Executive Summary
This RAND report examines the challenges and facilitators for siting and constructing AI data centers on U.S. federal lands, focusing on energy constraints and public-private partnerships. It evaluates potential sites using a cost model and qualitative framework, identifying key barriers like grid interconnection and permitting complexity. The report recommends a pre-RFLP federal site-screening framework to accelerate data center deployment.
Why It Matters
This document is crucial for defense and geopolitical analysts as it addresses the strategic implications of AI data center infrastructure development on federal lands, impacting national technological leadership and energy security. It highlights the challenges and policy recommendations for integrating critical AI infrastructure within national defense and energy strategies.
Key Takeaways
- Seventeen federal sites, including those under the Department of War, have been selected for AI data center development, but only three have progressed to development agreements, indicating slow progress.
Source evidence · PDF page 6
Seventeen federal sites have been selected across the U.S. Department of Energy (DOE)10 and the U.S. Department of War (DOW)11 for data center development, with no publicly announced U.S. Department of the Interior sites to date. Only three of the 17 sites have progressed to development agreements with private companies; given planned operational timelines as early as 2027, several of these projects will require accelerated permitting, construction, and energy deployment schedules.
- Persistent challenges for data center development on federal lands include grid interconnection constraints, complex permitting, fragmented multiagency coordination, and political uncertainties.
Source evidence · PDF page 26
Evidence from prior coordinated infrastructure acceleration initiatives suggest that grid interconnection constraints, permitting complexity, fragmented multiagency coordination, and political uncertainties are persistent challenges that could affect data center development on federal lands.
- A pre-Request for Lease Proposals (RFLP) federal site-screening framework is recommended to rank candidate locations based on energy costs, generation capacity, and grid feasibility to mitigate delays and accelerate deployment.
Source evidence · PDF page 58
For future site-selections, we recommend establishing a pre-RFLP federal site-screening framework, similar to the one introduced in Chapter 4, that ranks candidate data center locations based on levelized cost of energy, cost per MW firm generation, time-to-firm-power-delivery, grid interconnection feasibility, and infrastructure readiness.
Strategic Relevance
The strategic relevance lies in the critical need for robust AI computational infrastructure to maintain technological superiority and national security. Siting these energy-intensive data centers on federal lands presents both opportunities and significant challenges related to energy supply, grid stability, and regulatory hurdles, directly impacting the pace of AI development for defense applications.