Utilities across the U.S. Southeast are claiming that a massive buildout of data centers and factories will force them to construct gigawatts of new fossil gas-fired power plants over the coming decade — a fleet large enough and dirty enough to potentially put U.S. climate goals out of reach.
However, critics of these plans say that utilities have cleaner and cheaper alternatives to reliably manage surging new power demand, and that state utility regulators in Georgia, the Carolinas and Tennessee need to require them to explore those options.
For the moment, though, these utilities, which serve tens of millions of customers, appear set on a fossil-fueled power expansion that also promises them additional profits for years to come — profits that environmentalists and consumer advocates argue will be reaped at the expense of the climate and their customers.
“The problem we face now is that everyone is searching for power,” said Simon Mahan, executive director of the Southern Renewable Energy Association. “Utilities across the Southeast are scrambling to find every last megawatt they can get…. They are trying desperately to get these new large-load customers, because they make more money when they sell more power.”
In some regions, these potential new customers are big data centers to serve the skyrocketing demand for enterprise computing power, artificial intelligence and cryptocurrency mining. In others, they’re factories for electric vehicles, lithium-ion batteries and solar panels supported by billions of dollars of federal incentives from the Inflation Reduction Act.
The exact figures vary from region to region, but most of the utilities are now forecasting high single-digit percentage growth rates every year through the end of the decade. Demand for electricity over the past decade and a half has stayed flat or even declined, so growth on that order would be a sea change for utilities.
Whether this new electricity demand will emerge at the speed and scale these utilities are predicting is unclear; utilities have overestimated demand growth before. Some critics have accused utilities of seizing on hype around the rapid expansion of energy-intensive artificial intelligence technology to win approval for gas plants that are not really necessary.
But even if these projections are accurate, critics say new fossil gas plants aren’t the answer. They argue that gas plants are polluting, unreliable and likely to become stranded assets in the near future, as climate imperatives and cheaper clean-energy resources force them to close before utilities have recouped their costs.
“Are we facing a ‘grid crisis’ in the U.S. due to data center and factory expansion? No,” Tyler Norris, former vice president of development at independent power producer Cypress Creek Renewables and now a doctoral student at Duke University, wrote in a recent social media post. “But doomsday thinking appears to be spreading and increasing the risk of poor decision-making.”
There are more reliable and cost-effective ways to deal with an increase in electricity demand that must be explored further, Norris and others point out, like building solar and wind power — paired with batteries — or enlisting power-hungry corporate customers to use less electricity when demand is at its highest. These options also help achieve climate goals, rather than threaten them.
Experts say the only way to course correct is for state utility commissions to intervene — something each one has an opportunity to do in the coming months.
What the utilities want — and what it would cost the climate
The Southeast utilities’ current plans, if approved, could have a disastrous climate impact.
These utilities are also planning to add gigawatts of solar power, batteries and other carbon-free resources and to close down gigawatts of coal-fired power plants. But taken together, the carbon impacts of a large gas expansion would eclipse the gains of these projects, according to an analysis from the Southern Environmental Law Center.
Last month, Georgia Power, a business unit of multi-state utility holding company Southern Company, secured a preliminary settlement plan with Georgia regulators that would allow the utility to fast-track 1,400 megawatts of new gas-fired power plants in the next three years. Georgia Power sought permission last year to rush the plan through regulatory approval to meet what it now forecasts will be 17 times more power demand growth than it had predicted it would need just 18 months earlier, due to new data centers and factories being planned in the state. The settlement plan still requires the vote of the Georgia Public Service Commission, expected to be held on April 16.
Duke Energy, one of the country’s biggest utilities with operations in six states, recently added significantly more fossil gas plants to its plan for supplying North Carolina and South Carolina, boosting its request to a total of 9,000 megawatts. That’s nearly three times the amount it requested to build in a 2022 proposal, and would delay its ability to meet a commitment under North Carolina law to cut its carbon emissions by 70 percent from 2005 levels by 2030. Duke says the buildout is needed to meet a forecasted 12-percent increase in electricity demand by 2038, driven largely by dozens of economic development projects in both states.
