Nuclear Fusion with Lasers? - Nuclear Engineer Reacts to Megaprojects

Added: 2026-05-07

[00:00:00]Phosphate pass which is a thing. Each beam passes through these slabs multiple times.

[00:00:04]>> It's time for some more mega projects.

[00:00:07]Specifically fusions catch looks like a very sci-fi version of the national ignition facility. Let's see what all this is about.

[00:00:16]>> In the hills of San Francisco, the Lawrence Livermore National Laboratory sits behind fences and security checkpoints. Inside one of its buildings, 192 laser beams can fire simultaneously at a target smaller than a peppercorn. the building.

[00:00:29]>> So the key thing behind this focus, so this is inertial confinement fusion, is energy density, not just that it's small. So you're taking the energy of a power plant pulse and dumping it into something the size of a grain of sand in a few nancond. So extreme power density and an extreme precision coupling problem.

[00:00:52]>> Building covers roughly the footprints of three football fields. It cost about >> three football fields because American units of measurement and not the metric system.

[00:01:01]>> 3.5 billion dollars to construct. And on December the 5th, 2022, it did something no facility on Earth had ever done before. The National Ignition Facility known as NIF fired those 192 lasers into a tiny cylinder of gold. X-rays from the heated gold crushed a capsule of hydrogen fuel at the cylinder center. In a few billionth of a second, the fuel reached temperatures hotter than the core of the sun and compressed to roughly 100 times the density of lead.

[00:01:28]>> So, as far as never done before, it was not the first fusion that's been around since the 1950s, at least as far as a fusion reaction triggered by humans. And it was not even the first net energy system, but it was the first lab scale inertial confinement fusion ignition.

[00:01:48]That is to say a Q value or energy in greater than energy out a Q value greater than one at the target level. So not counting support systems or anything like that just at the target the thing that got hit with the lasers. If you compare this to other fusion this number would be kind of like Q plasma for a tokamac or really any device that uses magnetic confinement. But you're still talking within the primary system and really just the primary system that does the fusion, not what it takes to support the primary system. The fact that that's considered a monumental accomplishment, it really is, just goes to show how hard fusion is compared to vision or really any other form of power generation because that sort of aspect isn't even really measured. It would be like setting something on fire and surp and being surprised that it gives off energy. No >> fusion reactions rippled through the fuel and released 3.15 mega of energy about 50% more than the laser put in.

[00:02:58]>> Okay. So what he means by this is about 2 mega was delivered to the target and about 3 megajou was released from fusion. So that gives you a Q target value of 1.5. But in order to run all your support systems, well, that drops Q down to probably.001 or less. Another way of viewing this, it's a bit like saying your engine produced more energy than the fuel in the cylinder, but you burned 100 times more energy just to run the fuel pump. So, both statements are true, but only one of them matters for power generation. So very much still at the experimental scale. Not sure about inertial confinement fusion for power generation, at least compared to magnetic confinement fusion, but we'll see.

[00:03:50]>> That single result crossed a threshold scientists had pursued for over 60 years. For the first time in a laboratory, controlled fusion produced more energy than it took to trigger the >> Again, we're talking specific about inertial confinement. So he's not saying other types of fusion did work. If you want to get specific, first time more energy than laser energy coupled to the target, not more than total system energy. And that really is the difference between a scientific minds milestone and large scale engineering and production.

[00:04:28]>> United States government originally funded NIF for a very specific reason.

[00:04:32]After the country stopped detonating nuclear weapons underground in 1992, it needed another way to verify that its aging warhead still worked. NIF would recreate the extreme conditions inside a thermonuclear explosion, generating data for computer simulations that replaced live testing. The >> So, one thing to be clear, NIF does not recreate a full-on thermonuclear weapon.

[00:04:54]It recreates radiation driven implosion physics, which is the thing that causes modern nuclear weapons to go off. It's a modern two-stage design where you have a fision stage and a fusion stage. Plus, it validates data in these high energy density regimes in which those devices operate. And it does that by the equivalent of a snapshot, just one individual frame. If an entire nuclear explosion is the entire Game of Thrones series, what NIF doing is is looking at one frame, which is nice because you don't have to do full scale testing.

[00:05:36]>> Promise of advanced fusion energy research helped build political support, but the core mission was always national security. Getting from that mission statement to the 2022 result took decades of budget crises, leadership firings, congressional threats to cancel the project, and a long stretch where the physics just seemed not to be very cooperative. The facility was finished years late.

[00:05:59]>> That's annoying when your job insists on physics being cooperative and it's not.

[00:06:04]I mean, that's that's one of the nice things about working in nuclear power.

[00:06:09]the nuclear fision works and it's reliable.

[00:06:13]I mean, can you imagine a scenario? It's like, well, I guess gravity's not cooperating this time or something like that.

[00:06:20]>> Billions over budget and it spent more than a decade falling short of the goal it was named after. Why NIF had to exist.

[00:06:28]Livermore had been building giant lasers since the 1970s. Shiva, a 20 beam system fired its first shots in 1977 and demonstrated that laser-driven implosions could compress fusion fuel.

