Is America Running Out of Electricity?
Electric power demand is rising. But we have the solutions to meet it.
This story is part of a larger series exploring changes in America’s power grid. Check out our other stories on enhanced geothermal, battery storage, and rooftop solar. And subscribe to receive more stories like this in your inbox.
After 15 years of flat growth, demand for electricity in America is growing again.
Much of the news coverage of the country’s surging electricity demand has focused on the most dazzling cause: artificial intelligence. This year, Microsoft—just one of many companies trying to capitalize on the new technology—will build a new data center every 3 to 6 days.
It’s hard to read the news right now without running across a new projection for how much power data centers will consume over the next decade.
But AI is only part of the larger story of electricity demand.
Across the country, one of the primary drivers of growth is climate progress. The speed of this progress is evident in how quickly utilities and electric grid planners have had to update their assumptions about the future recently.
In 2020, Duke Energy Carolinas predicted that roughly 400,000 electric vehicles would be in operation in its service territory by the mid-2030s. Last year, they had to revise that number upward to 2 million EVs—4 times more vehicles that will need to draw power from the grid to charge their batteries every day.
Three years ago, the grid operator in New York, NYISO, projected that building electrification trends would result in 3,724 GWh of additional electricity demand by the year 2028. Two years later, they had to revise that forecast upward to 5,856 GWh—a near doubling of demand due to some of the country’s most ambitious building electrification policies.
In Georgia, which has become ground zero for the country’s industrial renaissance, the changes have been even more dramatic. $28 billion in clean energy manufacturing investments have been announced since Democrats passed their historic climate bill in the summer of 2022, more than any state. Companies producing everything from electric vehicles to heat pumps have flocked to Georgia.
With each new factory announcement has come a request to plug into the grid. And in the last year and a half, the requests have come like a tsunami.
Historically Georgia Power, the state’s utility, has signed up 100 MW per year of what they refer to as “large load projects”—basically big factories and commercial buildings that want to use a lot of power. In 2022 alone, the utility signed up 2,197 MW of new large load projects, a staggering 22 times the historical average.
Amid all of this, the Wall Street Journal has warned that growing electricity demand is creating an “electricity crisis.” The Washington Post recently warned that America is “running out of power.”
But make no mistake: These trends are signs of progress. Anyone who wants to see the world’s second-largest emitter of greenhouse gasses decarbonize its economy should celebrate them.
Zero growth in the power sector would indicate a continuation of the status quo. It would imply a future that looks much like our present—one filled with fossil fuel-powered vehicles and home heating equipment, one in which the materials necessary to decarbonize are all made in countries run by authoritarians with little regard for human rights or freedoms, one cooked by greenhouse gasses.
And yet there is a crisis brewing in the electricity sector, not because we’re running out of power, but because many of the companies and regulators in charge of building our electric grid are turning to fossil fuels to meet this growing demand.
Utilities plan to meet new demand with fossil fuels
In the Southeast, Duke Energy has said they want to build 8,900 MW of new gas plants over the next decade. To put the magnitude of that number in perspective, consider these two facts: Duke Energy is a single utility that serves about 5% of the country’s electricity customers, and 8,900 MW is as much natural gas capacity as the entire country brought online last year.
Duke Energy isn’t the only utility planning a massive gas build-out in the coming decade.
Georgia Power wants to meet its growing industrial electricity demand by building 1,400 MW of new gas plants. Tennessee Valley Authority, the federally owned utility, recently announced that it plans to build 8,850 MW of new capacity over the next decade.
But it’s not just new natural gas that is threatening to boost emissions in the power sector.
All across the country utilities have also delayed coal power plant retirements in the wake of growing demand. As Bloomberg recently reported, roughly two dozen coal plant retirements have been delayed everywhere from Kentucky to North Dakota.
In taking inspiration from oil and gas companies, many utilities who loudly announced climate targets just a few years ago have quietly rolled back those plans.
Earlier this year, FirstEnergy, which serves more than 6 million customers, dropped its plan to cut emissions by 30% by 2030. Duke Energy recently told regulators that it would be unable to meet its target of cutting emissions by 70% by 2030 and would instead need to delay that target by 5 years.
In defending their plans, utilities have said that America must choose between meeting new power demand and meeting its climate targets.
But this is a false dichotomy.
Burning fossil fuels isn’t the only way to solve the country’s power crunch. And many of the other solutions available would be cheaper than the gas plants utilities have proposed.
