Latest / First Trust ROI Podcast / Ep 75 | Nikki Hsu & Gabriela Privetera | The Road Back to Nuclear Power: Opportunities and Pitfalls Ahead | ROI Podcast
Transcript
- 0:08Ryan: For much of the past few decades, nuclear power seemed like it was on the decline.
- 0:13Ryan: Few nuclear power plants were being built.
- 0:15Ryan: It seemed like all you heard about was retirements, but all of that has changed recently.
- 0:20Ryan: Electricity demand has surged in the wake of reshoring of manufacturing, the build-out of AI data centers, and now nuclear seems to be gaining momentum in the other direction.
- 0:31Ryan: To help make sense of that, I've got two great guests on the First Trust ROI podcast today coming to us from Bloomberg Intelligence.
- 0:39Ryan: I've got Nikki Shu, who is a utilities analyst, and I've got Gabby Privatera, who is an equity analyst also at Bloomberg Intelligence.
- 0:49Ryan: Thank you for joining us on this episode of the First Trust ROI Podcast.
- 0:53Ryan: So, Nikki, thank you for being on the podcast.
- 0:56Ryan: Same with you, Gabby, thank you for joining us.
- 0:59Ryan: Are my perceptions correct when it comes to nuclear?
- 1:02Ryan: Have we begun to see a bit of a momentum shift towards building of new nuclear power plants?
- 1:08Nikki: Yeah, totally with you here.
- 1:10Nikki: So I actually started covering US utilities and power in um, I think around mid-2010.
- 1:19Nikki: And back then all I saw on the news headlight was nuclear shutdowns.
- 1:25Nikki: And now it's totally different, like you mentioned, right?
- 1:28Nikki: So I think we've been through um contraction and then stabilization and then hopefully expansion.
- 1:39Nikki: So where we are today is like almost expansion, but not exactly.
- 1:45Nikki: So I would say we have about 75 gigawatts of nuclear capacity today.
- 1:50Nikki: And but in the past 10 years, say that's from 2013 to 2022, um, we have 13 reactors shut down.
- 2:01Nikki: That's about 10 gigawatts altogether.
- 2:04Nikki: And the last one shut down was in 2022.
- 2:07Nikki: That's Palace, that's been restarted by um whole tech international right now.
- 2:13Nikki: So you probably will wonder like what happened to stop that shutdown, right?
- 2:20Nikki: So from 2020, um 2022, there was nothing shut down afterwards.
- 2:25Nikki: And I think it was because of the um passage of the Inflation Reduction Act.
- 2:30Nikki: It actually saved the existing nuclear power plants.
- 2:33Nikki: So before this AI power demand driving the momentum, we actually didn't pivot from contraction to accept to expansion.
- 2:42Nikki: There was stabilization, and that's because of the Inflation Reduction Act.
- 2:47Nikki: Um, it introduced a nuclear power production tax credit.
- 2:52Nikki: And so for that tax credit, it pretty much gives nuclear power plants um unlimited upside potential.
- 3:00Nikki: But it's inflation-adjusted tax credit.
- 3:03Nikki: So in today, you're looking at no matter how much the power market prices are, you are guaranteed to be receiving about $45 per megawatt hour.
- 3:14Nikki: And I think that stopped the nuclear power plant retirement.
- 3:17Nikki: And then with the AI demand, obviously, we're seeing a lot of headlines, especially in SMR gabby's coverage, right?
- 3:25Nikki: You're seeing a lot of development in that area.
- 3:28Nikki: And so we are gearing up for expansion for sure.
- 3:32Ryan: Interesting.
- 3:32Ryan: So you mentioned some of the tax credits from the uh the IRA and how that benefited stabilizing the space.
- 3:41Ryan: Um, if I'm not mistaken, wasn't the the plant that was built in Georgia that came online fairly recently, I believe that was supported by some tax credits as well.
- 3:50Ryan: Is that correct?
- 3:51Nikki: Yes, they are, because um, everything that's existing power plant that started construction before the II was introduced, um, they are all entitled for the same tax credits.
- 4:05Ryan: Okay.
- 4:05Ryan: Um I'm I'm curious, and we're we're gonna ask some SMR specific questions as well, but I want to stick to the sort of traditional power plants for a minute because um the plant that most recently that I'm aware of came online was that plant in Georgia.
- 4:21Ryan: And it took something like 14 years to get fully operational.
- 4:25Ryan: Um and it always makes me wonder why did it take so long?
- 4:28Ryan: Why does it take so long to build nuclear power plants?
