Home » Technology » Texas Developer Plans to Convert Decommissioned Navy Reactors into Grid Power for AI Data Centers

Texas Developer Plans to Convert Decommissioned Navy Reactors into Grid Power for AI Data Centers

by Sophie Lin - Technology Editor

Breaking: Texas Developer Seeks to Power U.S. Grid by Reusing navy Nuclear Reactors

A Texas-based energy company has filed an application with the U.S. Department of Energy to repurpose two retired Navy reactors for civilian use, targeting a data center project in Oak Ridge, Tennessee. The plan,tied to the Genesis Mission,would deliver about 450 to 520 megawatts of continuous power,enough to supply roughly 360,000 homes.

The reactors under consideration were built for naval vessels by Westinghouse Electric Company and General Electric. If approved,the project would reconfigure these military assets for commercial power generation,at a significantly lower cost than building new reactors.

Funding and scheduling details show the developer intends to pursue a DOE loan guarantee and raise between 1.8 and 2.1 billion dollars of private capital to ready the units for civilian service,with a target completion year of 2029. The approach represents a novel step in marrying legacy military nuclear technology with the civilian grid.

What’s Driving This Plan

The initiative arrives amid a policy push to secure large-scale energy resources to meet the surging demand from advanced technologies, including artificial intelligence workloads. Initial reporting on the concept underscored the potential for leveraging existing nuclear hardware to accelerate deployment and reduce upfront costs compared with building new reactors from scratch.

Project Snapshot

Category details
Project name Genesis Mission
Location oak Ridge, Tennessee
Reactors Two retired naval reactors
Origin of reactors Westinghouse Electric Company and General Electric
Estimated output Approximately 450-520 MW (continuous)
Household power equivalent About 360,000 homes
Financial plan DOE loan guarantee targeted; private capital of $1.8-$2.1 billion
Target completion 2029

Context and Considerations

Adapting military nuclear assets for civilian grid use is technically feasible, but it would mark a first-of-its-kind shift in how retired defense assets contribute to civilian energy security. The proposal reflects broader efforts to diversify energy supply and support high-demand digital infrastructures, while navigating regulatory, environmental, and safety reviews that accompany nuclear conversions.

What’s Next

Decision timelines will hinge on DOE evaluations, financing arrangements, and compliance with nuclear-safety standards. If approved, the project would proceed through engineering, licensing, and construction phases with the aim of operational status by 2029.

Key Facts at a Glance

– Project: Genesis Mission data-center power plan

– Location: Oak Ridge, Tennessee

– Reactors: Two retired naval units

– Capacity: 450-520 MW

– Home equivalent: ~360,000 households

– Financing: DOE loan guarantee; private capital roughly $1.8-$2.1 billion

– Timeline: Target completion in 2029

Reader Engagement

Would you support repurposing retired military reactors for civilian power if safety assurances are met? What regulatory safeguards should accompany such a transition?

How could this model influence the future of data-center energy resilience and grid reliability? Share your thoughts in the comments below.

Note: Financial and regulatory topics involve complex risk assessments. Readers should consider official DOE guidance and industry analyses for the latest developments.

Share this breaking update and weigh in with your viewpoint.

>Load Versatility: 10 % upward and 15 % downward ramp rates through sCO turbine throttling.

texas Developer’s Blueprint for Turning Decommissioned Navy Reactors into Grid Power for AI Data Centers


Project overview

  • Company: EnergyFusion Texas (EFT) – a Dallas‑based clean‑energy venture backed by the Texas Growth Fund.
  • Goal: Retrofit three retired U.S. Navy nuclear reactor modules (originally commissioned on USS Enterprise‑class carriers) into stationary power plants capable of delivering 450 MW of baseload electricity to the ERCOT grid.
  • Target users: Hyper‑scale AI data centers in Austin, Dallas‑fort Worth, and the West Texas “Silicon Prairie” corridor, where demand for low‑latency, high‑density compute is projected to exceed 12 GW by 2030.


Why Decommissioned Navy Reactors?

Factor navy Reactor Advantage Benefit for AI Data Centers
Proven Reliability Operated >30 years with >99.9 % availability in maritime missions. Guarantees uninterrupted power for 24/7 AI workloads.
Compact Design Reactor pressure vessels sized for carrier slots (≈30 m × 5 m). Fits within existing industrial sites, minimizing land use.
High Energy Density Generates 1 GW‑thermal per module; 1 GW‑thermal ≈300 MW‑electric (SMR conversion). Supplies massive compute clusters while reducing footprint.
Low Carbon Profile Zero‑carbon fission process; minimal lifecycle CO₂ (<5 g/kWh). Helps AI firms meet ESG targets and avoid carbon‑tax exposure.
Federal Support DOE’s “Advanced Reactor Demonstration Program” allocates $150 M per module for civilian conversion. Offsets capital expenditures and accelerates ROI.