We have cheaper, cleaner solutions available
Last month, the energy modeling nonprofit Energy Innovation released a report showing how utilities can meet growing demand without building new fossil fuel plants. They found many alternatives that don’t just result in fewer emissions, but would also cut utility bills across the country.
Some of those solutions have a long track record of success and simply need to be deployed at a greater scale. Energy efficiency—the breakout star of last year’s global climate conference—is one such solution.
Between 2006 and 2021, utility energy efficiency programs cut electricity demand by 220 TWh, as much power as the entire state of Florida consumes each year. Efficiency programs like these were one of the main reasons why power demand remained flat for the last 15 years. But many utilities are investing less in these programs than they were in 2019.
Duke Energy—the utility that wants to build a country’s worth of natural gas capacity—invests less in energy efficiency than most utilities in the country, according to the American Council for an Energy-Efficient Economy (ACEEE).
Efficiency programs can cut the total amount of electricity consumed each year, but much of the recent power panic has been over short periods of intense demand. Utilities like Georgia Power argue that the only way to meet these short bursts of demand is to build gas “peaker plants.”
But building gas plants to run for just a few hours on a hot summer day is inefficient and costly. Energy modelers have consistently found that it’s often more effective to just pay people to use less power during these times, a strategy known as demand response.
Last year, policymakers in Texas proposed spending $10-18 billion building new gas power plants to meet the state’s growing electricity demand. But when ACEEE ran its own independent model, they found that demand response and efficiency programs could achieve the same goals at a fraction of the cost. The programs ACEEE modeled would save $1.3 billion per year for seven years.
None of this research or technology is new. Demand response programs have already saved Americans billions of dollars. What is a bit more novel, is the idea of bundling a group of electricity consumers—say all the water heaters in a neighborhood or a bunch of EVs in a city—and paying them all to stop using power for a set amount of time. Do this and you have yourself a “virtual power plant” (VPP) that addresses peak demand by reducing power instead of generating it.
A recent Department of Energy report found that the country could meet the bulk of its new power demand between now and 2030 by deploying more VPPs. And they could do all this while saving consumers and businesses $10 billion per year in grid costs.
While the term “virtual power plant” might seem futuristic, this technology too has been around for years. Depending on how you define the term, VPPs in America are already cutting 30 to 60 GW of peak demand.
Of course, cutting demand isn’t the only way to solve the power crunch problem. Electrifying transportation, buildings, and industry will require doubling or quadrupling the amount of power generated over the next few decades. To meet this demand, we also need new sources of clean electricity.
Here, too, the ethos of efficiency can be employed to get new clean energy sources on the grid quickly.
In recent years, it’s become more difficult to build clean energy projects. One of the main causes of this slowdown is transmission congestion. Between 2012 and 2016, an average of 2,000 miles of high-voltage transmission lines were built in the United States. Between 2017 and 2021, that number fell to 700 miles. Last year, just 250 miles of high-voltage lines were built.
At a time when America needs to be building more transmission lines to connect new solar and wind farms to the grid, it’s building less than at any time since the New Deal era.
Many utilities have pointed to these trends in justifying why they plan to build new gas plants instead of solar, wind, and batteries. But the transmission bottleneck doesn’t have to restrict clean energy’s growth in the short term. There are plenty of ways to get more out of our existing system.
Building renewable and storage projects at the site of retired coal plants—locations with huge power lines extending into city centers—could unlock a massive 250 GW of new clean energy capacity, according to a study by RMI. That study found that the greatest opportunities for growth were in the Southeast, where many utilities are proposing gas plants.
We can also use new technologies to squeeze more out of our existing transmission system. Reconductoring existing transmission lines could roughly double the total amount of transmission capacity over the next 1-3 years, according to a University of California Berkeley study.
All of these solutions can help utilities across the country meet growing demand. And they can do it for a fraction of the cost of proposed gas plants.
There’s no doubt America has the technological and physical ability to meet surging demand. We don’t face a choice between electrification or clean energy manufacturing growth and cutting power emissions. There isn’t a choice between economic growth and climate action.
The choice we do face is over whether or not we let utilities—companies that require regulatory approval to enact any of these plans—have their way or if we force them to adopt cleaner, cheaper solutions.
Why didn’t transmission congestion affect new gas plants like wind solar batteries (Many utilities have pointed to these trends in justifying why they plan to build new gas plants instead of solar, wind, and batteries.)?
Any new load from AI should be required to be met with renewable build out and a *requirement* that build out not be fossil fuel based. It's 2024. AI, bitcoin, all these high-energy intensive industries need to be responsible in their energy load.