Nuclear Fusion Breakthrough Panel Discussion Transcript

Thank you all for staying on for this very exciting panel. I’m going to introduce the participants and then hand it over to our expert moderator. So these are the rock stars from Livermore who… Are going to tell you everything you ever wanted to know about what happened last Monday and beyond. First we have Alex Zylstra, who is the principal experimentalist for this experiment. Annie Kritcher, the principal designer and team lead for integrated modeling. Then we have Jean-Michel Di Nicola, who’s the chief engineer for the NIF laser system, Michael Staterman, target fabrication program manager, and then art pack, the team lead for stagnation science and the lead for diagnostics on this experiment. And Tammy Ma, who’s the lead for Lawrence Livermore’s Inertial Fusion Energy institutional Initiative. And now to moderate the panel, I’m going to hand the mic over to Mark Herman, who’s our program director for weapons physics and design and longtime fusion guy.

All right. Well thanks. We’re going to talk some more about today’s exciting announcement, an inertial confinement fusion experiment that yielded more energy in fusion than we input with the laser. We got out 3.15 megajoules, we put in 2.05 megajoules in the laser. That’s never been done before in any fusion laboratory anywhere in the world, so it’s super exciting. And this team and the many, many hundreds of colleagues back at Lawrence Livermore are really pumped about that. I want to be clear, ultimately this experiment, as Kim mentioned, drew about 300 megajoules from the grid in order to fire the laser. The laser wasn’t designed to be efficient. The laser was designed to give us as much juice as possible to make this incredible conditions happen basically in the laboratory. So there are many, many steps that would have to be made in order to get to inertial fusion as an energy source.

(02:27)
So in this experiment we use the world’s most energetic laser, the National Ignition Facility, to create x-rays that cause a tiny capsule to implode and create a very hot, very high pressure plasma. And that plasma wants to immediately lose its energy, it wants to blow apart, it wants to radiate. It’s looking for ways to cool down. But the fusion reactions are depositing heat in that plasma, causing it to heat up. So there’s a race between heating and cooling. And if the plasma gets a little bit hotter, the fusion reaction rate goes up, creating even more fusions, which gets even more heating. So the question is, can we win the race? And for many, many decades, we lost the race. We got more cooling out than we got the heating up. So we didn’t get to this ignition event. But last Monday, that all changed and we were able to win the race and get very significant heating of the fusion plasma up to very high temperatures, which resulted in that yield of 3.15 megajoules.

(03:48)
We study fusion ignition to keep our nuclear deterrent safe, secure, effective, and reliable, and do so without the need for further underground nuclear weapons testing. Fusion ignition is a key process in our thermonuclear weapons, and in addition, the very extreme environments created when the fusion plasma ignites enables testing that ensures we can maintain and modernize our nuclear deterrent. We hope that someday, and we’ll talk some more about this, fusion could provide a base load carbon-free source of energy to power our planet. So I’m very delighted to be on stage with several representatives of the incredible, incredible team that made this possible, and now we’re going to hear more from our panel of experts. And we’re going to start with Alex.

Good morning. My name is Alex Zylstra. I was the principal experimentalist for the shot that we’re discussing and the campaign that led up to it. So I’m here on behalf of the experimental team. As Mark said, NIF is the most energetic laser facility in the world, is where we conduct these experiments. We take the laser energy and convert it into x-rays, so we need to decide how that process will work, use those x-rays to compress the capsule that contains the fuel. If we can compress the fuel by a factor of more than 10,000, it reaches densities, temperatures, and pressures that are higher than the center of the sun, and that’s what’s required for our approach to fusion to work. So the job of the experimental team is to put all of the pieces together and ensure that nothing goes wrong.

(05:23)
That actually starts months earlier for a shot. We begin working with the design team to decide on a concept that we want to test. Once that’s decided, then we work with our colleagues in target fabrication and the laser science in order to bring those pieces of the experiment in. And then a few days before the shot it’s handed off to the operations crew in the facility. We work with them to actually execute the experiment because they’re the ones who are actually hands-on in the facility making this a reality. The shot goes off, it takes only a few billionths of a second, so we need an exquisite suite of diagnostics to measure what happened. And with those measurements, we can then increase our understanding of the particular experiment.

(06:06)
I want to emphasize that each experiment we do is building on 60 years of work in this field and more than a decade on NIF itself. So we use that previous understanding in order to field design or experiments. Each experiment then contributes to the overall knowledge, both these main fusion ignition shots and also experiments that we do with specialized configurations and specialized diagnostics, all building the understanding that allows us to make progress. All that work led up to a moment just after 1:00 AM last Monday when we took a shot, and we started to see… Hopefully that’s not representative, but we took a shot just after 1:00 AM last Monday, and as the data started to come in, we saw the first indications that we had produced more fusion energy than the laser input. So I’m very excited to be here on behalf of the experimental team and with my colleagues on behalf of the entire NIF team to discuss it with you.

