Diurnal, physics-based strategy for computationally efficient capacity-expansion optimizations for solar-dominated grids

Modeling energy storage for a renewables-driven grid using every hour of the year gives more insight and higher accuracy, but can be computationally demanding. In this study, we propose a novel and straightforward Critical Time Step technique, independent of the load and generation profiles, to reduce computational requirements with little loss in accuracy for a solar-dominated system. Using the diurnal cycle, specifically the critical times of an hour after sunrise and an hour before sunset, results in an excellent tradeoff of accuracy and computational complexity compared with fixed-time-step approaches. The accuracy and value of the technique are evaluated by comparing results for three weather years. The technique systematically underestimates the capacity expansion needed but differentiates the three weather years with results correlating well with the hourly simulations. Overall, the results show a high accuracy for the Critical Time Step technique in predicting the power expansion of the resources and the energy rating expansion of the storage system for a grid with more than 35% share of solar in the total operational power. The highest error occurred for storage power buildout but did not exceed 10% relative to the 1-hr-resolution simulation for the studied case.