Taking a closer look at the investment plans of GE Vernova, the U.S.’s leading gas turbine manufacturer, suggests that, even as energy demand ramps up, these delays will persist. Rather than potentially overinvest in the face of rising demand and suffer the consequence of falling prices, GE Vernova and its competitors are committed to capital discipline, lengthening their order book, and defending shareholder value. Their reluctance to invest, while justified in some part by the nature and history of the industry, will threaten policymakers’ push for energy abundance ― to say nothing about economic growth or innovation.
Meanwhile, supply chain shortages will constrain the growth of clean energy generation. Inadequate investment in gas and an insufficient buildout of renewables in the face of unprecedented demand growth ― these are a toxic cocktail for the American energy system. Forget visions of an all-of-the-above energy strategy. How about none of the above?
Energy project developers, utilities, and investors have already started adjusting their gas buildout expectations and timelines. NextEra CEO John Ketchum stated in an earnings call that new gas projects “won’t be available at scale until 2030, and then only in certain pockets of the U.S.” That’s due not only to turbine queues, but also to an historically sluggish and increasingly expensive gas project development environment. “The country is starting from a standing start,” he added. “This is an industry that really hasn’t seen any active development or construction in years … all of that puts pressure on cost.”
Even in Texas, where lawmakers created the Texas Energy Fund to provide $10 billion of concessional financing to new gas power plants, delays are biting developers’ balance sheets. Just last week, private developer Engie withdrew two loan applications for gas peaker plant projects due to “equipment procurement constraints.” There’s no other way to spin it — the turbines are the problem.
Given that wait times and reservation payments drain developers’ liquidity and increase their financing costs, energy giants are trying to cut the line. Chevron is partnering with GE Vernova to develop up to 4 gigawatts of gas power plants for data centers. NextEra also announced a partnership with GE Vernova, through which the two companies will co-develop and co-own “multiple gigawatts” of natural gas power plants.
It’s safe to say that GE Vernova’s power division is riding high. The company’s investor materials suggest a heady growth trajectory. Gas turbine equipment orders rose 66% between 2023 and 2024, from 41 turbines to 68 turbines. Those 68 turbines represented about 20 gigawatts of capacity, double 2023’s order book. Developers reserved 9 gigawatts more of turbines; those reservations will turn into contracted production orders by 2026. At this point, 90% of GE Vernova’s total order volumes are in its backlog; for its power division, that represents almost $74 billion of equipment delivery and service contracts.
The company plans to invest $300 million into its gas power business in the next two years. And CEO Scott Strazik is pitching investors on continued growth. “Given our expansion plans to produce 70 to 80 heavy-duty gas turbines per year beginning in the second half of 2026, up from 48 this year, we are positioning to meet this demand. We expect to grow our gas equipment backlog considerably in 2025, even as we ramp to ship approximately 20 gigawatts annually starting in 2027, and expect to remain at that level going forward,” he said on the company’s Q4 earnings call.
That last sentence should give readers pause: GE Vernova has plans to build no more than 20 gigawatts of turbines per year, and developers that miss the cutoffs will just have to queue up for the next year’s order book. Why the limit?
Strazik laid out two key reasons. First, he’s looking for developers’ “receptivity to pay for what I will call premium slots” in 2028 and 2029, to “capture every dollar of price with the precious slots available,” as he told investors during a different presentation in December. GE Vernova’s annual report, which it released in February, refers to this strategy ― inviting desperate developers to bid up the price of scarce turbines ― as “expanding margins in backlog.” Second, the company remains hampered by supply constraints, particularly on ramping up its new heavy-duty and H-class turbines. There are real limits to how much more GE Vernova can build, and how quickly.
But over the longer term, it looks like GE Vernova is intentionally committing more to capital discipline rather than to broader capacity expansion. The company has $1.7 billion in free cash flow, a third of which it will return to shareholders through dividends and stock buybacks. And Strazik wants to avoid using the rest to underwrite what he sees as dangerous overcapacity that could threaten GE Vernova’s profitability. “I think we have to be very thoughtful to make sure that we don't add too much capacity, even though we are starting to sell slots into 2029,” he said during the investor update. “We're going to continue to be very sequential on how we invest.”
Strazik’s current strategy prioritizes productivity and efficiency improvements at GE Vernova’s existing plant in South Carolina over building new manufacturing facilities. Some capacity expansion, sure ― but no new plant. “Concrete's expensive, cranes are difficult,” he told investors. The company’s main competitors abroad, Mitsubishi and Siemens, have the same backlogs, and Mitsubishi, at least, is responding with a similarly measured strategy. Mitsubishi CFO Hisato Kozawa is open to some degree of capacity expansion, but maintains that Mitsubishi can only increase capacity “in a very planned manner with discipline. And if we need more capacity, we may want to first improve the rotation of the capacity.”
To the CEOs of all three companies, history would likely seem to justify this discipline. In 2017 and 2018, years of investment into capacity expansion coincided with a near-total collapse in global demand for gas turbines. This market crash was most likely the combined effect of low energy demand growth, energy efficiency improvements, continued use of coal power across Asia, the growing share of renewable energy on the grid, and investors’ realization that solar and wind energy could meaningfully undercut gas on price. All three companies laid off tens of thousands of employees, and the crash contributed to the complete breakup of General Electric and its partial spin-off into GE Vernova last year.