In South Carolina, state lawmakers are advancing legislation backed by utilities Dominion Energy and Santee Cooper to fast-track construction of a 2,000-megawatt fossil gas-fired power plant. The bill’s sponsor, Speaker of the House Murrell Smith (R), has cited a looming “crisis point” for the state’s grid as a result of rising demand from factories and growing population.
And Tennessee Valley Authority, the federal entity that generates power for 10 million people across seven Southeastern states, is developing a plan that could include 6,600 megawatts of new gas-fired power plants to replace coal plants and serve growing power demand. TVA delayed the release of its official plan last month, leaving uncertain just how much new gas-fired power it will propose.
“If all of these gas proposals across the Southeast do come to fruition, I think we’re going to have a huge confluence of issues between climate and reliability and affordability,” said Maggie Shober, research director of the Southern Alliance for Clean Energy.
What’s frustrating, Shober said, is that these utilities “were already proposing new gas before this load growth showed up. Duke and TVA have each flip-flopped on who has the largest gas buildout in the country, but they remain first and second by a pretty large margin.”
Gudrun Thompson, senior attorney and energy program leader at the Southern Environmental Law Center, agreed that “gas has been the answer to multiple problems” for Southeastern utilities.
“A couple of years ago it was the bridge fuel they needed to accommodate renewables. After Winter Storms Uri and Elliott” — major storms that led to catastrophic power outages in Texas and rolling blackouts in the Southeast, respectively — “it was what they needed for reliability. Now it’s what they need to meet data center load,” she said. “Whatever the problem is, it seems the reflexive solution is to build a new gas plant.”
The nonprofit environmental advocacy law firm estimates that utilities in Alabama, Georgia, North Carolina, South Carolina, Tennessee and Virginia are planning to retire about 25,000 megawatts of coal by 2038, while simultaneously rushing to build 33,000 megawatts of new gas plants over the next decade. At the same time, utilities across the country need to cut power-sector emissions 80 percent by 2030 compared to a 2005 baseline to meet U.S. Paris Agreement commitments — any new fossil-fueled power plants are likely to put those targets out of reach, Thompson said.
Why fossil gas plants are not the most reliable choice
Clean energy and consumer advocates in the Southeast are also worried that new gas plants wouldn’t even solve the problem utilities are citing to justify them: making the grid more reliable in the face of rapid demand growth.
That’s because of the nature of the power shortfalls Southeastern utilities face. Day-to-day, the utilities have few problems meeting the electricity needs of their customers. But they do struggle to meet demand during grid “peaks” — the handful of hours during the hottest summer afternoons and coldest winter mornings when customers need the most power.
However, gas-fired power plants failed to perform that critical task during Winter Storm Elliot in December 2022. Supply fell short by more than 70,000 megawatts of generation capacity across the U.S. Southeast at the time, forcing Duke Energy and the Tennessee Valley Authority to institute rolling blackouts. Much of the failures occurred at gas-fired power plants that were forced offline because equipment froze or pipelines couldn’t deliver, calling into question the assumption that building yet more gas infrastructure will solve future problems.
“We’ve been trying to make the argument that gas plants not only didn’t save the day during these winter storm events, they were a big part of the problem — whereas renewables pretty much performed as expected,” Thompson said.
What performed “really well was demand response,” she said, referring to programs that pay households and businesses to turn down power use or switch to backup power during grid emergencies. Many of the customers driving the boom in demand — primarily data centers, which can shift electricity usage and tap backup power to ride through outages — “could really participate in demand response programs, and deliver a lot of peak load reduction,” she noted.
