Could Europe's Data Centers Revive Nuclear Power?
TL;DR: Yes, and it's already happening. Google just signed a 22 year nuclear power deal in Finland at a roughly 60% premium to forward prices, per a report cited by Yahoo Finance. Europe's data centr…

What happened
According to Yahoo Finance, Europe's data centre buildout is strengthening the case for new nuclear capacity as tech companies chase electricity that doesn't depend on the weather. J.P. Morgan's analysis puts a number on the demand side: European data centre electricity consumption could rise from about 70 TWh today to roughly 115 TWh by 2030, citing European Commission estimates. The bank's own forecast is steeper. It projects an additional 89 TWh of annual European data centre demand by 2030 versus 2023, with Iberia and the Nordics accounting for about 45% of that increase.
The supply side is already lagging. Europe's announced data centre pipeline stood at 66.1 GW at the end of 2025, against just 10.8 GW of live capacity, the same report shows. That gap is the whole story: someone has to build the power to match the buildings.
Where does this go wrong before it goes right?
I spent years on a submarine, where I signed off on more than 1,000 QA jobs before anyone trusted the work near open water. I want to know how a deal fails before I care how it wins. The failure points here are already visible. Europe currently has no operating commercial SMRs, and the first projects aren't targeted until the early 2030s, the report notes. The €241 billion the European Commission says is needed through 2050 for new reactors and life extensions is an estimate, not a committed budget. And every deal named in this piece is a corporate power contract, not a retail investment.
That last point matters most if you're newly accredited and wondering how to get exposure. You can't buy into the Google-Fortum contract. What you can underwrite is the premium hyperscalers are already paying for certainty, a narrower, more concrete trade than "buy nuclear."
The deal that started this
The piece that triggered this round of coverage is Google's power purchase agreement with Fortum, the Finnish energy group. Google agreed to a 22-year PPA covering nuclear generation in Finland, alongside a separate agreement exploring new capacity, as the report describes it. The contract carries an estimated premium of roughly 60% to forward electricity prices. That premium is the number I'd focus on: hyperscalers are not shopping on price. They're shopping on certainty, and they will pay up for it.
This isn't Europe's first version of that trade. Microsoft signed a 20-year PPA with Constellation Energy in September 2024 to bring the dormant Three Mile Island Unit 1 reactor back online for its cloud business, agreeing to buy most of the plant's 835 MW output, according to datacenter.news. The Google-Fortum deal is the European sequel.
| Deal | Reactor | Term | Premium |
|---|---|---|---|
| Google–Fortum (2026) | Nuclear plant, Finland | 22 years | ~60% premium to forward power prices |
| Microsoft–Constellation (2024) | Three Mile Island Unit 1, ~835 MW | 20 years | Restart of a dormant reactor for cloud power |
Should I do anything about this as an accredited investor?
Not directly, not yet. Every deal named above is a corporate power contract, not a fund you can buy into. But deal activity around the theme is already picking up: M&A in Europe's nuclear sector hit a seven-year high in 2025, with 25 transactions announced, up from 17 in 2024, and $1.5 billion in deal value excluding the UK's Sizewell C project, according to Enlit World, citing Mergermarket data.
Ireland is the case study in what happens without the nuclear option. Data centres already consume roughly 21% to 22% of Ireland's total electricity, a figure projected to rise toward 31% by the early 2030s, per Euronews. Ireland banned domestic nuclear generation by law in 1999, and a government official cited in that report said "nuclear power could not practically be part of our electricity system until at least the 2040s." Ireland is stuck buying time with onsite generation mandates while its neighbors sign 20-year nuclear contracts.
Look past Europe and the shortfall only gets bigger. The IEA projects U.S. data-center power demand will more than triple over the next decade, from 34.7 GW in 2024 to 106 GW by 2035, according to EnergyReader.io. Nuclear can supply only part of that demand even under an aggressive build-out, and I'd be skeptical of anyone who tells you otherwise.
