Proceedings of the PowerUp Conference

Resource adequacy (RA) studies commonly repre- sent energy storage via economic dispatch, with the real-time state of charge (SOC) of storage units implicitly assumed to follow a predetermined schedule until scarcity conditions arise. This paper explores the implications of this assumption, by documenting how downward deviations in energy storage SOC in real-time operations affect reliability outcomes. Applying a probabilistic RA model to an ERCOT-like system with 5 GW of energy storage, we quantify the sensitivity of normalized expected unserved energy (NEUE) and marginal effective load carrying capability (ELCC) to SOC deviations across two-, four-, and eight-hour storage durations. We find that a downward SOC adjustment equal to 25% of maximum storage energy capacity increases system NEUE by a factor of 1.6–2.3 and reduces the marginal storage ELCC by 11–13 percentage points, both of which represent significant differences compared to reference cases. Metrics for shorter-duration storage technologies are most sensitive to SOC deviations on a relative basis, shaped by the nature of lost load within the specific test system. The results highlight the importance of accurately modeling energy storage operations when accrediting storage in RA modeling efforts to help ensure model outcomes are consistent with real-world operational behavior.