That view was echoed by Norris in a presentation to South Carolina regulators last September. Norris highlighted the cascade of problems — power plant failures and an inability to import power over transmission lines from neighboring regions — that forced Duke Energy to institute rolling blackouts on the morning of Christmas Eve during Winter Storm Elliot.
But the presentation also showed that the duration of the demand spike that triggered the grid emergency was a relatively brief three to four hours, he said in an interview. That’s a gap that can mostly be met by lithium-ion batteries, at a cost that’s competitive with gas-fired power, or by commercial facilities like data centers agreeing to reduce their power demand.
But Duke Energy’s most recently updated plan, released in January, doesn’t take that potential flexibility into account, he said. Instead, it assumes that “new industrial and commercial load is 24/7/365 — zero flexibility. And they’re taking the total draw of the new load and putting it right on top of their winter peaking load forecast. It’s a maximalist, worst-case scenario.”
Peakers versus baseload gas plants: A big difference in carbon and cost
At the very least, utilities struggling to meet peak demand could focus on building the type of gas power plant built for that specific purpose, Norris said. But that’s not what’s happening.
Instead, utilities are proposing to build huge numbers of power plants that are designed to run regularly. It’s a “solution” that’s not matched to the problem of managing infrequent, hours-long grid peaks — and the impact of such a decision could reverberate for decades.
There are two types of gas-fired power plants: single-cycle combustion turbines (CTs) that can be “ramped up” to meet unexpected surges in power demand within minutes, and combined-cycle gas turbines (CCGTs) that make up the majority of gas-fired generation capacity in the U.S. today.
CTs operate at lower efficiency than CCCTs, but they usually run between 10 and 20 percent of the year — a stark difference from CCGTs, which on average run over 50 percent of the time in the U.S.
Duke Energy’s most recent proposal to regulators is heavily weighted toward CCGTs. In between the 2022 version and its January update, Duke has doubled the amount of CTs it is asking regulators to let it build by 2035 — but nearly tripled the amount of CCGTs it wants to build by then.
The fear is that because these power plants are designed to run far more often than CT plants, they will crowd out lower-emissions resources — and delay the shift to a carbon-free grid.
“Once these large combined cycle units are on the system, it will be very tempting to use them long into the future,” Norris said.
Why gas plants force extra costs onto customers
A rash of new gas plants could actually increase costs for customers. Electricity from newly built wind and solar farms is already cheaper than power from newly built gas plants. And lithium-ion batteries have fallen in cost to the point where using them to store clean power and discharge it later is competitive with building new gas-fired peaker plants.
In other words, “gas plants are no longer the cheapest option,” Shober said. But for utilities, they may still be the most convenient — and, crucially, the most profitable — route.
Regulated utilities like Duke Energy and Georgia Power pass the cost of building power plants and other capital expenses on to their customers in the form of higher rates on their utility bills, meaning it’s customers, not utilities, who pay for new gas plants.
Gas prices can also spike unexpectedly, whether due to seasonal shortages during cold winter months or global supply shocks such as Russia’s invasion of Ukraine. But under current regulatory structures, utilities can pass the cost of those spikes on to customers as well.
These disconnects are rooted in what’s known as “cost-of-service” regulation, under which investor-owned utilities are allowed to make a guaranteed rate of return, typically set in the 8 to 10 percent range, on investments in capital assets like power plants and power lines. The aim is to encourage utilities to build the infrastructure they need to deliver energy to everyone they serve.
But that also means that “investor-owned utilities under cost-of-service ratemaking have incentives to maximize capital expenses and little to no incentive to improve efficiency,” Norris said.
In some cases, that can lead to utilities taking actions that run counter to their customers’ interests — including the large customers utilities are trying to attract.
That’s certainly true of the data center developers that are part of the Clean Energy Buyers Association (CEBA), a trade group representing more than 400 companies with clean energy goals, such as Google and Microsoft, that have pledged to attain round-the-clock carbon-free energy by 2030.
