Artificial intelligence operators turn to repurposed airplane turbines to overcome grid backlogs

Facing multiyear delays to connect to electric grids, artificial intelligence companies are generating their own electricity using converted jet engines. Elon Musk's xAI has deployed eight repurposed aircraft turbines in Mississippi to power its Colossus 2 data center in Memphis, according to state permits and satellite imagery. In April, Denver-based startup Crusoe also purchased 13 converted units from provider ProEnergy to power OpenAI's Stargate site in Abilene, Texas. [1][2]

Missouri-based ProEnergy supplies the hardware by taking General Electric engines commonly found on Boeing passenger planes and retrofitting them to run on natural gas. Each turbine produces up to 50 megawatts of electricity, which would normally power around 15,000 homes. An analyst at Rystad Energy noted that using these engines can bring a data center online in 24 months, whereas securing a 100-megawatt grid connection in Northern Virginia takes roughly seven years, according to the International Energy Agency. [1][2]

The workaround brings notable environmental drawbacks, as the converted engines are less efficient and generate higher pollution than traditional gas plants. Mississippi permits granted to xAI allow a 25% higher limit for volatile organic compounds, sparking challenges from the Southern Environmental Law Center over carbon monoxide, formaldehyde, and alleged undercounting of projected emissions. While the Trump administration encourages data centers to self-generate electricity, experts warn that growing efficiency and emission concerns may limit the long-term viability of the practice. [1][2]