Nuclear Reactors In A Box (Radiant Founder, Doug Bernauer)

0:00How much resources do I have command of to put towards this thing that I know absolutely is the most important thing for SpaceX to eventually have is that power source to do fueling. So not only would it help that mission more, but it would help the mission of humanity more. Hello, and welcome to Giant Ideas with me, Cameron McLean, and me, Tommy Stadlin. We're co-founders of Giant Ventures, which builds and backs purpose-driven companies. At Giant, we're lucky to meet extraordinary people with Giant Ideas that are changing the world. This podcast brings you behind-the-scenes access to those ideas and the inspiring stories of the people behind them.

0:35We explore how one giant idea can kickstart a billion-dollar company, shape culture, and transform life as we know it.

0:45Today we are joined by Doug Bernhauer, founder of Radiant, one of Giant's most exciting growth investments. We backed Radiant in their Series C round. Their vision is to mass-produce portable nuclear reactors. But their vision goes far beyond this. They want to first manufacture nuclear reactors for Earth, giving clean energy to remote areas. It's a giant idea, both on this planet and beyond. The founder, Doug, worked with Elon Musk for 12 years at SpaceX. He worked on their reusable rocket and the Mars colonization project. In this episode, we talk about why nuclear's time has finally come, how Radiance approach is unique, and their long-term vision to deploy nuclear power on Mars.

1:24Before we welcome Doug, we've got our partner at Giant, Will Dufton. He led the investment in Radiant. He's going to give you an insight into what he saw in Doug, why he believes we're on the brink of a nuclear renaissance. Will, over to you for a quick explainer. Why should people care about nuclear energy right now? I think the primary reason is AI, like everything else in the news right now. But the rapid build out of data centers by companies like OpenAI and Meta in pursuit of artificial intelligence is basically bottlenecked by energy. There's a very specific type of energy that is required.

1:58It's 24-7, 365 power. It's baseload power. And the amount of energy is just almost unfathomable. So the forecasts are saying that by 2030, we'll need about 950 terawatt hours of power for data centers globally. This is the same amount of power that Japan consumes annually. It's just enormous. So where is this power going to come from? What the US is doing today is mostly leveraging its kind of natural gas resources. But the supply chain for natural gas turbines is pretty broken. And building this power and connectivity is also really slow and very expensive. So actually, even though nuclear kind of has this reputation of being a very slow power source to build and a very expensive power source.

2:44It is actually like a very attractive solution for data centers. It's clean, carbon-free baseload power that can be built behind the meter to serve these monstrous data centers. So yeah, really, nuclear is an AI trend. Just to be really clear, we're talking about nuclear fission, not nuclear fusion. So nuclear fission is splitting atoms. We've been doing this for 80, 90 years at scale. Nuclear fusion hasn't happened yet. nuclear fusion, when it does happen, will basically break all other energy sources, but it's still probably a couple of decades out. There's a lot of pressure on expanding energy resources for AI, but why nuclear?

3:22Yeah, look, it's not obvious why nuclear. Nuclear, as I said, has this reputation for being a very expensive and very slow to build power source. And obviously the kind of overarching reputation of nuclear is nuclear bombs. It is Three Mile Island, Chernobyl, fukushima these horrific disasters and so there is this reputation that nuclear is an unsafe technology you know aside from ai being an incredibly power hungry catalyst a consumer of energy kind of just demanding that humanity looks for more energy sources there are other reasons why nuclear kind of looks more attractive now than ever before the first is um it has been burdened with extreme regulatory safety and bureaucratic loads it's understandable like obviously we've just come out of a 30, 40 year period in the advent of Three Mile Island and Chernobyl, especially where we have not been building nuclear.

4:12We've been scared to build nuclear. And so as a result of that, the supply chain, the talent has atrophied for the industry. And so what was costing about a thousand dollars per kilowatt electric to build in the US in the 60s is now costing like 10x that much. It's pretty staggering. Nuclear was a cheap energy source in the 60s and 70s, and that was true in the US, in France, in Sweden. But in 2010 and beyond, it's become one of the most expensive and troublesome power sources to build globally. Public sentiment for nuclear now is at like decade-long highs, and people realize that it's safe. There are a lot of active nuclear power plants in the West, and so countries are kind of throwing their weight slightly more behind nuclear.

