Greenland is pushing forward on two hydropower projects that, combined, could churn out more than 9,500 GWh of electricity a year. For scale, that is more than the roughly 7 TWh expected annually from the nuclear plant whose full output Microsoft Corp (MSFT) contracted to buy for 20 years.

That matters for Sarfartoq, the company's rare-earth project in southwest Greenland, where energy is one piece of a much larger infrastructure puzzle. Greenland Mines Ltd (GRML) President Bo Møller Stensgaard pointed to hydropower, wind, geothermal, nuclear or small modular reactors and hybrid renewable systems as potential sources for large industrial projects like that one.

Greenland's government is separately advancing competitive tenders for the Tasersiaq and Tarsartuup Tasersua hydropower sites. Government documents put the combined potential at more than 9,500 GWh annually. Estimates evolve as hydrological studies update.

Bigger Than Microsoft's Nuclear Deal

For perspective, Constellation Energy Corporation (CEG) says its Crane Clean Energy Center, the restarted Three Mile Island Unit 1, will add approximately 835 MW of carbon-free power. Microsoft has agreed to purchase the plant's output under a 20-year power purchase agreement. Constellation expects the plant to return to service in 2028.

At expected operating levels, the reactor is commonly estimated to generate roughly 7 million MWh, or 7 TWh, annually. That puts Greenland's 9.5 TWh potential about 36% higher on an annual-energy basis.

The comparison matters, but so does the caveat. Microsoft's nuclear deal is tied to restarting an existing power asset, while Greenland's hydropower resources still require development, financing, construction and an industrial customer.

Greenland's government expects prequalification for the hydropower tender in October 2026, with the broader competitive process extending into 2028.

The Customer Matters

Stensgaard told MarketDash that Sarfartoq's development will require dependable energy alongside infrastructure, patient capital, processing and separation capacity and long-term offtake certainty. In other words, having the electricity resource is only one piece of the equation.

Greenland has already designed a government-take framework intended to attract experienced developers and long-term investors capable of financing and operating the projects under Arctic conditions.

For investors watching the AI-driven power buildout, the bigger question is therefore not whether 9,500 GWh sounds large. It is whether Greenland can turn that potential into bankable power infrastructure, and find industries willing to make long-term commitments to use it.

For Greenland Mines, that could strengthen the case for developing energy-intensive mineral processing locally. For the broader market, it is another reminder that the next phase of the AI power race may increasingly be about finding new sources of reliable electricity, not simply building more data centers.