The Exchanges

Q you know, hopefully over the next few years, some big milestones are achieved. These metrics that we're using, uh, are going to start to get a little bit confusing. They've already started to confuse me, so we'll come back to that. But okay, so that's, that's Tokamak. Um, let's talk about some other Approaches. You mentioned, I think, inertial confinement before. Can you describe that one and how it's different?

A Yes. So inertial confinement is, is probably the, the furthest, uh, you know, in the span of different fusion concepts. In inertial confinement, what you're trying to do is compress a very small amount of fusion fuel to, to a very small size at very high density. And you're not really trying to hold it there at all. You're just compressing it to high density, and then it's gonna disassemble on its own, right? And the goal is that just in the time that it takes to disassemble, you're able to get enough fusion burn to exceed the energy that you took to assemble that in the first place. And, ah, the most mature way to do that is using lasers to compress that fuel. So laser-driven inertial confinement fusion.

AI assessment note: “In inertial confinement, what you're trying to do is compress a very small amount”

Q everybody is trying to build toward. Nobody's achieved it yet, unless I'm wrong. It's, it's yet to be seen in the world, but Is widely appreciated to represent a watershed moment for fusion. I want to talk about what it actually means, the various shades of it, and then we'll talk a little bit about what has to come after that. But let's start with what is scientific breakeven, actually?

A Yeah, so scientific breakeven, you can define it as the energy produced by the fusion, uh, Uh, divided by the energy you delivered to the fusion fuel. So, uh, maybe it's easier just to talk about an example, like a tokamak. Okay, so you have your donut-shaped plasma fuel, and you are applying some kind of heating, whether it's electromagnetic waves or a high-energy beam of, of neutral particles, neutral beam injection, they call it. And so, it's the, it's the power of those heating systems. For example, you can be creating 10 megawatts of fusion power at any instant of time, But yet you're applying 15 megawatts of heating power. So that ratio is what we generally call scientific, uh, the scientific gain, right? And if, when they become equal, that's a scientific breakeven. In inertial fusion, it would be the laser energy that you're shining on your capsule.

AI assessment note: “scientific breakeven, you can define it as the energy produced by the fusion”

Q So presumably getting to Q equals whatever we need is not in and of itself going to be sufficient to build economic fusion reactors that work. What are some of the other technical challenges that we'll face in beyond the energy gain in actually building these systems?

A Yeah, so there's three main areas, I would say. One is, of course, extending the, the duration of the operation, right? You, you get, even if you get high Q for a split second, that's not very useful. So you have to extend that out to, uh, operating at high average power, right? For, for months at a time or longer. So that's one challenge. Um, and then two other challenges are, The, what we call the first wall, the first wall sees this tremendous energy and particle flux coming from the fusion core, and so you need kind of an integrated way to deal with that heat exhaust, uh, and also the materials properties that will, um, allow for the, uh, handling that heat exhaust. And then the final thing is the fuel cycle. So especially with deuterium-tritium fusion, You have to breed the tritium, and you have to extract that tritium and put it back in the system. Um, there's something called a tritium breeding blanket. And so the blanket technology and all of the tritium processing has to be handled at a scale that we haven't done yet. So that's the, that's the final challenge.

AI assessment note: “there's three main areas, I would say. One is, of course, extending the, the duration”

Q Okay, so in the 19 sixties, so 60 years ago, we reached ten million degrees, heating up the plasma to ten million degrees in a tokamak reactor, which we'll come back to what that means. Uh, what's happened since then?

A Yeah, so once, uh, the ten million degrees is a, is a very important symbolic achievement, because it means you, you've kind of passed that first major milestone of getting a stable plasma that can hold the heat in, and once you do that, then it's, it starts to become, you, you need to ramp it up, right? You need to heat it more, continue to improve the heat confinement, and so there was very rapid progress after that in terms of the, the density and the temperature and the confinement time. So from the Seventies until the nineties, 20 years, uh, that triple product metric increased by five orders of magnitude, and people like to compare that with Moore's law as an example, right? That the rate of increase was similar to that of Moore's law, and that brought us to the cusp of what we call scientific break even in the mid nineties of the tokamak. So meaning that the fusion energy produced Was about, ah, almost equal to the energy that you delivered to the fuel.

AI assessment note: “that triple product metric increased by five orders of magnitude”

Q feedstock, um, but the capital cost of building these systems has turned out to be a lot higher than we expected, at least historically, and so the economics of operating these things Has been challenged, not to mention their profile, which is baseload and not very flexible. So what do we know about kind of the economic drivers of a nuclear fusion power plant at the end of the day?

