NEW YORK, March 26: As public concern grows over the massive electricity demands of expanding data centers, the U.S. technology industry faces pressure to scale back power use during high-demand periods.
The power sector and regulators urge tech companies to reduce energy consumption at their data centers when necessary. This is crucial to avoid blackouts and soaring power bills during peak demand times. According to the Electric Power Research Institute (EPRI), data center power usage could rise to 17% of U.S. power supplies by the end of the decade.
"It's about delivering power when demand is at a peak," said U.S. Energy Secretary Chris Wright at the CERAWeek conference. During a recent winter storm, the Department of Energy directed data centers in the PJM Interconnection grid to use backup generators, freeing up power on the grid.
PJM, covering the largest data center market globally, predicts supply shortages as early as next year unless demand slows. Implementing flexible energy management could save up to $150 billion in capital investments over the next decade, research from Duke University suggests.
To address rising costs and power outages, data center investors and energy suppliers advocate for "demand response"-adjusting energy use when utilities request. Google and Nvidia have announced initiatives to lower consumption during peak periods.
EPRI recently published a framework for making data centers more flexible, aiming to expedite connections and enhance grid reliability.
"You're seeing a transition period where everyone wants to participate, and we're figuring out how to make it happen," said Jennifer Cahill of Black & Veatch.