- 4:32Gabriela: So you're talking about southern companies, two new reactors.
- 4:35Gabriela: We're talking about Unit Three and Unit Four.
- 4:37Gabriela: They finally did come online, but there were several issues along the way.
- 4:41Gabriela: First, uh the design wasn't what we call shovel ready, meaning that the construction began, but the design still needed some adjustments.
- 4:50Gabriela: So that created delays and also some confusion.
- 4:54Gabriela: Also, there was work like for building these reactors, you need very specialized workforce, which uh unfortunately is unsort shortage then and probably still now.
- 5:07Gabriela: Uh, in addition to that, you also had the Westinghouse uh bankruptcy.
- 5:12Gabriela: So they had a contract of what is called a fixed cost contract.
- 5:17Gabriela: And when the costs started like ballooning, they had to absorb some of that overrun cost, meaning that they experienced like massive losses and then eventually had to file for bankruptcy.
- 5:29Gabriela: Also, the whole idea of the construction of the reactors was that some of the some of the parts or some of the modules were supposed to be uh construct or built in a factory and then assembled on site.
- 5:41Gabriela: However, some of those parts or some of those modules were like uh delayed or were even defective by the time they arrived to the site.
- 5:50Gabriela: So it required rework and of course then more delays and more costs.
- 5:56Gabriela: So everything combined created these you know massive uh cost overruns and these uh overly uh delayed project.
- 6:06Ryan: So um is that typical?
- 6:08Ryan: I mean, there haven't been too many plants built recently, but there were there was a period of time where there was a lot of nuclear power coming online.
- 6:15Ryan: Um was it typical when they were originally built?
- 6:19Ryan: And mainly we're talking in the US now, I'm curious about around the world as well.
- 6:22Ryan: But um, is that timeline, you know, 14, 15 years?
- 6:26Ryan: Is that is that what we experienced before in building out the nuclear power plants, or is that something that just a cascade of uh problems came with uh the Vogtol plants in uh Georgia?
- 6:38Gabriela: I think the problem is that yes, we were adding multiple gigawatts, you know, back back in the day, but then the US stopped, right?
- 6:45Gabriela: So we lost that repetition and we lost all that experience that had been built.
- 6:50Gabriela: And therefore, like every new project, I mean, in this case Volcle, was like a new mega project that it had to start from scratch.
- 6:57Gabriela: So then you're like facing like it almost like it is a first kind of kind of uh construction when when it shouldn't have, because if we would have continued that progression, that you know, never stopping building, then we wouldn't have lost uh that repetition and that experience.
- 7:14Ryan: Is has the timeline been similar in other parts of the world that have um added nuclear power capacity?
- 7:21Ryan: Um I mean I I'm not really sure if if Europe has added anything uh recently, but um the plants that are built there or in Asia and China and so forth?
- 7:30Gabriela: So it's my understanding uh that it's not as bad as it is in the US.
- 7:35Gabriela: I couldn't speak to be honest with you, like in detail about it.
- 7:38Gabriela: I don't cover uh the markets uh and I don't follow that in that deep in that detail.
- 7:42Gabriela: But even in other countries, like uh you've seen some delays as well and some cost of runs.
- 7:47Ryan: Okay.
- 7:48Ryan: Um so we've talked about the decommissioning or retirement of nuclear power plants.
- 7:54Ryan: Um we over the last couple of years have seen some recommissioning, if that's I don't know if that's the right term to use, but um the big uh announcement came from Constellation uh that they were going to restart one of the reactors at Three Mile Island that had been shut down, and there was a 20-year power purchase agreement with Microsoft.
- 8:12Ryan: Um is that something that we're going to see more of?
- 8:14Ryan: Are there other plants that can be restarted?
- 8:17Ryan: Do you have any uh any thoughts on the restarting or recommissioning of nuclear power?
- 8:23Nikki: Um I would say we should be seeing more of nuclear power purchase agreements signed with hyperscalers, but uh restart of nuclear power plants probably would not be um likely.
- 8:40Nikki: Because right now we have three active projects.
- 8:42Nikki: Um the um Three Mile Island that you mentioned, or we can just call it TMI easier.
- 8:49Nikki: So TMI is one, the other two is um Palisade, that was what we talked about, that was shut down in 2022.
- 8:59Nikki: Um, Holtech International is restarting that project.
- 9:03Nikki: The third one is actually Nextera's um Dwayne Arnold.
- 9:07Nikki: All these three power plants were shut down um, I think 2019 or afterwards.
- 9:14Nikki: So they are pretty much um early in the decommissioning process.
