RAND researchers analyzed U.S. power availability for large artificial intelligence (AI) data centers, assessing 17 U.S. Department of Energy sites, two private centers, and three retired plants using a multicriterion framework. Most sites showed moderate potential; Stargate and Pantex in Texas and the Kansas City National Security Campus in Missouri ranked highest. Reusing existing infrastructure is key to meeting AI power needs sustainably.

Evaluating Potential Artificial Intelligence Energy Capacity at Different Data Center Sites

Ismael Arciniegas Rueda, Austin Smidt, Robin Wang, Hye Min Park, David Gill, Henri van Soest

ResearchPublished Mar 12, 2026

As artificial intelligence (AI) models grow more complex, their computational demands are driving unprecedented increases in electricity consumption. Meeting this surge will depend on the U.S. power grid’s ability to deliver reliable, scalable energy to next‑generation data centers.

This RAND report—the third in a series on AI‑related power demand—explores where in the United States sufficient energy capacity might exist to support hyperscale AI development by 2030. Building on earlier analyses of national generation trends and grid constraints, this report focuses on the geographic dimension: which sites and regions offer the most-favorable conditions for large‑scale data center siting.

The report introduces a structured framework for evaluating potential locations based on five categories of characteristics: energy supply, the energy system, supporting infrastructure, environmental factors, and governance and community considerations. Using this framework, researchers assessed public and private sites across the country, including U.S. Department of Energy facilities and major industrial campuses.

The findings highlight how existing infrastructure, regulatory environments, and local resources shape the feasibility of powering frontier AI systems. Although no single site is without trade‑offs, several regions demonstrate strong potential for development through strategic investment and coordination among federal, state, and industry stakeholders.

This research provides policymakers, utilities, and developers tools to understand locational energy constraints and opportunities—informing decisions that will determine where and how the next generation of AI innovation can thrive.

Key Findings

No site was deemed inadequate for development, and most showed moderate potential

  • Three sites—Stargate and the Pantex Plant in Texas and the Kansas City National Security Campus in Missouri—were identified as high-potential locations because of their strong infrastructure, favorable regulatory conditions, and available energy capacity.
  • The maximum grid-connected power estimated to be available at a single site by 2030 is approximately 4.2 gigawatts at the Rockport Power Plant in Indiana.

Reusing existing industrial and power infrastructure offers the most practical path to meeting AI power demands by 2030

  • Constructing new grid infrastructure is unlikely to be completed by 2030 because of permitting and regulatory delays.

Recommendations

  • Target investments to address site-specific barriers.
  • Leverage existing transmission and distribution assets.
  • Mitigate natural disaster risks.
  • Foster community engagement to ensure sustainable and timely AI data center development.

Topics

Document Details

Citation

Chicago Manual of Style

Arciniegas Rueda, Ismael, Austin Smidt, Robin Wang, Hye Min Park, David Gill, and Henri van Soest, Evaluating Potential Artificial Intelligence Energy Capacity at Different Data Center Sites. Santa Monica, CA: RAND Corporation, 2026. https://www.rand.org/pubs/research_reports/RRA3845-3.html.
BibTeX RIS

Research conducted by

This research was independently initiated and conducted by the Center on AI, Security, and Technology within RAND Global and Emerging Risks using income from operations and gifts and grants from philanthropic supporters.

This publication is part of the RAND research report series. Research reports present research findings and objective analysis that address the challenges facing the public and private sectors. All RAND research reports undergo rigorous peer review to ensure high standards for research quality and objectivity.

This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.