House Requests Federal Energy Regulatory Commission AI Report
Tech companies are reporting higher greenhouse gas emissions.
Teralyn Whipple
![House Requests Federal Energy Regulatory Commission AI Report](/content/images/size/w1200/2024/07/250508605_411246360379300_5390534206752177176_n-1.jpg)
July 18, 2024 – House Energy and Commerce Committee Chair Cathy McMorris Rodgers, R-Washington, and Energy, Climate, and Grid Security Subcommittee Chair Jeff Duncan, R-South Carolina, requested details on how the Federal Energy Regulatory Commission is addressing growth in electricity demand in a letter dated Tuesday.
“After years of minimal growth, electricity demand in the United States is projected to grow nationally at a significant pace through the end of the decade. It is anticipated that much of this demand growth will come from a surge in the number of data centers and the growing uses of artificial intelligence,” read the letter.
It added that estimates show annual growth of 5 to 6 percent through the end of the decade. Data centers could consume as much as 9.1 percent of all electricity in the U.S. by 2030.
This surge in demand “comes at a time when the North American Electric Reliability Corporation has repeatedly raised concerns over the adequacy and reliability of the grid,” said the letter. “These risks are due to a confluence of factors, including state and federal policies that have forced premature retirements of reliable generation without adequate replacement generation resources and electric infrastructure.”
Google reported in its annual environmental report that its greenhouse gas emissions rose 13 percent in 2023 and 48 percent since 2019 primarily due to data centers used to power artificial intelligence. This is despite Google’s stated goal to achieve net-zero emissions across all operations by 2030.
“As we further integrate AI into our products, reducing emissions may be challenging due to increasing energy demands from the greater intensity of AI compute,” said Google. Similarly, Microsoft reported an emissions jump of 29 percent since 2020, pointing to AI as the culprit.
Rodgers and Duncan requested that the FERC commissioners explain what the agency is doing to assess challenges of new demand growth, the potential for co-locating data centers, and how growing demand will add costs and strain to the grid, among other things.
AI requires significantly more energy than other processes, with a single Chat-GPT request using approximately 10 times more energy than a Google search, according to the International Energy Agency.
Image generation uses more than 60 times the energy used for text generation, found a study by Carnegie Mellon University and Ai startup Hugging Face.
The IEA estimates that global electricity consumption by data centers could double between 2022 and 2026, adding the equivalent energy consumption as “at least one Sweden or at most one Germany.”
Goldman Sachs Research estimates that data center power demand will have grown by 160 percent by 2030.
The House Energy and Commerce Subcommittee on Energy, Climate and Grid Security held a hearing in June to discuss energy demands of emerging technologies. Policymakers called for the strengthening of the electricity grid to keep up with growing needs and stay ahead of innovation in China.
“We should be striving to assure our innovators can provide a more prosperous future for all Americans and help secure our energy and technological leadership for the next century,” said Rodgers in her opening statement.
In May, the White House launched the Federal-State Modern Grid Deployment Initiative in an effort to bring together state, federal agencies and power sector stakeholders to drive advancements in the U.S. power grid.