- 9:20Nikki: Most of their reactors are largely intact.
- 9:23Nikki: So on top of it, um they are supported by power contracts.
- 9:29Nikki: So they have the economic um competitiveness to compete with new built, new power plants, right?
- 9:38Nikki: Because obviously restarting a retired nuclear power plant will be faster, cheaper.
- 9:46Nikki: So it makes sense.
- 9:47Nikki: But if you are too deep into the decommissioning process, you'll be almost like the refurbishment will be too much.
- 9:55Nikki: You will be almost like costing as much or takes almost as long as building a new one.
- 10:00Nikki: And building a new one, you could get up to 80 years of lifespan, right?
- 10:06Nikki: But the existing, but for the re to restart the project, it depends on which part which power plant you restart.
- 10:13Nikki: If if you are restarting something built in um 1980s, so 80 years you you will be what, 20, 60, right?
- 10:24Nikki: So the economic might not make sense.
- 10:27Nikki: So right now all three projects are all shut down largely because of economic reasons.
- 10:34Nikki: They were not able to compete with the lower cost natural gas and the abundance of renewable.
- 10:42Nikki: We built a lot of renewable in the past.
- 10:45Nikki: So that's the reason for the three power plants to shut down.
- 10:48Nikki: So the rest of the, like I mentioned, there were 13 shutdowns since 2013, right?
- 10:54Nikki: Most of those are either deeper into the decommissioning process or there could be some damage to the reactor itself.
- 11:04Nikki: I just don't think it's easy to restructur, to restart those projects and to be as competitive as building a new one.
- 11:13Nikki: Economic-wise, it doesn't make sense, right?
- 11:16Nikki: I think there were talks about um Crystal River in New Jersey, I think.
- 11:23Nikki: Um Jersey.
- 11:28Nikki: And then another one, I think, is Kiwani in Wisconsin.
- 11:34Nikki: Um, people were talking about potential to restart these.
- 11:38Nikki: But then again, these are really, um, in our view, they are already progressing in the decommissioning process, progress, right?
- 11:47Nikki: Um, to be able to really be competitive compared with New Bill.
- 11:52Nikki: I just don't see how you're gonna be able to attract any data center to be willing to support or sign a contract to make it um competitive enough to actually go through this process.
- 12:08Nikki: And obviously that's our view.
- 12:10Nikki: Yeah.
- 12:10Ryan: Yeah.
- 12:10Ryan: Well, and it makes sense um uh just pragmatically, I would think if it did make economic sense to restart those reactors, I would have heard an announcement already.
- 12:20Ryan: Um, because there is so much demand for electricity and and so much projected demand growth.
- 12:25Ryan: Um, it just seems like that was um, I mean, the constellation announcement was a couple years ago now, right?
- 12:32Nikki: Probably two, two, two years, two years, yeah.
- 12:35Ryan: And I I remember um not too long ago, uh there was a headline that suggested that it was ahead of schedule.
- 12:42Ryan: Is that um is that still true?
- 12:45Nikki: Um originally they were saying they are targeting 2028, but now they are targeting 2027.
- 12:51Nikki: But the problem is, so I guess um maybe I'll add on to what we talked about, why it probably is not likely to have more, because uh it's very um complicated once you go into the um the restarting um progress, right?
- 13:13Nikki: So for cancellation, the problem, the new risk that we heard about is not and is not technical risk, it's more of a monetization risk.
- 13:23Nikki: So cancellation is saying that they actually could have it online by 2027, but the problem is that the grid operator, BJAM, is not able to um interconnect this um power plant to um to supply power.
- 13:43Nikki: So the transmission and interconnection um is the problem right now.
- 13:48Nikki: So by the time 2027, cancellation might have the power plant ready to go.
- 13:53Nikki: But you might not be able to get 100% of the capacity out of this power plant just because transmission and interconnection, because the power plant was shut down in I think 2019, I think.
- 14:06Nikki: And from now to from now to then it's what six, seven years.
- 14:10Nikki: So there are a lot of renewables being built.
- 14:13Nikki: So transmission might be used somewhere else.
- 14:16Nikki: So so the capacity of transmission is not fully available for um for TMI.
- 14:23Nikki: So when it restart, you're probably not gonna get 100% out of it.
- 14:27Nikki: So that would be an issue because back then, when they run these numbers, it was based on 100% of capacity, I'm sure.
- 14:35Nikki: So now that's the challenge, right?
- 14:37Nikki: Then you you have a different challenge for Palace.
