Kairosのビジネス開発・財務担当副社長、ジェフリー・オルソンとのファイアサイドチャット

[0:00] Welcome back. I'm Vijay Vaitheeswaran. Global Energy and Climate Innovation Editor at The Economist. We're live from the Global Conference on Energy and AI. hosted by the International Energy Agency in Paris. I'm moderating a series of conversations with leaders from government, business, the tech industry, energy industry, and civil society. to talk about how AI will transform the energy system and the global economy in the years ahead. I'm pleased to have with me now Jeffrey Olson, head of commercial at Kairos Power.

[0:32] Welcome, Jeff. Thank you, Vijay. Pleasure to be here. So your company has already had a shout out in our series of live streams today from Google, a partner. We'll get into that in a little bit. But for those who are unfamiliar, you're developing novel advanced nuclear technology. Is that fair to say? That's fair to say, yeah, correct. Excuse me. Carbos Power is developing advanced reactor technology. Our power plants will be 150 megawatts when they come online starting in the 2030s. Great. We're going to get a little bit into your technology as it relates to our challenges that we're dealing with here at the nexus of AI and energy. But first, maybe we can start.

[1:09] with a broader frame. If you talk about the problem that you're trying to address here. Thank you. It's well established, including in our conversations today, that we're entering an era of electrification, rapid growth. We heard that the growth in electricity will be perhaps six times the rate of growth of energy growth in coming years, according to the IEA's analysis. And yet we've had a period of stagnation in terms of investment in grids. The U.S. has seen flat growth for the last 10 years, and now we're seeing a sudden wrap-up for multiple reasons, among which data.

[1:43] centers in AI are an important factor. utilities aren't used to this rate of investment, this rate of growth of late, and sometimes are risk-averse when it comes to new technologies. So can you tell me about a little bit of the how you persuade sometimes conservative regulators of utilities, as well as the utilities themselves, to be ready for a new era of growth and to maybe try some novel and exciting new technologies. Excuse me, absolutely. I think very well said. There's been great discussion over the past couple of days about how the demand for energy is moving up and to the right. Regardless of whichever precise curve one would want to pick, I think the general understanding is that is happening. Absolutely. And that's putting unprecedented pressure on both grid operators and the consumers of energy, right? And where are they going to find the energy they need to run their businesses?

[2:37] energy to us and we hit the light switch, we expect that to always be there, expect that electricity to be there. It is another level of challenge to take on the new risk of investing in new technology. And many utilities are doing that in ways. But I think oftentimes those procurement processes are set up to favor more conventional. less risky forms of generation. And so one of the things that we're doing at Kairos Power is we are working to work with those utilities through our iterative development, iterative build process to show them that the demonstration, the credibility of our ability to build, to demonstrate cost.

[3:16] and to demonstrate delivery schedule certainty. walk them through the development process and rather than as in perhaps some traditional large-scale nuclear plants, and our viewers should understand your small modular approach. This is not the traditional giant nuclear power plants of your grandfather's age as it were. that tend to be very lumpy, take decades to build, and have lots of inherent risks in the way that they're one-off built. Your concept is more modular, am I right? That is to spec out the process and then to have replicable plants. Yeah, that is correct. The Kairos Power plant is 150 megawatts when we come online. And like I said, those will be coming online the end of this decade, 2030 timeframe.

[4:07] The ability to build them side by side to scale up to a gigawatt or so is that modular flexibility. It's different than a solar module, of course, but it does allow that modularity. In terms of how the process works, for nuclear, what one must start with is the licensing process, and that's a sovereign process governed by each country. In the U.S., we go through the Nuclear Regulatory Commission, the NRC, and so we kick that process off. We gain our license based on a specific site, and then we go out, we build, and then ultimately operate. The time horizon that we're looking at through that whole process is more along a five-year horizon, a couple years for licensing and then a few years to build.