[00:06:39]Its successor, Nova, came online in 1984 with 10 beams and significantly more energy. Nova ran for over a decade and produced real fusion neutrons, but it could not reach ignition. The laser simply >> So fusion neutrons is a real thing that's distinct from fision neutrons.

[00:06:55]And that is one of the key ways of how you detect if fusion is occurring by detecting neutrons and seeing what the energy level is. Like a dutyium tridium fusion reaction releases a neutron with an energy of 14.1 million electron volts. So a fast neutron that is to say that neutron is carrying a lot of energy and it's actually a big chunk of what is released from that reaction because the total energy is 17.6 six me million electron volts. So it produces some energetic neutron. Compare that to a uranium 235 fision reaction. The entire energy release from that reaction is about 200 million electron volts. So a lot bigger. But uranium 235 is a lot bigger than dutium and tridium which are hydrogen 2 and hydrogen 3. So fusion is still more energetic in the case of dutium tridium versus uranium 235 fision. It's still more energetic per unit mass. But those neutrons, most of them are thermal neutrons. And by thermal, that means slow and less than one electron volt. Not 1 million electron volts, one electron volt. But those thermal neutrons are what's needed to sustain a fision reaction. Here, these neutrons are what is given off as energy from these fusion reactions. So you can quite easily with a fast neutron detector detect hey these fast neutrons are coming from f from fision or coming from fusion for that matter. In this case pretty indistinguishable that these neutrons are coming from fusion. Not that there would be any fision going on but there is a difference between fision neutrons and fusion neutrons. In some fisions you can have high higher energy neutrons but that specific 14.1 million electron volts from duty tridium fusion is one fingerprint and one piece of evidence that fusion is occurring.

[00:08:52]>> It did not deliver enough energy to push the fuel over the edge. By the late 1980s, Livermore physicists had a clear picture of what ignition would actually require. They needed a laser in the range of 1 to 2 mega roughly 10 times what nova could produce. A 1990 National Academy of SERS review endorsed that target and recommended the country pursue it. An earlier concept calling the 10 megajles had been shelved as simply impractical. The 1 to2 mega range looked achievable with existing glass laser technology that would be scaled up dramatically. Then the world shifted in a way that actually made the project politically urgent. The Soviet Union collapsed. The cold war ended and in 1992, President George HW Bush signed a moratorium on underground nuclear testing. You see, the United States had conducted over a thousand nuclear tests since 1945. But suddenly that was off the table and the moratorium created an immediate problem. Nuclear warheads get old, their components degrade. Engineers who designed those weapons, they retire.

[00:09:46]Without live tests, the country needed some way to verify that its stockpile remained safe and reliable.

[00:09:52]>> So, in addition to weapon degradation, because just like anything else, if they're not maintained, they're not routinely tested, you don't necessarily know that they work. Um, within nuclear power plants, there are tests that are required to be performed known as surveillances that it's the federal law that you have to verify those safety systems work under these specific test conditions. So, not being able to do that and the other part of this, the generation gap between people that know how the devices work and know how to do these tests versus people that don't.

[00:10:28]out to Quest. Not to say that that can't be taught in any sort of uh training program, cuz it can, but training can only get you so far. There's still some on the job aspects that need to be trained just like when you do anything else. I encountered that a lot through my career and had nothing to do with weapons. But within nuclear power plants, there was a similar generation gap. When I first hired in when I was in my 20s, most of the people I learned from were in their 50s and 60s. There weren't that many 30 and 40 year olds because of that generation gap associated with >> was a program called stockpile stewardship, a broad effort to maintain confidence.

[00:11:06]>> That right there was the Z machine >> in the arsenal through laboratory experiments and advanced computer simulations. NIF fit directly into that framework. It could recreate the extreme temperatures and pressures found inside a thermonuclear detonation, generating experimental data to anchor the computer codes that would now stand in for underground tests. In 1993, Energy Secretary James D. Watkins approved key decision zero, the formal authorization to begin the project. The justification was deliberately broad. NIF would serve weapon stewardship, advanced high energy density physics, and push toward demonstrating fusion ignition in the laboratory. Each of those missions demanded a different >> So that's kind of your classic um objective creep or scope creep, if you will, when you have something and then you're going to say, "All right, now you're going to not just test weapons, but you're going to try to do fusion as well." That's it's interesting. And you know what? I appreciate it's a you could argue it's a more noble cause using this technology for peaceful purposes but that can also explain the long learning times because when you shift your objectives like that or add additional objectives and you pursue them simultaneously often not with a proportional increase in personnel or funding >> deliverable for the weapons program NIF had to produce data under conditions no other facility could reach for the fusion energy community. It had to produce proof that laser-driven ignition was physically possible. Both goals pointed toward the same requirement, a laser powerful enough to compress and heat a tiny fuel capsule until fusion became self- sustaining. And it has to be tiny because you need absurdly high pressure and temperature plus confinement time. So that's why you had to go small because the loss in criterion of those three things is quite unforgiving. The sun can get away with it relatively easily. It also uses inertial confinement fusion because of its relatively high gravity. Gives you all the pressure and temperature and confinement time that you need.

