Bold start: The AI rush is colliding with a gas turbine crunch, and that tension could slow the entire technology sprint. And this is the part most people miss: the electricity needed to power AI may not be readily available at the pace demand is rising.
As AI hubs around the world push the boundaries of computing, the appetite for reliable power has surged in tandem. The problem isn’t just that demand is high; it’s that supply—specifically gas turbines to convert natural gas into steady, scalable electricity—has not kept up. The result could force the AI revolution to either decelerate or rely more heavily on coal to keep the lights on.
Natural gas has long been marketed as a transitional fuel between coal and oil and the era of wind and solar. Yet as concerns about emissions grew, some activists portrayed gas as a villain even more damaging to the atmosphere than coal. Then the AI wave arrived, shifting the spotlight once again.
Big Tech is accelerating its AI ambitions, and electricity consumption is exploding. For decades, growth in power demand was modest, which reflected in flat production for gas turbines. Now manufacturers must scale up quickly, while the industry is still catching up. In the meantime, electricity consumers are turning to expedient options, including repurposed jet engines being repurposed as gas turbines.
Siemens Energy, among the top three gas turbine makers, reported a record quarter for its gas services business, with 102 new turbines added to the backlog. About 40% of these orders came from the United States, and 35% from Europe. Siemens also signaled plans to invest about $1 billion in grid equipment production.
GE Vernova plans to invest roughly $600 million to expand turbine manufacturing capacity, targeting as many as 80 heavy-duty turbines per year, translating to around 20 GW of generation capacity. The manufacturer framed the investment as a means to help customers meet rising demand and to bolster the United States’ energy security and global competitiveness.
Mitsubishi, the third major turbine producer, announced last year that it would double turbine production capacity in response to soaring demand. The CEO noted that a 30% capacity increase was not enough to satisfy growing needs, and fulfilling orders remains the top priority.
Even with these expansion plans, materialization takes time. Meanwhile, industrial electricity users need immediate solutions, leading some to repurpose jet engines for power generation. The Wall Street Journal reported that converting jet-engine turbine tech into power turbines is drawing substantial investor interest. One player, FTAI Aviation, saw its stock rise about 42% after announcing this line of business, which can convert a Boeing 737 engine into a power turbine in roughly 30–45 days.
In this high-stakes race, waiting lists for major turbine manufacturers stretch for years. If momentum lags or supply tightens, investors may rethink commitments, potentially cooling the AI acceleration. Analysts warn that even with new capacity, equipment shortages will continue to constrain electricity supply growth.
Wood Mackenzie’s Ed Crooks pointed out that gas-fired plants will be a meaningful part of US power expansion through 2030, but heavy-duty gas-turbine availability will likely remain a bottleneck. He notes a current five-year wait for new turbines, a timeline far too long for the needs of data-center operators and AI developers. Replacing large turbines with jet-engine derivatives isn’t a viable full substitute, given the much smaller per-unit capacity.
That leaves a difficult fork: either Big Tech slows its AI momentum to adapt to grid constraints, or it seeks electricity from other sources. Some analysts argue for a push toward solar, paired with storage and backup generation, but this path requires substantial investment in batteries and reliability measures to cover baseload needs. Another possibility is a renewed willingness to rely on coal for baseline power, which could force reconsideration of coal plant retirements and affect energy-transition plans.
Bottom line: even with abundant baseload generation, the underlying physics and supply gaps could force AI growth trajectories to bend. The race’s tempo may bend or pause as the world negotiates between powering innovation and maintaining a sustainable, reliable grid.
By Irina Slav for Oilprice.com
