Federal Sites IDed for Large-Scale Data Center and Power Projects

The U.S. Department of Energy (DOE) recently announced it had identified 16 federal sites across the country that could be used as potential locations for developing large-scale data centers and associated power generation projects.

This initiative, aimed at supporting the quick-rising energy demands of artificial intelligence infrastructure, marks a significant push from the federal government to integrate advanced computing with broad U.S. energy infrastructure resilience.

The Trump administration earlier this year issued an executive order promoting leadership in AI development to solidify the U.S. at the forefront of this technology. That order called for the establishment of an action plan for accomplishing the goal, and revokes any policies or directives that would act as barriers.

The selected sites, spanning regions from California to Virginia, include DOE-owned lands and facilities such as national laboratories, power administrations, and even areas adjacent to military installations. Notably, the National Energy Technology Laboratory (NETL), which operates in both Pittsburgh, Pa., and Morgantown, W.Va, is among the identified sites. This inclusion illustrates the potential for AI infrastructure development in large energy research hubs, aligning advanced data processing demand with existing energy innovation and generation projects.

According to the DOE’s Request for Information (RFI), issued in April, the department is probing interest from private developers and investors to construct, own, and operate data centers and co-located power generation assets on these federal lands. The goal is to take advantage of underutilized spaces to meet the escalating power requirements of AI data center use and development, which are projected to significantly increase national electricity demand over the coming decades.

The rapid rise in development of energy-intensive AI data centers is causing concern that the country’s power generation supply will not be adequate to maintain a reliable electric grid, unless quick action is taken to build more generation facilities and transmission infrastructure. One government estimate is that U.S. power demand will double within five years, rising from 176 terrawatt hours (TWh) is 2023 to as much as 580 TWh in 2028.

This has led many data center developers to consider co-locating their own energy supply in close proximity. The DOE has emphasized a flexible approach in the RFI, encouraging proposals that integrate renewable energy, nuclear power, and natural gas to support these energy-intensive facilities. The RFI aims to “gather information on potential development approaches, technology solutions, operational models, and economic considerations associated with establishing AI infrastructure,” a release states.