[00:13:08]>> Proving an if meant committing to something far beyond a scaled upnova.

[00:13:12]The new facility would need 192 beams all arriving at a millimeter scale target within picos seconds of each other with pointing accuracy measured in the tens of microns. That is some close tolerances. Closer than the tolerances even used in nuclear weapons. POC 10 - 12 versus nano only 10us 9. Every beam had to deliver precisely the right amount of energy at precisely the right moment. Any significant imbalance would distort the implosion and kill the shot.

[00:13:43]How the laser makes a star. Of all the fusion analogies that talk about stars, a lot of them had to do with magnetic confinement, which it's not really the same thing, but this one's at least closer because it uses inertial confinement, which is the same type of fusion that stars use it, albeit on a way, way smaller scale and a way shorter period of time.

[00:14:08]So, an NIF shot begins with a weak pulse of infrared light. It starts in an oscillator and measures about 1 billionth of a jewel. Over the next few micros secondsonds, that pulse gets split, routed, and amplified across 102 separate beam lines.

[00:14:22]>> It's actually a pretty good explanation.

[00:14:24]But it's not just the amplification part. It's stored energy release. Cuz those glass slabs are pumped by flashlights and they store huge amounts of energy. Laser pulse is just the trigger. Kind of similar to your neutron source for instance in a fision reactor.

[00:14:40]And in this case, you have 192 of them and they have to be in sync.

[00:14:44]Magnification happens in slabs of neodymium doped phosphate glass which is a thing. Each beam passes through these slabs multiple times picking up energy at each pass. By the end of the amplification chain the >> 92 beams collectively carry about 4 to 4.8 mega of infrared light converted to about 2 mega of ultraviolet light at the target.

[00:15:08]>> That's brutal. 50% energy loss just in frequency conversion cuz that's all what happened. You're still dealing with electromagnetic radiation. You just shifted from the infrared side of the spectrum all the way up to the ultraviolet side of the spectrum. All you did was shrink the frequency and that killed half of your energy. This might be the reason why NIF will never be a power plant cuz that's just brutal.

[00:15:35]>> That infrared light is powerful but poorly suited for the target. Shorter wavelengths couple energy to the target much more efficiently. So before the beams enter the target chamber, they pass through crystals of potassium dihydrogen phosphate known as KDP. These crystals convert the infrared light to ultraviolet at 351 nanometers, roughly onethird the original wavelength. The >> at 351 nanometers, you're just slightly ultraviolet outside of the uh visible range of regular violet. So why are they doing this? Why are they sacrificing this much energy? Well, shorter wavelength, better absorption, less preheat, and better implosion symmetry, which is crucial to how this thing even works. So, the physics is so demanding, you have to cut that efficiency in half like that.

[00:16:23]>> Process is called frequency tripling. It sacrifices some total energy, but produces light, >> some that the target can actually absorb, which is useful. All 192 ultraviolet beams then converge on the target chambers center. They are time to arrive within a few picos seconds of each other aimed with the pointing accuracy on the tens of microns. Oh, by the way, a picoscond here. That's a a trillionth of a second. A micron is a thousandth of a millimeter. The precision required is in >> a micron is a thousandth of a meter.

[00:16:53]Yes, a micron is a thousand of a millimeter. So very small, very narrow tolerances.

[00:16:58]>> Insane. And every shot has to be that precise. Now the beams don't strike the fuel directly. They enter a small hollow cylinder called a hole room, roughly 1 cm long, made of gold or sometimes depleted uranium. The laser energy hits the inner walls of the hologram and heats them to about 3 1/2 million° C, which is I don't know a lot Fahrenheit.

[00:17:21]At that temperature, >> just multiply it by 1.8. The 32 is negligible at this point. The walls radiate a powerful bath of X-rays inward from all directions. This approach is called indirect drive. The X-rays provide a more uniform squeeze on the fuel than direct laser illumination could achieve. So, so here is your next trade-off here. This X-ray indirect drive improves the symmetry. What you're giving up is control controller. Not saying it's going to cause some sort of explosive fusion Chernobyl if such a thing were, but you're giving up direct control for smoother boundary conditions. Now, for scientific experimentation, but one thing about operating any sort of power plant, yes, this is another reason why this isn't a power plant is you like precise control.

[00:18:05]There's not a fusion equivalent of control rods here. This is it's either on or off and you're at the mercy of the X-rays.

[00:18:15]There's not a X-ray reducer control system that you you stick in there.

[00:18:23]Suspended at the center of the whole ROM sits the fuel capsule. It's a hollow sphere roughly 2 mm across with a thin outer shell and a frozen layer of dutyium tritium fuel on its inner surface. Dutium and tritium are both isotopes of hydrogen. When the X-ray bath hits the capsule's outer shell, the surface material explodes outward. Now, that outward explosion actually drives the rest of the capsule inwards, kind of like an explosion in reverse. The >> So, an implosion. You could also say it's a bit like the rocket equation in reverse in that mass ablates outward and a reaction force compresses inward. I guess kind of like that.

