Targeted strategies for accelerating US coal plant retirements
Based on Gathrid, S., Wayland, J., Wayland, S. et al. Strategies to accelerate US coal power phase-out using contextual retirement vulnerabilities. Nat Energy 10, 1274–1288 (2025). https://doi.org/10.1038/s41560-025-01871-0
US coal power has been on the decline over the past decade, but there is no path forward for a complete phaseout in alignment with climate goals. Targeted early retirement strategies are now made available for major groupings of coal plants using their key group characteristics and similarity to plants with announced retirements.
The Policy Problem
US coal plant retirements have progressed steadily since peaking in 2011, driven largely by economic factors. Yet, 105 GW of coal capacity (114 plants) lack retirement plans and will remain operational through 2035. Even a hypothetical age-based retirement at 50 years will still leave 37 GW of coal capacity by 2035. These numbers highlight a gap between current coal retirement trajectories and what is needed to meet decarbonization targets, underscoring the need for strategic planning and policy innovation.
With no federal mandate for coal phaseout, bottom-up solutions are essential. Yet plant-by-plant retirements are often slow and fragmented, and national one-size-fits-all policies often miss local context. We set out to address these gaps by identifying which coal plants are most vulnerable to early retirement and which strategies are most likely to accelerate it. By grouping plants with similar characteristics and developing a plant-level ‘retirement vulnerability’ metric, we aimed to support more strategic, efficient resource allocation. This group-based approach enables targeting of common drivers across multiple plants, offering a more scalable alternative to plant-by-plant action.
Key Findings and Proposed Solutions
If a coal plant retires in one place, it raises the question of whether the same factors can drive retirement elsewhere. Coal plant retirement decisions are complex: factors such as ownership, grid role, economics, policy exposure and local opposition interact differently at each site. To address this complexity, we used graph theory and topological data analysis to develop an open-source, objective and robust method for clustering 198 operational US coal plants based on 68 technical, environmental, financial, health, social and geopolitical factors. By grouping plants with shared retirement risk characteristics, we can identify where proven strategies might transfer and where localized approaches are needed.