So good morning. I’m Annie Kritcher. I’m the principal designer for the experiment and the campaign and also team lead for integrated modeling of these experiments, and I’m here to represent the entire design team. So the role of a designer is to define the input conditions to the experiment to achieve the desired plasma conditions, and that includes the target geometry, dimensions, materials, et cetera, as well as the laser pulse. So we don’t just smack the target with all of the laser energy all at once. We define very specific powers at very specific times to achieve the desired conditions. And we do this using a variety of tools, including complex plasma physics simulations, analytical models, and we benchmark them against experimental data. So the ultimate goal, as Alex mentioned, is to create a design that can reach the extreme conditions required for fusion ignition on a NIF, and in doing so, we were able to reach pressures more than two times the center of the sun on the last experiment and about 150 million degrees. And this requires a great deal of finesse and design optimization to reach these conditions as well as continual learning over the many years of designs and results that we’ve collected at NIF. And specifically, there were two design changes that were made leading up to the most recent result that fed into it. So this most recent result is part of a new campaign or a new effort to make use of the new laser capability that JM will talk about next.

(08:50)
And that is making the capsule that holds the fusion fuel a little bit thicker. And that does two things. That gives us more margin for achieving ignition when we have non-optimal fielding conditions, as well as it lets us burn up more of the DT fusion fuel. So the first experiment in this campaign was performed in September, and the experiment we’re talking about today made intentional design changes from the September shot also to improve the symmetry of the implosion. And with that, I’m happy to be here and to take questions and very excited about our team’s work on this.

All right, good morning. I’m Jean-Michel Di Nicola, chief engineer for NIF laser systems. I’m representing the team of technician, engineer, and scientists who design, build, operate, and improve this unique laser facility. Together our team is standing on the shoulder of multiple generation of optical material and laser physicist who design and optimized ever increasing performance in terms of the laser delivery, following the seminal paper from John Nichols, Livermore basically laid down the concept for nuclear fusion using lasers. However, we also want to acknowledge numerous collaborators, other national labs, university, industry, as well as international partners, and of course continued support from DOE and NSA in Congress for this mission so important for national security.

(10:26)
Thanks to outstanding science and technology and our team, Livermore with collaborator, made the impossible possible. The NIF laser is the largest laser in the world. It has the size of three football fields and it delivers an energy in excess of 2 million joules, with a peak power of 500 trillion watts. For a very short amount of time, we are talking only about a few billionth of a second, it exceeds the entire US power grid. Achieving ignition, however requires more than brute force. As Annie described, there are extremely fine tunings that needs to be performed to match the condition for ignition. Precision has been our focus for the last few years and we have been delivering more symmetric implosion and more reproducible experiments.

(11:24)
In addition, thanks to the continued investment from the nation, we have been able to deliver 8% more energy on the experiments last week compared to the one last year. It goes along with additional efforts that will be brought during the summer, and in next summer, we’ll be able to design experiments and field shots with additional 8% of laser energy, providing more margins for ignition. In the future with a sustainment and upgrade investment, the NIF laser could produce even higher energies and power and give the promise to larger target gains. That would be enabling additional extreme conditions for science-based stockpile stewardship program and maintain the United States of America leadership in this field. I’ll pass it over to Michael.

Morning. I am Michael Staterman. I am the program manager for target fabrication at Livermore. I’m here to represent a team of over a hundred members that is responsible for making almost all targets shot on NIF, and that comprises members of General Atomics, of Diamond Materials in Germany, and of Lawrence Livermore. The most important component that we deliver for fusion targets is the fuel capsule. Fuel capsule is a BB point sized shell made of diamond, and that needs to be as perfect as possible. We’ve been working over the last 16 years on continuously improving the quality of these shells to get to the state where we are today, and in turn, that effort has been based on decades of prior capsule development activities that have been done at Livermore and elsewhere. Today’s shells are almost perfectly round. They’re a hundred times smoother than a mirror, and they have a tiny tube attached to them that’s about a 50th, the diameter of a hair through which the fuel is filled into the shell.

(13:18)
As you can imagine, perfection is really hard, so we’ve yet to get there. We still have tiny flaws on our shells smaller than a bacteria that could be, for example, pits on the surface or holes in the wall. And despite their small size, these flaws still have the potential to affect the experiment. So not only do we make the shells, but we also characterize the shells and share the results with our colleagues so that we can select the best shell for each experiment. How strongly an experiment is affected by the flaws depends on the design input that Annie was talking about, and it looks like the result from Monday is a more robust design that is less affected.