5:00And what has also been a catalyst in Europe especially is the Russian invasion of Ukraine and the knock on effects of energy security that that had. And so countries like the UK and France, having said originally that we're going to kind of decommission a lot of their active nuclear power plants, are now exercising a kind of life extension program to get more out of them. But, you know, really like what is happening right now in nuclear, the most aggressive regime change is in the US. The Trump administration has been extremely bullish on nuclear power. And I think that was kind of captured most clearly in some executive orders that were issued in spring this year, which basically paid the bureaucratic burden on reactors.

5:48It said they were going to resupply the nuclear supply chain in the US, including fuel supplies. And there have been inking deals that are remarkable. I mean, just last week, the US government signed an$80 billion deal with Westinghouse to build more reactors in the USA. So this is ultimately what has triggered some people to call for a nuclear renaissance. And I'm very glad that nuclear is getting all of this positive attention. Statistically, it is one of the safest power sources in the world. is safer even than wind power. It's clean. It is baseload energy and it should be highly deployable, but we still have a long way to go, but we're moving in the right direction.

6:27And what's got you excited about Radiant and Doug, particularly as a founder? Yeah, well, this is kind of a funny story. We were looking at nuclear power and I was pretty convinced, candidly, that we weren't going to invest in a reactor startup. You know, the shibboleth that I've talked about in nuclear loomed very large. Capital intensity, regulatory uncertainty, slow learning rates, a broken fuel supply chain, inferior levelized costs. You know, it just felt too difficult. But then I met Doug and his team at Radiant and everything changed. He completely changed my mind. You know, I'll let Doug himself tell you a bit more about what they're building.

7:03But there were two reasons, basically, that got me really excited and made me change my mind on investing in reactors. The first is that rather than building a big piece of energy infrastructure, Doug and Radiant are building an energy product. And there's a subtle but important difference, which I'll just tell you a little bit about in my mind. Radiant has basically made design choices that are very attractive to a very specific customer base. These customers want rapidly deployable power. And in most cases, their only alternative is a diesel generator. So diesel generators are still expensive, still cumbersome, dirty, and you have to refuel them, you know, every so often.

7:42Radiant X customers in some cases are paying like hundreds of dollars per gallon to get diesel to these remote locations where they need power. So Radiant Kaleidos, a one megawatt portable meltdown proof reactor that needs refueling once every three to five years, is an amazing product and it can be offered at a competitive price to this niche market. So in essence, they were kind of in a way that no one else that I had seen in nuclear was doing, they were thinking about product market fit. And then the second reason is Doug himself and the team that he's built. I'll leave our listeners to form their own opinions, but in my view, Doug is one of the most special founders I've had the privilege of working with in my career.

8:25Welcome, Doug. We're here today to talk about your giant idea of building nuclear micro reactors to give earth clean power and eventually create a multi-planetary species i'm not sure there is a more giant idea than that so give us a bit of color on how you came to work on this giant idea your background your story and what this giant idea is to you yeah thanks for having me on uh so yeah how did i come to be working on nuclear reactors that you can mass produce in a little box. So I was a SpaceX employee for about 12 years, electrical engineer. I joined when they had two rockets that had failed and they had no successful ones.

9:07And I thought it was just a great opportunity, great company, something very cool to work on just to completely remake rocketry, essentially. It wasn't working yet, but I didn't need that part. So I came there to make that happen. I got to work on the first rocket with legs, just four of us. I was reporting directly to Elon, and I went to work on First Grid Fins after that, and then Hyperloop and Boring Company and a Mars colony. And in playing around with this Mars colony design, what we were trying to really do is make a refueling station, more than a colony initially, because it's very expensive to take a starship and land it on Mars.

9:43And if you do that, what you want to do is refill it and bring it back. So you're not wasting a ship and wasting all that fuel every time and needing to bring in a fuel with you to then try to return. So you really want to make it there. To do that you need power. I was looking at solar power to do that. We had a solar power team. We knew what it would take to do it, but it looked like maybe four or five miracles in sequence. You need multiple rockets, about four football fields worth of solar. You need little robots to go deploy those things and connect them. The ground is all dusty and rocky in mixed conditions.

