Nuclear is suddenly everyone’s favorite again—presidents, prime ministers, think tanks, even data center bros. But underneath the renaissance branding are three very old questions: who gets the industrial spoils (UK vs US), who gets a veto (tribes vs DOE), and who gets paid to stay flexible (Constellation vs everyone still pretending kilowatt-hours are a one-way street).

UK Picks Rolls-Royce. Washington Sulks.

Britain picked Rolls-Royce SMR for three reactors at Wylfa in Wales — a £2.5B bet that the UK can mass-produce nuclear modules and turn them into an export line. The Trump administration wanted Westinghouse in that slot and made its disappointment unusually public.

The real takeaway: The global nuclear race is shifting from PR to procurement. The U.S. is trying to revive Westinghouse at gigawatt-scale. The UK is standing up its own factory-built SMR fleet. And both are chasing the same AI-driven demand wave with tech that still hasn’t been proven at scale.

This is less about diplomacy and more about who actually gets steel in the ground before the load curve breaks your grid operator’s spirit.

DOE Pushes to End Tribal Vetoes on Hydro Permits

DOE wants FERC to remove tribal governments’ ability to block preliminary hydro permits at the earliest stage. Tribes say this erodes sovereignty; hydro developers say the veto power blocks even basic site studies.

Here’s the grid-reality angle: permitting timelines are now the single largest obstacle to adding firm power in the U.S. You can’t license, study, or build your way into reliability when every early-stage “no” kills a decade of potential supply. DOE’s move is a blunt instrument, but the motivation is obvious — the country needs more dispatchable power yesterday, and the current process practically guarantees never.

Removing the early veto doesn’t greenlight projects; it just forces the real arguments to happen once actual designs exist instead of at the hypothetical stage.

Constellation Wants a Nuclear-Sized Block of Demand Response

Constellation’s new strategy: build ~1 GW of demand response and customer-side storage — effectively a virtual power plant the size of a full nuclear unit.

Quick tech explainer:
Demand response = paying customers to drop load during peak hours, coordinated by software. Add smart controls, backup generators, onsite batteries, and AI forecasting, and you can schedule reductions with near-plant reliability. The utility treats customer behavior like dispatchable capacity.

Constellation sees huge interest from data centers and large commercial users. The pitch: don’t build a new peaker — contract with a few thousand megawatts of flexible load and let algorithms pull the levers.

If they pull it off, it rewires the old model: not “generate more,” but “use smarter.”

Upgrade to Grid Brief Premium to get extra deep dives into energy issues all over the world.

Conversation Starters

• The Next Nuclear Renaissance? – Cato
  A libertarian autopsy of past “nuclear renaissances” that never quite materialized, and a sober look at whether Gen IV, SMRs, and uprates can clear the cost and complexity bar this time.
  Why read: It’s the rare nuclear piece that loves the tech but still asks the only question that matters: who actually pays if the promises miss… again.

• What Can States Do to Embrace Advanced Nuclear? – C3 News (Nick Loris)
  Q&A on how states can clear the path for SMRs—siting, regulation, and cost control—while protecting ratepayers from becoming involuntary venture capital for first-of-a-kind projects.
  Why read: A useful playbook if you want more reactors but less “socialize the risk, privatize the hype.”

• Data Centers Are Wrongly Taking the Blame for Electricity Price Hikes – Kansas City Star
  David Mastio pushes back on the popular story that AI data centers are single-handedly driving bills through the roof, pointing instead to inflation, grid hardening costs, and the unwinding of old cross-subsidies.
  Why read: It’s a reminder that blaming the new shiny thing is easier than reading a rate case—and that in many states, more demand has actually correlated with lower prices.

Good Bet: AutoGrid (Schneider Electric)
AutoGrid is one of the leaders in AI-driven demand response and virtual power plant management — the kind of software Constellation and every other utility will need to turn flexible load into firm capacity. If the “virtual nuclear plant” model takes off, AutoGrid’s stack is perfectly positioned.

Bad Bet: Nuclear Nationalism as an Economic Strategy
The idea that you can industrial-policy your way into cheap, fast nuclear without fixing costs, workforce depth, or permitting realities. Whether it’s Rolls-Royce, Westinghouse, or anyone else — the flags don’t build reactors. Schedules and budgets do.

Share Grid Brief

We rely on word of mouth to grow. If you're enjoying this, don't forget to forward Grid Brief to your friends and ask them to subscribe!