Transcription
And that would be commercial power plant.
>> No, it would not.
Today we're going to be looking at the genius solution to the nuclear fusion problem. Question is, which problem? This video is by Two Bit Da Vinci. For those of you who don't know me, I'm Tyler Fols. I'm a nuclear engineer with a little over 10 years of experience in the commercial nuclear power industry, from engineering operations to emergency response. I don't think everything there is nuclear, but I can certainly share some moments. Let's see.
The world's very first failed attempt at nuclear fusion might have just had an overhaul over [music] a century later. A revival that could beat all other fusion technologies to the grid with the most compact reactor in the world, without the billion-dollar price [music] tag.
>> Okay, so talking a little over a century later. So a lot of early fusion research was explored in the 1950s. So we're saying this thing's not going to be remotely feasible until 2050 at the absolute earliest. At least they're being reasonable with the timeline.
>> Without using magnets. [music] No need for superconductivity, no heating, no crazy lasers, just a good old zap of electricity. Did we just figure out how to control Zeus's?
>> So, we're talking Z-pinch. >> Lightning in a reactor core on a shoestring budget. Let's figure this out together.
>> So, if that's the case, if we're talking Z-pinch, yes, it's compact, potentially lower capital intensity, but fusion cost isn't just the reactor core. It's power electronics, heat extraction, tritium systems, shielding, maintenance, regulatory overhead, and balance of plant, just like you would see in any other sort of power plant, especially nuclear power plants, 'cause fusion is categorized with the regulators as a nuclear plant. Despite it being fusion versus fission, it's still the Nuclear Regulatory Commission at work. So that is to say, small plasma does not always mean cheap power plant in the context of fusion. But when it comes to "without using magnets," Z-pinch still uses magnetic confinement. Think what he means to say is without external magnets. It's from self-generated magnetic fields. No heating is not the case. Ohmic heating is still heating. Resistive heating of plasma is still heating. This would be a bit like saying a nuclear power plant doesn't heat fuel, it just fissions. Well, they kind of go hand in hand.
>> I'm Ricky, and this is Two Bit Da Vinci. Right now, a company called Zap Energy is operating a machine that defies the conventional wisdom of nuclear physics.
>> Ooh, no. [laughter] I mean, I think I get what he's saying. He's talking about challenging engineering pessimism around instability, but nothing here is defying nuclear physics. It is nuclear physics. Really cool nuclear physics.
While the rest of the world is building fusion reactors the size of stadiums, like the massive $22 billion ITER project in France, Zap Energy is building one you could fit in your garage.
>> So ITER is large because it's designed for steady-state confinement, very long pulse duration, and diagnostic access. Those are its goals. A tokamak-style reactor optimizes confinement time. That is the key parameter in the Lawson criterion. That is to say, you need heat, pressure, and confinement time to achieve self-sustaining fusion. They're going for confinement time. Z-pinch, you're looking at density, so pressure. Plus, it's a much simpler design. So, it has a different goal as to what they're trying to go for. Now, as far as one you could fit in your garage, if he's just talking about the plasma column, yeah, maybe. But if you have to take into consideration the power electronics, shielding, tritium systems, safety systems, heat exchangers, yeah, that's uh you're just going to need a full-on facility for that, even with a small reactor portion.
They're using a technology called the Z-pinch. It's a method of nuclear fusion that compresses hot plasma using a self-generated magnetic force. This technique lets you strip away all the complexity. You get rid of this.
>> That's a pretty good succinct description there. But as far as stripping away the complexity, you're removing some complexity and you are getting different complexity. That's engineering tradeoffs in a nutshell. This is essentially what happens all the time with competing technology. So you're losing those superconducting external coils, but you're getting ultra-high current pulse power systems that don't last as long. So you're giving up what ITER has going for it in terms of its confinement time. This is going to add fast switching requirements and erosion-resistant liquid metal handling. So yeah, like anything else in engineering, they're trade-offs.
Superconducting coils that need to be cooled to near [music] absolute zero. You get rid of the natural beam injectors that TAE uses for its field-reversed configuration machine. They rely on a simple, brutal pulse of electricity. They run a current 10 times stronger than a bolt of lightning through a column of hydrogen gas.
>> Okay. All right. Um, that's silly. Lightning current is anywhere from 10 to 200,000 amps. What really matters is current density, rise time, and repetition. This is a bit like saying a pulse laser tattoo removal that you can get on eBay or Alibaba. That handheld device has power density in the megawatt range, meaning it has more power than your car, which is technically true in the definition of power, but that's not because it's some crazy high-performance vehicle you have in the palm of your hand. It's because you're dividing by a really small number because those are pulse lasers. Completely different animal from sustain lasers. Just like lightning current isn't really a good benchmark for your reactants.