- 14:41Nikki: Palace was supposed to come online last year, then they push it to this year.
- 14:46Nikki: And it's because once they started to bring the power plant back online, the NRC, the Nuclear Regulatory Commission, will obviously do more review.
- 14:59Nikki: And then they realize actually it needs more refurbishment than what they expected for on the nuclear, the nuclear island itself.
- 15:07Nikki: So now they have to push it back to 2026, and we are already about mid of the year, so we'll see how it goes.
- 15:16Nikki: So that's a technical challenge, right?
- 15:18Nikki: So obviously it's something more difficult than what we expected to actually restart a power plant.
- 15:27Nikki: And for Next Terra Strain Arnold, um the power plant itself, um, it does the nuclear island is fine.
- 15:35Nikki: They just need to um rebuild their cooling tower, which is a more simplified process.
- 15:40Nikki: It's just an industrial project, right?
- 15:43Nikki: So not nuclear island itself.
- 15:45Nikki: So it seems to be easier than the um Palisades issue.
- 15:51Nikki: But I think um Nextera was saying that they are trying, they the timeline might shift a little bit into early 2029 rather than um 2028.
- 16:03Nikki: So most likely it's just an execution risk that we have to be monitoring there.
- 16:09Nikki: So that's why, like all these new risks, um, we didn't see this coming when these projects were announced.
- 16:16Nikki: So that's why there's always something that you might not see when you make the announcement.
- 16:23Nikki: And to be able to go through this process for the other 13 reactors, I just feel like there's a lot of risk there.
- 16:31Ryan: So um that that brings me to another question.
- 16:33Ryan: I'm curious about um you mentioned the the technical issues of restarting a reactor.
- 16:38Ryan: And um for those of us who are laymen and don't really understand how nuclear power plants work, um, can maybe just give us a brief description because then I'm I'm curious how that translates into a small modular reactor.
- 16:52Ryan: So, how do nuclear power plants generate electricity?
- 16:56Gabriela: Both traditional uh nuclear power plants and and small modular reactors are you know based on this on the same premise.
- 17:03Gabriela: Like you create nuclear fission that you know creates uh heat and then steam and then eventually produces electricity.
- 17:12Gabriela: The concept is the same regardless of how the large, how large or how small the reactor is.
- 17:18Gabriela: Um but if the real difference comes that you're not talking about a multi-gigawatt project, you're talking about potentially a couple of hundred megawatts or even smaller.
- 17:32Gabriela: And the whole premise is that this can be scalable, meaning that you can have one module of 75 megawatts, for example.
- 17:40Gabriela: And while your industrial process, your AI data center or whatever it is, is ramping up, then you are also adding more modules and meeting that demand that is going to the whatever you have, aid data center, whatever needs.
- 17:54Gabriela: Um but the premise is is, you know, basically in terms of how it operates, is basically the same.
- 18:00Gabriela: It uh it's just a matter of the scale and what is the purpose of the power plant.
- 18:07Ryan: So is the timeline to build those smaller modular reactors um uh in terms of and I know they're not commercialized yet, but uh the the expectation, my assumption is it would be uh quicker to build some SMRs than it would be to build a full-blown nuclear power plant.
- 18:26Gabriela: So we're hoping that we're going from, you know, um to replace that huge scale of one uh gigawatt, you're producing massive amounts of uh electricity from one facility.
- 18:39Gabriela: That's a traditional power plant.
- 18:40Gabriela: And then you are going to offset that scalability with multiple uh modules that are going to be uh hopefully produced efficiently and is going to create economies of scale and repetition, meaning that the costs are going to start really high, but eventually they're going to uh be lower and it's going to be uh competitive in the market.
- 19:03Ryan: Okay, so if we're thinking about adding SMRs to the to the grid or or maybe behind the meter for certain data center projects or things like that, um it seems to me that um again, I'm based on headlines and things that I've read.
- 19:23Ryan: I don't cover it as an analyst, that this is something that is not expected in the next couple of years, but in the not until like the 2030s.
- 19:31Ryan: Is that uh is that timeline realistic?
- 19:33Ryan: Where do you when do we actually begin to see the commercial commercialization of some of those SMR projects?
- 19:40Gabriela: Yeah, so it depends like where how we define, you know, commercialization of these SMRs.
- 19:46Gabriela: Uh, I'm sure you're referring because of the headlines, like a lot of demonstration projects are probably going to be online in the early 2030s.
- 19:56Gabriela: But when we talk about commercialization, we're talking about like being at scale, meaning deployment, meaning that you can order and it can be delivered relatively quickly and resolve the AI data center problem.