This is very encouraging for us because we know that the shell that we shot had flaws in it. And this gives us confidence that we can make shells of equal quality or even better quality in the future, that we’ll be able to reproduce this experiment or even improve on it. So with that, I’ll pass it on to Art.

Thanks. My name is Arthur Pak. I’m the team lead for Stagnation Science, which is really focused on trying to understand how to create the conditions to achieve ignition at the NIF. So I’m responsible for coordinating the portfolio of integrated experiments, so [inaudible 00:14:38] ignition as well as overseeing the analysis groups that take these observations we get from the experiments and try to understand the conditions of the fusion plasma. And this has been really critically important to understand these conditions, as I’ll talk about. And it’s really been enabled by sort of state-of-the-art diagnostics, optical x-ray nuclear diagnostics that have been developed over decades of work by an exceptional team of engineers and scientists from Lawrence Livermore and from other national laboratories in the US, the UK and France, in addition with major contributions from our domestic and international academic and industry partners.

(15:18)
So for each experiment, we field over 50 scientific measurements to diagnose key quantities of the reacting fusion plasma such as the temperature, the fusion energy yield, the duration over which this occurs. And to give you a sense of scale and just how remarkable these diagnostics are, the plasma we’re trying to measure is a 10th of a millimeter in diameter, so that’s the thickness of a strand or two of human hair. As was mentioned, the temperatures of these plasmas are over 100 million degrees, so that’s 10 times hotter than the center of the sun. And the entire reactions occur in a fraction of a nanosecond. So that’s about a billion times faster than you can blink your eye.

(16:02)
And then as Alex mentioned, we’re only able to get these great measurements due to this sort of tireless work of technicians and operators who field them at the facility and allow us to get these amazing observations to help us understand what’s going on. And these observations have been really critical for our progress. So they’ve helped to identify, quantify, and mitigate degradations or loss mechanisms, which have impeded our progress. They allow us to test hypotheses and design changes to understand what are the sensitivities of the system. And so I’m here today on behalf of these diagnostic groups to help answer any questions and talk about what we can learn from these experiments.

Good morning. My name is Tammy Ma, and I am the lead for the Lawrence Livermore Institutional Initiative in Inertial Fusion Energy, or what we call IFE. Developing an economically attractive approach to fusion energy is a grand scientific and engineering challenge. Without a doubt, it will be a monumental undertaking. However, the potential benefits are enormous; clean, carbon free, abundance, reliable energy capable of meeting the world’s energy demands, and furthermore, providing for the energy sovereignty and energy security of the US.

(17:32)
As the secretary mentioned earlier today, this spring, the White House hosted a summit announcing a bold [inaudible 00:17:40] vision for fusion energy, building on great advances we had in both inertial and magnetic confinement fusion. These recent results on the NIF are the first time in a laboratory anywhere on earth we were able to demonstrate more energy coming out of a fusion reaction than was put in. This lays the groundwork. It demonstrates the basic scientific feasibility of inertial fusion energy and sets the roadmap for us to move forward to even higher gains and towards fusion pilot plants in the coming decades.

(18:15)
The Department of Energy Office of Fusion Energy Sciences recently commissioned a basic research needs report in inertial fusion energy. This report will help lay out the framework for a new IFE program here in the US, and that report should be coming out imminently. Such a program will inevitably require participation from across the community, both the public sector but the private sector as well. So of course, new fusion startup companies, their investors, national labs, universities, academia, public utilities, and more. We look forward to working with the Department of Energy to leverage and capitalize on these great results for a fusion energy future. The time is now. And I will hand it back to Mark.

Thanks, Tammy. So we’re going to ask each panelist a question here and then we will open it up to the floor. Annie, this experiment improved on an experiment, as you mentioned, that we had done in September of 2022, and it was a modified design from the one in August of 2021. Can you explain how you used physics insights and simulations to make changes in the design for this recent experiment?

Sure. So we do rely on our detailed radiation hydrodynamic simulations to design many aspects of our implosions. However, we know that there’s areas of parameter space where we might not be as predictive. So we have developed experimental playbooks and analytical models, and our design team used a combination of both to make the adjustment from September, going into this latest experiment. There was two sort of different flavors of symmetry adjustments that were made. One is during sort of the second half of the laser time history and one is during the first half of the laser time history. During the second half of the laser time history, we transferred more energy between laser beams to control the symmetry. So that’s actually quite a useful tool, and if you haven’t heard about it, it’s really awesome. You can move energy between beams and control symmetry that way.