10:14So it's very hard to go, yeah, this is an easy mission that's going to work. And so I presented this information. Elon went, that's too many miracles, probably do nuclear. And I went, let me go look at that. And it turns out that when you look at nuclear, it's, you know, five miracles become one. And you go like, oh, wow, that actually works. That actually is, well, it's maybe, not only does it work, it's the only way you'll actually make life multi-planetary. And so I took that with me, started to learn. I knew almost nothing about nuclear, right? I was an electrical engineer. Yeah, I knew almost nothing about it.

10:47I mean, I knew it existed, but nothing about the tech or could you package that. I knew, you know, reactors are very large. But as I learned about it, I found that, you know, there were space reactors. The U.S. had launched one in the 60s. It's called Snap 10A. Russia had launched actually a huge quantity of them in the past into space. So I started to look around, okay, who can make this reactor so we can put it on a rocket so we can go get this mission done? and as I looked around and learned about the companies and national labs in the U.S. and other entities, I found there were really no good options.

11:18There was nobody making anything, but all the exciting stuff happened before I was born. But then about three years after this event, I became nuclear aware, and I went, well, that's the only way you colonize, and that's the whole point of me joining SpaceX. That was my mission. Three years later in the U.S., this program came out called Pele Program, and it was the US military going, we need an under 100 ton C-17 transportable reactor. And so I got really excited because they wanted something around a megawatt and I knew what that could do on Mars. It's about the right size to refill a rocket every two years, which is the interval you have between Earth and Mars.

12:02And so anyway, I tried to convince Elon and others there to let me just create nuclear at SpaceX. How was that conversation? More like a series of conversations. Like going to people close to Elon and going, how can we do this? And then we go, yeah, it'll never. It was not something we're going to touch. And it makes sense. I mean, SpaceX at the time was working on Starship and on Starlink also, on a bunch of other giant ideas, arguably, right? But I talked to enough people that I met some folks who convinced me it was actually much easier than I imagined. And so I jumped out of that company, founded Radiant, and started to, I went and learned how to become a nuclear engineer.

12:47I basically, you know, went and talked to people who were smart and started modeling reactors almost right away. Did that for about a year. Put together a pitch deck and, you know, pivoted the idea to go, well, really my customer is the military. Like I have this eventuality of, you know, I like space. I think it's very cool for humans to just dream beyond the planet we were born on. You know, we're not stuck here. But the product idea really was, well, I want to make a tiny reactor. I want to pick the smallest the military would accept. And I want to mass produce it because mass production is the story of how things become cheap and reliable.

13:22Okay, we're going to get into that. You make it sound so logical to just leave SpaceX and start a micronuclear reactor company as if it's going for a walk. But was it a hard decision to leave SpaceX? and I mean, it's such an exciting company. It wasn't. It wasn't a hard decision at all. I was immediately convinced that, you know, the mission that I was going to do actually would benefit SpaceX more than me staying at SpaceX. So not only would it help that mission more, but it would help the mission of humanity more. And it is sort of a simple calculus of like, well, how much resources do I have command of to put towards this thing that I know absolutely is the most important thing for SpaceX to eventually have, is that power source to do fueling.

14:02And I thought, you know, maybe it's on the order of a couple million. If I can do, you know, if I can raise more than that, I can make more of an impact. I read a story that when you were doing kind of your early research, you started pretending as if you were farming land on Mars in your back garden, just to get a sense of what that would be like. I did. I did. Because I kind of, I go all in on every idea. So when I did, when I say colony design, I really, I did go above and beyond. I figured out not only power, but I figured out a number of missions you would have that you could take a rocket, put it on its side, cover it with regolith.

14:38There's radiation on Mars, but if you were to go outside on Mars one hour every day and then go back inside of these pressurized structures, like going in your house, right? The rest of the time, that's the same radiation amount you'd have on Earth. And so it's not actually a challenge, not a problem. So I went, okay, here's the radiation solution. I got that one. I went, okay, now how do you actually make food? Like if you want to go from that just fuel producing base to then a real colony. And I had the spreadsheets and I went and identified like the plant that grows the most food, which is this variety of wheat where it doesn't grow very tall.

15:10So it doesn't waste a lot of its energy growing a stalk or growing these leaves. But it just grows lots and lots of edible content. And you can get it from the have USDA has this agricultural research. You know, the mission that I was going to do actually would benefit SpaceX more than me staying at SpaceX. So not only would it help that mission more, but it would help the mission of humanity more. And it is sort of a simple calculus of like, well, how much resources do I have command of to put towards this thing that I know absolutely is the most important thing for SpaceX to eventually have is that power source to do fueling.