This current generates a magnetic field that crushes the gas into a tiny [music] super-hot filament where fusion happens. But
>> Yep, that is exactly how the Lorentz force works. Magnetic field scales quadratically with current. So hence, I could tell he that's what he was going for when he brought up the high current, but again, what matters is sustain.
Here's the catch, and it's a big one. According to physics history, this machine shouldn't work.
>> He's being really absolute when talking about physics and things not working. So the theory never said Z-pinch couldn't work. It was all about instabilities associated with the magnetic field and plasma dynamics. That is to say, it's an engineering stability problem. Nobody's trying to do anything that violates the laws of physics.
For 70 years, scientists have known that trying to squeeze plasma like this is [music] trying to compress water with your hands. The plasma should tear itself apart in a millionth of a second. And for
>> Okay. Yeah. So it's a stability and microsecond is the right time scale when you're talking about working with plasma.
For decades, that's exactly what it did. But Zap Energy's found a way to control it. The solution is called shear flow stabilization, and it's the single reason we are talking about this company today. Zap's founders spent 20 years at the University of Washington perfecting this. The idea is to stop treating the plasma as one solid block. Instead, they manipulate the flow so that the plasma moves in layers, kind of like a highway. Imagine [music] being in the center lane of a freeway at 50 mph. The lane to your left is moving at 80 mph, and the lane to your right is moving at 100 mph. Can you change lanes? No. The difference is speed. The shear locks you in place. You're effectively trapped in that center lane. Zap applies this logic to the plasma.
>> That was actually a very good analogy. This shear disrupts coherent modes the same way velocity gradients disrupt vortices. This is also like flow-assisted stability in reactor coolant channels. And the idea of velocity shear suppressing instabilities has been pretty well known in fluid mechanics and plasma mechanics. What's really impressive though is achieving this consistently in an energy-dense regime relevant to fusion. That's where the real innovation comes in and why this
You create a flow profile where the outer layers of the plasma are moving faster than the inner layers. When the central plasma tries to bulge or kink, which is a specific instability that killed the Z-pinch in the 1950s, the fast-moving outer layer shears it back into alignment. This smooths out the instability purely through fluid dynamics. No external magnets required.
>> Okay, I'm glad he brought that up 'cause he talked about no magnets. Yes, it's no external magnet. You're this is still magnetic confinement fusion, just internal transient magnetic field.
>> Allows them to keep the plasma stable thousands of times longer than previous attempts, long enough to fuse by
>> Okay, so that means stability time doesn't need to be long, just long enough to satisfy the Lawson criterion for that particular density and temperature regime. Okay.
The physics with software and fluid dynamics rather [music] than brute-force hardware, they unlocked a massive advantage.
>> Okay, saying software did it is like saying reactors don't melt down like Chernobyl did because we have better computers. No, control systems solved it. Plasma flow control did. I mean, the software helped develop good simulations, good modeling software to help you design things, but you're still ultimately limited by what you can physically do.
Speed because their reactor doesn't need those massive superconducting magnets [music] or building-size lasers. They can design, build, and commission a new machine in just one year. And they can
>> That sounds extremely impressive. Now again, I think they're just talking about the reactor, if you will, not the whole facility, but still, that that's still very impressive.
>> Do it for a cost of just a million dollars per prototype. [music] This allows them to iterate faster than anyone else in the industry. While their projects take decades to build a single test reactor, Zap is churning out prototypes. They have already managed to generate thermonuclear fusion in deuterium-tritium fuel, a world first for this technology. And they're incred.
>> So, clarification here, they have achieved measurable fusion reactions with deuterium-tritium. They have not done net energy gain. Still an amazing milestone, but just want to be careful here 'cause a lot of these milestones, especially with stuff like fusion, are tricky.
Incredibly close to reaching net energy gain. Before we crack open the
>> Okay, so what do you mean by net energy gain? The letter here thrown around within the fusion community is Q. There's Q plasma being around one. That's your scientific milestone as far as net energy in equals net energy out for Q plasma equal one. For a power plant, you need Q electric. So that's not just plasma energy in, plasma energy out. Your plasma energy, basically your heat input, must greatly exceed the electric cost of all of your support systems, including your balance of plant systems by a significant margin. If nuclear fission or really any other power plant operated with this Q factor for a power plant, the number is going to look something like 50 or 100 for it to be a power plant. And that's not the final step. The final step is looking at capacity factor. How long can this facility stay online providing electricity? That is to say, its availability. And then you can look at levelized cost of energy. And this is ultimately what you look at when you get into adopting this sort of reactor technology. And each one of those three steps is significantly harder than the previous one. So I'm thinking what he means by close is if you're at the stage where you're getting measurable fusion reactions, your next milestone is going to be Q plasma.
