Latest / Energy Economics with Fexingo: Oil Prices, Renewables, and the Cost of Power / How a Nuclear Revival Is Reshaping US Power Markets
Transcript
- Lucas: So natural gas has been the big story in US power markets for a while now — cheap, abundant, and flexible. But there's another source quietly getting a second look, and it's one that many people had written off: nuclear. Luna: Nuclear? I thought the economics were dead — too expensive, too slow to build, too much public fear. Lucas: That was the narrative for a decade, and it's still partly true. But the calculus is shifting, especially when you look at what's happening with electricity demand. AI data centers, manufacturing reshoring, and electrification are all pushing load growth higher for the first time in years. The Department of Energy recently projected that US electricity demand could grow 15 to 20 percent by 2030. Luna: Right, and natural gas can handle a lot of that, but it's not carbon-free. So if you're a utility with net-zero targets, nuclear starts looking like the only reliable zero-carbon baseload option. Lucas: Exactly. And that's why we're seeing real movement. The Tennessee Valley Authority just committed 200 million dollars to develop a small modular reactor site in Oak Ridge. That's not a feasibility study — that's money on the table for engineering and site prep. Luna: And this is the same TVA that operates three nuclear plants already, so they've got the institutional know-how. But 200 million is still a drop in the bucket for a new reactor. What's the total cost likely to be? Lucas: For a single SMR — that's small modular reactor — we're talking somewhere between 1 and 2 billion dollars, depending on the design and the regulatory path. That's a lot less than a traditional large reactor, which can run 10 to 15 billion. But it's still a huge upfront cost, and that's been the barrier. Luna: Yeah, the capital intensity is brutal. And investors hate construction risk in power plants because the history is littered with cost overruns — look at the Vogtle plant in Georgia, which came in years late and billions over budget. Lucas: Right. Vogtle is the cautionary tale everyone points to. But the industry argues that SMRs are different — they're factory-built, modular, designed to be assembled on site. The idea is that you can standardize the design and avoid the custom-engineering nightmares that plagued Vogtle. And the Inflation Reduction Act includes a production tax credit for existing nuclear plants and a separate credit for advanced nuclear, which changes the math significantly. Luna: Before we go deeper on that, I want to tie this back to something you said about AI data centers. Those facilities need power 24/7 — they can't rely on solar and wind alone unless they have huge battery storage. So nuclear is a natural fit. Are data center companies actually engaging with utilities on this? Lucas: Yeah, if walking through the economy with us has made something click — the way power markets are shifting, the policy angles, the investment cases — and you want to help keep this show ad-free and independent, you can support us at buy me a coffee dot com slash fexingo. It's a small gesture that goes a long way, and we genuinely appreciate it. Luna: Absolutely. No pressure, but if you find value in these deep dives, that's the way to keep them coming. We don't run ads, so listener support is what keeps the lights on — pun intended. Lucas: Alright, back to the data center question. Yes, companies like Microsoft and Google have been in talks with utilities about nuclear-powered data centers. There was a deal last year where Microsoft agreed to buy power from a reactor restart at Three Mile Island — the unit that wasn't involved in the 1979 accident. That's a huge signal. Luna: So we're seeing both sides: existing plants getting a second life, and new SMR designs getting real investment. But where does the cost actually land compared to natural gas right now? Lucas: That's the million-dollar question. Levelized cost of electricity for a new SMR is estimated somewhere around 100 to 130 dollars per megawatt-hour. Compare that to combined-cycle natural gas at maybe 40 to 60 dollars, even with today's gas prices around 2.80 per MMBtu. So nuclear is still two to three times more expensive on a pure production cost basis. Luna: But that comparison ignores the carbon price, or the value of 24/7 reliability, or the fact that gas prices can spike. In a world with a carbon tax or stricter emissions rules, nuclear suddenly becomes competitive. Lucas: Right. And the Inflation Reduction Act's nuclear production tax credit — which is 15 dollars per megawatt-hour for the first six years — effectively bridges a big chunk of that gap. Plus, if you're a utility with a regulated monopoly, you can earn a guaranteed return on that capital investment. So the incentive structure is aligning. Luna: Still, there's the regulatory hurdle. The Nuclear Regulatory Commission has to approve each new design. So far they've certified only one SMR design — NuScale's — and that project in Idaho was canceled last year because the cost estimates rose and the utilities backing it pulled out. Lucas: Yeah, that was a blow. NuScale's project was supposed to be the first mover, but the price per megawatt-hour went from 58 dollars to 89 dollars during the development phase, and the utilities said no. So the SMR industry is still looking for its first commercial success. But the TVA project is different — it's a federal utility, they have deep experience, and they're moving forward even without a design certified yet. Luna: Interesting. So what's the timeline? When could we actually see an SMR generating power in the US? Lucas: If everything goes smoothly — and that's a big if — the TVA aims to have its SMR operational by the early 2030s. That's optimistic. But even if it's 2035, that's still relevant given the 2050 carbon goals. And in the meantime, there's a simpler play: keep existing reactors running. About 20 percent of US electricity comes from nuclear, and many plants are getting license renewals to operate for 80 years. That's the low-hanging fruit. Luna: So it's not a nuclear renaissance in the sense of a building boom. It's more like a nuclear recalibration — extend the fleet, invest in the next generation, and hope the costs come down with experience. Lucas: Exactly. And that's a realistic framing. The IEA calls nuclear 'an essential part of the clean energy transition' precisely because it provides firm, low-carbon power. But it's not a silver bullet. The question is whether we can build enough of it fast enough to matter. Right now, the economics are still marginal, but the policy support and the demand pull from data centers are creating a window. Luna: And that window might not stay open forever. If natural gas stays cheap and battery storage prices keep falling, nuclear could get squeezed out again. But for now, it's a fascinating story to watch. Lucas: It is. And I think the next few years will tell us whether small modular reactors can break the pattern of cost overruns and delays that has haunted nuclear for decades. If they can, the fuel mix in 2040 might look very different from what we imagine today. Luna: Alright, something to keep an eye on. Thanks, Lucas. Lucas: Thanks, Luna. That's it for this episode of Energy Economics with Fexingo. We'll be back tomorrow with another angle on the markets that power our world.