15:44But at one point I went like, wait a minute. I don't know how to farm. I don't know what I'm talking about. And so I went, I'm going to go become real. And just in my little tiny, like California front yard, like 15 by 15 feet, I just dug up all the grass and I didn't ask my wife for permission. And I planted all these seeds. I grew a wheat field just to do a cycle of that. Just, yeah. Yeah. So I don't know. I like, I liked it. And I like to try to be real, go to do something right and not be stuck in that spreadsheet. Yes, incredible. I mean, in some ways makes it sound like science fiction, but actually I think the thing that really strikes me about Radiant is that of all the work that's been done on nuclear, so much of it feels like it is a distant future, whether it's on Earth or Mars.

16:31I think the striking thing about Radiant is that it's happening very soon, right? You've got this big test coming up next year, which we're going to ask you about. Could you maybe talk just about the initial go-to-market, very much grounded on Earth, working with the US military and with data center companies? What is the actual idea for the first version of the Radiant reactor? Yeah, great question. So, yeah, we are doing this test, but that's a test unit. And then by 2028, we want to have one unit out to a customer, which would be likely licensed through the NRC. although we do actually have the option of using a Department of Energy authorization.

17:06So there are these two different licensing pathways. With a DOE authorization license, we can put this at a military base in the U.S. We can put it on government land, essentially, through this other pathway, which really only became available through these executive orders that came out. There were four of them all targeting nuclear, different aspects of it. And this is one of the details in there. but uh so idea really for the first one is you know we want a partner who uh knows how to operate these units who doesn't who likes that it's mobile but doesn't need it to be mobile um and where we can we can do testing on it basically go and try to do the the maximum power and go you know switch it between being on and off um and and to do that many many times uh and so the ideal partner for us is a domestic military installation.

17:53Not only because of those reasons, right, those technical reasons, but there is security at that facility. So we don't have to worry about someone coming up and doing something that they shouldn't be doing to the unit. But so the early customers, we have actually signed a contract with the Air Force earlier this year. There is also a DOD program called Janus that was just announced. This is, I think, to the tune of a billion, a billion and a half-sized program. We don't know how much will go to the developers, but there'll be likely three, maybe two developers in that that we'll down-select on later.

18:32So those are pretty exciting. And our goal is really to make maybe 15 reactors to put into those programs. So the initial pull from the Department of the Defense, but I think you've had some great traction with the likes of Equinix. We'd love to hear about that. That's right, yeah. So we have a deal signed with Equinix, which is a great data center company. Actually, they have a data center right across the street and they're in a ton of major cities and that's for 20 units. So we've signed a deal and there's a down payment. It's, I think, one of the more real deals in nuclear. Sometimes they're very handshake.

19:10But we're really excited about working with them and we didn't really understand the interest data centers would have. We initially thought a one-megawatt reactor is fairly small. People are talking about data centers that are very large today, that are 30 megawatts or 100 megawatts even for a single building. But they're actually extremely excited about a one-megawatt size, and it's because our reactor can go in about a 1 ,500-square-foot area for one reactor. And with more, it's even better. So we could put 4 megawatts on 4 ,000 square feet is the kind of easier thing to remember. But their interest is in having power immediately.

19:49And in a lot of cases, there's a lot of new technology. There's AI. All this new technology needs power. It needs power immediately. You can't go to the utility. And when you do that, you go to the utility and say, I need more megawatts or I need even one more megawatt. And they'll say, get in line. And that line, depending where you are, it's two to four years in the U.S. That's what's typical for a major power upgrade. And you're working with a utility, so there's no competition, right? You're just, you have to do what they say. And people are willing to do that, and so you'll find buildings filled with diesel generators, you know, 10 or 20 of them, or however many they need to go and make that power that they need immediately to power their business, to compete against their peers.

20:30And so that's where the interest and excitement is. because our unit, you put on a truck, you drive it to the customer site, you insert it into a docking bay, which is like a, it's just big blocks of concrete that we set up ahead of time. But you put that reactor in, it powers on the same day, you're making full power the next day, and then that power lasts for five years with no refueling. And if you don't have that, you need about 1 ,200 gallons of diesel per day per reactor. At four reactors, that's a full 5 ,000-gallon fuel truck every single day. at your facility okay so in terms of cost you guys have the advantage of competing against diesel as opposed to you know on grid uh other power sources are on grid so that's a major advantage let's just maybe break down the the product into a few different areas so size you've mentioned it's one megawatt compared to you know traditional nuclear plants are about a gigawatt right powering millions of homes this is powering you know one remote data center or one uh military base right So it's small, it's portable.

