Data Centers Turn to Rebuilt Jet Engines as Grid Connection Queues Stretch to Seven Years

Data center operators facing multi-year waits for grid connections are increasingly turning to rebuilt aircraft engines as a stopgap power source, according to state permits and satellite imagery reviewed by the New York Times. Eight converted jet engines are now generating electricity in Mississippi for xAI's Colossus 2 data center in Memphis, Tennessee, marking a growing trend of behind-the-meter generation that sidesteps overwhelmed utility interconnection queues.

The engines, retired from commercial aircraft and converted to burn natural gas by ProEnergy, are reworked versions of a General Electric engine commonly fitted to Boeing jets. OpenAI adopted the same approach in 2025 at its first Stargate campus in Abilene, Texas, where Crusoe, the Denver-based developer building that site, ordered 13 of ProEnergy's converted units in April. The arithmetic behind the shift is stark: connecting 100 megawatts to the grid in Northern Virginia takes roughly seven years, according to the International Energy Agency, while a converted jet engine can have a data center running in 24 months, Rystad Energy analyst Audun Martinsen told the Times. "It's all about getting power as fast as possible," Martinsen said.

The supply crunch extends beyond grid connections to the turbines themselves. Most new generating capacity is not expected to arrive until 2031, an executive at Mitsubishi Power told Bloomberg last week, and orders placed with large equipment makers now run to the end of the decade. The geography of the Mississippi-Memphis arrangement illustrates another advantage: power does not have to be generated where it is consumed, and a converted engine can be placed close enough to a campus to avoid transmission constraints altogether. The queue has already reshaped corporate decisions elsewhere—Nscale's data center in Loughton, England, lost anchor customer Microsoft after the local utility said it could not deliver sufficient power until the early to mid-2030s.

Industry experts caution that the converted engines are not a permanent solution. They are less efficient than turbines built for conventional gas plants, emit more per unit of electricity, and require more units to produce the same output. Some experts told the Times that efficiency and emissions will matter more in a year or two, once the first wave of sites is running and comparisons with grid power become concrete. The commercial history of recent months suggests the market is still searching for the right machine: in September, Crusoe walked away from a $1.25 billion agreement to buy 29 natural gas turbines from Boom Supersonic, a startup developing a supersonic passenger jet that has also begun selling a version of that engine as a data center power source. A Crusoe spokesperson said at the time that the company still intended to use turbines.

Operators who deploy converted engines take on responsibilities normally handled by power companies, including fuel supply, maintenance intervals and emissions permitting, at sites originally planned as computer buildings. The New York Times' count of eight units at one site is a data point rather than an industry total, and nothing in the public record specifies cost per megawatt, carbon emissions compared with grid connections, or total installations across the sector. How long the engines stay in service depends on two clocks running at different speeds: the time required to energize a substation and the time for a rival to ship a model that renders an existing cluster obsolete. If the latter is shorter, the engines may never be replaced at all—simply decommissioned alongside the hardware they were purchased to power.