The U.S. is nearing the end of one of its hottest summers on record. Across the nation, heat waves have driven peak electricity demand on some days to levels far exceeding seasonal averages.
Grid operators rely on so-called “peaker” plants to ensure they will have enough supply to meet these demand surges. Peaker units can start up quickly and at relatively low cost, but they typically burn more fuel per unit of electricity produced than other types of fossil fuel units.
Because they are less efficient than other plants, peakers typically run only during high-demand periods. Historically, peakers have run for less than 10% of the year, often for just a few hours at a stretch.
Nonetheless, their higher emissions per unit of electricity produced raise environmental and health concerns. As of 2021, there were 999 peaker plants across the U.S., in all 50 states. About 70% of these plants burned natural gas, and the rest were powered by oil and coal.
To reduce air pollution and combat climate change, the U.S. is shifting away from fossil fuels and increasing its use of renewable energy sources such as wind and solar power. Ironically, though, as climate change generates more frequent and intense heat waves, many electricity systems are increasingly relying on peaker plants to balance fluctuations in renewable power generation. Proposals to build new peakers or extend the lives of old ones have stirred controversy in states including Wisconsin, Massachusetts, Texas and New York.
My research focuses on the economic and environmental costs and benefits of producing electricity. Here’s how the clean energy transition is changing the role of peaker plants and some other options for keeping the lights on during peak demand periods.
South Bronx residents call on New York state to shut down local peaker plants in 2022. These plants are disproportionately located in low-income and minority neighborhoods.
Balancing the power supply
For system operators, one key characteristic of a power plant is whether it can produce power on demand. Many renewable resources, including wind, solar and certain types of hydropower, are known as nondispatchable resources because they are governed by nature, producing energy when conditions allow. The cost of generating electricity with them is low, so they are typically used to their maximum capacity.
Power plants that run on fossil fuels or nuclear power are known as dispatchable resources because they can produce power whenever it’s needed. They have higher operating costs than renewables, however, mainly because gas, coal, nuclear and oil plants must buy fuel in order to operate.
Some of these plants – historically, those that run on coal or nuclear fuel – are called baseload plants. They generate power relatively cheaply but take time to start up and ramp up to full power. Intermediate units produce power at a higher cost for each additional megawatt-hour produced, but they can cycle up and down more quickly than baseload plants. Peakers have the highest costs per megawatt-hour, but they can adjust their output very quickly.
Historically, baseload units operated year-round, with intermediate units adjusting output to meet short-term demand fluctuations. Peaker plants were used only during rare peak demand periods.
But as power suppliers add more wind and solar energy to the grid, they are using dispatchable fossil fuel units more frequently to balance changes in renewable generation – for example, to run air conditioners when the sun goes down but temperatures are still high. This favors units that can quickly change production levels, even if they are less fuel-efficient. The result is a growing role for peaker plants.
The Texas A&M University System has taken its first steps toward creating a peaker power plant network on land it owns across the state of Texas to help stabilize the state’s power grid during peak demand: https://t.co/sTDtzazDwF pic.twitter.com/WXknlN9EaU
— Texas A&M System (@tamusystem) May 22, 2024
Environmental justice flash points
Electricity production from fossil fuels in the U.S. has decreased with large-scale investment in wind and solar generation. But fossil fuel-burning power plants still produce about 60% of U.S. electricity – and those plants emit pollutants that contribute to climate change and degrade local air quality.
Exposure to sulfur dioxide, nitrogen dioxide, particulate matter and ozone is linked to respiratory and cardiovascular illnesses and premature death. While overall air pollution has decreased in the U.S. in recent decades, low-income and minority neighborhoods still suffer disproportionately from poor air quality.
One 2022 report estimates that 32 million Americans live within 3 miles of a peaker plant. In 2024, the U.S. Government Accountability Office reported that historically disadvantaged racial or ethnic communities were statistically more likely to be located closer than average to peakers.
Other ways to meet peak demand
How else can electricity supply and demand be balanced? One option is using batteries to store electricity when wind or solar output is high, then discharging it when demand exceeds supply from conventional resources.
Although battery investment costs currently are high, they are projected to decrease significantly in the coming decades. In 2023, the U.S. had a total of about 15 gigawatts of battery storage capacity – equivalent to 15 large nuclear power plants – and that figure could double in 2024.