For years, CEBA has pushed Southeastern utilities to expand clean energy to help large corporate customers meet their goals. In a recent Georgia Power public hearing, Priya Barua, CEBA’s director of market and policy innovation, noted that overbuilding fossil fuel capacity in Georgia “would result in higher costs for existing customers and make it more difficult for existing customers to meet their sustainability targets.”
“If you don’t have solutions to empower customers to bring clean energy to the system, there’s no guarantee that those customers are going to site there,” Barua told Canary Media. “I think that is something that Georgia Power and regulators have to factor in when making decisions.”
What can regulators do?
While other utilities across the country are also seeking permission to build new gas-fired power plants, the largest buildout is slated for the Southeast. The confluence of climate, cost and reliability concerns over these utilities’ plans puts a burden on utility regulators to carefully examine them, said Mike O’Boyle, senior director for electricity policy at think tank Energy Innovation.
Under the regulatory compact that allows utilities to operate as monopolies in the territories they serve, “utilities only recover costs that are prudently incurred,” he said. “That prudency standard is rooted in whether the utility fully examined alternatives.”
In a March report, O’Boyle and Energy Innovation colleagues laid out several reasons why the plans of Southeastern utilities may not meet that standard. While demand for electricity is almost certain to grow over the coming decade, “the exact pace of the growth in the short term remains uncertain, particularly with the addition of factories and data centers,” it noted. “Therefore, short-term investments by utilities should prioritize low-regrets, flexible options that avoid locking in expensive and potentially stranded assets.”
The report highlighted that utilities in the regions with the largest projected growth by percentage — the Northwest, Southwest, and California — are “markedly not moving to add gas to their resource plans.”
Environmentalists’ fight against new gas plants has been complicated by the shift in power demand growth patterns, however. Grid reliability is an increasing concern among regulators and industry groups, as coal retirements accelerate and it becomes clear that — whatever the magnitude and speed — electricity demand is set to rise in the years to come. This uncertainty presents utilities and regulators with a conundrum, said Danny Freeman, senior partner, energy and utilities with consultancy West Monroe.
“They have to pull together a credible projection of what the load is going to look like,” he said. Data center developers “are looking across the country and trying to find the cheapest possible energy supplier across any number of states. When these deals will be done, and when they’ll kick in is a huge question mark.”
At the same time, “there are realities to serving this growing load that have to be dealt with,” he said. Data centers may not be willing to commit to shutting off their power during hours of peak grid demand as a precondition of being able to connect to utility grids, he said. And renewable energy, with its variability tied to weather, “presents a challenge to grid operators,” he said.
But just because cleaner and cheaper options are more complicated than building new gas-fired power plants doesn’t absolve utilities and regulators of the responsibility of examining them, O’Boyle said.
“Regulators have to start by asking the right questions,” he said. “Are there viable projects that can use your existing interconnection points from retiring coal plants? Are there bids in prior RFPs that are still viable, and have you considered them as alternatives to your new gas plant? Have you considered energy efficiency or flexible load?”
Public utility commissions must insist that utilities examine these options more thoroughly as an alternative to new fossil gas-fueled power plants, and compel them to share their assumptions and methods for assessing their relative merits, he said. If they don’t, it’s very hard for all parties involved to find a mutually acceptable path forward.
“I don’t know if there’s a smoking gun here for utilities acting in bad faith,” he said. “I think they were caught off guard, as many analysts were, by this load growth — and they’re looking for solutions that can work. Their number one incentive is that the lights don’t go out. And it’s a lot easier to say ‘one plant solves my problem.’”
“But the stakes are really high,” he said. “You can’t just jump into a billion-dollar expenditure — whatever it costs to build a gigawatt of new gas — especially when these large consumers are coming to the table and saying, ‘We want something else, and we can help.’ It’s worth taking a breath and working collaboratively on solutions that are lower risk and lower cost, and actually meet customers’ needs.”