Google’s First European Nuclear PPA Funds €1bn Loviisa Life Extension

A Silicon Valley technology company has just become the financial backbone of a nuclear plant that first fired up in 1977, and it has done so by offering something the European energy market has struggled to provide: two and a half decades of guaranteed revenue.

On 9 September 2026, Fortum and Google announced a 22-year power purchase agreement (PPA) contracting up to 50% of the Loviisa nuclear plant in Finland. The deal underpins an investment programme of roughly €1 billion to extend the plant’s life through 2050.

The pressure point behind the agreement is stark. Without this contract, Loviisa’s fossil-free output would have ended in 2030, with roughly €700 million of upgrade spending stuck on the shelf, unable to secure a final investment decision.

This piece lays out what the deal actually finances, why Google is writing the cheque, and what the arrangement signals about how existing nuclear capacity gets funded across the next decade.

What Google and Fortum actually agreed to

Start with the term, because the length is the whole point. This is a 22-year commitment, not a rolling annual supply arrangement, and its duration is deliberately matched to the plant’s renewed operating licence running to the end of 2050.

Electricity delivery begins in 2028 at a reduced capacity level. From 2030 through 2049, Google takes a full 50% of Loviisa’s total generation output.

That structure makes this Google’s first nuclear PPA outside the United States. It is also the anchor of a wider Finnish clean-energy partnership, which includes a Memorandum of Understanding (MOU) covering new nuclear development, renewables, flexibility solutions, and portfolio management.

The key contract parameters break down as follows:

• 22-year PPA term
• Electricity delivery commencing 2028 at reduced capacity
• Full 50% offtake across 2030-2049
• Explicitly tied to the operating licence running to end-2050
• Google’s first nuclear offtake agreement outside the US

Now the number that tells you what the deal actually does. At announcement, roughly €700 million of the life-extension capex was still awaiting final investment decisions. The PPA supplies the revenue visibility to advance it, converting a stalled investment programme into a financeable one. For anyone tracking how nuclear life extensions get funded, that distinction is the entire story.

The balance sheet effect is disclosed. Fortum expects the contract to lift its Group comparable return on net assets (RONA), a measure of profit relative to capital employed, by approximately 1.4 percentage points once the full 50% is contracted. No €/MWh strike price or escalation formula has been made public.

What the money finances: Loviisa’s €1 billion life extension

Loviisa is Finland’s first nuclear power plant, and it carries a load out of proportion to its age. It runs two VVER-440 pressurised water reactors, a Soviet-era design, with Unit 1 in commercial service since 1977 and Unit 2 since 1981.

The plant supplies more than 10% of Finland’s total electricity and employs around 580 people. In February 2023, the Finnish government extended its operating licence to the end of 2050, the regulatory foundation the entire deal rests on.

The operating licence extension process is the regulatory foundation every life-extension investment programme rests on, and the sequence of safety review, public consultation, and conditional approval that precedes a multi-decade licence renewal shapes the investment timeline as decisively as any financing structure.

The plant’s baseline specifications:

• Two VVER-440 pressurised water reactors
• Unit 1 online 1977, Unit 2 online 1981
• More than 10% of Finland’s total electricity supply
• Approximately 580 employees
• Operating licence extended to end-2050

Here is the reference point that gives the deal its weight.

Fortum has stated that without the substantial lifetime-extension investments enabled by this agreement, Loviisa would be unable to sustain fossil-free electricity output beyond 2030.

Four years. That is the size of the window without this capital. With it, the plant runs for another two and a half decades and is projected to deliver up to 177 TWh of emission-free electricity over the full licence extension.

The 300-project investment programme and what the PPA unlocks

The life extension is not one large project. It is more than 300 individual projects organised into portfolios, each evaluated and decided on separately under Fortum’s standard investment governance.

At announcement, roughly 80% of those projects, the €700 million tranche, lacked the revenue certainty to proceed. The Google contract is what changes their status from pending to bankable.

Two capacity uprates sit inside the programme. A 38 MWe low-pressure turbine modernisation is already underway, phased through annual maintenance outages between 2026 and 2028. The Google PPA enables an additional 10 MWe increase on top of that, extra clean output that would not otherwise have been financed.

Why hyperscalers are buying nuclear output, and why existing plants beat new builds

The Google-Fortum deal is not a solo move. By mid-2026, total announced hyperscaler nuclear arrangements, spanning PPAs and partnerships, reached approximately 13 GW.

The logic starts with the load profile. AI and cloud data centres draw continuous, heavy power, and wind and solar cannot reliably serve that around-the-clock demand on their own. Dispatchable baseload nuclear, generation that can be relied on at any hour, is the preferred fit for technology companies with firm 24/7 carbon-free commitments.

