The boom in energy-hungry data centers in the US has led to a revival of interest in nuclear power. Small modular reactors, or SMRs, are often held up as the way forward, but the technology remains in development, which has led some big tech firms – or ‘hyperscalers’ – to explore other pathways for using nuclear power. One alternative is restarting some of the 11 nuclear facilities that have been shut in the US in the last 15 years, an option made all the more attractive given the regulatory hurdles, cost concerns and schedule overruns that tend to plague new nuclear projects. On today’s show, Tom is joined by Chris Gadomski, BloombergNEF’s lead nuclear analyst, to discuss key findings from his report “Hyperscalers’ Energy Appetite Boosts Nuclear Prospects”.
Complementary BNEF research on the trends driving the transition to a lower-carbon economy can be found at BNEF<GO> on the Bloomberg Terminal or on bnef.com
Links to research notes from this episode:
Hyperscalers' Energy Appetite Boosts Nuclear Prospects - https://www.bnef.com/insights/35223
This is Tom Rowland's Reese and you're listening to Switched on the BNF podcast. Today we're discussing the revival of nuclear power in the United States, driven by the demand of hyperscaling tech giants led by the boom in AI. The rapid growth of energy hungry data centers has resulted in an ever increasing outlook for electricity demand, which the US Energy Information Administration forecasts at three percent growth for twenty twenty four alone. With the need for a secure twenty four hour energy supply, nuclear power fits the bill. The nuclear path isn't as straightforward as it would seem, however, Small modular reactors or SMRs are still in a development phase, Regulatory hurdles remain in place, and cost concerns coupled with project overruns have been a persistent problem. Regardless, the need to secure energy to power their ever expanding data center fleet has led to companies such as Amazon, Google, and Microsoft to announce a flurry of deals with nuclear energy companies. To discuss the details of this nuclear revival further, Today, I'm joined by BNFEED nuclear analyst Kris Kodomski to review findings from his recent report Hyperscaler's Energy Appetite boosts Nuclear Prospects. Bn F clients were able to access this and other related research at BNF go on the Bloomberg Terminal and at BF dot com. Now let's talk to Chris about data centers and nuclear energy. Chris, lovely to have you here. Thank you for joining us. It's great to be back. You have been covering the nuclear industry at BNF for a number of years. Were you covering it before BNF as well? You've been at the company longer than I have, so.
Yes, I was. As a matter of fact, I was first intrigued by the nuclear industry by doing studies following the collapse of the former Soviet Union and understanding what market opportunities existed in the power sector in those countries that were affected, Poland, Slovakia, Czech Republic and in Hungary. And then I went into the yucka mountain repositor that was being built or had been built in Nevada, and we had spent six billion dollars to do that and we weren't going to use it. So I kind of intrigued, why would we spend six billion dollars being in nuclear repository and not use it. So that's what really got me intrigued.
I mean, the reason I asked this question, you know, how long you've been covering nuclear is I am imagining that you have seen nuclear come in and out of fashion in the sort of both the public and government and private domain. You know, kind of being cool and uncool, and then maybe cool again and then sort of in some form uncool. In terms of the sort of that cycle, where do you see nuclear right now? You know, both in terms of how it's perceived and in terms of what the state of the actual global nuclear fleet is right now.
First of all, we need to understand that the nuclear industry moves like molasses, and it's that quick moving. It just goes very very steadily forward or backwards. What has happened in the last twenty four months or so is that there is a tremendous interest in nuclear power for two reasons. The first reason is for trying to address unresolved issues with regards to climate change. It's a very strong and powerful technology, and I think the world is starting to recognize that it is something that we need as part of a portfolio of technologies to solve climate change. But what really pushed nuclear forward recently has been the Russian event in the Ukraine, when all of a sudden energy security has become a very very high priority for countries in Europe, in the United States, and realizing that nuclear does provide a tremendous sense of security and it's not intermittent, and it's a very very powerful technology that should be part of the equation for climate change and for national security. Got it.
And then I suppose to that point about climate that manifested itself twenty eight, which was in Dubai with this tripling nuclear pledge, which I mean, I know you've written about, and I think that your view is that it's going to be very difficult to achieve, if that's a polite way of putting it. But at least it shows like everything you're saying is sort of again this technology that maybe at some point, you know, if we go back to Fukushima and what happened there with the earthquake and then the tsunami, and then you know what we saw then the reaction in Germany that was very much a technology that was out of fashion to an extent one could even describe as irrationally so and it seems like now that you know from everything you've described, nuclear is beginning to have a bit of a moment again because of again of global events.
