구글은 핀란드 사업의 기저전력 확보를 위해 포르툼과 22년 장기 전력구매계약(PPA)을 체결했다.

Google announced on September 9 that it would invest at least €13 billion (approximately $15.1 billion) in digital infrastructure across Finland in 2027 and 2028, marking the company’s largest single investment in Europe. The commitment encompasses four data center sites, a portfolio of clean energy contracts totaling 629 megawatts (MW) of new wind capacity, a grid-scale battery system, and a €31 million (approximately $36 million) community investment fund. Yet the deal’s most consequential element appeared almost without ceremony in the energy section of the press release: a 22-year power purchase agreement (PPA) with Finnish utility Fortum tied to the Loviisa nuclear power plant. According to both parties, this arrangement will prevent a reactor complex supplying 10 percent of Finland’s electricity from shutting down in 2030.

That detail reframes the announcement entirely. Google is not merely purchasing computing capacity or sourcing power opportunistically. It is providing the long-term revenue certainty that enables Fortum to proceed with a roughly €1 billion (approximately $1.2 billion) modernization program and to justify operations at Loviisa through 2050. Without the PPA, Fortum stated, the plant could not have continued beyond 2030. For the first time on European soil, a data center company’s demand for AI compute has become the financial mechanism sustaining a nation’s nuclear energy asset.

Finland Won the AI Infrastructure Race — Here Is the Technical Reason

The selection of Finland is deliberate, with competitive advantages deeply embedded in the engineering of data center operations.

Power usage effectiveness (PUE)—the ratio of total facility power to power actually delivered to computing equipment—is the industry’s primary operating efficiency metric. A PUE of 1.0 represents theoretical perfection; most facilities operate between 1.3 and 1.5, meaning at least 30 percent of consumed electricity powers cooling rather than computation. At Hamina, Google has demonstrated for over a decade that cold-climate, free-cooling architectures can achieve PUE ratios approaching 1.1. The facility relies on seawater cooling drawn directly from the Gulf of Finland. While technically unique and not universally replicable, this mechanism illustrates the performance ceiling that cold geography enables.

The Hamina cooling system operates as follows: Cold seawater is pumped from the Gulf through tunnels carved into solid granite bedrock, originally part of the paper mill Google acquired in 2009. The water passes through a two-stage heat exchanger—seawater on one side, server cooling fluid on the other—transferring thermal energy without the seawater ever contacting the server circuit. This separation is critical; the dissolved salts and oxygen in seawater would corrode metal components within months if introduced directly to the equipment. After absorbing heat, the warmed return water is blended in a temperature-equalization building with incoming cold seawater before discharge, minimizing thermal impact on the marine environment. Mechanical chillers—the energy-intensive refrigeration compressors relied upon in warmer climates—are largely eliminated. The result is a data center that dissipates the heat of a major city’s computing infrastructure using gravity-assisted seawater flow and dual heat exchangers.

Rather than being wasted, that thermal energy is put to productive use. Google’s heat recovery system at Hamina channels thermal energy from the server cooling loop into the district heating network operated by Haminan Energia, the local municipal utility. Once fully operational, this waste heat will cover 80 percent of Hamina’s annual residential and commercial heating requirements at no cost to local customers. In a country where district heating serves roughly half of all buildings and winter heating costs represent a significant household expense, the impact is both material and visible. Google has confirmed that all new European data center sites will be designed from the ground up with heat recovery readiness, rather than retrofitted later.

The three newly announced sites—Kajaani, Muhos, and Vaala, all in north-central Finland—were selected partly because they reside within what Finnish grid operator Fingrid classifies as a production-dominated northern zone. This region generates substantially more electricity than it consumes locally, with surplus flowing south to population centers. Locating data centers at points of excess generation, rather than consumption, reduces the need for new transmission infrastructure and lowers overall system costs for all Finnish electricity users. Fingrid President and CEO Asta Sihvonen-Punkka explicitly endorsed this strategy: “These locations at strong grid points both help the whole electricity system and ensure cost-effective network development.”

Finland’s grid holds another structural advantage growing increasingly rare: it is already overwhelmingly carbon-free. Nuclear power generated approximately 38 to 40 percent of Finland’s electricity in 2024 and 2025, wind contributed roughly 24 to 27 percent, hydropower supplied 15 to 17 percent, and bioenergy accounted for another 11 to 13 percent. The country’s overall carbon intensity sits at approximately 88 to 95 grams of CO₂ equivalent per kilowatt-hour, well below the European average and dramatically lower than markets still heavily reliant on gas or coal. For a company facing investor and regulatory pressure to align AI compute loads with genuinely carbon-free generation—not merely offset credits—Finland offers a ready-made grid profile that most European nations cannot replicate.

How the Loviisa Nuclear Deal Actually Works

Located on a small island off the southern Finnish coast near the town of Loviisa, the Loviisa nuclear power plant houses two Soviet-designed VVER-440 pressurized water reactors. They have operated continuously since 1977 and 1981, respectively. Combined, they deliver 1,014 MW of nameplate capacity and maintain an availability rate of approximately 89 percent, placing them among Europe’s most reliable nuclear assets. The plant supplies a disproportionate share of Finland’s baseload power. Because it operates at a high capacity factor around the clock regardless of weather, it provides the grid stability necessary to integrate large volumes of variable wind and solar capacity without compromising reliability.

The challenge Fortum faced prior to the Google PPA was straightforward: existing electricity sales contracts expire around 2030. Without long-term revenue certainty, Fortum could not justify the roughly €1 billion (approximately $1.2 billion) in modernization investments—replacing low-pressure turbines, renewing turbine automation, and swapping out seawater pumps and motors—required to operate the facility safely through its government-granted license period ending in 2050. Each investment project requires separate financial justification. Without a buyer committed to purchasing a substantial share of Loviisa’s output for two decades, that justification did not exist.

The Google PPA provides it. Initial volumes under the agreement are modest when deliveries begin in 2028, scaling to approximately 50 percent of Loviisa’s output between 2030 and 2049. This timeframe aligns precisely with when the modernization investments must be operational and when the plant would otherwise face decommissioning. Fortum President and CEO Markus Rauramo described the arrangement as providing “a strong foundation for the continued development and reliable operation of our Loviisa power plant for decades to come.” Fortum’s shares rose sharply following the announcement.

The agreement extends beyond the immediate PPA. Alongside the contract, Google and Fortum signed a memorandum of understanding to explore cooperation on new power generation capacity, including the potential deployment of new nuclear reactors at Loviisa. In this context, the PPA functions not merely as a bilateral transaction but as a proof of concept for a novel financing instrument in European energy. If a hyperscaler’s AI compute revenue can sustain the economic viability of a legacy nuclear plant, the same template could theoretically be applied to aging nuclear assets in France, Belgium, the Netherlands, and elsewhere across Europe. These are markets where operators have struggled to justify continued operation against lower-cost intermittent renewables that cannot themselves deliver the 24/7 baseload profile required by AI data centers.

Why AI Workloads Specifically Demand Nuclear Baseload

Understanding why the Google-Fortum structure makes engineering sense requires understandi