AI Power Shortage Sends Data Centers Searching for Used Turbines and Retired Jet Engines
Brownstone Worldwide | November 9, 2026 | AI & TECHNOLOGY
The artificial-intelligence boom is colliding with a less glamorous reality: new data centers need dependable electricity, and power-generating equipment cannot be manufactured as quickly as developers want to build. That mismatch is sending companies into the market for refurbished turbines and even retired aircraft engines.
A supply-chain scramble
Reuters reported November 9 that worldwide orders for gas turbines exceed 100 gigawatts, while estimated annual production capacity is only 60 to 70 gigawatts. U.S. delivery times can average six years, and some negotiations now contemplate deliveries in 2030 or 2031. The figures describe industry conditions, not the capacity of any single company.
Developers and utilities are consequently exploring used machinery, restarting idle equipment and converting former jet engines into power-generation systems. The secondary market is tightening as more buyers compete for a limited supply.
Why an old turbine is not an instant solution
Refurbished equipment may cost less than newly manufactured units, but operators must assess maintenance history, efficiency, remaining service life, emissions permits and suitability for continuous duty. Electricity connections, environmental approvals and fuel availability can delay a project even after a turbine has been purchased.
What the energy race means for communities
Large data-center campuses can increase demand on local grids. The effect on consumer bills depends on utility rate design, who pays for upgrades and whether new generating capacity arrives on time. Residents and elected officials should ask for clear public accounting of infrastructure costs, water use and emissions rather than assume every new data center has the same footprint.
What creators should know
AI services appear to users as apps and subscriptions, but behind each prompt sits expensive physical infrastructure. Power scarcity could influence service prices, availability and the economics of smaller AI providers. Businesses building their operations around these tools should maintain backups and track costs.
What happens next
Watch new turbine orders, used-equipment prices, grid interconnection approvals and alternatives such as gas engines and battery-backed power systems. Reuters notes that Finland’s Wärtsilä has secured about three gigawatts of U.S. data-center orders for gas-engine equipment.
Source: Reuters, November 9, 2026.
