FutureLens
Forecast intelligence
Forecast dossier

⚛️ Google picks Tennessee nuclear reactor to power AI data centers, TVA seals first PPA

Google, TVA, and Kairos plan a 50MW Hermes 2 reactor in Oak Ridge to support Google's Tennessee and Alabama data centers by 2030. The deal includes clean energy attributes and marks the first U.S. utility PPA for Gen IV nuclear power. The agreement sits within a broader plan to develop up to 500MW across multiple deployments. Licensing milestones are in place, but operating approvals and timelines remain pending. Community benefits and grid integration will shape outcomes.

Verdict: A first-of-a-kind PPA positions TVA to buy power from Kairos's 50MW Hermes 2 for Google's data centers by 2030 (TVA Inks First U.S. Utility PPA for Gen IV Nuclear Power in Landmark Three-Way Deal with Google, Kairos, 2025-08-18). Google frames this as part of a 500MW multi-reactor plan with clean energy attributes (Our first advanced nuclear reactor project with Kairos Power and Tennessee Valley Authority, 2025-08-18). NRC construction permits for Hermes 2 were issued in 2024, but operating approval and delivery risk remain (Hermes 2 - Kairos Application | NRC.gov, 2024-11-21).

Back to board
Date
Aug 18, 2025
Reliability
80
Harm potential
Medium

Scenario odds

Best Case

15%

Hermes 2 meets the 2030 date and delivers reliable 50MW around the clock. Google procures attributes that raise hourly CFE scores across both sites. The project unlocks financing for additional Kairos units and anchors a skilled nuclear workforce in Oak Ridge.

Baseline

50%

Construction progresses with manageable delays and cost increases. TVA integrates Hermes 2 on schedule for initial testing, then phased output. Google's 24/7 CFE improves modestly while supplementary solar, storage, and market purchases fill gaps.

Adverse Case

25%

Licensing amendments for the uprated design trigger review delays. Supply chain or HALEU fuel issues push operations past 2032. Local opposition and cost overruns erode support and dampen additional deployments.

Wildcard

10%

Federal incentives or a major reliability crisis accelerate nuclear timelines. Competing firm resources, like geothermal or long-duration storage, outcompete on cost. Google shifts sites or offtake structures to faster technologies in the TVA region.

Timeline projections

1-Year

🔧 Year 1: Contracts, engineering, and site mobilization

Developments: Detailed engineering packages mature and long-lead components are ordered. TVA publishes interconnection and transmission upgrades for Oak Ridge. Public engagement increases as workforce programs launch with local universities (Our first advanced nuclear reactor project with Kairos Power and Tennessee Valley Authority, 2025-08-18).

Risks: Design changes from 28MW to 50MW require licensing clarifications. Supply chain tightness for HALEU and specialty materials introduces schedule risk. Cost estimates drift as inflation and labor constraints persist.

Outlook: Stakeholders clarify scope and timelines. Community benefits and training efforts start. Early schedule risk emerges but remains manageable.

2-Year

🏗️ Year 2: Civil works and component delivery

Developments: Groundworks, foundations, and auxiliary buildings proceed. Major components arrive as fabrication slots open. Grid upgrades progress to support testing and eventual output.

Risks: Weather and contractor turnover slow civil work. Vendor quality issues require rework and inspections. Community concerns rise over traffic and noise during peak construction.

Outlook: Construction advances with typical friction. Quality controls tighten. Communications mitigate localized disruption.

3-Year

🧪 Year 3: Systems installation and pre-operational testing

Developments: Primary systems install and integrated testing begins. Digital controls undergo verification and validation. TVA readies operational procedures and staffing for initial fuel load.

Risks: Commissioning discovers valve and sensor defects that require fixes. Cybersecurity reviews add steps and time. Insurance requirements raise costs for first-of-a-kind units.

Outlook: Readiness improves across systems. Testing reveals solvable issues. Schedule cushions narrow but hold.

5-Year

⚙️ Year 5: Initial criticality and staged power ascension

Developments: Reactor achieves initial criticality and enters phased power ascension. TVA synchronizes to the grid and validates 24/7 delivery claims. Google begins receiving clean energy attributes at volume (Tennessee Valley Authority agrees to buy nuclear power for Google data centres, 2025-08-18).

Risks: Unexpected thermal performance limits reduce net output. Regulatory findings delay full commercial operation. Market volatility affects PPA economics and certificates' valuation.

Outlook: Commercial milestones arrive. Measured output supports data center growth. Financial signals remain mixed but positive.

10-Year

🏭 Year 10: Expansion and replication decisions

Developments: Performance data supports decisions on additional Kairos units. Financing structures evolve to lower capital costs. Regional supply chains for components and fuel mature.

Risks: Alternative firm resources undercut cost projections. Policy changes alter nuclear credit eligibility. Aging grid assets constrain additional nuclear interconnections.

Outlook: Evidence guides scale choices. Costs trend down with replication. Grid limits shape siting choices.

20-Year

🛰️ Year 20: Portfolio of firm clean assets

Developments: Multiple reactors or hybrid plants operate in the TVA region. Data center growth stabilizes and aligns with firm clean supply. Workforce pipelines sustain operations and maintenance.

Risks: Waste handling and decommissioning funding pressures grow. Extreme weather stresses cooling and grid reliability. Public sentiment shifts after unrelated nuclear incidents.

Outlook: A regional firm clean portfolio emerges. Long-term stewardship needs increase. Public trust requires constant work.

50-Year

🌐 Year 50: Legacy impacts and technology turnover

Developments: First units approach mid-life refurbishments or retirement. Advanced designs and storage dominate new builds. Lessons from Hermes 2 inform safer, cheaper designs worldwide.

Risks: Decommissioning costs strain budgets if reserves lag. Fuel cycle geopolitics remain sensitive. Climate and water constraints require new cooling solutions.

Outlook: Early projects shape industry norms. Lifecycle planning proves decisive. Technology turnover resets competitive baselines.

Planning prompts to verify

  1. FOIA and request PPA term sheets and interconnection studies from TVA and Kairos
  2. Interview Oak Ridge community leaders, grid planners, and Google energy executives on siting and workforce
  3. Model hourly CFE, congestion, and reliability impacts under growth scenarios through 2035