Long-duration energy storage (LDES) can shift energy across days or seasons, adding flexibility to power systems. However, incorporating LDES in resource planning tools poses challenges due to the various simplifications used to manage computational run-times. This research aims to better understand how to incorporate LDES (storage with duration exceeding 10 hours) in long-term planning. The focus is on understanding what modeling approaches can be used to accurately model long duration storage resources along with other shorter-duration storage technologies, and what system factors and characteristics are conducive to LDES deployment.
This research uses an open-source, unit-level capacity expansion model and a test system that approximates the Western Interconnect to conduct a series of numerical storage modeling experiments. Storage with durations of 4, 10, 24, and 100 hours are included, each with associated capital costs and round-trip efficiencies derived from publicly available sources estimates. A range of scenarios with differing modeling choices and system characteristics are run to determine their effect on storage deployment.
Results indicate that the deployment of LDES and energy storage technologies more broadly is influenced by several factors, including the chronology and temporal resolution used in models, as well as system characteristics such as carbon emission targets, the type and level of renewable resources in the systems, and potentially extended and frequent periods of low variable resource generation or high net demand, especially when other flexible resources are limited.
Authors Karen Tapia-Ahumada and Sean Ericson