AI data centres: geological CO₂ storage could absorb up to 74 per cent of their US emissions by 2030

A study published in *Energy & Fuels* predicts that emissions will increase more than fourfold by 2030 and suggests that gas-fired power stations fitted with carbon capture technology could help to limit the impact.

Two researchers, Hon Chung Lau (assistant professor at Rice University) and Steve C. Tsai (Low Carbon Energies), have identified the data center projects announced in the United States to estimate their carbon footprint. According to their calculations published in the journal Energy & Fuels, the installed capacity would increase from 40 GW in 2025 to 169 GW in 2030. If this electricity continues to come largely from fossil power plants, the associated emissions would rise from about 90 million tons of CO2 per year to more than 404 million.

Saline aquifers under future campuses

The authors defend a short-term solution: gas-fired combined-cycle power plants coupled with carbon capture and storage. They cross-referenced the project map with that of deep saline aquifers and conclude that 34 states have sufficient capacity for more than a century of emissions. Storage could cover 59 million tonnes by 2025 (66% of emissions) and 299 million tonnes by 2030 (74%), or even more than 90% by transporting CO2 from one state to another.

Texas, which alone is expected to add some 25 GW of demand, is in the lead with Virginia, Pennsylvania, Ohio, Arizona, Colorado, Utah and Illinois. The researchers point out that this avenue must be in addition to renewables and other decarbonization levers, not a substitute for them.