The Great Uprate Rush: Big Tech and Existing Nuclear Fleet Capacity

Zach Martin

Nearly one gigawatt of firm power, coming from reactors that already exist. That's the Great Uprate Rush, and it's the story I want to unpack today on NUCLEAR CAPITAL, here on I R Hub Radio. Quick note before we start: companies we discuss are for illustration of public developments only. This is informational, not investment advice, and not a buy or sell recommendation.

Emily Martin

And the facts are worth laying out cleanly. On October 6, 2026, Google and Constellation announced a 20 year power purchase agreement for 890 megawatts of new nuclear capacity in PJM. It spans 11 Constellation units in Illinois, Pennsylvania, and New Jersey. Constellation says it represents more than 4.3 billion dollars of new investment, and the first uprate is expected by 2028.

Zach Martin

What I like is the discipline in it. Think of it like compounding. You don't need a brand new building if you can make the one you own work harder every year. Turbines, steam generators, digital control systems. Small improvements, stacked up, adding real megawatts.

Emily Martin

The engineering logic is strong. Uprates expand output at plants that are already interconnected, so they sidestep the greenfield transmission work and the interconnection bottlenecks that new builds face. And there's a second piece people skip over. Google also signed a separate 15 year agreement covering 2,700 megawatts of existing PJM supply, which Constellation describes as keeping those operating assets economically viable.

Zach Martin

So the uprate is the headline, and the supply deal is the foundation underneath it.

Emily Martin

Yes. Then Amazon. On September 30, a 20 year agreement tied to Calvert Cliffs in Maryland. Amazon contracts 690 megawatts, including 190 megawatts of new uprate capacity, supporting over 3 billion dollars of infrastructure investment. Those uprates are expected online between 2030 and 2032, and the revenue certainty lets Constellation pursue a 20 year license extension, pushing operation toward 80 years.

Zach Martin

Eighty years. That tells you what this is really about. It's life extension and capacity growth wearing the same jacket. And Meta had already contracted for a 30 megawatt uprate at Clinton in Illinois, so all three hyperscalers are underwriting incremental megawatts.

Emily Martin

Add federal financing. On October 27, the Department of Energy issued a conditional loan commitment of up to 5 billion dollars to Constellation, supporting uprates across eight facilities and targeting nearly 1.1 gigawatts under the UPRISE program. Conditional matters here. It isn't final money yet.

Zach Martin

Now the part that confuses people. Landmark deals, billions committed, and the stocks sold off. Sell the news. Emily, is the market wrong?

Emily Martin

I'd say the market is pricing timing, not value. These megawatts ramp gradually between 2028 and 2032. Outages have to be scheduled, work happens during refueling windows, and the capital is heavy. Four point three billion dollars for 890 megawatts is no small check. Anyone who bought the rumor had already priced the headline.

Zach Martin

I'd push on that a little. Profit taking after a run is normal, but I'd caution against reading too much into one reaction in either direction. Execution is the whole game. A contract is a promise. A turbine upgrade delivered on schedule is a result.

Emily Martin

Agreed on execution, but I'd defend the structure. Long term revenue certainty from creditworthy buyers is what makes the capital allocation sensible. And both deals were set up so the power flows into the broader PJM market, which speaks to the ratepayer cost shift concerns.

Zach Martin

Fair enough, but that cuts both ways. If you're the utility, you're betting on flawless outage scheduling for years. That's a lot of moving parts.

Emily Martin

It is. But compare it to the alternative. Small modular reactors still face fuel supply questions around TRISO and HALEU, first of a kind engineering hurdles, and licensing cycles stretching into the 2030s. Uprates use conventional technology, swapped in during planned outages, in a two to five year window.

Zach Martin

So it's not either or. Uprates are the near term bridge, SMRs are the long game.

Emily Martin

Right, and firm power is the prize for AI data centers. They need around the clock supply, and the existing fleet is the fastest place to find it.

Zach Martin

My takeaway is steady beats flashy. Squeeze more from what you own, extend its life, and let the long term builds mature on their own clock. That's the real asset thesis in one sentence.

Emily Martin

One reminder for listeners. Everything we covered is informational, companies are mentioned only to illustrate public developments, and none of it is investment advice.

Zach Martin

That's NUCLEAR CAPITAL on I R Hub Radio. Real assets. Real ideas. A brighter tomorrow.