>> Reactor to see the genius inside. We need to understand why this approach is so promising compared to everything else. In fusion engineering, we live and die by the triple product. It's a simple equation: density times temperature times confinement time. To get fusion reactions that produce energy, you need to get the product of these three numbers high enough. Every fusion approach compromises one of these variables. One, magnetic confinement tokamaks. These machines, like ITER, are like slow cookers. They reach moderate temperatures for a very long time, seconds or minutes. But the plasma density is very low. Two, inertial confinement using lasers. These machines, like the NIF, are like explosions. They reach incredibly high temperatures but for a tiny fraction of a second. Zap Energy's Z-pinch sits in the Goldilocks zone. They reach moderate temperatures and densities for moderate amounts of time.
>> Yeah, that was a very solid explanation. And hence the appeal of Z-pinch is less of the extreme requirements. But Goldilocks zone does not mean it's easy, just means it's more balanced.
And the data shows that they are remarkably close to the finish line. This is truly a golden.
>> No, I'm sorry, but as someone who sees the finish line as this thing is powering my house, powering cities, driving human progress, it's still got a ways to go. Maybe more like 2050, though. I mean, he initially said a century past the initial theory behind fusion and all the early experiments which took place in the 1950s. So I figured he was talking 2050.
Age of sustainable tech. Their current machine, Fuse 3, is operating at 500,000 amps of current. At this level, they are already fusing atoms. But the magic number is 650,000 amps. At that specific current, their models predict they will reach Q=1, or scientific break-even. That is the
>> Okay, so 500,000 amps, that's some current. At those levels, electrode erosion, magnetic pressure, mechanical fatigue, switching losses, all become dominant engineering challenges. Just a lot of And so Q equals 1. So we're clearly talking Q plasma at 650,000 amps. Getting a model to predict Q=1 at a certain threshold is encouraging, but fusion history is paved with models that died on contact with hardware. So again, back to the whole thing about software. Great for developing, but you still have to deal with the physical limit.
Holy Grail, where the energy coming out of the fusion reaction equals the energy put into the plasma. There
>> If just Q plasma is the Holy Grail, what does that make utility production? The Holy of Holies.
>> Just a 30% power boost away from a milestone that eluded scientists for decades. And unlike the National Ignition Facility, which claimed net energy but ignored the massive inefficiency of their lasers, Zap's efficiency calculations are direct electricity in versus electricity out. If they push that current to roughly
>> So that's a fair criticism of NIF. Laser inefficiency matters a lot when you're talking about power production.
>> 2 milliamps, their model predicts a Q of around 30. That means for every one unit of energy they put in, they get 30 units back out. And that would be a
>> Q plasma of 30.
>> Commercial power plant.
>> No, it would not. It would be a necessary condition, meaning you better be getting at least Q30 for plasma, but still not a sufficient condition by yourself. You still need that high duty cycle. You still need high availability, affordable maintenance, a self-sufficient supply of tritium, and the non-trivial task of getting your license accepted by the NRC. And that's not even taking into consideration competitiveness because 30 is situationally competitive at best. The real key parameter is capacity factor, which is why you don't hear anyone talk about Q in the context of nuclear fission or the context of coal or natural gas because it doesn't really mean a whole lot. It's like you've essentially advanced beyond the stage where you're concerned about Q and you're talking about how long can I keep this thing online and producing competitively relative to uh everything else.
So, how do they do it? How do they control a miniature lightning bolt without it blowing apart? Let's put on our lab coats for a minute and break down the physics because this is where the engineering insight really shines. The reactor itself is a vertical column about 3 meters tall. It looks a bit like a vortex cannon. You inject the fuel mixture, deuterium and tritium gas, at the top. Then you hit it with a massive pulse of electricity. This effectively turns the gas into a plasma and accelerates it down the column. The central rod in the machine ends in a nose cone. So the plasma shoots out the end into the empty part of the vacuum chamber, forming a long, thin column. This is where the Z-pinch happens. It unfolds in a specific sequence of phases. Phase one, an electric field is applied along the gas column. Phase two, this causes electrons and positive ions to flow in opposite directions. Phase three, this flow of current generates concentric circular magnetic lines that wrap around the plasma column. Phase four, these magnetic field lines interact with the moving charged particles. This creates a Lorentz force. Using the right-hand rule, we can see that this force points directly inward toward the center of the cylinder. Phase five.