[00:19:00]>> The implosion is violent and fast. It compresses the fuel to roughly 100 times the density of lead.

[00:19:06]>> Yeah, this figures out to about a kilogram per cubic cm. That's getting you near stellar core density, which I mean that's the part where they talk about it being similar to a star, albeit very small and very briefly. But yeah, you're starting to get into that territory on a very small scale with this experiment >> and heats the center to over 100 million° Celsius, which is many, many hundreds of million Fahrenheit as well.

[00:19:36]It's really hot. Under those conditions, detium and tritium nuclei slam together and fuse, releasing energy and fastmoving helium nuclei called alpha particles. goes out.

[00:19:44]>> So the reason why you need it on the order of 10 times hotter than the sun's core, not quite 10 times, the sun's score is about 15 million and this is 100 million. But the sun can operate relatively colder because it has all that gravity confined. NIF has to compensate for it because we don't have any equivalent of the sun's amazing gravity even on a really small scale. So just have to make it hotter to compensate for the shorter confinement time as well. It's probably that's actually the bigger constraint is the confinement time that is lacking in this relative to the sun.

[00:20:18]>> Alpha particles are key to ignition. As they collide with surrounding fuel, they deposit their energy locally and heated further. If the implosion is symmetrical enough and the density is high enough, this alpha particle heating becomes self- sustaining. A burn wave propagates outward through the compressed fuel and the fusion reactions generate more energy than the laser originally delivered to the target. Now >> that was a pretty good explanation.

[00:20:40]Fusion makes alpha particles. I mean that's a key fusion product which gives you a sense of how small scale fusion is relative to fision that you do this with small things hydrogen 2 and hydrogen 3.

[00:20:53]You make an alpha particle as a product and yet which is the normal decay mode of something like uranium. It's kind of it's kind of funny how that just naturally works out. But the an alpha particle is a very stable chunk of nuclear material which is why it shows up twice. So those alpha particles heat the fuel hotter fuel more fusion. But the opposing forces are radiation losses and hydrodnamic disassembly.

[00:21:20]>> That threshold is what the NIF defines as ignition. Reaching it depends on near perfect conditions. The implosion has to be spherically symmetrical to within a few%. The fuel >> Yeah.

[00:21:34]So that undersells it a little bit because these pertibbations grow, these asymmetries grow by what's known as the Raleigh Taylor instability. And what that looks like is imagine trying to compress a balloon evenly while the surface is wrinkled and every wrinkle is going to grow exponentially in a matter of nanconds and has to be perfectly even. Yeah, that's why this is hard. All layer has to be frozen uniformly with surface roughness measured in fractions of a micron. A slight bump on the capsule or a small imbalance in the X-ray drive can seed instabilities that tear the implosion apart before the fuel gets hot enough. Every target is handmade and inspected under microscopes.

[00:22:18]>> And again, the failure mode here, nothing dangerous per se. It's more that just ruins your fusion reaction. Every shot is a single attempt at a goal where the margin for error is absolutely tiny.

[00:22:33]Building the beast.

[00:22:35]Now the building that houses all of this 704 ft long, 43 ft wide, and 85 ft tall.

[00:22:40]Its concrete foundation was engineered for seismic stability because a facility that aligns laser beams to within tens of microns cannot tolerate the ground shifting beneath it. and you put it near the San Andreas fault.

[00:22:55]Yeah. So, when you think of seismic stability and seismic design criteria in a nuclear power plant, it's all about safety. Ensuring that if cabinets and electrical equipment tip over, it's bolted so that it doesn't tip over or if it's in a scenario where it does, it doesn't damage anything. And you use things like snubbers to hold things in place. Here you're talking about I mean that so that was talking about seismic stability in case of earthquakes. Here you're talking about seismic stability with these micro earthquakes to even call them earthquakes. So sub threepointer things that you can barely even feel. But when you're talking about that level of precision, these sensitive instruments or instruments that require to operate on these very close tolerances simply won't cooperate. Cuz after all, you're aligning 192 beams over hundreds of meters to a millimeter target where you care about micron scale drift.

[00:24:03]>> Structure sits on deep foundations designed to dampen vibrations from earthquakes and even heavy truck traffic on nearby roads. Inside the 192 beams are grouped into 48 clusters of four called quads. These >> Yeah. So those those trucks things that are we're talking sub one on the RTER scale. A big truck driving past.

[00:24:22]>> Quads run through two enormous laser bays that stretch the length of the building, one on each side. The beams travel through their amplification chains in these bays and then converge inward toward the target chamber at the building center. It's >> like a giant pencil. Installing that target chamber was actually one of the most dramatic moments of the entire construction process. The chamber is a sphere 10 meters in diameter weighing 264. Anything when you're trying to design spheres is hard. That's one of the reasons why reactor vessels are not spheres. I mean, that would be ideal for a uniform fuel assembly, uniform neutron flux, uniform thermal flux profile, but yeah, it's just not worth it.

