As AI infrastructure expands at unprecedented speed, data centre operators are facing a growing power generation crisis. With conventional gas turbine production unable to meet soaring demand, several companies are now repurposing retired aircraft engines into industrial grade aeroderivative turbines, a solution that could keep the lights on for the world’s AI boom.
Industry experts warn that this is not a short term problem. The International Energy Agency (IEA) estimates that AI data centres will consume over 10% of global electricity by 2030, putting extraordinary pressure on grids already strained by electrification and renewables transition delays.
Why Jet Engines?
Large scale AI and cloud data centres require tens or even hundreds of megawatts of continuous power, comparable to small cities. Yet, new heavy frame turbines from GE Vernova, Siemens Energy, and Mitsubishi Power now have lead times stretching three to five years or more due to supply chain backlogs, high nickel prices, and parts shortages.
“There just aren’t enough gas turbines to go around, and the problem is probably going to get worse,” says Paul Browning, CEO of Generative Power Solutions and former head of GE Power & Water and Mitsubishi Power.
That bottleneck has forced operators to innovate. Energy engineering firms like ProEnergy are now refurbishing decommissioned aircraft engines such as GE’s CF6 80C2 and LM6000 models, originally designed for Boeing and Airbus jets, into ground based power systems branded as PE6000 units. Each system can generate up to 48 MW of electricity, enough to power roughly 40,000 homes or sustain a mid sized AI data centre.
These mobile, modular units can be delivered, installed, and activated within four to six months, compared to years for new installations.
Technical Edge: How Aeroderivative Turbines Work
Aeroderivative turbines are essentially jet engines turned power plants, compact, high speed systems that burn natural gas to spin turbines and generate electricity.
Their key advantages include:
- High power density, more megawatts per square foot than conventional gas turbines.
- Fast start up, capable of reaching full load in under 10 minutes.
- Modularity, units can be trucked, swapped, or maintained individually.
- Lower emissions, advanced combustion technology keeps NOx emissions as low as 2.5 ppm, versus 10 to 25 ppm for legacy industrial models.
These characteristics make them ideal for AI driven workloads, which demand instant scalability, stable voltage, and uninterrupted uptime. According to ProEnergy, 21 such units were sold this year alone to two hyperscale clients, collectively providing over 1 GW of interim capacity, roughly the energy needed to power a metropolitan district.
Bridging the AI Infrastructure Boom
AI driven data centres now form the backbone of cloud computing, language models, and generative AI services. But powering them has become a race against time.
Traditional utility grid connections can take up to five years due to regulatory reviews and local permitting. Aeroderivative turbines act as bridge power plants, giving operators immediate, independent capacity while waiting for permanent grid hookups or renewable microgrids. Once permanent solutions are online, these portable turbines can be relocated, resold, or reconfigured as emergency backup units, offering a unique flexibility that fixed infrastructure lacks.
Environmental and Sustainability Considerations
Although these repurposed turbines rely on natural gas, their lower emissions and reuse of existing hardware make them relatively sustainable compared to new builds. Refurbishing retired aviation cores avoids the high carbon cost of manufacturing new turbines, while extending component lifespans by up to 15 additional years.
Some operators are experimenting with biomethane and hydrogen blending, allowing partial decarbonization of fuel use. According to the Delft University of Technology, gas turbine systems are essential to maintaining stability during the global transition to renewables because of their fast response capabilities, something solar and wind still lack.
Still, environmental groups caution that over reliance on natural gas could lock in fossil dependencies, delaying the shift toward zero carbon data infrastructure.
The Wider Implications for Energy and Technology
The AI sector’s insatiable energy appetite is already reshaping global power strategies. Microsoft, Google, and Amazon collectively announced more than 20 GW of new data centre capacity under construction, much of it awaiting dedicated power sources.
In the US, states like Virginia, Texas, and Iowa are fast tracking turbine powered substations near hyperscale campuses. Meanwhile, in Asia and the Middle East, hybrid solutions pairing gas turbines with grid scale batteries are being deployed to balance renewables and meet AI uptime guarantees.
This convergence of aviation technology and AI computing marks a new chapter in energy engineering, one where innovation is driven by necessity, not convenience.