- 20:08Gabriela: That is probably not going to happen until after 2035.
- 20:11Gabriela: There are several things that need to be resolved.
- 20:14Gabriela: First is we have an issue with fuel.
- 20:18Gabriela: So most of the small modular reactors, in particular generation four, which are non-light water reactors, they need what we call helium.
- 20:29Gabriela: And that's uranium that is unreached between uh 5% to 20%.
- 20:33Gabriela: That's practically non-existent at this time in the US.
- 20:37Gabriela: Uh so there is a lot that needs to be done for that supply chain to be developed so it so we can then deploy uh the SMRs.
- 20:47Gabriela: Then that's one issue by itself.
- 20:50Gabriela: It's probably one of the more important bottlenecks.
- 20:53Gabriela: But then you also have regulation.
- 20:55Gabriela: Uh so all these uh small modular reactors need an NRC uh license to be uh able to operate.
- 21:03Gabriela: And although we've seen uh some uh incentives uh from the federal government in streamlining that process, is still, you know, we still don't really know how effective and how quick uh that can be done in order to have uh you know uh an industry that is going to be at that scale.
- 21:24Gabriela: And then uh you also have the technology is still very new in the sense that we don't really have uh many, we don't have one online.
- 21:35Gabriela: We're still working on the demonstration projects.
- 21:38Gabriela: Uh and so there's a lot of uncertainty, and because of that reason, uh if you go to the bank, it's very unlikely that you're going to get uh any financing, right?
- 21:50Gabriela: Because then you don't know when it's going to be online, you don't have a power purchase agreement, you don't have fuel availability.
- 21:57Gabriela: So there's so many things that need to happen.
- 21:59Gabriela: That's why we are what we will say culturally optimistic.
- 22:04Gabriela: And we do think that technology, it has, you know, it's promising and can provide 24-7 firm power.
- 22:14Gabriela: But there's a lot of work that needs to be done and a lot of stakeholders that need to cooperate for that to happen.
- 22:21Ryan: So one of the one of the questions that uh I've always had, and and actually I asked another uh friend of mine who's a uh utility manager uh was about um the fact that there's the military has a version of small modular reactors that have been on nuclear submarines and aircraft carriers for years.
- 22:39Ryan: Um and so it exists in some way.
- 22:42Ryan: Why is that not um just ready to be commercialized?
- 22:47Gabriela: Those are just different purposes, right?
- 22:50Gabriela: The military ones that are focused on like being very small and focused on national security.
- 22:57Gabriela: They need uh also they're looking not to be refueled too often because they're in extreme conditions.
- 23:04Gabriela: And like opposite, we have the SMRs, we want it to be commercial available for anyone to be able to acquire them.
- 23:12Gabriela: So we're being we're focused on cost efficiencies.
- 23:15Gabriela: We're trying uh to make it scalable so like it, you know, it can be commercialized easily.
- 23:20Gabriela: And also fuel.
- 23:22Gabriela: I believe the military ones use highly enriched uh uranium.
- 23:30Gabriela: Uranium.
- 23:30Gabriela: Thanks for having me.
- 23:31Ryan: Yeah, that's well that's what that's what my other friend uh we were uh talking actually on the podcast, and he uh he thought it was something like 95% enriched uranium, correct, and which seems like that would be a problem uh out in the wild.
- 23:43Gabriela: Absolutely, correct.
- 23:44Gabriela: So that right right there, yeah.
- 23:46Gabriela: You you just don't, you just have to use something different.
- 23:49Ryan: Yeah, so there's the economics where the federal government um can spend what it wants to spend.
- 23:54Ryan: There's the scalability, there's the engineering related to using um uh you know highly enriched uranium, you gotta use less enriched uranium, and then there's the supply chain.
- 24:04Ryan: So those are all uh some of the hurdles.
- 24:06Ryan: Yeah.
- 24:07Ryan: Yeah.
- 24:09Ryan: Um so the other thing that I I remember um was a big concern, and I haven't heard about as much about it lately, um, is related to um the waste that that's produced by nuclear power plants.
- 24:25Ryan: Um is that an issue?
- 24:26Ryan: Is I mean how much nuclear waste is actually produced by running a nuclear power plant and how is that mitigated?
- 24:33Ryan: Is this an issue that people should be concerned about?
- 24:37Nikki: Well, obviously people are concerned about it, but if you look at uh one gigawatt of uh nuclear power plant, which is the average size of a nuclear power plant, each one gigawatt of large-scale nuclear power plant, they probably produce two 20 to 30 metric tons of waste, right?
