ENERGY SECURITYFramework Reveals a Smarter and Faster Way to Retire U.S. Coal Plants

Even as coal power continues its steady decline in the United States, more than a hundred plants still have no retirement plans—a gap large enough to derail national climate goals. A new study led by UC Santa Barbara researchers offers a way forward, showing how targeted, data-driven approaches could help accelerate the transition.

Published in Nature Energy, the study tackles a critical question: if market forces have already driven many coal plants to close, why are so many still running? Despite years of decline, roughly 105 gigawatts of coal capacity—representing 114 plants—are still slated to operate through 2035, even though a complete phaseout by that date is widely considered essential for meeting U.S. net-zero emissions goals.

“Coal is complex—there’s no single right way to deal with it,” said Sidney Gathrid ‘22, the study’s lead author. “Our goal was to build tools that reflect that complexity, so different actors can take on different facets of the problem. There’s no one straightforward path, and we wanted to do research that represented that reality.”

Working with Grace C. Wu, an assistant professor in the Environmental Studies Program and senior author on the paper, Gathrid and his team show that reaching those goals will require policymakers to move beyond age-based or one-size-fits-all approaches—and instead focus on the specific contexts that accelerate the retirement of certain coal plants.

To do that, the researchers—including Jeremy Wayland, Stuart Wayland ‘22, and Ranjit Deshmukh, an associate professor in the Environmental Studies Program and the Bren School of Environmental Science & Management—developed a new framework combining graph theory and topological data analysis to classify the entire U.S. coal fleet into eight distinct groups based on 68 technical, economic, environmental and sociopolitical factors. They also introduced a “contextual retirement vulnerability” score that measures how susceptible each plant is to early retirement by comparing it to facilities that have already announced closures.

The framework goes a step further by identifying “retirement archetypes”—patterns that explain why plants in each group are retiring. These range from regulatory and health-based drivers to unfavorable economics or political pressure, offering a clear set of levers that can be applied to similar facilities elsewhere.

“Instead of asking only why coal plants retire, we asked how we can make retirements happen faster—and in ways that are efficient and grounded in data,” Gathrid said. “Our framework helps policymakers and advocates identify where they can have the biggest impact.”