[4:49] And you're an American company. We are, yes. And the U.S. government has been supportive of SMR technology, not just your company but others as well. Correct. As sort of an effort to lead this race. Is that fair to say? The United States Department of Energy has been supportive in the form of the Advanced Reactor Development Program, and Kairos Power was a recipient of that program. That has certainly benefited the development of our first demonstration unit, which will come online in a few years. But now you're beyond that. You're looking at what I recall from your recently announced deal with uh... with google a deal that could be for multiple plants and it's sort of an advanced market commitment. Can you tell us a little bit about what that order book approach to doing this means, what it means for you, what it means in terms of de-risking or sharing of risk, and what it might set as a template for others trying to bring novel energy technologies to bear so that some of the demands from AI can be met and other sources can be met. Yeah, so absolutely. The

[5:49] AI developers and technology companies have the ability to play an incredibly catalytic role to bring new energy to market. And many are. Google absolutely is playing that role with us through our Multi-Plant Development Agreement. And the reason a Multi-Plant Development Agreement is so important is because it increases the number of iterations that we and Google have to demonstrate to the world and to utilities that these projects, these new technologies are viable. Right? So Google helps to step in to give that off-taker surety so that the utility doesn't have to convince a larger consortium of their ratepayers to do so. It helps Google to have the benefit of knowing they have a firm baseload carbon-free technology coming online. And then we get the iterations to drive down our cost and increase the certainty. And today has been a lot of discussion about data. And if we look at any new, almost any new energy technology curve, there's a learning curve that drives costs down over time. And that is an important part of this process and this deal structure that we have. Just to unpack that a little bit for our viewers.

[6:53] some of the advantages of thinking about this order book concept. the end of one, the first plant of any new technology that one builds, will probably be more expensive and more complicated. Absolutely. Probably run into some supply chain problems because you're building with novel kit and so on. By the time you get to end of six, let's say, the hope is that you've come down a learning curve. Is that what we're talking about here? 100%. Yes, you should come down that learning curve. And are you working with your supply chain? Again, there are some issues even in the small modular reactor space. We have seen a company that has fared badly in recent months, in part, they say, because of their supply chain and maybe market conditions changed. So there are uncertainties. How does that work?

[7:34] having an order book concept or your modular approach help you deal with some of those uncertainties? So the order book helps to demonstrate the commercial reality and the commercial commitment of an off-taker to suppliers. Carbis is also handling our supply chain a little bit differently than a lot of SMR companies, and we've got a commitment to vertical integration for those things on the nuclear island that touch the fabrication of our fuel, the production of our salt, and other nuclear island critical components at risk.

[8:04] Right, so you're dealing with some of these risk factors. Presumably, an approach like this, there are other approaches too, but this idea of order book or advanced market commitment or bulk. Yeah. Orders, sharing of risk. would help with financing as well, because it shows your potential financiers that they can get in on some sort of pooled risk model? Yeah, I think there are at least three legs of the stool that you need. You need to demonstrate a capability of technology, a capability to license, and Kairos had accomplished those things. There's still work to be done to ultimately deliver your reactor online, but there's good progress there. And this commercial commitment that we have with Google is that third important leg of the stool for investors to see. Absolutely.

[8:41] And you also need to show that you can actually execute. We've seen in Europe, Northvolt, a major battery consortium, a European champion, fall apart. not necessarily because of technology flaws, but perhaps management and strategic decisions. of over expansion and how capital was allocated. So I imagine that's also something you'll have to prove. Yes, we have to be careful there. And there are a few ways that we're handling that. We've mentioned the iterative build process, continuing to build that fitness to deliver projects is imperative, we view, to drive cost and delivery certainty. And the other piece is there is a global market for nuclear out there, and we respect that and we look forward to that. But our near-term focus is a U.S. focus because one can too quickly expand, and then you can lose control if you move too quickly beyond a market scope. Very briefly, you mentioned the globality of this. We're in France. We're a long legacy of nuclear, a significant amount of energy is produced from traditional nuclear plants.

[9:37] rapid builder and prolific builder at the moment. Which market in the world do you think will lead on small modular reactor technology implementation? Well, that's a great question. I think we'd like to see ourselves having an important, meaningful leadership role within the United States. But that said, I think we're seeing traction across the globe in the discussion rooms over the past two days. So I'd like to see a healthy competition that raise for all of us. Great. Well, hopefully that competition will benefit not only AI, but the rest of society as well. We're going to have to leave it there, but thank you for joining us. Thanks to our viewers. This conversation, live from the IEA's Global Conference on Energy and AI from Paris.

[10:18] will be available for replay on the IEA's website. Up next, I'll be speaking with Pankaj Sharma, who is executive vice president for secure power at Schneider Electric. We hope you'll join us then. Thank you. ♪ ♪ you you you you you you