The Google-Fortum deal sits within a broader wave of hyperscaler nuclear procurement that has reshaped how utilities think about long-duration revenue, with technology companies now representing a structurally new class of offtaker for generation assets that spot markets cannot adequately reward.

The second driver is cost. Analysis from the OECD Nuclear Energy Agency (NEA) and the International Energy Agency (IEA) points in the same direction on why existing plants win over new construction.

The IEA concludes that extending the operating lifetimes of existing reactors avoids the massive upfront capital, lengthy deployment timelines, and execution risks tied to new-build nuclear projects, making life extension the lowest-cost route to additional low-carbon generation.

The third driver is risk transfer. A multi-decade PPA gives the utility the revenue visibility to make a capital-heavy programme bankable, shifting financial exposure onto the corporate offtaker. The proximity of Google’s Hamina data centre to Loviisa gives the Finnish version of this arrangement a clean geographic rationale.

The peer group shows how established the pattern has become. Terms cluster around 17 to 20 years, with deliveries largely commencing between 2027 and 2032. US market pricing for comparable deals is estimated in the US$95-115/MWh range, though these are analyst estimates rather than confirmed contract figures.

The 13 GW figure tells you this is no longer a run of one-off experiments. Hyperscaler procurement has reached a scale where it is restructuring how existing nuclear capacity gets financed across multiple markets.

For readers wanting to trace the commodity and supply chain implications of scaling nuclear capacity for data centre demand, our full explainer on uranium supply chains for AI infrastructure examines how uranium procurement, enrichment constraints, and fuel cycle security factor into long-duration corporate nuclear offtake.

What the deal signals for European nuclear financing and the risks it leaves unresolved

For Europe, the precedent is the headline. This is one of the first major corporate nuclear PPAs explicitly tied to a life-extension licence inside an EU member state, and it demonstrates that a private counterparty can supply the long-duration revenue stability that volatile spot markets cannot.

The national interest dimension matters too. Finnish authorities and Fortum have both framed the agreement as supporting energy security, electricity self-sufficiency, and price stability for households and businesses, given Loviisa supplies more than 10% of the grid. This is not being sold as a purely private transaction.

For investors, the takeaway is that private capital can close the financing gap for nuclear life extensions without state subsidy. But the model carries real constraints, and honesty about them matters.

The structural risks fall into four categories:

• Price lock-in: Fixed pricing over multi-decade horizons carries risk if renewable-plus-storage costs fall significantly
• Regulatory exposure: Energy policy, grid rules, safety mandates, and carbon pricing can all shift across a 20-year term
• Scalability limits: The pool of buyers able to underwrite billion-dollar, multi-decade commitments is small
• Equity and public interest: Dedicating significant national nuclear output to private technology firms raises fairness questions

The scalability ceiling is the sharpest of these. The roughly 13 GW contracted globally is only a modest fraction of nuclear capacity, which is why analysts frame these deals as exceptions rather than the norm. EU market reform debates continue over whether public-interest tools such as contracts for difference are a more durable financing route for strategic generation than private PPAs.

European nuclear investment has not been confined to life extensions: France’s state-led EPR2 programme represents a parallel track in which new-build capacity, backed by government capital, pursues the same long-duration clean baseload goal that the Google-Fortum corporate PPA achieves through private offtake.

The read for investors is that this model works best as a complement to public policy frameworks, not a replacement for them.

What determines whether this model spreads across Europe’s ageing reactor fleet

The relevant question now is replication. Three conditions made the Google-Fortum deal possible, and they will govern whether it becomes a template or stays an outlier:

1. A secured operating licence providing regulatory certainty, as Loviisa’s end-2050 extension does
2. Physical proximity between the generation asset and the corporate buyer’s load centre, as with Hamina and Loviisa
3. A corporate counterparty with the creditworthiness to underwrite a 22-year commitment

The timing across Europe is favourable to replication. Many EU plants face similar licence-extension decisions across the late 2020s and early 2030s, an ageing fleet approaching the same fork Loviisa reached.

The detail that separates this from a one-off is the MOU. Its scope covers new nuclear development, renewables, flexibility solutions, and portfolio management.

The broader partnership signals that the PPA may be a first step toward a more integrated long-term energy relationship between Google and Fortum rather than a standalone transaction.

If the two are planning new capacity together, the offtake deal is evolving into something closer to a vertically integrated supply relationship, a structurally different signal than a single contract.

The variables to watch are specific: EU energy market reform outcomes, European nuclear licensing pipelines, and whether other utilities with ageing fleets and pending licence decisions adopt the Fortum approach. Scaling beyond today’s 13 GW ceiling will require either a wider buyer pool or policy mechanisms that broaden the model.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Financial projections are subject to market conditions and various risk factors, and forward-looking statements are speculative and subject to change based on market developments and company performance.