In energy systems that rely substantially on solar and wind power, rare periods of especially low renewable resource availability could challenge efforts to meet demand reliably. As such, there is considerable interest in designing and deploying technologies that can mitigate the impacts of these low resource periods. Proposed solutions include batteries with very low costs of energy capacity, technologies that store energy in and generate power from chemical fuels, transmission systems that cover broader areas, and processes that modulate their operation to match power availability. In this research, we characterize the size and dynamics of fluctuations in solar and wind power resources in the context of large-scale energy systems. Using forty years of high-resolution weather data, empirical time series of electricity demand, and an energy system model, we identify cost-minimizing compositions of possible future energy systems given high uncertainty regarding technology costs and the overall level of reliance on renewable resources and energy storage. We then simulate the operation of high renewables systems at 816 locations spread across the coterminous United States, as well as systems that rely on resources integrated over broad regions. We use these results to identify periods of low resource availability and quantify their temporal and spatial characteristics. We demonstrate how these low resource periods influence requirements for technological strategies designed to ensure energy demand is met, including deploying low-cost energy storage; using dispatchable non-solar, non-wind power generators; expanding transmission infrastructure; and managing demand. The results provide insights that can guide efforts to develop new chemical and energy technologies and plan systems that can help address climate change.