>> I love when people bring up right-hand rules. I mean, it's just classic stuff that I remember learning about in high school that are still relevant for this kind of stuff. It's just shows you how powerful that is.
Every particle in that column feels a crushing force from all sides. Phase six, the plasma collapses. This magnetic vice grip compresses the plasma into a filament that is only 1 to 3 millimeters thick. All in all, a pretty good high-level explanation, but this filament length is impressive and a little scary from a control standpoint. So, after all, small plasmas are way easier to compress, but a lot harder to stabilize and diagnose. It is hot, it is dense, and this is where the fusion happens. But, as I mentioned earlier, this is where physics usually fights back. In a normal Z-pinch, that tiny filament creates sausage and kink instabilities. [music] Tries to bulge out or twist.
>> I love the term sausage instability. Funny, there are certain terms that are hilarious that you only hear within nuclear engineering. [laughter]
This is where Zap Energy's secret sauce comes in: shear flow stabilization. But the engineering genius doesn't stop at the plasma. It extends to the reactor wall. If you are generating a mini-sun inside a metal can, how [music] do you stop the walls from melting? And how do you harvest that energy? Zap uses a liquid metal wall. The reactor chamber is lined with a
>> Yeah, it's simple. You skirt the issue by having something that's already molten. It's a bit like saying in a molten salt reactor, you can't have a reactor meltdown. [laughter] As I said in Super Troopers, you can't be more pulled over, right?
Flowing river of molten lead and lithium. This is the brilliant engineer's choice for four [music] distinct reasons. One, it's the electrode. The liquid metal acts [music] as the outer electrode to complete the electrical circuit. Two, it's the shield. It absorbs high-energy neutrons and radiation so they don't leak out and irradiate the facility. Three, it's the coolant. It acts like a heat sink. It carries the intense heat away from the core to a heat exchanger where it can generate steam and electricity. Four, it's the fuel factory. This is the coolest part. Deuterium is abundant in the ocean, but tritium is radioactive and scarce. When the fusion neutrons hit the lithium in the liquid wall, the interaction breeds tritium. Zap can extract that tritium [music] and pump it right back into the machine as fuel. And
>> That is probably my favorite part about this design when you have neutron shielding, tritium breeding, self-healing, and heat extraction all from one thingy. That reminds me of the old Generation 4 fission reactor design. I I say old 'cause people don't really talk about them as much anymore, but I remember that was the nuclear renaissance that happened in the late 2000s when I was in college. [laughter]
Because it's liquid, it's self-healing. A solid wall would get brittle and crack under neutron bombardment, but the liquid simply flows. They even optimized the chemistry by using a eutectic mix, roughly 16 lithium atoms for every lead atom. They lowered the melting point of the lead from 327.5° down to 235°.
>> That's going to be good, especially if they could scale this up. Lower parasitic losses like that are going to help.
This means they waste less energy just keeping the metal molten. To really appreciate how revolutionary this is, we have to look back at why this revival is so historic. The Z-pinch isn't new. It's actually ancient. The phenomenon was first observed way back in [music] 1905 when a lightning bolt struck a hollow lightning rod in Australia [music] and crushed it flat. In the 1950s, the Z-pinch was actually the very first approach scientists tried to control fusion. It was Project Sherwood's
>> It was one of them. There were a lot of weird things that were thought of back then, but sure. Preferred model. It was the original dream, simple, compact fusion. But that kinking issue was just catastrophic. The plasma would pull off and die in nanoseconds. [music] After 7 years of failure, the scientific community gave up. They pivoted to using massive external magnets to force the plasma to behave. The
>> And that was actually probably the best decision they made with the tools they had available to them at the time.
Thus, the tokamak was born. Costs skyrocketed, and the Z-pinch was forgotten. Zap isn't inventing new physics. They are fixing the oldest one. They're redeeming a 70-year-old failure.
>> And that is perhaps its greatest strength. If you're trying to do an engineering project by inventing new physics, you should probably step back and question if this is a dumb idea.
With modern fluid dynamics. Now, [music] I'm an engineer, not a hype man. And we have to look at the numbers and the hurdles and be real about this. Tokamaks, like ITER, [music] have budgets that have ballooned to over $22 billion. The National [music] Ignition Facility is roughly around $3.5 billion, and Zap Energy, well, they raised about $327 million in total. So how is it so cheap? It's the magnets in a tokamak. The magnets alone cost around $400 million. Zap's magnets are the plasma itself. This means their capital cost for a power plant would be two orders of magnitude lower than a tokamak for the same output. We're talking about a reactor core that costs just a few million. A typical 1-gigawatt nuclear fission core costs over a billion. Zap could slash the capital costs by over 90%. But what's the catch, right? What's stopping them from doing it right now? The barrier isn't the reactor core anymore. It's the switching. To make this a viable power plant, you can't just fire it once. They [music] need to fire it repetitively.
