ThinkPatternGet the app
Story
TECHNOLOGY · JUN 22, 2026

US Data Center Power Demand Projected to Double by 2030

Lawrence Berkeley National Laboratory and independent researchers report that AI-driven power demand will surge, though data centers previously helped lower retail electricity rates.

The Lawrence Berkeley National Laboratory, with backing from the Department of Energy, reports that U.S. data centers could more than double their share of electricity use by 2030. The analysis estimates these facilities will account for 9.5 to 15 percent of total U.S. electricity consumption by the end of the decade. Researchers found that while hardware efficiency has improved, the massive growth in computational demand driven by artificial intelligence has offset those gains.

Parallel research by Asa Watten, John Bistline, and Geoff Blanford indicates a different historical trend. Their pre-print paper suggests that data centers actually caused average retail electricity rates in the United States to fall modestly between 2015 and 2024. This decrease is attributed to economies of scale in transmission and generation, which spread fixed costs across a larger volume of kilowatt-hours.

Despite previous price drops, future stability faces significant headwinds. Researchers cited supply constraints such as transformer shortages, backlogged natural gas turbine orders, and tariffs on Chinese photovoltaic arrays. In response to rising demand and grid instability, the Federal Energy Regulatory Commission has launched a probe to determine if regional grids are fairly allocating the costs of necessary upgrades. Separately, Chase Bank raised its forecast for data-center capacity growth to 138 gigawatts through 2030.


Reported across 4 outlets
Actors
United States Department of EnergyFederal Energy Regulatory CommissionLawrence Berkeley National LaboratoryJohn Bistline

Keep reading in the app

The full story and every source, free in the app.

Download on the App StoreComing soonGoogle Play