AI, Data Centers, and the U.S. Electric Grid: A Watershed Moment
Summary
Analysis of the impact of AI-driven data centers on U.S. electric grid demand, projected growth, regulatory challenges, and policy recommendations.
Key quotes
"Data center demand will grow from 176 TWh in 2023 (4.4% of total U.S. electricity) to 325-580 TWh (6.7-12.0%) by 2028."
"In July 2024, a voltage fluctuation in Virginia triggered disconnection of 60 data centers, causing a 1,500-megawatt power surplus."
"Future data center energy usage projections range from 200 TWh to over 1,000 TWh by 2030 due to uncertainty and data opacity."
"Texas and Virginia are developing distinct regulatory responses to data center growth, highlighting tensions over cost-sharing and grid reliability."
This report by the Belfer Center examines the surge in electricity demand driven by AI-powered data centers, estimating future growth and analyzing its potential strain on the U.S. grid. It highlights regulatory challenges and case studies from Virginia and Texas, where rapid data center expansion is prompting divergent policy responses. The authors emphasize the need for standardized load profiles, transparent financing, and equitable cost-sharing mechanisms to mitigate risks like stranded assets and grid instability.