruptions or include irrelevant ones. Consequently, these indices may not adequately identify areas critically vulnerable to water-related disruptions. This study develops a water-specific SVI to classify counties based on their vulnerability to water service interruptions. Twenty-one relevant factors were selected and grouped into three categories: health-related, socioeconomic, and housing-related vulnerabilities. Principal Component Analysis (PCA) was used to reduce dimensionality, followed by Pareto ranking to construct the index. Pareto ranking complements PCA by enabling a balanced, multi-dimensional comparison without imposing arbitrary weights, thereby preventing dominance by any single factor. For comparison, an alternative version of the index was also developed using percentile ranking. The resulting water-specific SVI was compared with FEMA’s existing SVI to highlight the importance of a tailored index. Significant differences were observed. For example, unlike FEMA’s classification, our index identifies some counties in the Southeast and Midwest as highly vulnerable due to factors such as aging infrastructure, high rental occupancy, and inadequate plumbing facilities—factors not captured by FEMA’s SVI. The proposed SVI provides a more targeted tool for policymakers, urban planners, and emergency managers to prioritize investments, allocate resources effectively, and enhance the resilience of water infrastructure systems. 

Data and Codes: https://github.com/amsalman1987/NSF-CAREER-Research-Products