- 24:57Nikki: Yeah.
- 24:58Nikki: So and then uh it doesn't seem like a lot.
- 25:02Nikki: Um, and they are actually stored on site in a pool first to cool off.
- 25:08Nikki: And then once the it's a few years later, I think three to ten years, once it's cool off, they were stored it in a dry cask.
- 25:18Nikki: And then after that, the um I don't know, maybe 30 years or 40 years, they are there, then been transport to um Yuca Mountain.
- 25:31Nikki: Um, I think that's what it's called.
- 25:33Nikki: So um I believe we do have the space to be able to store these uh waste so far as of now.
- 25:44Nikki: Yeah.
- 25:44Ryan: So so it's in a mountain, so the idea is you put it where it's not gonna have any um radioactivity or it can't it's far enough away from everybody that it won't be harmful.
- 25:55Ryan: Yeah.
- 25:56Ryan: Yeah.
- 25:57Ryan: Um interesting.
- 25:59Ryan: So am I correct in my assessment that the sentiment has shifted?
- 26:04Ryan: And and if so, is it um you you were talking about some of the stabilization as a result of maybe some more positive economics from uh tax credits and things, but it seems to me that um it has a lot to do with data centers and the ratcheting up of high quality electricity.
- 26:22Ryan: And you know, nuclear power plants have a long shelf life, 80 years.
- 26:28Ryan: If you can produce 80 years with a capacity factor of, you know, 90 plus percent compared to wind and solar, which are you know going up like crazy, but their capacity factor is 25 to 30 percent.
- 26:40Ryan: Um is and they're weather right, they're weather dependent.
- 26:44Ryan: So um is it all about data centers and and has the um has the sentiment shifted because of that?
- 26:53Gabriela: So I think um in part, yes, because the demand is there.
- 26:58Gabriela: And a lot of these, especially A data data centers, you know, all these tech companies have like uh climate goals that they want to reach.
- 27:08Gabriela: So they are very interested in that 24-7 clean power.
- 27:12Gabriela: Uh so definitely there's some of that.
- 27:14Gabriela: But you also have uh other factors like electrification, uh reshoring, that are also impacting uh the power demand.
- 27:24Gabriela: Uh so I think it's just a combination of everything along with the policy um support that we've seen, uh, especially uh lately, that is coming also, you know, to support a new nuclear.
- 27:38Ryan: So I I have kind of been watching the demand grow for electricity, and all those things have contributed data centers, reshoring, electrification of vehicles and electric heat pumps and all the building codes and things.
- 27:52Ryan: Um I'm curious where we are in that cycle because it it has only ratcheted higher year by year.
- 27:59Ryan: Uh, I think uh, you know, the the five-year window of time we're in right now, the forecasts in 2022 when Chat GPT was first announced were something like you know, on trend 3% uh over that five-year period.
- 28:12Ryan: Uh last year the figure I saw was it's gonna be like six or seven times that, so like 20%.
- 28:18Ryan: Um, do you do you um think that it is going to continue to ratchet higher, or have we sort of reached the maybe the peak amount of expected growth for electricity demand?
- 28:30Nikki: Um I would say that um the actual electricity sales probably would grow at a slower pace than what hyperscaler CapEx has just suggested, because you could have um invest as much in building a data center, but if you don't have power to support it, you're not gonna be able to run.
- 28:52Nikki: So based on our analysis, we actually did an analysis, a scenario analysis.
- 28:58Nikki: And in the scenario, we actually see um data centers um contributing um from now to 2032, because we look through 2032 only, probably two-thirds of the incremental um electricity demand coming.
- 29:19Nikki: But um with that um it it's that way just because um we expect there will be a capex to power lag of about three years.
- 29:33Nikki: And that's actually pretty um aggressive.
- 29:36Ryan: Already so is that lagged because of just the backlog of demand for some of like you were talking before about the the grid not being able to keep up with um the constellation project.
- 29:46Ryan: Is that why there's the backlog?
- 29:48Nikki: Interconnection is uh, yeah, transmission interconnection are probably the main um bottlenecks.
- 29:54Nikki: And then if you look across different regions, some regions are even worse, right?
- 30:00Nikki: PJM, which is um it's it covers northeast mid-Atlantic states, 13 states plus DC, right?
- 30:09Nikki: This area actually um has the longest wait time, um up to seven years are likely, right?
- 30:17Nikki: So from what we based on our analysis, um data center demand um would not translate into one-to-one right away.
