This study examines the resilience of electric power systems, which is defined as the capability to reliably meet societal electricity needs under normal and disrupted conditions. The study uses comparative assessment approaches tailored to peacetime and wartime contexts. Specifically, it evaluates the impact of systematic missile and drone attacks on Ukraine’s electric power system in 2024. The categorization of damaged facilities is based on severity, and restoration efforts are characterized by the duration of repairs. These parameters are incorporated into a dynamic equation that models the availability of generating units over time.A resilience assessment is conducted using clustered models of power facility load modes. In these models, functionally similar facilities are grouped into clusters, and the discrete operational states of startup, loading, and shutdown are represented by integer variables instead of the traditional binary formulations. The modeling results demonstrate strong consistency with operational data reported in open sources, confirming the adequacy of the cluster approach.Key difficulties in evaluating resistance during ongoing major attacks are pinpointed, such as changing attack plans (frequency, scope, and targeting), the changing effectiveness of air defense systems, the staged implementation of protective structures, the performance of backup systems, and the adequacy of repair resources. The proposed clustered electric power system models' effectiveness in addressing these challenges stems from their ability to reduce the number of decision variables, thereby improving MILP solution efficiency without compromising accuracy. Quarterly simulations of Ukraine's power system for 2024 were conducted, and it was shown that computational time is reduced by 66–85 % with clustered models while maintaining an objective function deviation below 0.12 %, thereby underscoring their practical value for strategic resilience planning under extreme conditions.
Saukh et al. (Mon,) studied this question.