(20:28)
In doing so, you have to go back and readjust the symmetry during the first half of the laser pulse. And we did that, but we also did that making an additional adjustment with improved models based on data that were collected just in the last few months. So in that piece, we really do rely on our models to benchmark against tuning data and then extrapolate out to the design space. And just for perspective, we’re talking about compressing something the size of a basketball down to the size of a pea, controlling that compression symmetry to about 1% in drive and also, the final symmetry to just a few percent. So it’s quite a challenging problem and we have great tools to get there.

Thanks, Annie. Jean-Michel, for many years, the maximum laser energy on NIF was 1.9 megajoules, but in this experiment, you turned the laser up to 2.05 megajoules. How did you and your team make sure that this would be safe to operate the laser at these higher laser energies?

Thank you, Mark. First of all, I would like to point out that the National Ignition Facility is one of the first generation of the inertial confinement fusion to meet and exceed its requirement. So it’s by itself quite a grand challenge to do that because those laser architectures are extremely complex and are a marvel of engineering and laser physics and non-linear optics. Second, since the commissioning of the National Ignition Facility, we have continuously increased the energy and power over a few years to make sure that we were obtaining the regime where ignition experiment could take place in favorable conditions.

(22:15)
In parallel, there was a tremendous amount of investment in terms of optics, science and technology as well as laser physics to have optics that would be higher performance compared to the previous generation. And we gained orders of magnitude compared to previous generations. No questions about it. We perform also a lot of work trying to control the shape of the laser in space and time to make sure we can transmit the maximum throughput through the optics with maximum reliability.

(22:46)
Last but not least, we also figured out that some of the debris are migrating back into the laser from the target experiment and ignition. When it takes place, debris are transported back near the optics. And finding and mitigating those debris sources was also a key instrument. So very minor modification to the laser architecture, and we published a paper in 2019 in Nuclear Fusion, if you want more detail about that.

Thanks Jean. Alex, many, many elements have to come together to enable a successful NIF experiment. We need a good target, a cryogenic ice layer, we need the lasers to be specified and we need to make sure the diagnostics are going to be taking pictures at the right time. So how did you prepare for this experiment knowing it had the potential to be a really exciting shot?

Yeah, so one of the things that’s so challenging with these experiments is that any one thing going wrong can be enough to prevent ignition, so everything has to be right, and so we really have to sweat the small stuff. To kind of give you a sense for that, Annie discussed our tuning of the symmetry. We had a debate over a laser setting equivalent to five trillionths of a meter going into this experiment. We had a discussion with the laser science team over a timing discrepancies of 25 trillionths of a second. With the target, each target we look at all of the flaws that Michael was describing that are the size of a bacteria to decide if they’re acceptable or not. They then grow a cryogenic DT ice layer, which has the same sorts of requirements. Bacteria size defects can be problematic. And then we’re setting up the diagnostics where small timing errors, a billionth of a second, would be an eternity for us in this experiment. We’re trying to image something that’s the brightest thing on earth for a hundred trillionths of a second. And so getting all that right is a team effort going through an intense review process.

Thanks. Art, how do we know this experiment passed the threshold of ignition and got more fusion energy out than laser energy in for the first time in history?

Yeah, thanks. So in these experiments, we use deuterium and tritium as our fusion fuel, which are isotopes of hydrogen. And when these two ions fuse, they release a helium ion, or an alpha particle, and a neutron, and they’re released with very well known energies. Now, the alpha particle, it stays in the plasma, further heating it, leading to more fusion reactions, while the neutrons largely escape. And so if we can measure the number of escaping neutrons, then we know how many reactions took place, and then we just multiply the number of reactions by the energy released for each fusion, and that’s how we measure the fusion energy. So in this experiment, to get a target gain one, we used 2.05 megajoules of laser energy. So to exceed one, we need to make at least that much fusion energy.

(25:47)
So the way we measure the number of neutrons, we do that multiple ways using independent diagnostics. But one of the methods we use is to place a high purity metal sample close to the reactions. That gets irradiated by the escaping neutrons and becomes radioactive. So the unstable states in this foil will then decay, and because we know very well the rate of activation and decay, we can then measure how many neutrons went through that foil, what the total number of neutrons were and how much fusion energy was released.

(26:23)
Have other methods. So another independent method is to turn the escaping neutrons into a charged particle spectrum, disperse that using a magnetic spectrometer and sweep that onto a piece of film and so we can actually measure the number of incident neutrons directly. So all these different methods give us very, very high confidence that we produced 3.15 megajoules of fusion energy, which corresponds to a gain of roughly one and a half. So that’s how we know.

Thanks, Art. So Michael, you talked about the targets. They’re incredible marvels of engineering and manufacturing. Can you briefly describe how you’ve been working to actually make them even better, the thin spherical diamond shells that contain the fusion fuel?