Nuclear is having a moment if you rely on politicians, statements, NGOs who are freaking out that we're not really resolving addressing net zero aspirations in climate change. But when you look into the nuts and bolts and you realize how difficult and challenging it is to build these nuclear power plants, at least in the West, now that it becomes a much harder and more difficult slog than there are in other places in the world. We've lost the ability in the West to build nuclear power plants on time and on budget. And that's manifested by the fact that the Chinese can build a nuclear power plant, a large one, for maybe one fifth of the court that we can do it in the United States. So it's understandable that the Chinese are building twenty nine nuclear power plants right now and the US is not building any simply because every utility in the country is freaking out by the realization that the next nuclear reactor may be slightly cheaper or maybe not got it.
It's sort of like even if maybe nuclear is becoming a little bit more popular again, it doesn't magically make it easier to do or solve the fact that, you know, in the US, at least, the skilled workforce required to build these reactors isn't there.
We assembled a skilled workforce for building the Vogel three, for the last two reactors that were completed, that was completed several months ago. Finally, what has happened to that workforce? There is no second large nuclear project for them to go to. In China, you'll have six reactors built on the same site, and the workforce will go from one to the next, to next to next. Where are all those workers that built the Vogel three and four. There's no new nuclear project being proposed or actually constructed now, so they're being dispersed and pursuing other opportunities.
Seems like a missed opportunity because a lot of maybe you could characterize the cost overruns with that project, is the cost of learning again, training up those workers, you know, developing those skills in the country. And then the US is not taking advantage of that. Right.
The US government is suggesting that the next ap one thousand, the reactor or the next large reactor that theoretically would be built in this country will come in at a cost advantage of say thirty percent. I'm not exactly sure that will happen. I'd love to see it happen, but I'm questioning whether or not that can happen in an environment where we have persistent inflation and higher interest rates than we had over the last couple of years.
So this is kind of, you know, tease up the sort of getting onto the main topic today. So, as I mentioned, we have maybe a more favorable view developing globally for nuclear and in the US, but maybe an industry that is not in a state to step up to where it needs to be to kind of ride that wave. Mixing my metaphors. And then suddenly there's this new factor you mentioned climate, you mentioned Russia and energy security. And then there's this third thing that everyone is talking about, which is demand growth for electricity coming from data centers. So do you think that's going to shake things up in any way?
I certainly think it will, and we're seeing evidence of that where the large Hyperscalers, the Amazons, the Googles, the Microsofts, the meta are all negotiating contracts with advanced reactor companies and also suggest that we bring back some of the older reactors that have been prematurely retired because of inexpensive natural gas or intermittent renewables. And these are very valuable assets that I will continue to serve for for several more years. So it is a combination of we need reliable twenty four to seven electricity that is carbon free, and nuclear power is a great source of that.
And it's interesting because you mentioned both the advanced reactors and then this idea of bringing back some of the older reactors. So let's focus for a second on the advanced reactors. Why are they being spoken about? You know, why is there so much interest in those parsically because unlike the conventional reactors, they're not proven. So there's this new challenge and we're looking to a new solution rather than an existing solution.
So one has to realize that there's five pathways to nuclear power. There are large reactors, small modulear reactors, advanced reactors now explain the difference, microreactors less than fifty magua, and finally fusion. Each one of these technologies have different risk profiles, cost uncertainty, and timelines to commercialization. So when you look at a large reactor, it took fifteen years from filing of the initial application license until electrons are flowing onto the grid. For many of these hyperscalers waiting for fifteen years, we've missed the opportunity. But will the SMRs or the advanced reactors cross the finish line sooner and be able to deliver electrons to the grid before that, And the vendor suggests yes, they will, they're looking at the end of this decade. I suggest that's probably more likely after twenty thirty, given the molasses movement historically of the nuclear power industry, so that we're looking at twenty thirty. But there are several other reasons why you'd like to look at a SOMR. There's a smaller emergency planning zone, so you don't have more area that needs to be excluded unpopulated around these nuclear power pl Allegedly, there's a quicker time to market, as we've talked about, there's lower upfront capital costs, so you're talking thirty four billion dollars for two large reactors. You can get a three hundred megawote reactor an unspecified cost, but it's going to be less than the costs of the large reactors hypothetically. And then there's also the ability to add capacity as you need it. As the data center demands grow, you can add incrementally to the power supply. And finally, you avoid the problem of single shaft risk if the reactor. If you lose a thousand megawatts of capacity, if something goes on, it's a problem. If you have three three hundred megawatt reactors operating, if you lose three hundred megawats, you may be able to find a work around that's not going to be so catastrophic.