21:32How have you managed to make the economics of that work, where fuel costs for such a small reactor would be normally a massive proportion of the cost? You do something very clever on fuel, which, as you said, allows it to be refueled every five years. So fuel costs are under control. And it's also safe. We'd love to hear about the safety. We'd love to hear about the smart things you've done on fuel to keep the cost down. Yeah, great questions. The fuel is the largest proportion of cost at this really small scale. and so there are some deals that we've signed we are actually working with with the UK with a Yarenco have signed a deal for enough fuel for five units I think that is the first commercial halo deal ever signed also so it's pretty exciting and so securing the where the supply is gonna be coming from before everybody's out there trying to get fuel is important to help control costs but the so the The smartest thing technically going on is really not about the fuel, but about the nuclear core.

22:30We use a material called zirconium hydride, and we actually make that here in our facility. It took us about three years to develop that. And what it does really, we have enriched fuel, right? We use this partially enriched. So it's not a high enriched, it's not a weapons grade material, it's what you'd use in like a research reactor. So 20 % enrichment, which is more than what you'd use in a normal reactor, which is about a 5 % enrichment in one of those big gigawatt scale reactors. But we actually use this material as zirconium hydride, and if you put that in a reactor, you can actually use less fuel then.

23:06And the reason you can do that is because when the fission occurs, you've got neutrons moving at about 7 % the speed of light. They're extremely fast. And when they're moving that fast, they don't interact. They don't cause fission. And so you want to slow them down. this zirconium hydride slows them down and it does so over a really short distance inside the reactor so you can control the size of the reactor and you can actually load less fuel and still get a relatively high burn up and why is it so safe two two reasons well there's really three two main reasons so uh we use a fuel called triso and this this fuel is from the 1960s um like like pretty much all nuclear tech uh and you know there's not really much that humans have not explored and gone and built.

23:48So it's a fuel that's been around for a while. It is tiny poppy seed sized kernels of nuclear fuel that is overcoated with ceramic layers. Those ceramic layers prevent the fission gases from getting out. Not only that, but those ceramics have an extremely high temperature rating, right? They can go to almost 2000 degrees Celsius. So what that means is you can't have a meltdown. It's a fuel form that just will not melt down. So that traditional, typical problem thing that occurred in a couple of famous reactor incidents. So that can happen. The other thing is helium gas is what we're using as the coolant, and helium uniquely does not become radioactive in the reactor.

24:30And if it leaks, it's gaseous, it will just dissipate into the atmosphere, and it's not radioactive. So this is a really great thing to have, especially because we know that our customers may need power off a grid. They may be out in the middle of nowhere, right? And so you want to have a reactor that is totally safe, even if it's, you know, not in a, it's not babied, right? If it's not in a really, it could be out in the outdoors in a rugged environment. And the third thing is, is a passive safety. And part of that is just due to the size of the reactor. If you make a reactor extremely small, it does, when you turn a reactor off, it keeps producing a little bit of heat.

25:05And then that slowly decays away. And you need to cool it during that period normally. So for a big reactor, they run these cooling water pumps and they have to have redundant cooling water pumps and they have to have a backup generator to run those things. With a tiny reactor, you don't need that. And it's because you just don't have much fuel. And what we do is the air, just outdoor ambient air at whatever temperature, even if you're out in the middle of the Sahara, that will flow up along the wall of the pressure vessel on the outside of it. And that carries away enough heat to prevent any overheat that incident.

25:36So is this foolproof, Doug, or somewhat theoretical? Is there absolutely no chance of something like Chernobyl happening with the Radiant product? In that reactor incident, they removed 224 control rods to the maximum extension. That is a thing which should not have even been possible by design. So there's a bad design element here. And they talk about this in that really great show, HBO show Chernobyl, which I found awesome, fascinating. And I'm a pretty harsh critic. But the thing they don't tell you, they tell you they removed all the rods to the full extension. They don't tell you it's 224 of them.