Assessing the impact of extreme weather events on power grid failures

: a data-driven analysis in Sicily

Abstract

This thesis investigates the impact of extreme weather events on electric transmission networks in Sicily. As climate change increases the frequency and intensity of such events, and as decarbonization and integration of renewable energy introduce new operational complexities, the need for resilient grid infrastructure has become increasingly urgent. Sicily presents a compelling case due to its central Mediterranean location, geological and island characteristics, aging infrastructure, and growing renewable capacity, all of which increase vulnerability to weather-induced disruptions. A three-stage modeling framework is developed to predict grid failure occurrence, severity, and economic impact using daily panel data from 2014 to 2023, combining high-resolution weather data with transmission outage reports. The first stage employs binary classification to forecast outage occurrence; the second, multiclass classification to assess severity; and the third, regression analysis to estimate unserved energy and associated economic loss. While acknowledging that outages may have multiple causes, the analysis identifies weather, especially wind, as a key driver. Results also show that interpretable models, such as logistic regression, can provide operational value for prioritization and planning. The study contributes to climate-resilient grid planning by demonstrating the utility of supervised learning for predictive diagnostics and highlighting the importance of improved data integration in critical infrastructure monitoring.

Date of Award	23 Jun 2025
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Pedro Afonso Fernandes (Supervisor)