So there's various facts. I think the one that's really interesting is this kind of time to market that you mentioned. You know, because this demand growth in relations data sensors is so pressing. So are you saying that we know that conventional nuclear takes a long time. Even though there's all the issues we mentioned with the workforce in the US, it is still a known quantity to some extent. And what we know is that from a regulatory perspective, it's going to be working through molasses. And so some of these small modular reactors they're kind of what they're banking on is that that process is so slow. If they have a technology that does not need to have such tight regulatory oversight, they can maybe commercialize that technology and get it through the regulatory process in the same kind of time frame that it would take for conventional nuclear to just get through the regulatory process. Is that what their their role of the dice is.
Well, the role of the dice is that we will go through the regulatory process and then we will be able to deliver and build these reactors in a faster timeframe. They're sorter timeframe than the larger reactors. And very, very importantly, we talk about the difference between the first of a kind cost and the nth of a kind cost. How fast can a reactor or vendor proceed down the learning curve. If you're making small reactors and you have a long pipeline of potential customers, you can invest in the manufacturing infrastructure with which to make these reactors, build these reactors more efficiently, and then get down the learning curve. Think what it's like to sort of manufacture reactors in something similar to a commercial Boeing jet manufacturing facility.
And what's interesting there as well? I mean you mentioned, yeah, they're smaller so you can come down the cost curve faster to get to end of a kind. And the other thing that is essential to doing that is having a kind of a pipeline of customers. And if I think about wind and solar, the tech industry has been really essential in various different ways in developing the market in the US for those technologies. And so it's interesting then that you know, now we're talking about data centers creating this new demand, this new interest from the tech industry in new nuclear technologies. I would characterize that nuclear probably can't succeed without the interest of that particular sector. Maybe I'm wrong. I mean, you can push back on that if you don't agree. But to that point, which companies have we seen engaging with nuclear and in what particular technologies?
So the four major hyperscalers are all very much involved. We have Amazon has invested in one of the advanced reactor companies and now it's ex Energy. They can participate in the Series C five hundred million dollar financing, and that was there. Google has committed to buy a certain amount of energy from several chiros multen salt reactors technologies that they're developing. And Microsoft has contracted to buy electricity from some of the large reactors that are being brought back to life, specifically the Three Mile Island in Pennsylvania. And just the other day Meta got into the act and asked for a request for prosals for between one to four gigawatts of nuclear power to power their data centers.
That's so interesting, and I remember in your the note that you wrote about this topic. The thing that was also interesting. I think it was even companies that were not necessarily engaged with the tech companies directly saw their share prices increase with each of these announcements. I suppose it's not just reflecting how much influence these tech companies interest has on the sector's prospects as a whole.
The two companies that you're talking about Oklo trades on the New York Stock Exchange Oklo and New Scale trades, I think it's over the counter. SMR is the symbol. New Scale suffered a setback at the end of twenty twenty three when their principal project selling four hundred and seventy two megawate plant to uamps based on the Idaho National That project was canceled because of allegedly high costs that stock try to below two dollars or close to two dollars. I checked today that stock is up to twenty six dollars. New Scale has the only company so far that has gone through the licensing gauntlet of the NRC with their fifty megawatt reactor. They're upgrading they're designed to seventy seven megawatts and they hope that they will be out of the gate with that sometime in the middle of twenty twenty five. And they have many negotiations with potential data center companies, and so there's a tremendous amount of interest for them. They seem to be ahead of the market. Olklow as well has signed various different contracts with data center companies. They're having an approach which are much smaller up to fifty megawatch reactor which can be deployed very quickly, very easily. They're in the process of talking with the US Air Force for building a reactor that will sell electricity to an Air Force base in Alaska, and that's a very interesting project. Their technology is also well suited to providing and displacing electricity from diesel generators in the far north of Canada.
For example, when you're talking about New Scale. You mentioned the NRC, which I presume is an abbreviation for Nuclear Regulatory Commission.
Exactly.