[00:25:02]>> £4,000. It needs to sit 21 ft below ground level in a specially excavated concrete well. To lower it into place, Livermore brought in a 900 ton crane from the Nevada test site. In June 1999, crews guided the sphere down into position in a careful, slow operation. A drop or a hard impact would have damaged the chamber's precisely machined ports where the laser beams and diagnostic instruments connect. There was no spare, and believe me, this is very expensive.

[00:25:28]Once the chamber was sealed, the real complexity began to build the building around it. The laser system required roughly 3,100 amplifier slabs of neodymium doped glass. Each one manufactured to strict optical specifications. About 7,500 large optical components, including lenses, mirrors, and the KDP frequency conversion crystals had to be installed and aligned along the beam lines. Beyond those large pieces, another 30,000 smaller optical components went into place. An automated control system tied together approximately 60,000 control points, coordinating everything from beam timing to amplifier power to target positioning, assembly and commissioning stretched across most of the 2000. Beam lines were activated in stages with engineers verifying alignment and performance on each quad before moving to the next. By early 2009, the full >> Yeah, you're going to want all levels of independent verification to make sure everything's properly aligned. I mean, we do that in nuclear power plants.

[00:26:22]Well, for a whole mess of applications, but one of the more critical aspects is moving fuel during an outage to make sure everything's properly aligned.

[00:26:30]>> System was ready for its first integrated test. March the 10th, 2009.

[00:26:33]All 192 beams fired together and delivered about 1.1 mega of ultraviolet light into the empty target chamber, reaching roughly 2 terowatts of peak power. The shot confirmed that the laser worked on a single coordinate >> terowatt of peak power. That is to say over a thousand times more than a nuclear power plant to stay. Now granted, this is peak power and pulse power. So not at all the same energy regime because after all with pulse lasers, pulse laser tattoo removers, hey, you're on the order of megaww. And that's because you just divided by a really, really small number.

[00:27:09]But to say that your handheld laser pointer has more power than a high performance sports car, well, no, because sustained is what matters when you compare stuff like that. But yeah, terowatts of peak power. The main thing about peak power is designing things to withstand those peaks. I mean, peak pulse power, that's essentially how Styr Pyro's crazy 400 batteries wired together experiment operated. ated system. A decade of construction and commissioning had produced a functioning facility. Building NIF, however, turned out to be the more straightforward half of the problem. Straightforward Jesus.

[00:27:46]Making the targets actually ignite would prove far harder and take far longer than most anyone anticipated.

[00:27:54]The mega project crisis.

[00:27:56]So, when NF received formal approval in the mid1 1990s, the estimated construction cost was roughly 1.1 to$ 1.2 billion. The schedule called for about 8 years of building with completion targeted around 2002 or 2003.

[00:28:10]Those numbers helped sell the project to Congress. They also turned out to be absolutely wrong. For years, Livermore reported to the Department of Energy that NIF was on track. Then in mid 1999, a different picture emerged. Independent reviews and congressional inquiries revealed that the project was badly behind schedule and hundreds of millions of dollars over its baseline. Well, when you're building any big new projects with something new at scale, especially stuff that involves nuclear, well, I'm not even remotely surprised that this happened.

[00:28:40]>> Budgets whoops at Daisy and were still DOE and Congress had not been told the full extent of the problems as they developed. Multiple review panels investigated what went wrong, and their conclusions pointed consistently at management failures rather than impossible physics. The project had been planned with insufficient contingency reserves. systems engineering and integration have been poorly coordinated. Oversight within the laboratory had failed to flag growing problems early enough.

[00:29:06]>> So I mean there's a saying that goes for some of these sort of things that I've heard in various engineering departments is that physics is hard but systems engineering is harder. I mean when you scale things up like this. Yeah. Add in that it's first of kind for this particular type of target type of laser array. I mean, when it comes to the system engineering, gets so much harder when you scale everything up.

[00:29:29]>> One review used the word hubris to describe the institutional culture around NIF's original estimates. Energy Secretary Bill Richardson delivered a public rebuke that became the defining quote of the crisis. Bad management, he said, has overtaken good science. In September 1999, NAF project director Michael Campbell resigned. The project came dangerously close to cancellation.

[00:29:50]Senator Tom Harkin of Iowa called VNIF out of control and openly questioned whether it would ever work or was even worth completing. Congressional committee sure might still be out.

[00:30:00]>> Funding was at risk and Liverour was forced to reorganize. George Miller took over NIF programs and Ed Moses was appointed project manager. The new team conducted a full rebaseline of the project. That's also part of it when you have these turnovers in in management and then you essentially start over, especially with something like this where on the uh on the waterfall to agile scale of project management, this one is all the way to the waterfall style in terms of everything being highly customizable and very strict construction related deadlines.

[00:30:40]and orders of operation and critical path methodology and yeah any sort of change any sort of management issue.