Another alternative is expanding transmission systems, which make it possible to draw on electricity from lower-cost units in distant areas instead of relying on nearby peaker plants. Building new transmission lines, however, comes with significant regulatory, permitting and land use challenges.
A third option is demand response programs, in which electricity consumers pay higher prices during higher demand periods. This could help reduce peaks and valleys in demand across the day, benefiting more efficient but less flexible baseload units designed to run around the clock.
Most consumers, however, don’t currently pay prices that reflect short-term changes in wholesale electricity costs. Moreover, it’s uncertain whether residential customers would alter their consumption based on short-horizon price fluctuations. Technologies such as smart thermostats and energy management apps could help by taking the burden off consumers to manually adjust their electricity use in response to price fluctuations.
View this post on Instagram
Finally, power plant owners can invest in technologies to reduce emissions from fossil fuel units. Peakers typically lack pollution control technologies because they aren’t used very often. Retrofitting older plants to make them more efficient could also help, since they would produce fewer emissions for each unit of electricity.
These investments are costly, so policymakers have to weigh the health benefits of reduced air pollution against the investment costs for power plant owners.
Increasing investment in wind and solar energy is reducing local air pollution from electricity production. But it’s also shifting production away from thermally efficient baseload units that can’t respond quickly to shifts in demand or renewables output. I believe it is increasingly important to explore policies that create incentives for investing in alternatives such as battery storage and transmission infrastructure, as well as in power plant upgrades to reduce pollution exposure.![]()
This article is republished from The Conversation under a Creative Commons license. Read the original article.






Comments
Welcome to the new Resillience commenting system!
We have moved from the former Disqus commenting system to a new system as of September 10, 2026.
We made this change because of privacy concerns regarding Disqus. It's also part of a set of changes to bring more services to the Resilience community under a single login: commenting, (coming September 2026), and online courses (coming 2027). Unfortunately, your old Disqus comments may not reliably link to your account in this new system. Please contact us if you have any questions or concerns.
This quote is just a bold faced lie
" the U.S. had a total of about 15 gigawatts of battery storage capacity – equivalent to 15 large nuclear power plants"
NO we don't have 15 gigawatts of battery storage, we have 15 gigawatt hours of storage.
I am pretty sure that the author knows that but goes with the deception.
15 gigawatt HOURS of storage is the output of ONE nuclear power plant for 15 hours.
Not the output of 15 nuclear power plants for a year like the author implies.
There is a general confusion between GW, which is the measure of instantaneous power output, and GWh which is the measure of energy stored.
Typically, utility scale batteries store enough GWh to produce their maximum output GW for about four hours. This means that if the US has 15 GW of battery output power capability, the energy stored would be around 60 GWh. Unfortunately, I have been unable to find data for current utility scale GWh in the US, but 15 GWh seems very low for 15GW of power capacity.
Lets see…
The size of the US power grid is ~1,300 GW (instantaneous power).
15 GW hours or 60 GW hours is ~1 to ~4 minutes worth of stored electricity.
By comparison a typical coal power plant has 60 days worth of coal stored on site, the oil industry typically has a 90 day supply of oil in the system and there are huge underground caverns filled with natural gas have many days worth of natural gas usage.
To have anything like our current economy you need several orders of magnitude (10,000 or 100,000 ) more storage for renewable energy. (far more likely is the current economy can not be sustained)
I never said otherwise. My only reason for replying to your first comment was that there were reasons why "just a bold faced lie" was an overstatement.
Nothing is going to be more than a tiny partial replacement for fossil fuels. When the carbon pulse ebbs, so will industrial civilization.
I tried looking it up from original data – and I'm so confused as the government spreadsheet didn't have summaries of the MWh (Column X). I had to go in and add the SUM function myself on 26,000 lines – about a third of which (from a fast pan down) were empty of data. From that I got 56 GWh storage. Can anyone else run it in case I stuffed something?
EFFICIENCY FIRST: It hardly needs to be said on a blog like this, but if we built state of the art passive solar ecocities with enough thermal mass – much of our heating and cooling requirements would be met by thicker walls and the right window siting – not electricity.
TOTALLY AGREE with more powerlines and batteries – especially new sodium batteries that do not require any copper or other critical minerals.
DISPATCHABLE: We can make W&S dispatchable by building enough batteries to cover the first few hours, then enough OFF-river pumped hydro for days to weeks! The world has over 100 times more potential sites than we could need.