Sure, Mark. So the main problem that we have right now with getting perfect capsules are small flaws. And so to be able to improve on the flaws, the first thing that we have to be able to do is actually see them, measure them, count them, and quantify how many of the flaws are on a shell. The primary tool that we use today to characterize a shell is x-ray tomography. X-ray tomography generates a lot of data, and in the past, it has required trained experts a lot of time to go through these images and basically find the needle in the haystack. So now what we’re working on is software that helps us in this process and hopefully will be more accurate than a human in the longer run that will help us quantify exactly to see how many flaws there are, how big they are, and whether when we’re making a change

… to the deposition, to the fabrication process, where that actually improves the number of flaws in their size. So that’s the characterization part. On the actual improvement of the material, we are working very closely with our collaborators in Germany at Dye Materials to look at the fabrication conditions. We’re working as a team using the other’s capabilities as a reference to our own so that we can isolate the problem to machine, operation or surroundings. And this process has been very fruitful for narrowing down a set of conditions that has allowed us to improve the target beyond where we are today.

Thanks, Michael. Okay, Tammy, so to power the laser for this experiment, to get that 2.05 megajoules, we had to pull about 300 megajoules from the grid. And then we got out 3.15 megajoules of fusion energy, right? So why do we think it’s possible to turn this into an energy source in the future?

That’s a great question. Thank you, Mark. The NIF is a scientific demonstration facility. And when we built the NIF, the considerations that you have to build that facility are a little bit different than if you’re going to build a fusion power plan. So for example, maximizing the efficiency of every single component is not necessarily the most important thing. And so for example, the wall plug efficiency of the laser was not necessarily a very important requirement for this science facility.

(29:33)
However, as Kim alluded to, the NIF is now over 20 years old. The technology inside the NIF is ’80s, ’90s technology, and things have progressed quite a lot since then. We have new laser architectures that cannot not only run at rep rate, but are far more efficient up to something around 20% wall plug efficiency. There’s also been enormous advancements in target fabrication, new materials, computation and simulations, the application of machine learning. So it’s a really exciting time because we get to incorporate all of these emerging technologies with this new scientific understanding that we’ve developed to really push towards inertial fusion energy.

(30:16)
And I’ll also note the perseverance it took us to get here. As has been mentioned, it’s been 60 years since Ignition was first dreamed of using lasers. And it’s really a testament to the perseverance, dedication of the folks that made this happen. It also means we have the perseverance to get to fusion energy on the grid.

Thanks, Tammy, and thanks, panel. And then now we’re going to open it up to questions.

First off, I think, a round of applause for our panelists [inaudible 00:30:53]. Testing. All right. Hi, everybody. I’m NNSA public affairs deputy director, Shayela Hassan. I want to thank those of you who stayed for this more technical discussion with the folks that actually made this happen. It’s a very exciting day for NNSA. So we’re going to start the Q$A portion, we have a little bit more time, with media, so if there are any media folks who came late and need to move up, this is where we have media seated just to make sure you get your questions in. And my colleague Jeremy will be there on the other side to get a mic if you’re a little too far and feel like you need to reach. So if you could please state your name and your outlet before your question, otherwise I will remind you.

Hey, it’s Ari Natter from Bloomberg. What was the cost of the targets you’re using? And also, what fraction of the fusion fuel was actually converted to energy during the shot? Thank you.

You want to take the… I’m happy to jump in, Michael, you want to.. I think the way to think of the targets is how long does it take for someone to build them, right? Because the predominant cost is the labor of the people, and so they take months to build, right?

So the components that comprise a target take a variety of different times to make. The actual assembly of the target only takes about two weeks. To make the fuel capsule, however, takes seven months from the inception of what you want to the delivery of the capsule. The targets that we are using today are designed for science, right? They’re designed to be flexible so that we can reconfigure them as we want to learn different things. And they’re designed to be able to see what’s going on. So these targets are not designed for fusion energy.

Is there a dollar amount though, a monetary figure?

No, I don’t think… It’s kind of like we’re sized, we have a staff that can produce the targets we need per year on NIF, and so it’s really about a workforce more than a dollars per target. The components, the actual amount of gold or carbon or whatever are actually really, really cheap. I mean, the actual cost of the materials in the targets is next to nothing. It’s all the labor that goes into it that makes it time consuming.

To answer the second part of your question, it was about 4% of the DT that was burned on the shot.

All right, thank you. And before we go back to media, we have a question from our administrator, Jill Ruby.

Well, thanks. I often talk about how special the National Laboratories are because we do team science, and there’s no better example of that than what we just heard from you. But I’m guessing maybe you didn’t come expecting to do exactly what just happened, so I’d be very interested in just short stories about why’d you come and why do you stay at the lab? Thank you.