[00:30:50]It's just going to cost you both in terms of time and in terms of >> producing revised cost and schedule estimates that were meant to be credible rather than insanely optimistic. The new completion date slipped to 2008 or 2009 and the construction budget rose to roughly $2 to $2.2 $2 billion. When the Government Accountability Office factored in research and development costs along >> Government Accountability Office Construction, its projections ran between 3.3 and $4 billion. The final frequently cited figure for the total build settled at approximately 3.5 billion. The rebaseline saved NIF politically, but it did not end the scrutiny. And good thing, too. GAO auditors later noted that the laboratory had declared construction complete while deferring certain safety related working including the installation of shielding doors in the laser bays. Something most scientists probably wanted. That deferred work later disrupted early operations and drew additional criticism about where >> and that's the other thing you can you start dropping scope and then only for it to be picked up later because you're concerned about one very specific budget cycle which could ultimately end up costing you more in the end. But if this is all you have, this is all you have.

[00:32:04]And then you start dropping stuff or deferring. I mean, you see this in nuclear projects with when you start dropping scope and deferring things to the next outage cuz a lot of maintenances can only be done during refueling outages because the reactor needs to be cooled down and depressurized in order to do certain types of maintenance and certain types of inspections. And if your outage starts to run long or you get over budget, then well, a lot of times scope stops starts dropping hard and then you end up having to wait till the next one.

[00:32:36]It's a business decision >> whether completion milestones have been defined to meet reporting deadlines rather than actual readiness. By the time all 192 beams fired together in March 2009, NIF had surveyed budget revolts, leadership purges, and repeated congressional threats. It was years late, billions over its original estimate and deeply scrutinized. The facility existed, the lasers worked, and the only way to justify what it had cost was to deliver the result printed on the building's name.

[00:33:06]The long climb from first experiments to a decade of almost >> with the laser working, NF shifted experiments in 2009 and 2010. Weapons physics shots began generating data for the stewardship program almost immediately. In September 2010, the facility conducted its first integrated ignition experiment.

[00:33:24]>> I like that it's shot director. It's I mean, it's kind of like a NASA-ish control room except flight director.

[00:33:30]It's shot director. It's kind of fun.

[00:33:31]>> Barring all 192 beams at a cryogenically cooled target with a dutium tritium fuel layer. The following month, a shot set a record for neutron yield from a laser-driven implosion. These early results were encouraging enough to support an ambitious timeline. The national ignition campaign running from roughly 2010 through 2012 aimed to reach ignition by pushing the >> like anything else with fusion. I mean some milestones are easy to achieve but then you can kind of underestimate how long it takes them to get from one milestone to another >> laser to its full design specs. On July the 5th 2012 NF fired a shot delivering 1.85 mega of ultraviolet energy at a peak power of 500 terowatt. The laser had essentially met the performance targets set >> I love peak peak power. You can make numbers just sound crazy high just because you're dividing by nancond.

[00:34:18]You're dividing by a billion. So in the 1990s >> a or you're dividing by a billionth. So you're multiplying by a billion rather.

[00:34:27]>> The fuel didn't ignite. The implosions were producing fusion neutrons but falling far short of the self-sustaining burn that defines ignition. Shot after shot, the yields came back below expectations. The National Ignition Campaign ended without achieving its goal, the laser done its part. The problem simply lay in the physics of the implosion itself.

[00:34:48]>> Well, and that just goes to show even if the laser worked perfectly. There's still the target fabrication which is not easy. There's laser plasma interactions which are not predictable, at least not on the tolerances that you desire. and the rapidly growing instabilities with this inconsistency. I mean, it's a couple nonlinear systems.

[00:35:12]So, the failure modes are complex.

[00:35:15]Smaller symmetries in the X-ray drive, imperfections in target fabrication, and instabilities during compression were all sapping energy from the hot spot before it could reach the conditions needed for a burn. The gap between where the experiments landed and where ignition required them to be was enormous. The institutional consequences arrived in 2013. The Obama administration's budget proposal scaled back NAF's funding and cut dedicated time for discovery science experiments.

[00:35:40]Staff were laid off or reassigned. The shot schedule was reduced. Federal overseers pushed the facility more tightly toward its weapon stewardship mission, which had always been the primary justification for its budget.

[00:35:51]Operational reforms followed, reorganizing NIF along the lines of a user facility, >> getting all those defense dollars >> to increase the number of shots per year and extract more value from each operating day. Progress after 2013 came slowly and through painstaking iteration. In February 2014, a shot achieved a milestone called fuel gain greater than one, meaning the fusion energy produced by the fuel exceeded the energy the laser actually deposited into the fuel capsule. The overall laser energy was still far larger than the fusion output, but something important had changed inside the target. Alpha particles from the fusion reactions were depositing enough energy to meaningfully heat the surrounding fuel. self heating was real, even if it was not yet enough to trigger a runaway burn.

[00:36:35]>> And as far as experiments, this proved that hey, you can heat things using alpha particles from fusion cuz before that that was theoretical and specifically from lasers. I mean again any sort of fusion reaction would have been using magnetic confinement before then. Vansaw used a strategy called the high foot pulse which shaped the laser's energy delivery to reduce instabilities during the early stages of implosion.

