The Role of Extended Horizon Methodology in Renewable-Dense Grids With Inter-Day Long-Duration Energy Storage

Abstract: This study addresses the challenges in optimizing long-duration energy storage (LDES) dispatch within future power systems featuring high integration of variable renewable energy (VRE). The research focuses on conducting a comparative analysis between traditional and extended horizon methods for the optimization of LDES dispatch, using open-source and commercial production cost models (PCMs), tested on a futuristic Electric Reliability Council of Texas (ERCOT) grid. The findings indicate that, despite its complexity and longer solution times, the extended horizon approach demonstrates superior performance in LDES dispatch and effectively reduces the impact of degenerate solutions in sequential simulations. This study underscores the trade-offs between computational efficiency and improvement in storage dispatch, which is crucial for future energy systems. The analysis highlights the necessity of addressing the degeneracy issue in storage dispatch in grids dominated by zero operating cost VRE generators and low operating cost energy storage devices. Additionally, the research reveals revenue discrepancies for LDES operators across different models, a consequence of the persistent presence of degeneracy in high VRE systems. These findings suggest an urgent need for refined modeling techniques in the planning and operation of future energy systems.