A Yeah, great question. And in fact, ARPA-E has, um, funded, uh, some costing efforts in, in recent years, you know, based on earlier methodologies. And there's, there's some commonalities with, with, Uh, nuclear advanced reactors. It is dominated by capital cost. In fact, the particular ARPA-E funded study showed that maybe 65% of the LCOE might, might be coming from capital cost. But it's the balance of plan. It's the, it's the buildings. It's the infrastructure, right? So, so even though you can optimize your fusion core and get to high gain, honestly, at the end of the day, those things are a smaller driver of the ultimate total cost. And so some of the lessons being learned in, uh, from advanced reactor development, um, I think will apply to fusion. So reducing construction time, um, maybe some modular construction, um, reducing operation and maintenance costs and things like that. Those in the end will have bigger drivers, uh, on the economics for fusion as well.

AI assessment note: “ARPA-E funded study showed that maybe 65% of the LCOE might be coming from capital cost”

Q hope for a long time. People have been working on nuclear fusion for, for a very long time. And it's been one of these things that I think, you know, we've heard about for decades. So let's do a brief history lesson. Can you kind of walk us through the history of nuclear fusion research and when there've been major milestones hit that leads us to where we are today?

A Absolutely. So Really serious efforts to do controlled fusion started perhaps very early in the 19 fifties, and in the first almost 20 years worth, people were trying very hard simply to make a stable plasma, and I should introduce what the word plasma means. Plasma is a collection of charged particles, and basically when you heat up the hydrogen fuel, it necessarily becomes a plasma. When you, when you heat it to high enough temperatures, millions of degrees or more, um, the, the fuel becomes a plasma. The, the electrons are stripped off of the atoms. Um, and so again, in the first 1520 years, people were struggling and working toward making the plasma stable and being able to heat it up to the millions of degrees and higher. Um, it wasn't until about the late 19 sixties That people succeeded, uh, to about ten million degrees of temperature, and that was in the now well-known tokamak concept. So that was kind of the first really major inflection point.

AI assessment note: “Really serious efforts to do controlled fusion started perhaps very early in the 19 fifties”

Q that before, but we didn't actually talk about it. What was that finding? It made a little bit of news, but actually less news than I would have anticipated given how big a deal it was. Um, but it came right in the middle of like a bunch of other fusion announcements, largely from private companies raising a lot of capital. So what, what happened at the National Ignition Facility?

A Yeah, what happened was it got 1.3 megajoules of, of yield compared to 1.9 megajoules of laser energy. And I think, uh, maybe to everyone except for the, the team doing it, it kind of came out of nowhere. Um, It was a dramatic improvement, right? I mean, usually these things happen a little incrementally, but this was, you know, if you look at a chart of, of the performance over the past year, um, this record shot in August just was like almost 10 times bigger than, than anything beforehand. So it was a really dramatic increase in performance, and it brought it to the cusp of, of what people call ignition, right? In inertial confinement fusion, You're trying to heat up a tiny little bit of fuel, uh, and, and that tiny little bit of fuel generates enough fusion to burn a surrounding layer of more fuel. Um, and that's, if you're able to burn up that surrounding layer of greater fuel, that's called ignition in, in inertial confinement fusion. So this NIF shot basically got to the cusp of ignition. I mean, it started to burn that surrounding fuel, which, which is a big physics, uh, you know, uh, accomplishment.

AI assessment note: “what happened was it got 1.3 megajoules of, of yield compared to 1.9”

Q electricity market globally, because you have a fairly long way to go, I think, between just hitting that point, and then, one, getting to ratios that actually make it attractive, but two, actually, like, engineering and building fusion, right? Power plants. So what does it look like to go from Q equals one to Q equals 10 or whatever it's going to need to be to build economic fusion reactors?

A That is a great question, and you're indeed right. There's more work to be done. You have to get to higher gains, um, plasma gains, right? So you, you might need to get to plasma gains of 10 or 20 for magnetic confinement devices. Again, that's because magnetic devices have somewhat good efficiencies, uh, heating efficiencies. For inertial fusion, you might need to get to plasma gains of a hundred because of the poorer efficiency of lasers. Um, and frankly, it's, it's not entirely known what that's going to look like going from gain of one to gain of 10 or a hundred, right? People, I think, have an idea from, from the theory and from modeling, um, But ultimately, we have to do the experiments to, to prove it out.

