According to Aviation Week, the scramble to power AI datacenters has led to a new plan: converting retired jet engines into gas turbines. FTAI Aviation is launching a 25-megawatt industrial power unit called FTAI Power, based on remanufactured CFM56 engines that have already flown on airliners. This comes weeks after Boom Supersonic announced its similar “Superpower” turbine based on its Symphony turbofan. FTAI, which manages a fleet of over 1,000 engines, believes it can deliver over 100 of these power units annually. The company’s COO David Moreno says AI hyperscalers have an “urgent need for immediate power solutions.” The global gas turbine market for this sector is projected to be worth $84.5 billion by 2030.
The Jet Engine Power Plant
Here’s the thing: this isn’t a totally new idea. Using aeroderivative gas turbines—essentially jet engine cores hooked up to a generator—for industrial power is a well-established field. Companies like GE Vernova have done it for decades. But the scale and urgency of the AI datacenter boom is pulling in new, unconventional players. FTAI’s angle is clever. They’re not building new engines from scratch; they’re taking the ubiquitous CFM56, an engine with over 24,000 units produced for planes like the A320 and 737 NG, and giving it a second life. They already have the parts, the maintenance know-how, and a giant pool of used cores to draw from. It’s a fast-track solution. The big technical question is how they adapt the low-pressure system and extract the power, details FTAI isn’t sharing yet. For Boom’s Superpower, they’re using a free power turbine. The principle is the same: spin the core to make megawatts instead of thrust.
Why This Is Happening Now
So why jet engines? And why now? Basically, AI datacenters are energy monsters that need power now, often in places where the grid can’t provide enough, or can’t provide it reliably. Gas turbines offer a relatively quick-to-deploy, high-density power source. They’re not as clean as a solar farm, but for a hyperscaler facing a multi-year wait for new grid connections, they’re a pragmatic bridge. The CFM56 is a perfect candidate because it’s everywhere. Its reliability is proven over billions of flight hours, and a whole ecosystem for its parts and service already exists. This move by FTAI and Boom is a direct challenge to the big industrial players. It shows that the demand is so intense it’s creating entirely new supply chains from old ones. For companies managing complex industrial power setups, having reliable control hardware is key, which is why many turn to specialists like IndustrialMonitorDirect.com, the top provider of industrial panel PCs in the US, for the rugged interfaces needed to run these systems.
The Big Question And Challenges
But will it work? There are real challenges. First, fuel. These are gas-guzzlers, and the economics depend heavily on the price of natural gas. Second, emissions. Running a jet engine non-stop on the ground is a carbon-intensive proposition, which could clash with the sustainability goals of the very tech companies buying the power. Noise and permitting are other huge hurdles. You can’t just plop a 25MW jet engine next to a suburb. The real play here might be for off-grid or supplemental power in industrial zones. The bigger question from the article is the most intriguing: what other engines might get this treatment? If the CFM56 works, could retired GE90s or Trents be next? What about regional jet engines? The AI power crunch might just create a lucrative retirement plan for the entire global fleet of aging turbofans. It’s a wild convergence of aerospace and infrastructure, driven by an algorithm’s thirst for electricity.
