California produces lots of renewable energy. More than the state can sometimes use.

Researchers have offered varying ideas for what to do with this “curtailed” energy. Next 10, a nonpartisan nonprofit centered on innovationaround environmental and economic issues, has proposed siting data centers near renewable energy farms and timing high-energy computations to the generation of surplus renewable energy.

Industrious Labs, a climate nonprofit based in Ohio, has suggested powering the numerous industrial activities in areas like the California Central Valley with renewable energy.

“There’s a lot of opportunity to put curtailed renewable energy to work, and while data centers are certainly part of the conversation, we believe the bigger opportunity lies in industrial applications that deliver broader public benefits, particularly for communities in the Central Valley,” Teresa Cheng, Industrious Labs’ California director, said in an email.

Next 10 recently released a report exploring the scope of curtailed renewable energy in California, and what to do with it. The reasons for curtailment are numerous, and are generally due to imbalances with supply, demand and the capacity to deliver the energy to where it’s needed.

Some 3.4 terawatt-hours of renewable energy were curtailed in California in 2024, enough to power about 500,000 homes, according to the report, “Curtail to Compute: Siting Datacenters to Leverage California’s Stranded Renewable Energy.” This was a 29 percent increase from 2023 curtailment levels.

It’s not that this excess energy is not needed, experts say. But transmission lines are often at capacity, limiting how much renewable energy can effectively move from remote solar and wind farms to population centers. And given that this infrastructure is so costly to construct, the Next 10 report proposes situating data centers near renewable energy where a combination of battery storage and managed operations could be synced with renewable energy production.

To be clear, the siting of data centers “depends on a complex combination of many factors,” Benjamin C. Lee, a professor of electric and systems engineering with the University of Pennsylvania, and one of the authors of the report, said.

“Our study focuses primarily on energy availability, prices, and the role of batteries and finds that data centers in the Central Valley could benefit from curtailed energy that is otherwise unused, low electricity prices for a large share of their computation, and economically viable energy storage,” Lee said via email.

The growth of AI and its energy-intense computations is generating new streams of electricity demand across the country, leading to a congested grid, higher costs for utility customers and increased dependence on fossil fuels. In the next 10 years, demand for electricity is set to increase 20 percent to 50 percent, according to research by the American Council for an Energy-Efficient Economy.

Industrious Labs has proposed incentivizing industrial operations to move away from fossil fuels, with cheaper electric rates tied to renewable power production. Industry ranks as California’s second most polluting sector, according to Industrious Labs, making operations like food processing, coffee roasting and other Central Valley activities primed for using clean, renewable energy.

“Redirecting cheap, curtailed electricity toward electrifying these processes could reduce emissions, improve local air quality, support good-paying job preservation and creation, and put downward pressure on rates for all customers,” Cheng said.

A recent study by Industrious Labs, “Unlocking industrial electrification in California,” noted the industrial electric rate in California in March 2025 was 19.84 cents per kilowatt-hour, twice the national average. The state is also home to the largest number of gas-fired boilers in the country, according to Industrious Labs. Transitioning just low-temperature boilers to cleaner electric models, powered by competitively priced renewable energy, could displace 5.5 million metric tons of CO2 annually, according to the study, and help to reduce air pollution in an area notable for high rates of asthma and respiratory problems.

“One of the biggest barriers preventing facilities from taking advantage of this cheap power is California’s current outdated utility rate design, which makes switching from gas to electric more than five time more expensive for manufacturers,” Cheng said.

Legislation to modernize the electricity rate structure for industrial purposes, allowing for cheaper rates when using surplus renewable energy during off-peak hours, passed the Senate Standing Committee on Energy, Utilities and Communications this week.

Data centers are known for being large users of not just energy but water, a precious resource in many parts of California, particularly the Central Valley where farmers are idling more than 1 million acres a year, according to researchers at the University of California Merced — driven in part by decisions related to water scarcity.

Data center water use must be taken into consideration when siting these facilities, Lee said.

“There are ways to design data centers that use less water, but more power — and vice versa,” he said. “Exploring the right trade-off based on each geographic location’s constraints would be a natural next step in the research.”