The race to curb emissions from the AI-driven data center boom is on, and a new study suggests that carbon capture and storage (CCS) could be a game-changer. As artificial intelligence accelerates the demand for computing power across the U.S., a study co-authored by Hon Chung Lau, an adjunct professor at Rice University and founder of Low Carbon Energies LLC, highlights the potential of CCS to limit the climate impact of data centers.
The study, published in Energy & Fuels, predicts a massive growth in data center power capacity, from 40 gigawatts in 2025 to 169 gigawatts by 2030, a more than fourfold increase. Without new strategies, emissions could soar from 90 million metric tons per year to over 404 million metric tons by 2030. This is a critical issue, as data centers are becoming a major energy challenge in the AI era.
The research, conducted by Lau and Steve C. Tsai, analyzed U.S. data center data, including power capacity, energy sources, and locations. They estimated emissions based on state electricity mixes and examined the potential for CCS in saline aquifers. The findings reveal rapid growth in states like Texas, Virginia, Pennsylvania, and others, with Texas alone needing an additional 25 gigawatts of power capacity by 2030.
The study highlights the role of natural gas combined cycle power plants with CCS as a practical near-term solution. Natural gas is abundant, emits less CO2 than coal, and many data center growth areas are near saline aquifers for long-term storage. Interestingly, 34 states have enough storage capacity to handle projected emissions for over a century.
When considering out-of-state storage, the researchers found that over 90% of emissions could be mitigated through CCS. However, Lau emphasizes that CCS is not the sole solution, but it offers a significant impact, especially in high-growth states.
The study's conservative estimates, including only publicly announced data centers and constant state energy mixes, provide a useful framework for balancing digital infrastructure and climate goals. As the AI economy demands massive energy, this research helps identify emission hotspots and potential CCS solutions, offering a glimmer of hope in the fight against climate change.