[00:36:59]The trade-off was lower final compression, but the implosion held together more cleanly. It pointed toward a principle that would guide the next several years of work. A slightly less extreme implosion that stayed symmetrical could outperform a more aggressive one that fell apart. From 2014 through 2019, campaigns refined nearly every variable in the system.

[00:37:18]Scientists tested changes to whole room geometry, including rugby-shaped designs meant to improve X-ray uniformity. They shrank the fill tubes used to inject fuel into the capsule, reducing the defect each tube left in the shell. They adjusted pulse timing capsule thickness and the precise temperature at which the dutaterium tritium layer was frozen.

[00:37:35]Every adjustment was incremental. Each required dedicated shots and months of analysis. By the late 2010s, NIF was conducting roughly 300 to 400 shots per year, far more than during the ignition campaign. Then on August the 8th, 2021, a single shot produced a result that stunned the fusion community. The target yielded 1.35 mega of fusion energy, roughly 70% of the 1.9 megajou the laser delivered. The fuel had dented what physicists called a burning plasma regime where alpha particle self-heating was the dominant source of energy in the hotspot. A propagating burn wave had begun.

[00:38:09]>> So that's important and that shows that ignition is indeed feasible. You could argue that that was just as important as when they achieved ignition, which I'm sure he'll get to in a little bit.

[00:38:20]>> Moving through the fuel, about 2% of the dutarium tritium was consumed before the implosion flew apart. It was by far the highest yield any laser fusion experiment had ever produced. It was also agonizingly close to the threshold.

[00:38:34]Attempts to reproduce the result in the months that followed fell short. Small variations in target quality and laser conditions produced yields well below the August peak. The facility could reach the edge of ignition, but it couldn't reliably stay there. Every capsule was still handmade. Every shot still depended on conditions aligning within razor thin tolerances. The next step required a shot where everything came together just well enough to push the output past the energy the laser delivered.

[00:39:02]The December 2023 brick.

[00:39:03]>> Yeah, this was the one I was thinking of. Yep. Well, on December the 5th, 2022, NIF fired shot N221204, 192 beams delivered 2.05 mega of ultraviolet light into the whole room.

[00:39:16]The X-rays compressed the fuel capsule, the implosion held together, and when the diagnostic instruments finished recording. The fusion yield read 3.15 mega. Approximately 1.1 * 10 18th neutrons had streamed out of the target in a few billionth of a second. The fusion energy output exceeded the laser energy input by more than 50%. The result was reviewed internally for 8 days. On December the 13th, energy secretary Jennifer Granholm stood at a press conference in Washington and announced the achievement. She compared it to the Wright brothers first flight to Kittyhawk, a demonstration that something long theorized was physically real. The comparison was carefully chosen. The Wright brothers didn't build an airline. NAF did not produce usable electricity. That's probably a decent comparison and one to state to that most people would realize. Yeah. Again, this is scientific ignition nowhere close to an engineering break even.

[00:40:09]>> What happened on December the 5th was simply proof. For the first time in any laboratory anywhere, a controlled fusion experiment had generated more energy from the fuel than the laser delivered to it. This was the ignition criterion NIF had been built to meet. After 13 years of falling short, the facility had finally crossed the line and the result deserves prec >> the goal of ignition. Sure.

[00:40:34]>> Precise framing. The 2.05 mega that entered the whole room represented the energy of ultraviolet light reaching the target. Generating that light required the facility to draw hundreds of megles from the electrical grid to charge the capacitor banks that >> and there it is. So 1% efficiency, and this is just to get to the point where you're heating stuff, not including what it would take to produce electricity, something like that, that you'd get well below 1%. Now, this isn't a criticism of NIF necessarily, cuz again, it's not a power plant. The goal isn't to be a power plant. It looks nothing like a power plant. doesn't act anything like a power plant, but it should just give you a sense of how far away inertial confinement is from being on the grid, if if ever.

[00:41:18]>> Power the laser amplifiers. The overall wall plug efficiency of NIF measured as fusion energy out divided by total electrical energy consumed remained around 1% or less. A fusion power plant would need to produce far more energy than its entire facility consumes, not just more than the laser delivers. By that standard, NF was still nowhere close. Each shot also carried a significant marginal cost, often cited in the range of half a million to a million dollars when accounting for target fabrication, laser maintenance, and operations. NIF fires roughly once a day at most during active campaigns.

[00:41:52]Power generating facility would need to fire at its target several times per second. The December 2022 shot demonstrated physic.

[00:41:58]>> And even if you did, you would need perfect target injection every time and optics that survive this continuous operation. I mean, it's a precision experiment. You're not going to have a I don't know machine gun style beltfed NIF targets going in there. X it did not demonstrate engineering, economics or scalability. And those caveats really mattered in the months that followed as headlines around the world announced a fusion breakthrough without always specifying what had and had not been achieved. Within the scientific community, the reaction was more measured, but it was still significant. Ignition had been an open question for decades. Skeptics had argued that laser-driven inertial confinement might never reach it. That argument was now settled by data. In 2023, NIF conducted additional shots that produced fusion yields in the range of 3 mega, confirming that the December result was reproducible rather than a one-time alignment of conditions.