I think I can give a quick answer to that. So I actually toured NIF while it was under construction as an undergraduate. And so seeing the scale of what was possible at the National Laboratories is what drew me in, as well as being able to see the ability to do great science that has an actual impact on national security and energy applications. So I hope that anyone watching who’s inspired by this current moment comes to work with us in the future and take this to the next level.

And I actually have a very similar story, so you might notice a trend here. I came as an undergraduate in 2004, and I saw NIF and I thought, “This is a once in a lifetime opportunity, to be part of this large interdisciplinary teams working together, grand challenge on problems that move humanity forward and are so important for the nation.” Also, it’s just fun to tell your friends you shoot big losers. And my dad really liked it as well, so…

All right, we’ve got one question over here.

Hi. Michael Greshko, National Geographic. Thank you all for being here, and congratulations. What time on December 5th did the shot occur? And for each of you, in the minutes and hours after that, how did you first learn that this shot was special?

I can start with that. So the shot went off, I believe, at 1:03 AM last Monday. And so I was up to look at the data. And as the data started to come in, we started to see some indications that this had happened. And one of the first things I did was call one of the diagnostic experts to double check the data, and we kind of went from there.

So I had vivid dreams of all possible outcomes from the shot. This always happens before a shot from complete success to utter failure. And then I woke up. Thankfully, Alex had sent me a message, so by the time I woke up, I saw that it wasn’t a failure. And then, of course, I start texting you right away. Yeah, so just an amazing feeling. You see one diagnostic and you think, “Well, maybe that’s not real.” And then you start to see more and more diagnostics rolling in pointing to the same thing, and it’s just a great feeling.

So for the laser team, the work started earlier in the night because we had to check and make sure during the preparation stages of the experiment we’re on track to a successful delivery. So numerous phone calls, emails, checking of the results real time, giving the green light to proceed. And then the relief when we see the result and the laser didn’t screw up the experiment, that was relief number one. And then we start to receive the text with potential yields, and that’s a eureka and hoorah moment.

Well, for target fabrication, our work usually ends somewhere about a week before the shot, when we deliver the target to the facility. We had high hopes for this one, so I was checking email as soon as I woke up to see how the shot did and was very excited about the numbers that I saw.

So as Annie mentioned, we’re building off of prior experiments, and we had good reason to be optimistic for this experiment, so I was very keyed-in and curious. But as it started going later and later, I got to get the kids up in the morning, so I was like, “Okay, I’m going to go to bed.” But I woke up at 3:00 in the morning just out of sleep and I went to check my email, and I saw the number. And of course, I had done the math before the experiment so I knew what number I wanted to see for getting greater than one. And so the preliminary data came out, and I was like, “Holy Smokes.” So that’s how I… And then I went back to bed or tried to go back to bed, which was…

I actually didn’t find out until the next morning. I was at the airport at SFO about to board a plane to come to DC, actually, to attend the Fusion Power Associates meeting, which is an annual meeting where we have fusion leaders coming together to really work together on how to push forward fusion energy. But anyways, I got a call from my boss saying, “I think we got ignition.” And I burst into tears. And I was jumping up and down in the waiting area, the crazy person. And yeah, the tears were streaming down my face. After all these years, every time I walk into the National Ignition Facility, I still get goosebumps. And so it is a wonderful place to work and I’m so proud of this team.

Thank you all.

Hi. James Rearden, Science News Magazine. Since the ignition, the prior one that was an important event, have there been events like this that have been failures, that have been disappointments? And why on earth would you do it at 1:00 AM? Is there a technical reason?

Yeah. So yeah, following last year’s experiment, we tried to repeat that experiment three or four times. And what we learned from those experiments was that this design was still very sensitive to the target defects that were present. And so we were able to quantify the impact and understand the origin of these defects. And so knowing that, we went for improved designs to make ourselves more robust, we used that knowledge to try to pick the best capsules to minimize the impact. And so that’s what led us to this event here.

(39:52)
And why 1:00 AM? So NIF runs 24/7, so to grow these ice layers, it’s a multi-day process. And it’s not fully deterministic, so you’ll try, it will fail, you’ll try to grow again. And so it just happens that the shot cycle goes that way, and we take the shot when we’re ready to take the shot, so…

James Osborne, Houston Chronicle. There’s two fusion technologies, magnetic confinement and inertial. Can you sort of explain where your breakthrough… Does that sort of put inertial at the forefront now? Is that the technology society is going to pursue, or does it have implications for magnetic that might allow that to continue? I think it was implied earlier that magnetic had sort of seen more advances of late. Obviously, though, this is the big breakthrough. So if somebody could sort of explain that. Thank you.