What to watch next
- Whether the €241 billion European reactor and life-extension estimate gets a funded pipeline behind it, or stays a Commission projection.
- Whether more hyperscalers follow Google and Microsoft into long-dated PPAs at a premium, the clearest signal this is a real capital cycle, not a headline.
- Whether SMR projects targeted for the early 2030s actually reach construction, given Europe has zero operating commercial units today.
This power buildout keeps showing up in the same three shapes: debt against the infrastructure, equity into the utility doing the building, and sovereign capital following the demand overseas. I covered the debt version in Infrastructure Debt Investing: How Accredited Investors Can Fund the AI Data-Center Boom. It's the closest thing to an accessible entry point today. KKR's EDF Power Solutions deal is the equity version, built around clean power rather than nuclear specifically. CPP Investments' $741 million bet on Indian data centers shows the same capital chasing the same demand outside Europe entirely.
Common mistakes
Investors hear "nuclear revival" and jump straight to buying uranium miners or SMR startups on hype alone. Downside first: most SMR designs targeted for Europe haven't broken ground, the €241 billion figure is a Commission estimate of what's needed, not a committed budget, and the PPA premiums hyperscalers are paying tell you power is scarce today. They don't tell you which company profits from the capacity that gets built years from now.
FAQ
Is Europe actually building new nuclear reactors because of data centres? Not yet in large numbers. Europe has zero operating commercial SMRs today, with first projects targeted for the early 2030s, Yahoo Finance's coverage explains. What's happening now is life-extension deals and long-term power contracts on existing plants.
Why did Google pay a 60% premium for nuclear power? Google's 22-year Fortum deal carries an estimated premium of roughly 60% to forward electricity prices, per that same coverage. Hyperscalers are paying for guaranteed, dispatchable, low-carbon power, not shopping on spot price.
How big is the actual power gap in Europe? Europe's announced data centre pipeline stood at 66.1 GW at the end of 2025, against 10.8 GW of live capacity, a gap of more than 6 to 1, as Yahoo Finance reported.
Is this an accredited-investor-accessible theme yet? Most of the named deals above (Google–Fortum, Microsoft–Constellation, KKR–EDF) are corporate or institutional transactions, not retail offerings. The accessible angle right now runs largely through infrastructure debt funds and public utility and power equities. See the infrastructure debt piece linked above.
Why can't Ireland just build nuclear plants like Finland did? Ireland banned domestic nuclear generation by law in 1999, and a government official said nuclear "could not practically be part of our electricity system until at least the 2040s," per Euronews. The legal and infrastructure lead time is the constraint, not demand.
The Bottom Line
Nothing here is retail-accessible yet. What's real is the premium hyperscalers are already paying for certain power. That premium is dated, contracted revenue. You can underwrite it today instead of trusting a projection. Pull up your own utility or infrastructure-fund exposure and check whether any of it sits inside a company that has signed, or could plausibly sign, a hyperscaler PPA like Fortum's or Constellation's.
Want AIN watching deals like this before they hit your feed? Sign up for the free AIN briefing. It's the fastest way to see the next one before the headline does.
Educational content only. Not investment, tax, or legal advice. Not an offer or solicitation to buy or sell securities. Past performance does not guarantee future results. Private-market investments are illiquid and involve risk of loss, including total loss of capital. Consult qualified advisers. Angel Investors Network is not a broker-dealer or investment adviser.
Part of Guide
Looking for investors?
Browse our directory of 750+ angel investor groups, VCs, and accelerators across the United States.
About the Author
Jeff Barnes, MBAContinue Reading

Iridium Holders Approve Rocket Lab's $54 Deal

Anduril Isn't the Only One Building Air Force Drones

US Debt Hits $40 Trillion. Yields Make It Worse

Can You Buy a Business With Your IRA? The UBIT Trap

IRA Investing: What a Self-Directed IRA Allows