- 30:28Nikki: So if you assume the three-year delay, we see the electricity demand growth, um sales growth somewhere around, I think 4.5% keger from um over the seven years.
- 30:41Nikki: Yeah, if I remember correctly.
- 30:43Nikki: Yeah.
- 30:44Ryan: Uh we won't hold you to it.
- 30:46Ryan: Um so um, if I'm not mistaken, maybe two or three weeks ago, maybe a month ago, um, the Trump administration declared a national emergency and invoked the Defense Production Act related to some of the grid issues.
- 31:02Ryan: And um, I'm, you know, I never know what to think about some of the announcements that are made by the federal government and what impact they'll have.
- 31:10Ryan: Um does that figure into um, I mean, should we be thinking about that?
- 31:14Ryan: Will that actually, you know, if you the Defense Production Act, I think the federal government can provide more financing and maybe lower some of the regulatory barriers.
- 31:23Ryan: Um, is that something that would have a material impact, do you think?
- 31:29Nikki: Um, I think on the regulatory side, maybe Gabby can speak better, but um a lot of these executive orders definitely help with market sentiment.
- 31:38Nikki: So, and President Trump has been very supportive of um nuclear power.
- 31:44Nikki: So um obviously the big beautiful bill, right?
- 31:48Nikki: It introduced a new um tax credit, so it extended the nuclear tax credit because the inflation reduction act, I think, ended in 2032.
- 31:58Nikki: But then with the new um bill that passed by President Trump in July last year, I think, it it could go all the way through 2039 or even beyond for the new nuclear debt.
- 32:12Nikki: A lot of them probably will be likely to be SMR.
- 32:16Nikki: But the regulatory process probably will be very different because nuclear obviously is some there's a safety concern everyone is worried about.
- 32:24Nikki: So what do you think?
- 32:25Gabriela: I think we'll have to see.
- 32:27Gabriela: Uh we're very uh, I guess, hopeful after uh TerraPower obtain a construction permit uh from the NRC in a relatively short time uh compared to what uh we have seen in the past.
- 32:42Gabriela: Uh but it but it remains to be seen, right?
- 32:44Gabriela: It's still uh very early.
- 32:46Gabriela: The NRC is going through a process of you know revising uh their processes and how they're going to approve uh and all that.
- 32:55Gabriela: And they have uh you know uh come out with uh new ways of approving, especially for non-water uh uh reactors, non-light water reactors, I'm sorry, uh, that gives us hopes that it will be a more efficient uh process.
- 33:11Gabriela: But then again, we don't have that many uh not that many uh reactor developers have uh entered a new uh application, so it will be hard to test if that is going to be uh if you know effective or not.
- 33:27Ryan: Yeah, that the one you mentioned, the Terapower one, is that the one in Wyoming?
- 33:31Ryan: That's um somewhere around 2030, I think, is when that's starting to be.
- 33:35Ryan: Correct.
- 33:36Gabriela: 2030, exactly.
- 33:37Gabriela: Yes.
- 33:37Gabriela: That's the demonstration, the first demonstration project, correct.
- 33:41Ryan: Okay, so um my other again, I'm thinking about data centers and and uh reshoring and big chip factories and things that have a a lot of demand.
- 33:50Ryan: Um maybe that'll be satisfied by um nuclear in some form, but um I am curious, um, and maybe this is a broader utility question, but do we do we expect to see more behind the meter building of power plants?
- 34:06Ryan: Um and that seems like an SMR thing would be great, but they're not around for the next 10 years, as you as you mentioned.
- 34:13Ryan: So um are are we gonna see more behind the meter um building of power plants?
- 34:19Ryan: And what does that mean for um do they have to be connected to the grid?
- 34:23Nikki: You actually bring up a very good point.
- 34:25Nikki: So obviously, with the interconnection delay, right?
- 34:29Nikki: Data center, they want power, they want power now, right?
- 34:32Nikki: If they cannot wait for nuclear to be online by 2038, they don't want to wait to be um connected it for five to seven years.
- 34:40Nikki: So they have to find a way to get power.
- 34:44Nikki: So a lot of them, what they are doing right now is they are looking for bridge power.
- 34:48Nikki: So it could be natural gas.
- 34:51Nikki: Mostly of these will probably be natural gas and renewables.
- 34:57Nikki: But behind the meter, um I think some states probably would have more.
- 35:05Nikki: Well, like we mentioned earlier, PJM, it has the longest wait time, but then at the same time, it also has the most complicated regulatory process.
- 35:19Nikki: So um Ferg and PJM has to have a reform to actually try to um speed up the interconnection queue.