Amazing, So, apart from clarifying what that abbreviation means, I'm kind of curious because it sounds to me like from the way you're describing it is that if we're looking at the progress these companies are making in getting real credible projects close to commercialization, we should be looking at how they're getting through the NRC's process. The NRC is a US agency, Is that correct?
Is yes, But there is the Generic Design Assessment in the UK, the Canadian Regulatory They all have very rigorous processes to make sure that the nuclear technology that's being developed is going to be safe opera R correctly, and it's in our interest that we have a very rigorous process for that. The Biden administration passed the Advanced Act, which is going to suggest to the NRC to streamline the process so instead of taking three to four years to license, it should be done in twenty four miles. And also, New Scale spent five hundred million dollars on its own staff preparing a license application and paying the fees that the NRC would charge for reviewing all of these applications. That's a pretty formidable obstacle for many of the upcoming startups in the nuclear power industry, and so there is some provisions in the Advanced Act to sort of lessen that financial burden.
So it sounds like there's a lot of tailwinds to brewing for particularly SMRs, Advanced Nuclear, all of these companies that we're talking about. You know, for all the reasons we discussed this on nuclears back in fashion, there's also this new need of meeting data sent some demand and very specifically the tech industry that can really make things happen are showing so much interest. So with that sort of very optimistic picture in mind, I suppose it's also your job as an analyst to pour cold water all over every time someone gets too excited about any new technology. That's that's kind of what we do. So what are the potential stumbling blocks for this upscaling?
When I look at the opportunities presented to the nuclear power industry and all the vendors associated with it, I look at where they sit in the licensing process. Have they engaged with the NRC. There's twelve companies that are out in pre application discussions with the NRC, so they're having conversations as we're having conversations, say this is what we're proposing to do. Does this make sense? Yes, back and forth. So if you're not in the NRC engagement process already, I don't consider you a serious company that's going to come to the finish line sometime in the next few years. The second thing that is very very important is that you have the financial backing that it takes to go ahead and cross the finish line. We've seen the first company who was trying to build a small microreactor. The first project was going to be in Canada. They went bankrupt because they didn't sort of succeed in raising the cash that was necessary to go forward. That's very very important, and do they have a technology that seems to be responsive to the needs of the utility marketplace and also the hyperscale opportunity. There are big distinctions between small module reactors, which are light water reactors, which are just innovations of just shrinking the size to the reactor, but the same type of technology, and the idea is we can build them smaller for less money and quicker. Then the advanced react to technologies, the terror Powers which is developing a sodium cooled fast reactor, x Energy which is developing a helium cooled high temperature gas reactor, and Chiros which is developing a molten salt reactor hybrid type of technology. All of these represents innovations in the nuclear space. That's a departure from historical technology that has been deployed. If you think about the oil and gas industry, the renewable energy industry, we have seen a tremendous amount of innovation in all of those industries come to the marketplace and the result has been significantly reduced costs for the production of gas, oil and renewable wind power and solar power. Where has the innovation come in the nuclear space. We are now seeing it come in the form of a fast reactor, a high temperature gas reactor, and a molten salt reactor. These are new technologies that afford the promise of being a safer they promise to be more efficient, they promise to have less spent fuel, which is a problem, and they're very very exciting to see all of these moving forward. And it's a decision that utilities and hyperscalers need to sort of make a decision large reactors, small reactors, advanced reactors, and there's different risk and challenges that each of these face. And then if you even take it a step further, you have microreactors if we talked Oakloh And also we have fusion technology, which is something that we cannot ignore because a fusion industry has raised seven point one billion dollars for private sector investment among thirty or so companies, and having been to many fusion facilities in the US and Europe, I can say that there's something that about that technology that presents some very interesting possibilities for starters. The prototype that's being built in Devon's, Massachusetts by Commonwealth Fusion, it's located within three hundred meters of residential housing. You couldn't put a large reactor within three hundred meters or a small fission reactor. And this technology is safer, it will follow a different regulatory path, and so the path way to commercialization once they prove their technology, once they can commercialize their technology, will be much simpler and easier and faster.
This kind of raises an interesting question to me. So we've got these five different pathways, including conventional nuclear and there is this kind of need to be met, and we talked about you knows so challenging in different ways. But the fact that there's five different ways to get there. On the optimistic side, does that multiply the chances is that something will be able to meet the moment by five because there's five different, quite independent pathways to get there. But on the flip side, does that mean that if one of these technologies can get their first to meet the moment, does it massively squash the chances of relevance for the other four.