Yeah, absolutely. There’s definitely pros and cons for each different approach, and there’s different technology developments that need to occur. And so both magnetic fusion and inertial have made great advances in the past couple of years. There’s also been enormous private sector investment, actually more on the magnetic side than the inertial side in recent years. I think what we’ve been able to demonstrate on the NIF is a burning plasma, we’ve gotten gain. However, like Kim alluded to, we’re a little bit farther behind in some of the technology developments because that’s just not what we’ve been focused on the past few years.

(41:42)
That being said, there’s a lot of commonalities between the two where we can learn from each other. There’s burning plasma physics, material science, reactor engineering, and we’re very supportive of each other in this community. A win for either inertial

… inertial or magnetic confinement is a win for all of us. And we really just want to see fusion energy happen. And the point is though that there are different technological risks for both of them. Right now, fusion is so incredibly impactful and important for humankind that what we really want to do is maximize the potential pathways to success. So we want to carry these different approaches and see what will really work.

Peter Bear with India News. Could you give us your vision of what a commercial fusion plant would look like using this technology? Jump ahead in a couple of decades, what might this production of energy look like?

So right now, there are a number of different approaches to target physics that could actually get us to high gain, high margin. All of those have slightly different drivers. So for what we’re doing on the NIF, we use high energy lasers, but there are designs using heavy ions, pulse power, et cetera, et cetera. And there’s been a number of different integrated studies to try to pull all of this together. Really where we are right now is at a divergent point. We’ve been very lucky to be able to leverage the work that the NNSA has done for inertial confinement fusion, but if we want to get serious about IFE, we are at a point where we need to invest in those technologies. We need to figure out what that integrated system looks like, because the target, like Michael said, is complex. It takes a long time to build.

(43:56)
And what we need for a power plant, it has to be simple. It has to be high volume. It needs to be robust. And there are trade-offs. If you can get to higher gain with your target, you can turn down your laser energy or your laser efficiency a little bit. There are decisions to be made about the materials that you would use for your reactor. Those would impact your design as well. So what we’re doing now with the Department of Energy is actually embarking on doing these integrated system studies, again with our best known information about technology as it’s evolved today and to figure out where the biggest gaps are, where we need to invest, where we need to buy down risk. And that will be an ongoing activity for quite a few years. But with this bold decadal vision, we are really trying to accelerate and put together these designs to see what is the most viable, feasible design and all come together to work on it.

Hi. Dave Nyczepir, FedScoop. The director mentioned earlier that machine learning played a key role in sort of the in-between, between reaching the threshold to ignition and then this latest test. So I’m curious, how have advances in machine learning and computing helped you in your work? And how do you anticipate they might help you moving forward?

Yeah, I can take that question. So we have made quite a bit of advances in our machine learning models to kind of tie together our complex radiation hydrodynamic simulations with the experimental data and learning. Specifically going between the August 8th, 2021 experiment and this latest experiment, we used it more for the predictability phase versus the design phase for this change. So for this change, we sort of used our traditional design methods of not running thousands of simulations, but then we fed the design to the team, the CogSim machine learning team, and have them do an analysis of the design. And they did find that it had a higher probability of achieving gain of one. So with the CogSim’s models, we’re able to look at if the design is more robust as some of the issues that we had been having last year.

Thanks, Annie. Now I believe we have time for one or two more questions, so before I go back to Nat Geo.

David Crandall, retired from Department of Energy, where I spent 30 years and most of it related to fusion and most of it related to NIF. The target. The target for 08/08/21, the physics ignition shot, was deemed to be the most pure target you ever shot. How did this one compare?

Yes. As you mentioned, the target for 21/08/08 was probably the most pristine shell that we’ve ever had. It even compared favorably to the other shells in the batch, as we learned later on as we re-analyzed the data for all of those shells. This target here had a substantial number of flaws compared to that one, specifically, at higher… It has tungsten inclusions in a large number.

All right. Time for our final question.

James Riordon, Science News magazine. So this experiment is specifically interested in the search for ignition, but most of your research is focused on stockpile stewardship, is it not? And in that case, what fraction is dedicated to this sort of research? What fraction is dedicated to stockpile stewardship? And do these sorts of shots have any important information in them for stockpile stewardship?

Sure. I’m happy to take that one. So the ignition work we’re doing is for stockpile stewardship. Our thermonuclear weapons have fusion ignition that takes place in our weapons. And so studying fusion ignition is something we do to support the Stockpile Stewardship Program. In addition, fusion ignition creates these very extreme environments that we have no other way to access on Earth. And in fact, in this experiment, for the first time ever, we were able to put some samples of materials that are important for our future stockpile modernization efforts that are going on at Lawrence Livermore today in very close to this intense neutron burst and then see how did they respond to that intense neutron burst. So we’re actually using the output from these really cool science experiments, which we’re also trying to understand, but we’re also using it to actually test materials for stewardship applications.

