Apr 26, 2024 - Energy & Environment

AI advances may frustrate U.S. climate goals as electric demand surges

Illustration of an electrical outlet with a 1 and zero for holes

Illustration: Sarah Grillo/Axios

Artificial intelligence's thirst for electricity is conspiring with other factors to trigger a sharp spike in electricity demand that the U.S. is only beginning to address.

Why it matters: How lawmakers, utilities, regulators and tech companies manage this trend may determine whether America's emissions reduction goals are met, with more fossil fuel-powered plants under consideration in the near-term.

At the same time, the rise of generative AI, which requires far more compute power than typical cloud computing functions, is also contributing to unheard of growth rates in electricity demand for utilities — particularly ones that serve large numbers of data centers.

  • However, while generative AI requires more electricity, this technology and others could be enlisted to help make data centers more flexible in how they use electricity, rather than constantly running near peak demand.

The big picture: Call it the dark side of the push to electrify everything.

  • From electric vehicles, swapping natural gas water heaters with electric units, plus other efforts to decarbonize as new manufacturing facilities get added, these trends have been incentivized by President Biden's landmark climate law.
  • While beneficial in the fight against climate change, such shifts are adding to demands on a grid that was already showing signs of strain.

Zoom in: Take Dominion Energy, for example, which provides electricity to the hundreds of data centers comprising "data center alley" in northern Virginia.

  • The utility is still fully committed to generating 90 to 95% of its electricity from carbon-free sources, according to a company spokesman.
  • But ravenous data center and residential needs call for installing more baseload power stations that can be reliably dispatched when renewable sources ebb, the company official said.
  • This is leading to planning for new natural gas plants, which burn fossil fuels and can add to global emissions, at a time when many such facilities are being retired and presented with new challenges.

By the numbers: Other utilities are facing the same quandary. Georgia Power, for example, has increased its projected load growth from about 400 megawatts in January 2022 to 6,600 MW more recently.

  • Duke Energy has quintupled its projected load growth from a previous estimate made two years ago, partly owing to data center load increases.
  • According to figures from the Electric Power Research Institute, a nonprofit that researches electricity needs, data centers currently utilize about 2.5% of the total energy generated in the U.S.
  • That is expected to increase to 6% by 2030.

Threat level: Generative AI is unlikely to bring down the grid anytime soon, but it is forcing data center companies, the Energy Department, utilities and utility regulators to get creative when it comes to bringing online new generating sources, transmission lines and wringing more efficiency out of the current system.

  • Energy Secretary Jennifer Granholm told Axios in March that electricity's AI-driven boost is something that worries her, given the potential to undermine the administration's climate goals.
  • She tasked the department's science advisory board with presenting their latest thinking on this issue, which it did on April 9, showing the data center industry as being on the cusp of its third wave of growth.
  • The presentation, from Arshad Mansoor, the head of EPRI, and Maria Pope, president and CEO of Portland General Electric, emphasized the need for utilities to deploy a range of new generating technologies in the near term, including solar and wind along with natural gas. Meanwhile, there's a need to wait for options like small modular nuclear reactors to become available.
  • They noted that data center operators are moving quickly to secure land for building centers with a capacity of between 100 MW to 5 gigawatts.

The intrigue: David Porter, vice president of electrification and sustainable energy strategy at EPRI, told Axios the biggest challenge facing utilities is building new infrastructure to meet rapidly growing energy demands.

  • This is partly because it can take a decade in the U.S. to approve new transmission lines, whereas data centers can be approved and built within 18 to 24 months.
  • "Those two things do not align very well," Porter said in an interview.
  • Another option to pursue is grid-enhancing technologies to better utilize the existing transmission system, he noted.

Context: Many of the major tech companies advancing generative AI have committed to lofty sustainability targets, and they are signing deals with clean tech startups for advanced carbon-free energy options, such as SMRs and even fusion plants, to cut emissions from their own data center fleets in the next few years.

What they're saying: With AI compute converging with other demands, "We're almost in the perfect storm for the demand for green electrons," Christopher Wellise, VP of sustainability at data center firm Equinix, tells Axios.

  • "Construction spending for new manufacturing facilities [has] more than doubled since 2021. In the U.S., we have broader electrification of the grid taking place." He notes the overall "strong emphasis on decarbonization," with generative AI coming on top of that.
  • He cautioned though that the demand growth numbers given by many entities are worst-case scenarios.

The bottom line: Expect to hear a lot more about this topic in the next few years, as AI's compute demands grow and new manufacturing plants come online.

Go deeper