[00:42:52]Details remained limited because >> that's an important part of the scientific method.

[00:42:56]>> Much of NIF's target and shot data falls under classification restrictions tied to the weapons program. But the pattern was clear enough. The facility could reach ignition more than once, and the scientific teams were learning which target and pulse parameters kept them in the right regime. The data readout on December the 5th, 2022 landed after 30 years of political crises, engineering marathons, and incremental physics gains. It confirmed that a controlled fusion burn could ignite and sustain itself in a laboratory target. That had never been done before. After ignition, what changes now for the stockpile stewardship program? Ignition open >> not fusion power plants. I can tell you that much.

[00:43:36]>> Opened a new category of experiments.

[00:43:38]Before December 2022, weapons physicists could study implosion dynamics and measure material properties under extreme pressure. After ignition, they can study burning plasmas where alpha particle heating dominates. That regime is far closer to what happens inside an actual thermonuclear weapon. The experimental data feeding into the simul >> compared to what you would need in the power >> codes that replaced underground testing became substantially richer within the weapons community. This likely cements NIF's political standing for years for the fusion energy question. Ignition settled a foundational debate and immediately exposed the next one. The physics works. Laserdriven inertial confinement can produce a self-sustaining burn. That answer had been genuinely uncertain for decades.

[00:44:21]And it is no longer uncertain that a functioning power plant based on this approach would need fusion gains in the range of 10 to 100 times laser input.

[00:44:31]>> Oh, it would need more than that.

[00:44:34]It would need 10 to 100 times total facility input. That's about how much a nuclear power plant is. If you talk about the the house loads of a nuclear power plant are on the order of 15 to 20 megaww. And the nuclear reactor is on the order of 3,000 which figures out to about 1,000 on the grid. And that is something that is competitive. So fusion needs the equivalent of that. That that would be like saying a fision reactor or a fision power plant has a Q total value of 150. And now no one no one talks about that. What is talked about is capacity factor and that's its ability to stay online and stay producing power and minimize its downtime which nuclear fision reactors that is one of their strengths is very high capacity factor.

[00:45:36]And you can tell you're very far off when you're measuring it in terms of energy in versus energy out versus capacity factor. That is a completely different way of thinking and it just goes to show how how far away we are from that.

[00:45:53]>> Not one and a half times. It would need to fire targets several times per second, not once per day. The lasers would need wall plug efficiencies many times higher than NIF systems achieve.

[00:46:02]And the targets, currently handmade spheres inspected under microscopes, would need to be mass-produced by the millions at extremely low cost with consistent quality. NIF was designed to demonstrate ignition for weapon science, not to serve as a prototype power source. It cannot be upgraded into a high repetition rate, high gain facility. If inertial fusion energy is going to advance beyond this proof, it would require a successor facility designed from the start around different engineering priorities. That is a policy decision.

[00:46:29]>> It would basically be the NI the inertial confinement equivalent of ether to go further down that direction. But even eater is not a power plant. But that would be the next step down that path. And is it worth it? I can't really speak for the scientific the the hard science aspects of it. But in terms of power generation, I mean there's just there's just better ways to do it.

[00:46:53]>> No one has yet made and the price tag for such a facility does not currently have a serious estimate attached to it.

[00:46:59]The third consequence runs through NIF's role as a science facility. During the years when ignition remained out of reach, budget pressure squeezed time for basic science and academic research on the system. Discovery science campaigns were cut, access for university researchers was limited. Ignition strengthens the argument that NIF is a singular scientific instrument with capabilities no other facility can match. Researchers in astrophysics >> That's true. It's the only facility that's designed that specific way.

[00:47:26]>> Planetary science and material science all want beam time to study matter under extreme conditions. But NF's budget is limited and its shot schedule is governed primarily by weapons program priorities. Greater scientific demand for access collides with an unchanged funding structure. France operates the only peer facility in the world, the laser mega jeul near Bordeaux. It serves a similar stewardship role for the French nuclear arsenal. The two programs share some unclassified research but operate independently on classified work. The UK contributed cooling equipment to NIF under collaborative agreements. One indicator how expensive even incremental improvements to shock capability.

[00:48:03]>> Bringing star power to Earth. That's cute.

[00:48:06]>> D can be for partner nations. Ignition left an open question that the December 2022 result cannot answer. Whether NIF's achievement represents the final milestone of a cold war era investment or the starting point for a new generation of fusion facilities depends on decisions the governments have not yet made. The proof exists. what to build next does not.

[00:48:28]>> So ultimately, yeah, it proved that fusion ignition is definitely achievable and inertial confinement is viable, but that does not mean it's economically viable or power generation viable. It's just a completely different engineering problem to work with. Thanks so much for the recommendation and thanks so much for watching.