(48:45)
But to your other question, roughly speaking about 15% of the experiments we do on NIF are indirect drive experiments of the type that this experiment was. We do another roughly 15% that are other types of fusion experiments, inertial confinement fusion, but using different approaches. And then the rest of them look at things like the behavior of materials at high pressures and getting data that’s important to use in our simulations for our nuclear weapons, understanding the behavior of radiation in very complex environments and geometries, understanding how hydrodynamic plasmas mix at very high temperatures. All of those things help us benchmark our simulation tools, learn new things about how matter behaves in these really extreme conditions and underpins the confidence we have in our deterrent.

Thank you. And we had a question over here.

Well, thank you. How many questions do I get to ask?

As many as you want. [inaudible 00:49:48].

No. This is really exciting. My understanding was that the experiment in 2021 was a little unexpected result, which was really good news, and then it was kind of hard to repeat that result. So was it more the target? And something that Annie brought up that I hadn’t been aware of is the time history of the laser pulses is also important. So first question is what’s more dominant, the time history or the target? And a question for Arthur, how many orders of magnitude of neutron flux do you have to monitor to get a full picture?

So I can answer the last question first perhaps. So as Alex pointed out, many things have to happen, have to go right basically for these experiments to really reach very, very high yields and gains. So we have to set up our diagnostics to capture that event, but we also have to set up the diagnostics to capture lower orders of magnitude so that if it doesn’t go right, we understand what happened, what was the failure mechanism, so that we can fix it in the future. And so right now, I think we have something like three orders or so of dynamic range that we can measure. So we have predictions of how we set up the diagnostics and we try to go above and below it to make sure we get good measurements. Does that answer your question?

Yeah.

Yeah?

[inaudible 00:51:13].

Yeah, So I can take a crack at the other one. So going into 21/08/08 or August 8th of 2021 experiment, we predicted about half the yield increase that we got. We got about eight times the yield increase and we predicted about four times. And the reason for that is because we had improved the capsule quality, like was mentioned earlier. So going into the pre-shot prediction, we’re assuming the same capsule quality with some other design changes that were being made. We did perform a set of experiments after that to try to piece apart which is quality and which is the other design changes. And it was pretty consistent with that sort of split between the two. Following that experiment, we haven’t been able to replicate the same capsule quality and that is a main driver, because these defects are extremely difficult to model and predict. It’s been a main driver in the performance, as well as the predictability. So then the thought was to try to design our way around some of these stringent requirements.

(52:19)
And what was the other part of your question?

[inaudible 00:52:21] the time history. [inaudible 00:52:22].

The time history. Yeah. Thanks. Yeah. Thank you. So the time history, there’s different forms of asymmetries and the one that I’m talking about, the change that was made between September and this most recent experiment, was a slightly different type. That’s an intrinsic symmetry correction. During the repeat experiments, we did have two experiments that had an anomalous laser deviation, which is a mode one. So that means, for example, if the laser delivered differently on top versus bottom, that would push the implosion in one direction. So something that’s not designed or expected that did impact two of those experiments. But ever since then, we’ve been quite good. And maybe JM can talk about the improvements made there, that that hasn’t been as big of an issue.

Yeah. I can elaborate if you like. We basically are in the process of modernizing what we call the front end of the laser, which is based on the fiber optics technology. The NIF laser was first commissioned at low level in 2001. So this part of the laser was literally 20 year old. And as you can imagine in telecom industry over 20 years, there has been many revolutions. And so we were able to catch up and capitalize on the latest technologies to improve this delivery.

All right. Time for our final question. I think we have one over here.

James Osborne, Houston Chronicle. Again, I think you’ve probably answered this, but maybe just for the layman among us. So if the shell quality wasn’t as clean on this experiment, what in fact made the difference where you were able to achieve this breakthrough this time? What do you attribute it to?

Yeah. So we shot basically the same capsule quality in September, so that part remained constant between the two experiments. And in September, we achieved about 1.22 megajoules. So being able to first achieve megajoule yields again with worse capsule quality than August 8th, 2021, that was the first stop, and that was attributed to the enhanced laser capability and the design change of the thicker target. And then moving between September and the most recent experiment, the only changes to the input conditions were to improve this intrinsic low modes asymmetry. So make the implosion more symmetric as it’s coming in, you can better couple your driver energy to the hot plasma.

All right. I’d like to thank everyone for coming and I’d like to thank Dr. Herman and the panel for taking time to share this amazing achievement with us after what must be a whirlwind week. One more round of applause, please. And with that, that concludes our NIF Press Conference event. Thank you for coming. [inaudible 00:55:29].