AI assessment note: “frankly, it's, it's not entirely known what that's going to look like”

Q term. How do you hold both of those things in your mind at the same time? Is it just that, like, the technical progress has been so accelerated in the past five to 10 years that we can't look at this history around fusion timelines and costs and so on as we're considering what happens in the future, or is it just that we're getting overly excited at the moment?

A It could be, it could be some of all the above, ok? I'll explain a little further. I like to look at this a couple different ways. One is there's, there's a spectrum of risk involved, right? Eater is, um, first of all, recognize that Eater was designed with the knowledge and the technology know-how of about 20 years ago. Um, and, and that's what we knew how to build and what we needed to do. And of course, like I said earlier, the scientific advances have been happening over the last 20 years, and other technologies, uh, have Exists now. That didn't exist 20 years ago. So that's one reason of, uh, of the difference, right? Um, if, if you were to redesign, um, a device with the same goal of Eater, you, you would naturally come up with something very different. So I think that contributes already quite a bit to the difference in these, in these things. The other thing is, if you look at the private companies talking about very fast timescales, They're not designed to be a prolonged research endeavor, right? Eater is meant to provide a platform for us to really study high gain burning plasmas, and we can get into what that means. Whereas the private companies are looking to check a milestone and move on very rapidly, right? And not, not do a lot of scientific discovery.

AI assessment note: “It could be, it could be some of all the above, ok?”

Q we're in this horse race to get to energy break even. Tokamak has had a longer road and is much more mature, but, uh, inertial confinement sort of coming up fast, and they've, you know, at least in what's known publicly, they've, they're at sort of a similar place. Um, what's the, what's the sort of trade-off with inertial confinement? Relative to tokamak, what's more attractive about it? What's riskier?

A Yeah, that's a great question. Um, it really does present different challenges. So on the physics side, which is in the fuel itself, the, the way that you achieve, um, a burning plasma and self-heating, uh, is quite different. In inertial confinement fusion, uh, You're getting to this very high density, and you don't really have to hold it there, right? Like I mentioned earlier, you just want to get hold it there. You just rely on the inertia of the fuel itself, uh, to burn enough before it disassembles. Uh, and that, that can be seen as an advantage. As long as you can get to that point, you don't need to worry about sustaining it. Uh, so that's one advantage of inertial confinement fusion. The disadvantage, though, is perhaps a less efficient, uh, So even though they, they're both at the point of 70% of the input energy, the efficiency of tokamak heating is much higher than the efficiency of the laser. So you have a longer ways to go to get true wall plug gain, which is ultimately what you need. And there's other, there's other differences too, of course, and yeah.

AI assessment note: “that's one advantage of inertial confinement fusion. The disadvantage, though, is perhaps a less efficient”

Q All right. Any other pathways that we should run through?

A Yeah, and then, of course, each of the three, right, magnetic, confinement, magneto-inertial, and inertial, they all have a zoology of variance, right? Each of those has, has a bunch of variance, and, um, with magnetic fusion, it's really a trade-off between the strength and the, and the geometric complexity of the magnetic field, right? So, you know, the less strength and the less complexity, of course, that is, those are engineering advantages, but then it's harder to, to hold the fuel in place. So that's the trade-off you're making. Uh, on the, on the inertial side, uh, you're making some trade-offs as well. Again, it's the, the degree of complexity in your driver, the laser, or the target. Um, and again, the control of, of the fuel, right? It's the same thing a magneto-inertial. So in general, you're trading off some degree of complexity in, in how you assemble and hold the plasma, uh, versus the physics performance of the plasma.

AI assessment note: “each of the three, right, magnetic, confinement, magneto-inertial, and inertial, they all have a zoology”

Q Right. We'll come back to the energy breakeven point, because that's, I think, the sort of nearest term, most substantial milestone that a lot of folks are looking toward. But again, notably, we're talking about the mid-nineties now, right? So we were nearing, uh, energy breakeven 30 years ago, 25 years ago. So what has happened since the mid-nineties?

A Yeah, so what's happened since the mid-nineties, well, a couple things. One is, The science and the understanding and the capabilities, the modeling, the diagnostic capabilities have continued to, to improve substantially. Now that's less visible to the public, of course, right? You're, you're not making a big, big major milestone, uh, achievement, but that's what's been happening in the background. There's also been progress in other fusion concepts, like inertial fusion, which I'm sure we'll get to as well. But what, so most of the time from the nineties up to say five, 10 years ago, there was a lot of scientific advances, probably a little bit out of the public's eye, right? But in the last five to 10 years, um, the situation has started to shift again. The landscape is changing.

AI assessment note: “so what's happened since the mid-nineties, well, a couple things. One is, The science”