Elon Musk의 xAI가 미시시피주 데이터센터의 전력 공급을 위해 개조 제트엔진 8기를 설치했으며, 해당 허가에는 오염물질 배출 한도를 25퍼센트 높이는 조건이 포함되어 있다.
Permits for xAI's Mississippi power plant authorize carbon monoxide emissions at twice the rate allowed for the site's other turbines. The arrangement is now drawing legal scrutiny in a region that has dealt with the company's permitting problems before.
The plant serves Colossus 2, xAI's artificial-intelligence data center near Memphis, Tennessee. Eight converted aircraft-engine turbines, supplied by Missouri-based ProEnergy, form the core of the installation. ProEnergy refurbishes General Electric engine cores for stationary power generation. According to the New York Times, those cores previously flew on Boeing 747s, though that detail should be confirmed against the original report.
Each turbine can generate up to 50 megawatts. Combined, the eight units represent a potential 400 megawatts of generating capacity.
**Permitted Pollutant Levels Exceed the Site's Own Benchmarks**
The eight ProEnergy units carry a carbon monoxide limit of 4 parts per million, double the limit set for the other 33 turbines on the same property. Volatile organic compounds (VOCs) are permitted at a level 25 percent higher than the site's other units.
Carbon monoxide interferes with the body's ability to transport oxygen. VOCs contribute to smog formation and can carry toxic effects depending on the specific compound and exposure level.
**A Legal Challenge Questions Whether the Permits Are Enough**
A permit appeal filed by the Southern Environmental Law Center on behalf of the NAACP, Young Gifted & Green, and the Safe and Sound Coalition contests Mississippi regulators' approval of the turbines. The challenge alleges that xAI failed to provide adequate guarantees that the turbines meet emissions standards for pollutants including formaldehyde.
The groups also allege that xAI underestimated or failed to account for projected emissions tied to the data center's power sources. They further argue that state regulators should have required pollution offsets or reductions elsewhere in the area. The status of the permit appeal remains unclear, and it is separate from earlier litigation over previously unpermitted turbines.
Neither xAI nor ProEnergy has publicly responded to these allegations. Details of any direct response were not available at the time of publication.
**Why Converted Jet Engines Are Appearing at Data Centers**
Onsite generation using repurposed aircraft engines can reduce a developer's reliance on slow-moving grid-interconnection queues, but the tradeoff is lower efficiency and higher emissions.
ProEnergy builds its units around refurbished aircraft-engine cores, adding cooling, emissions-control, and noise-reduction systems for stationary use. Even with those modifications, the New York Times reports, the converted turbines are less efficient and more polluting than turbines designed for conventional power plants.
Data-center developers are turning to this technology primarily for speed. Grid-interconnection queues are long, and utility infrastructure expansion may not arrive quickly enough to match the surging electricity demand of AI data centers. Onsite generation can reduce reliance on a conventional grid interconnection, though it still requires air permits, fuel supply, and construction approvals. This is a broader industry trend, not an arrangement unique to xAI.
**The Region Has Been Here Before**
This dispute follows an earlier legal fight over xAI's use of gas generators without required permits in a nearby area. That conflict has made cumulative pollution a concrete concern for local residents and environmental groups, not a hypothetical one. The current permit appeal, layered on that history, raises questions about whether successive permitting decisions adequately account for the total pollution burden on nearby communities.
**A Policy Precedent May Be Taking Shape**
How Mississippi resolves these competing permit and legal challenges could influence how other states and data-center developers approach onsite fossil-fuel generation. Regulators who consistently accept elevated pollutant limits set a replicable precedent, and other developers may seek identical terms as AI infrastructure demand grows.
A successful legal challenge, by contrast, could influence future permitting decisions by requiring additional emissions controls, pollution offsets, or more rigorous documentation. The outcome in Mississippi may serve as a reference point, in one direction or the other, for how communities weigh the electricity demands of AI infrastructure against public-health safeguards.