- 35:29Nikki: So, and they also um are actually debating uh behind the meter, um large um large load, that those are the data centers, the um chips factory, right?
- 35:44Nikki: They are considered large load.
- 35:46Nikki: And then for this process to pass, we'll give clarity on how much behind the meter can be done, right?
- 35:54Nikki: And I think part of it, a lot of like when we talk about the nuclear power contracts signed with Microsoft, um, Google, Meta, right?
- 36:04Nikki: We haven't heard a lot recently.
- 36:05Nikki: And part of the reason is because of the regulatory uncertainty.
- 36:09Nikki: Most of the nuclear power plants are in the Northeast, the PJM zone, most of them are there.
- 36:15Nikki: But they also are the ones lacking the clarity.
- 36:19Nikki: So no one is moving forward.
- 36:21Nikki: We are all waiting.
- 36:22Nikki: So for people to actually go with behind the meter, they have to wait for that, right?
- 36:28Nikki: So um, but in the integrated states, regulated states like Midwest, these areas, they they are just regulated states.
- 36:39Nikki: So the utilities are the ones providing generation, transmission, and distribution.
- 36:46Nikki: So when we talk about behind the meter, I do believe um they will move forward, but um how large of a scale, we don't know.
- 36:56Nikki: Again, coming back to the um model that we built, we did talk about this.
- 37:02Nikki: So for us, we expect um more behind the meter project, or you could call it off-sile, um, off-grid or on-site generation.
- 37:13Nikki: We just call it on-site generation.
- 37:15Nikki: So a lot of data centers probably will look to find whatever power or transmission capacity available with low cost.
- 37:24Nikki: So they'll find a state with low cost, supportive regulatory environment, and have the generation and transmission capacity and to get connected ASAP.
- 37:36Nikki: And then they will also look for on-site generation, right?
- 37:40Nikki: So um just so that they they can get their data center started.
- 37:45Nikki: Because right now, when you start a data center, you don't need full RAM-part, full capacity.
- 37:50Nikki: You just need something to start.
- 37:52Nikki: So you'll start your data center first.
- 37:54Nikki: And then in the meantime, you still want to be connected to the grid.
- 37:59Nikki: On-site generation, a lot of them call on-site because they are on site, but that doesn't mean they are not connected to the grid.
- 38:06Nikki: Most of data centers want their generation to be connected to the grid.
- 38:11Nikki: My thinking, and Gabby, you can correct me, I think we we think that um even when we have behind the meter generation, right, a lot of them will still be connected to the grid.
- 38:24Nikki: Because, first of all, um you want to have that guarantee.
- 38:29Nikki: Because building a pure on-site behind the meter generation is very expensive.
- 38:36Nikki: Because with the grid support, if there is, and if you need power, you can get it from power.
- 38:42Nikki: If you build it on site, for one gigawatt, you're gonna be worried about what if the power plant went down with something.
- 38:49Nikki: I don't have the reliability.
- 38:51Nikki: Yeah, I need backup, or maybe I need instead of one gigawatt, I may be building two gigawatt.
- 38:56Nikki: But with um utility, you have that kind of support.
- 39:00Nikki: So um I honestly feel like um even with behind the meter, you're not gonna be, most of them will still want to be connected to the grid.
- 39:10Nikki: It's most more cost-effective.
- 39:12Nikki: You also have more reliability um guarantee as well.
- 39:17Nikki: But at the same time, we do believe on-site generation will grow much faster.
- 39:21Nikki: So I think EIA was expecting close to 9% of growth over 2032.
- 39:29Nikki: In our model, we see a slightly higher, a little bit over 11%.
- 39:34Nikki: So 2-3 percentage higher than what EIA is is expecting.
- 39:39Nikki: Yeah.
- 39:40Ryan: Well, I can't believe that the time has flown by so much talking about nuclear power, utilities.
- 39:46Ryan: Um, you have found a way to make it very, very interesting for our viewers of the podcast.
- 39:51Ryan: So thank you very much, Gabby.
- 39:52Ryan: Thank you, Nikki.
- 39:53Ryan: Thank you for having us.
- 39:53Ryan: Um, maybe we can do this again sometime, but I've I've really enjoyed the conversation, and I think I've learned a lot.
- 39:59Ryan: So hopefully the viewers of the podcast have as well.
- 40:02Ryan: Um, so again, thank you for joining us on the podcast, and thanks to all of you as well for joining us on this episode, the First Trust ROI podcast.
- 40:09Ryan: We'll see you next time.