These gas turbine manufacturers are also some of the world’s leading wind turbine blade manufacturers, and a similar fate befell that sector in the past decade. Large-scale capacity expansion and competition for contracts drove down costs and margins across the supply chain — only for those to move sharply in reverse when supply chains froze up during the pandemic and interest rates shot up in 2023. Now offshore wind projects are plagued with problems and, at least in the U.S., President Trump’s de facto moratorium on offshore wind development has further reduced the sector’s ability to bounce back. These companies have been burned before. It only makes sense not to repeat past mistakes.
Combined-cycle gas turbines are complex machines, similar to airline engines in their intricacy and in the extensive global supply chains required to produce them. But their leading producers, afraid of getting over their skis, won’t undertake the massive upfront investments required to increase their long-term production capacity. Where does this leave the energy transition?
Bankers and energy project developers alike can see the writing on the wall. Beth Waters, managing director for project finance at Japanese bank MUFG, has insisted that “renewables have to be part of the electricity mix. It cannot just be gas-fired.” NextEra’s Ketchum has said the same: “Renewables are here today,” he stated during the latest earnings call — unlike gas. Jigar Shah, the head of the Department of Energy’s Loan Programs Office under President Biden, wrote on LinkedIn about his confidence that “batteries will be deployed at 10X the capacity of combined cycle natural gas units over the next 4 years.” Major utility companies, for their part, still have large clean energy procurement targets in their integrated resource plans. The smart money is clearly betting that an “all-of-the-above” energy deployment strategy will be better than eschewing any particular energy source.
They’re being optimistic. Not only does new utility-scale renewable energy take years to build, there’s also not yet enough transmission and longer-term energy storage on the grid to balance the variance in existing solar and wind resources. That prevents solar and wind from providing the kind of 24-hour stable power that corporate and industrial customers demand. Expanding energy storage and transmission resources will depend not just on regulatory reforms to permitting and interconnection, but also on resolving the severe bottleneck in grid transformers, where analysts believe capacity expansion has also failed to meet roaring demand, resulting in wait times of three to four years. (GE Vernova and Siemens build grid transformers too.) The status quo has left hundreds of gigawatts of clean energy projects across the country stuck in a regulatory and financing limbo, and the grid issues that tie up clean energy development will further constrain gas power growth.
To be sure, President Trump’s “energy dominance” agenda seems to favor the development of clean firm energy resources, such as nuclear and enhanced geothermal, to cut through the literal gridlock. The gas turbine manufacturers, all of which build steam turbines for nuclear power, stand to benefit from interest in restarting and upgrading now-shuttered plants. But building new nuclear projects currently takes at least 10 years, if not more. The singular new nuclear project built in the U.S. in the past three decades was completed seven years late and almost $20 billion over budget.
Enhanced geothermal might fare somewhat better ― its drilling technology comes straight from the fracking sector, and the pilot projects of companies like Fervo are achieving impressive heat and electricity production targets. Still, to turn heat into electricity, Fervo needs turbines, too. While enhanced geothermal projects need organic Rankine cycle turbines, as opposed to the combined-cycle gas turbines used in gas power plants, commodity market strategist Alex Turnbull theorizes that the commonalities between the two will threaten geothermal developers with the same delays and bottlenecks. (Fervo’s turbine supplier is an Italian subsidiary of Mitsubishi.)
The tech giants building data centers are already investing in new power ― but if neither nuclear nor geothermal can be deployed at scale in the absence of massive policy support, then that leaves tech companies paying for whatever energy sources their regional electricity grid relies on in the meantime. As Cy McGeady, a fellow at the Center for Strategic and International Studies, told Heatmap last year, “Nobody is willing to not build the next data center because of inability to access renewables.” But drawing so much from existing resources ― mostly gas, but also nuclear ― without building sufficient new power leaves less for every other energy consumer.
Policymakers on both sides of the aisle have their work cut out for them to avoid a crisis born of a failure to build any energy resource adequately: They must execute a thorough grid overhaul while also punching through the specific supply chain bottlenecks that prevent energy generation from being built quickly. Regardless of energy demand projections, these are goals worth pursuing. They advance grid reliability, energy affordability, and decarbonization, as well as accommodate any necessary energy supply growth.
Still, it’s worth questioning the prevailing narratives around load growth. It’s not clear how much energy data centers in particular will actually require. Not only have innovations like DeepSeek challenged market assumptions about tech companies’ investment requirements, but recent research also suggests that load growth projections could fall significantly if data centers’ energy demand were more flexible. Not to mention that data center developers often make duplicate interconnection requests with different utilities to maximize their chance of securing a power agreement.
Our energy grid will need a lot less hot air if data center demand goes up in smoke ― and that would be a relief for American consumers and the climate alike. But courting a gas turbine crisis should itself give policymakers pause. The fact that our energy system is at a point where neither turbines nor transformers nor transmission is available in sufficient capacity to meet any policymaker’s vision of energy abundance suggests that our leaders must reorient the government’s relationship to industry. During periods of economic uncertainty, capital discipline might appear rational, even profitable. But the power sector’s profits are, through rising energy bills and more frequent climate disasters, revealed to be everyone else’s costs. Between clean energy and fossil fuels — between what Americans need and what private industry can provide — the energy transition is shaping up to be, quite literally, a power struggle.
EIA
EIA