>> I was going to say, I would I would argue the ability to switch is associated with the core. But [laughter] yeah, when you have current that What number did he throw out there for 30? Was it 2 million amps? Yeah.
>> To match a small nuclear power plant, they need to fire roughly 10 times every second at 10 hertz. This means switching 650,000-plus amps on and off 10 times a second continuously for years. The power electronics required to handle that kind of load without melting are incredibly complex. Currently,
>> Yeah. All right. So, that's not good for switches or bus bars. You're going to have a lot of arcing that's going to degrade stuff really fast. And the failure mode's going to be a lot of heat and a lot of operational fatigue. Tokamak accelerators don't have this issue. They're either continuous or long pulse. As far as inertial confinement, well, there your pulse rate is limited by recharging your laser. And for a power plant, it's going to need to be continuous. Like you said, you can't get away with occasional fusion and call yourself a fusion power plant. And pulse systems, by definition, are a low duty cycle. So, it's clearly going after fusion with a different goal or approaching a the problem from a different angle than tokamak, pitcher, long pulse. The other extreme, the opposite of this would be the stellarator. They are inherently steady state, or at least trying to be.
>> The Fuse machine operates at a relatively mild 10 to 20 keV. That they need, they'd have to ramp it up to around 30 keV. These high voltages put a massive strain on the switching devices. However, compared to the challenges facing other fusion companies like TAE, which needs to increase neutral beam power by several orders of magnitude, Zap's hurdle is just a 50% increase in voltage. It's an engineering challenge, not a physics miracle. And they're already
>> I like that. That's a good thing to say right before your shareholders meeting. It's just an engineering challenge. We're not asking for a miracle. I like that. But this is why it's still relevant.
>> Working on it. Zap isn't just running science experiments. They are building a product. They're currently constructing Project Century. This is the world's first 100-kilowatt-scale repetitive Z-pinch device.
>> It's a fully integrated system designed to test the durability of the liquid walls, the electrode, and the switches. They've already
>> Okay, great. That is a good next step because you're going to want to see how all of your vaunted super control system and confinement system work when you scale up the wattage. That's That's great.
>> Proven they can fire the power driver once every 10 seconds for three hours straight without a single fault. That is reliability data that most fusion startups can only dream of. If they succeed, we are looking at a future where fusion energy isn't generated in massive, centralized plants that take a decade to build. We're looking at modular, factory-built reactors that can be deployed anywhere. It's a vision of a democratized power grid [music] powered by the most energy-dense fuel in the universe.
>> Okay. All right. Let's uh let's pump the brakes a little bit. Even small fusion plants are still going to be industrial infrastructure. You're not putting one of these in your backyard. I already know what the NRC is going to say when you ask them.
And Zap Energy is betting that the solution to our energy future isn't bigger. It's smarter. But what do you think? I know we've made videos about fusion. It's smarter. But what do you think?
>> So when it comes down to it, fusion isn't just by itself about what happens in the core. You take a couple of small things and make them into a bigger thing and you get all that glorious delta MC squared to uh make heat. I mean, that's a big part. It's also about turning these neutrons into heat in a safe, robust blanket. Capturing it reliably, running turbines efficiently and continuously, no matter what type of fusion you're making. Handling tritium safely. That's actually something you already look at in fission power plants because water is a coolant and water just needs to absorb a couple of neutrons in a high neutron field and then hey, you got tritium. You also have to interface with the grid and you also have to beat the uh current market shareholders on cost. Now, Z-pinch is one of the more interesting ones in that it's compact and potentially lower cost, at least as far as the reactor design, but it's got to achieve repetitive pulse operation, which again, we go back to the neutron damage material size, 'cause that's still a thing. Long-term neutron embrittlement of sensitive components, which is endemic through all fusion reactions, because the neutrons are actually higher energy than the neutrons you get from fission. So all the neutron embrittlement you have to worry about in the context of nuclear fission, well, it's less significant and the equipment is less sensitive. There's the duty factor issue with all the little pulses. And then there's long-term reliability and a competitive capacity factor. None of those are easy, and it's a big step between. And while they don't need new physics, that's great. But there's a big difference between physics getting you to do it once and engineering getting it to you. It runs every day for decades and it makes you a lot of money. Thanks so much for the recommendation and thanks so much for watching. I'll see you next time.