Lead â As powerâgrid connections and new turbine projects stall, AIâfocused data centers are installing converted jet engines to generate electricity on site, according to TechCrunch. The move reflects mounting pressure to secure power quickly for expanding compute workloads, even though the engines are less efficient and more polluting than purposeâbuilt gas turbines.
Powerâgrid bottlenecks push data centers toward selfâgeneration
According to TechCrunch, many AI data centers are encountering âyearsâlong linesâ for grid connections and for the construction of new turbineâbased power plants. The NewâŻYorkâŻTimes, cited by TechCrunch, reported that firms are now turning to retired aircraft engines that have been rebuilt to burn natural gas. These repurposed engines are not intended as permanent powerâplant solutions, but they offer a rapid path to electricity for facilities that cannot wait for conventional infrastructure.
Rystad Energy analyst AudunâŻMartinsen, quoted in the report, explained that speed is the primary driver: âItâs all about getting power as fast as possible.â He added that jetâpowered units can enable a data center to become operational within 24âŻmonths, whereas attaching 100âŻmegawatts of capacity to the grid in Northern Virginia typically requires about sevenâŻyears, based on International Energy Agency data referenced by TechCrunch.
Highâprofile adopters test the jetâengine approach
TechCrunch identified several highâprofile projects that have already installed converted jetâengine turbines. ElonâŻMuskâs xAI is reported to have placed eight such units in Mississippi to support its âColossusâŻ2â data center in Memphis. The engines are supplied by ProEnergy, a company that rebuilds GeneralâŻElectric jet engines commonly used on Boeing aircraft.
OpenAI followed a similar path in 2025, installing gas turbines at its first âStargateâ site in Abilene, Texas, according to the source. The campus is being built by Crusoe, a Denverâbased startup that specializes in AI data centers. Crusoe ordered 13 ProEnergyâconverted units in April, as reported by TechCrunch.
Earlier, Crusoe had considered a larger purchase from Boom Supersonic â a firm best known for its supersonic passenger jet â which was offering a version of its jet engine as a power plant for data centers. In September, Crusoe walked away from a $1.25âŻbillion deal for 29 naturalâgas turbines from Boom Supersonic, but a company spokesperson told TechCrunch that the firm still intends to use turbines in its projects.
Gridâdelay impacts on business decisions
The powerâsupply lag is already influencing commercial relationships. TechCrunch reported that Microsoft withdrew as the anchor tenant for Nscaleâs dataâcenter project in Loughton, England, after the local utility indicated it could not deliver sufficient power until the early to midâ2030s. The delay illustrates how insufficient grid capacity can jeopardize major investments in AI infrastructure.
Mitsubishi Power, a leading global supplier of powerâgeneration equipment, told Bloomberg â as referenced by TechCrunch â that most new power capacity is expected to come online by 2031. The timeline underscores the growing gap between AI compute demand and traditional electricity supply.
Policy environment and environmental tradeâoffs
According to TechCrunch, the current U.S. administration is adding pressure on data centers to generate their own electricity, encouraging onâsite solutions such as the jetâengine conversions. However, the source notes that these retrofitted engines come with notable drawbacks. They are less efficient and emit more pollutants than turbines designed specifically for stationary power generation. Moreover, achieving the same output requires a larger number of engines, which could amplify both cost and environmental impacts.
Brief analysis (opinion)
The shift to repurposed jet engines appears to be a stopgap measure born of necessity rather than a strategic longâterm energy solution. While the rapid deployment timeline offers a compelling advantage for firms racing to meet AI compute needs, the lower efficiency and higher emissions may become increasingly problematic as regulatory scrutiny and corporate sustainability goals intensify. If gridâexpansion accelerates or new lowâcarbon generation technologies become more accessible, the reliance on such shortâterm fixes could diminish.
Conclusion
TechCrunchâs reporting shows that a growing segment of AIâdriven data centers is turning to converted jet engines to bypass prolonged gridâconnection timelines. The approach delivers power quickly but sacrifices efficiency and environmental performance. As the powerâgeneration landscape evolves and policy pressures mount, the industry may need to balance immediate operational demands with longerâterm sustainability considerations.
Source: TechCrunch
Frequently Asked Questions
Why are data centers opting for converted jet engines instead of waiting for grid connections?
According to TechCrunch, the primary reason is speed; converted jet engines can be operational within about 24âŻmonths, whereas grid connections for comparable capacity can take up to seven years.
Which companies have installed these jetâengine turbines?
TechCrunch reports that ElonâŻMuskâs xAI, OpenAI, and the AIâfocused dataâcenter builder Crusoe have all installed or ordered converted jetâengine turbines from ProEnergy.
What are the main drawbacks of using repurposed jet engines for power generation?
As reported by TechCrunch, the engines are less efficient, produce higher emissions, and require more units to generate the same amount of power compared with purposeâbuilt gas turbines.