A review of probabilistic resource adequacy assessments in power systems: Methods, applications, and future challenges
Marius Tillmanns, J. Schöttler, Aaron Praktiknjo
Abstract
The increasing integration of renewable energy sources, storage, demand-side flexibility, and sector coupling introduces significant complexities and uncertainties into modern power systems, reshaping the very nature of resource adequacy (RA) assessments. The growing sensitivity to correlated weather-events such as heatwaves, droughts, severe storms, and abrupt cold snaps heightens the likelihood of common-mode failures and amplifies the imperative to revisit RA methodologies and frameworks. RA refers to the ability of a power system to reliably supply present and projected electricity demand levels under uncertainty. This paper presents a thorough review of probabilistic RA approaches, spanning analytical capacity-balance methods, non-sequential Monte Carlo simulations, and optimization models for unit commitment and economic dispatch embedded in sequential Monte Carlo frameworks. We contrast two overarching paradigms for assessing RA in modern interconnected power systems: (1) the system planning approach, which iteratively optimizes capacity expansion under exogenous reliability targets; and (2) the system evaluation approach, which evaluates the adequacy of projected resource portfolios and derives reliability thresholds via economic trade-offs. Drawing on examples from NERC's Long-Term Reliability Assessment and ACER's European Resource Adequacy Assessment , we summarize key metrics, regional standards, and best-practice implementations. We then identify prevailing challenges—including multi-dimensional uncertainty and correlation, time-coupled operations of energy-limited resources, market design and network representation, incorporating load shortfall characteristics into RA frameworks, electricity demand forecasting, and advanced acceleration techniques—and highlight recent methodological advances and research gaps. Finally, we propose future research directions and policy recommendations to strengthen probabilistic RA frameworks and system planning for the evolving landscape of large-scale, interconnected power systems.