Grid connection delays for AI data centers can stretch to years; a 100 MW connection in Virginia can take about seven.
Some operators are converting rebuilt jet engines to burn natural gas and generate power, a faster approach that is less efficient and more polluting than conventional gas generators.
The solution of improvised generators
Companies are taking aircraft engines and modifying them to function as generators. With this method, a data center can be up and running in 24 months, avoiding years of waiting for upgrades to public grids. The Trump administration is supporting this practice, encouraging data centers to generate their own electricity.
Who uses turbines?
Several companies have already started implementing this plan:
- xAI: Elon Musk's company is using eight such turbines in Mississippi for the Colossus 2 data center.
- OpenAI: It plans to use similar units at its facility in Abilene, Texas.
- Crusoe: The company ordered 13 units from ProEnergy in April 2026. In September 2026, Crusoe walked away from a $1,25 billion deal to buy 29 turbines from Boom Supersonic.
The effects of energy shortages
The difficulty of accessing electricity is already causing problems for large-scale investments. For example, Microsoft was forced to end a partnership in England, as the local grid could not provide the necessary energy before the period 2030-2035.
Environmental cost
Despite the speed advantage, jet engines converted into generators have significant disadvantages compared to conventional solutions designed specifically for power generation. They are less efficient and cause more pollution, highlighting the environmental risks of this temporary solution.
↳
READ ALSO
HACKS.GR / RELATED
AI Unlocked
Ballistic Ventures prepares new $490 million fund
↗
>Stay connected with Hacks.gr
Set us as a preferred source on Google and find us on social media.
Preferred source
↗
end of article
Top comments (1)
The operational question is not just whether turbines can cover demand, but what reliability model they create: fuel logistics, maintenance windows, emissions constraints, and how quickly the site can transition back to grid power. Generation close to the load changes the failure domains as much as the power source.