U.S. data center power capacity is projected to surge more than fourfold, from 40 gigawatts in 2025 to 169 gigawatts by 2030, as demand for artificial intelligence rapidly increases. A new study co-authored by Hon Chung Lau, adjunct professor in the Department of Chemical and Biomolecular Engineering at Rice University and founder of Low Carbon Energies LLC, reveals that without intervention, carbon dioxide emissions from powering these facilities could leap from approximately 90 million metric tons annually in 2025 to over 404 million metric tons by 2030. “Data centers are becoming one of the defining energy challenges of the AI era,” Lau said. The research, published in Energy & Fuels, suggests carbon capture and storage could mitigate a substantial portion of these emissions, with states like Texas facing a particularly acute need to add roughly 25 gigawatts of power capacity by the end of the decade.
Data Center Power Capacity Growth to 169 Gigawatts by 2030
The expansion of artificial intelligence is expected to dramatically reshape energy demands, and a new study reveals a projected surge in power consumption by data centers. Researchers estimate U.S. power demand will increase significantly. The study, co-authored by Lau and Steve C. Tsai, an energy transition consultant at Low Carbon Energies LLC, analyzed publicly available data on announced U.S. data centers, identifying states poised for the most substantial growth, including Texas, Virginia, and Pennsylvania. Texas alone is projected to require an additional 25 gigawatts of power capacity by 2030 to accommodate data center needs. The analysis further explored the potential for carbon capture and storage to mitigate these emissions, finding that 34 states possess sufficient saline aquifer storage capacity to accommodate over 100 years of projected carbon dioxide output from data centers beyond 2030.
In 2025, these aquifers could store an estimated 59 million metric tons, representing 66% of sector emissions, increasing to 299 million metric tons, or 74%, by 2030. Lau emphasized that carbon capture and storage is not a single solution, but rather a viable option given existing geological conditions. “This does not mean carbon capture is the only solution,” Lau said, “But it does show that the geology exists to make a meaningful impact, especially in states where data center growth is strongest.” The researchers acknowledge their estimates are conservative, based solely on publicly announced data center power requirements and a static state energy mix through 2030, yet still provide a valuable framework for aligning digital infrastructure expansion with climate objectives.
Data centers are becoming one of the defining energy challenges of the AI era.
Hon Chung Lau, adjunct professor in the Department of Chemical and Biomolecular Engineering at Rice University and founder of Low Carbon Energies LLC
The escalating demand for artificial intelligence is rapidly reshaping the energy landscape, and a new analysis reveals a potential, geographically-focused mitigation strategy for the resulting carbon footprint. Researchers, led by Hon Chung Lau of Rice University and Steve C. Tsai of Low Carbon Energies LLC, investigated the feasibility of utilizing saline aquifers for carbon capture and storage related to data center emissions. Their work, published in Energy & Fuels, demonstrates that these geological formations possess significant capacity to absorb carbon dioxide produced by power plants serving the burgeoning data center industry. The study estimates that by 2030, U.S. data center power capacity will surge to 169 gigawatts, a more than fourfold increase from the 40 gigawatts projected for 2025. The analysis considered publicly available data on announced data centers, factoring in projected power capacity, energy sources, and locations to estimate carbon emissions based on each state’s electricity mix. The researchers found that including out-of-state storage options could push the potential mitigation rate beyond 90%.
This does not mean carbon capture is the only solution.
Hon Chung Lau, adjunct professor in the Department of Chemical and Biomolecular Engineering at Rice University and founder of Low Carbon Energies LLC
See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
