Traditional planning reserve margin (PRM) constraints struggle to produce an adequate fleet of resources at least-cost in capacity expansion planning (CEP) models as renewable energy (RE) and storage are included in the plans. Refining how resource adequacy (RA) is handled in CEP models can improve system reliability and reduce cost. This report introduces an event-based RA approach that extracts net-peak stress events and energy-limited intervals from long weather histories and embeds those stress conditions as constraints in CEP models. Using a common CEP backbone and 71 years of weather-driven demand and RE availability across multiple scenarios (baseline, alternative capital costs, large loads, and electrification with winter peaking), the study compares CEP models with event-based, several accreditation-based variants (static ELCC, forecasted ELCC, ELCC surfaces) RA frameworks.
Key findings show the event-based RA method generally yields lower costs (0.4–4.4% savings), improved adequacy (40% lower LOLE), more consistent performance across the planning horizon and scenarios, and reduced bias in the resource mix. The event-based method is sensitive to extreme events in the weather data, which can lead to overbuilding and increased costs, but this issue can be mitigated by managing RA requirements in the extreme events; this may ultimately give modelers more control over resilience outcomes in event-based models compared to traditional models. ELCC surfaces also perform well compared to other accreditation approaches, but all the accreditation approaches struggle with storage duration and technology options. One reason the event-based models are successful is their ability to select storage options that work with the rest of the portfolio to improve adequacy at low cost.