I certainly would not want to be a utility right now trying to decide which horse to bet on in those five technology options. There's not going to be a winner take all situation. I think that the larger reactors have certain applications in the in the US market and North American market, and they're being looked at by a variety of utilities and hyperscale developers. And however, the SMRs seem to offer certain attractions, certain advantages, and certainly do the advanced reactors. So it's a very very difficult situation. I think that you will see, you know, niche markets develop and there'll be opportunities for all of these technologies going forward.
So kind of last question from me, you know, talking about all these new technologies that could meet the moment, including also building out of the conventional technologies. But then the other thing, and you alluded to it earlier, is, for example, the deal between Microsoft and Three Mile Island to bring back existing nuclear plants that have been shut down but can be reopened and you know, still have potentially life in them with a little bit of investment. Do you see that as a near term solution? And you know how significant could that be? Is a three Mile Island for example, is that a one off? Or is this it's not going to become a thing.
We've closed eleven reactors in the US in the last fifteen years since Fukushima, and of the eleven, I look and suggests that perhaps we can bring back three of them. The Palisades Nuclear Reactor is the first out of the gate that's trying to be developed by Holtech, the owner. That could be in NRC approval in the third quarter of twenty twenty five and starting to operate in twenty twenty six. Three Mile Island may be the second one a year or two later, and there's another reactor out west. There's also a possibility that has been recently closed. But of the eleven reactors we see, three of them are possible candidates. However, what is going to happen at Pallisades. The plan is for the existing reactor and we come back online, and then Haltech, which is developing a three hundred megawatt lightwater reactor, they will build two of those reactors on that site. It makes it a much easier commercialization pathway if you can build a new reactor at an existing site. Some of these other sites that have been closed may be opportunities for SMRs to be on the same site. Hall Take owns Oyster Creek facility in New Jersey that is a logical candidate for an SMR. There's discussions about bringing back nuclear to Indian Point, twenty five miles north of New York City. They have an interesting opportunity there. However, in all of these situations, you have nimbiism going on, local opposition and people are starting to complain, well, we don't really want to bring these reactors back. And that's a very very important thing that has to be looked at. Is that's one thing to talk about bringing a reactor back. As soon as you start the process that people come out of the woodwork and suggest, well, maybe we don't really want to do that, and that's something that has to be carefully managed and a lot of communications and cultivation of the local surroundings to go ahead and let them move forward with the project with local buy.
It sounds to me like even with these challenges, there has been a period of not much happening in the nuclear industry, particularly in the US, and then suddenly lots is happening all at once, even if the plants haven't yet arrived. It reminds me of there's a British poet called Wendy Cope who has a poem called Bloody Men, and I can't remember the entire poem, but the first line is bloody men are like bloody busses, and the point is is that you wait for one to come along and then to arrive at once. I sort of feel talking to you, like, this is what you're experiencing. We've waited for some activity on the nuclear front for a good amount of time, and we've had a bit of promise, and then now suddenly there's a lot happening all at the same time. We've got these five different pathways, and as you said, it's a nuclear so if you're a utility, you don't even know which one to back, which is the right bus to get onto. I don't know. Would you say that's a fair characterization of the situation now.
I think there's a lot of innovation, there's a lot of enthusiasm. There's a huge need for baseload, non intermittent and carbon free electricity, and so the stage is set. What we really need to see is regulatory approval, licensing, and construction to start. Once the first reactor is under construction, working, starting to work and move forward, then I think we'll see a lot of utilities and hyperskilles. I want the same thing, and we'll follow. So it's just a matter of breaking the ice, getting the first reactor licensed and operating. And it's the first time in the US since two thousand and seven that we haven't had a nuclear power plant under construction, and that's pretty sad.
Chris, this has been really fascinating. As always, Thank you very much for stimulating conversation.
Thank you, Tom.
Today's episode of Switched On was produced by cam Grad with production assistance from Kamala Shelling. Bloomberg NIF is a service provided by Bloomberg Finance LP and its affiliates. This recording does not constitute, nor should it be construed, as investment advice, investment recommendations, or a recommendation as to an investment or other strategy. Bloomberg ANIF should not be considered as information sufficient upon which to base an investment decision. Neither Bloomberg Finance LP nor any of its affiliates makes any representation
Or warranty as to the accuracy or completeness of the information contained in this recording, and any liability as a result of this recording is expressly disclaimed.