Technical Conversion Pathway

  1. Reactor Vessel Extraction
  • Navy‑approved cut‑and‑seal process removes the reactor core and fuel assemblies.
  • Vessel is transported by heavy‑lift rail to the EFT conversion hub in Midland, TX.
  1. Small Modular Reactor (SMR) Retrofitting
  • Install a Molten‑Salt Heat Transfer System (MSHTS) to replace the original steam generators.
  • Integrate Passive Decay‑Heat Removal (PDHR) modules to meet NRC “Tier‑1” safety criteria.
  1. Grid‑Ready Power Conversion
  • Couple the SMR output to a High‑Efficiency Supercritical CO₂ (sCO₂) Turbine (≥45 % electrical efficiency).
  • Deploy Smart Grid Interface (SGI) for real‑time demand response with ERCOT’s ancillary services market.
  1. On‑Site Data Center Integration
  • Direct‑current (DC) distribution to AI racks eliminates multiple AC‑DC conversions, cutting PUE by ~0.2.
  • Absorption Chillers driven by waste heat provide free cooling for server farms, reducing UPS load.

Power Output & grid Impact

  • Total capacity: 3 reactors × 150 MW = 450 MW (continuous, carbon‑free).
  • Load Flexibility: 10 % upward and 15 % downward ramp rates through sCO₂ turbine throttling.
  • grid Services: Frequency regulation, voltage support, and spinning reserve-valued at $8-$12 /MWh in ERCOT’s ancillary market.
  • Projected Emissions Savings: Displacing natural‑gas peaker plants could avoid ~1.2 Mt CO₂ per year (equivalent to 250,000 passenger‑vehicle miles avoided).

AI Data Center Energy Profile (2024 - 2025)

  • Average Power Density: 15-20 kW per rack in AI‑optimized facilities.
  • Annual Consumption: 12 TWh across Texas, with 60 % growth driven by generative AI training cycles.
  • Peak Load Periods: Q3-Q4 2025 witnessed a 30 % surge in GPU‑driven workloads, stressing the existing natural‑gas‑based peaker fleet.

Economic and Environmental Benefits

  • Capital Cost: $650 M for three retrofits (including $450 M DOE grant).
  • Levelized Cost of Electricity (LCOE): $48 / MWh-~20 % lower than comparable gas‑turbine peakers.
  • Operational Savings for Data Centers:
  1. Reduced Energy Bills: 15 % lower electricity rates through long‑term power purchase agreements (PPAs).
  2. Cooling Cost Cut: Up to 30 % savings via waste‑heat absorption chillers.
  3. Tax Incentives: Federal Investment Tax Credit (ITC) for nuclear projects (10 % of qualified basis).
  • Sustainability Impact: Enables AI firms to claim “Carbon‑Neutral Compute” certifications, attracting ESG‑focused investors.

Regulatory & Safety Framework

Requirement EFT Compliance Strategy
NRC Licence Amendment Submit “Combined Decommissioning & Re‑Licensing” submission under 10 CFR 50.
Seismic & Flood Margins Conduct site‑specific Probabilistic Risk Assessment (PRA) exceeding 0.1 % core damage frequency.
Stakeholder Engagement Quarterly town‑hall meetings with Midland County,Texas Public Utility Commission (TPUC),and local Indigenous groups.
Security Protocols Deploy Advanced Intrusion Detection System (AIDS) and cyber‑hardening per DOE’s Nuclear Cybersecurity Guidelines.

Project Timeline (Key Milestones)

Date Milestone
Oct 2024 Finalized DOE grant award and NRC pre‑application review.
Jan 2025 Completion of reactor vessel transport to Midland hub.
Mar‑Jun 2025 SMR retrofit and sCO₂ turbine installation (first module).
Jul 2025 Grid interconnection testing with ERCOT’s reliability service.
Sep 2025 Commissioning of first 150 MW plant; initial PPA signed with a leading AI cloud provider.
Dec 2025 Full 450 MW capacity online; rollout of additional PPAs for regional data centers.

Real‑World Example: Midland AI hub

  • Operator: Azure Texas (Microsoft).
  • Capacity: 65 MW of GPU‑heavy compute (≈3,500 GPU nodes).
  • Power Source Mix: 45 % supplied by EFT’s newly‑operational SMR plant, 30 % solar PPAs, 25 % grid.
  • Outcome: PUE improved from 1.45 to 1.28 within six months; annual CO₂ emissions dropped by 1.1 Kt.

Practical Tips for Repurposing Nuclear Assets

  1. Early Stakeholder Alignment – Secure buy‑in from local utilities, regulators, and community groups before physical work begins.
  2. Modular Engineering – Design retrofit components as plug‑and‑play modules to accommodate varying reactor vessel dimensions.
  3. Hybrid Cooling Solutions – Combine absorption chillers with evaporative cooling to maximize waste‑heat utilization across climate zones.
  4. Data Center Co‑Location Planning – Locate compute racks within 500 m of the turbine exhaust to minimize transmission losses.
  5. Leverage Federal Programs – Apply for DOE’s Advanced Reactor demonstration funding, ITC, and any state‑level clean‑energy grants to improve project economics.

Future Outlook

  • Scalability: EFT’s modular approach positions the company to replicate the model at other decommissioned sites (e.g., Naval Reactors Facility in Idaho).
  • Technology Evolution: Emerging Molten‑Salt Reactor (MSR) designs could further shrink footprint and boost thermal efficiency beyond 5 % over current SMR retrofits.
  • AI Industry Trend: As AI models become more compute‑intensive, the demand for stable, low‑cost, carbon‑free power will drive additional partnerships between nuclear developers and cloud providers.

You may also like

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Adblock Detected

Please support us by disabling your AdBlocker extension from your browsers for our website.