With Thomas Palley, Asa Palley, and Oguz Cetin.
May 2024.
Abstract: Policymakers worldwide have set ambitious goals to decarbonize electricity generation over the next few decades, which will require significant investments in renewable resources such as solar panels and wind turbines. Because these renewable generation sources are intermittent, utility-scale energy storage is seen as critical to meeting these targets. However, the amounts of generation and energy storage capacity needed to achieve a particular renewable portfolio standard (RPS)—a required minimum fraction of electricity produced by renewable sources—remains an open question. We develop a model that, by solving for the optimal short-term operational plan given any portfolio of generation and storage resources, can determine the optimal long-term plan for investment in wind, solar, and energy storage capacity to achieve a given RPS target. The model embeds several key features of the problem, including the nondispatchability of renewable generation, demand fluctuations, and intertemporal optimization of storage. We estimate the model for three electricity grids in the United States (Texas, California, and the Mid-Atlantic) using hourly data on demand and renewable generation from 2019-2021. We find that RPS targets of 70-90% by 2040 yield substantial gains in social surplus for a broad range of social cost of carbon estimates, and these targets can be met without significant investment in energy storage capacity. However, the total costs of meeting higher RPS targets increase at an increasing rate—a convex cost structure that we prove characterizes the problem more generally—which ultimately makes storage capacity a critical part of the portfolio.
The most recent version of the working paper is available as a pdf. (67 pages, 4024 KB)