Apparent nuclear fusion power breakthrough -- net energy gain (breaking news 11 Dec 22)

The Financial Times article is paywalled; above is a tweet that sums it up.

If true, and if I understand it correctly, this could be very big news for energy. AFAICT, fusion really could be the end-all be-all for clean energy, if it actually works like proponents say it can.

I think this is the key paragraph:

The fusion reaction at the US government facility produced about 2.5 megajoules of energy, which was about 120 per cent of the 2.1 megajoules of energy in the lasers, the people with knowledge of the results said, adding that the data was still being analysed.

Paging @Stranger_On_A_Train

I too will wait for more analysis from those more knowledgeable but I think it’s reasonable to see the money and effort going into this research and believe that smart people believe in its feasibility.

The rewards are just so incredibly high.

A practical fusion reactor is now only a couple of decades away!

(again)

There was a gap of over a century before the Savery Steam Pump and the start of the Industrial Revolution. Technology progresses at its own pace - sometimes it crawls, sometimes it sprints. At some point all the right pieces will fall into place, and the world will change overnight.

With all due respect, we’re decades past due for retiring this joke.

Here’s a news article that is not paywalled:

Still doesn’t say much more. I’m suspicious. Nothing on the Sci-news, Science Daily, or Science News web sites I follow, nor on Scientific American.

Well, don’t pop the champaign cork yet. First of all, an yield gain factor of Qp~1.2 (indicating a 20% increase in yield relative to the input power), even if real (we’ll get to that in a moment) is not enough to sustain a nuclear fusion reaction. Any fusion process will have substantial losses just in how much energy can be recovered even before you get into converting it into electricity or usable thermal energy. A Qp>10 is generally assumed to be the minimum gain for a self-sustaining fusion reactor of any kind, so this is still almost an order of magnitude away from that threshold.

Second, I can’t read the Financial Times article but I’m assuming that this is either the Nuclear Ignition Facility (NIF) at Lawrence Livermore National Laboratories (LLNL) or the Z Pulsed Machine (ZPM) at Sandia National Laboratories (SNL). Although both are designed and used to research fusion conditions, neither is intended or capable of producing sustained power. The primary mission of the NIF is actually nuclear weapon research and sustainment, although most of the papers you see coming from it are on solar astrophysics, and ZPM is really designed to study high energy plasma dynamics. Both produce interesting physics and may be of some value to nuclear fusion power researchers but neither are a concept leading to net usable energy production. In both cases, the yield produced is so transient that it is often difficult to get an accurate measurement of yield, and if these are preliminary results I could easily expect a ~25% uncertainty on the estimated yield.

Assuming this is the NIF, this would be deuterium-tritium fusion (D-T). The products of this reaction would be an alpha particle (ionized helium) with a kinetic energy of 3.52 MeV, and a ‘fast’ neutron with a kinetic energy of 14.06 MeV. Recovering much of the energy of the alpha particle is pretty simple since it is a charged particle; you can just run it through an electrostatic grid, or through an induction loop, or whatever and use it to directly produce electricity. The neutron, however, is not only electrically neutral but unlike the thermal neutrons that are used to heat water in a boiling water fission reactor, these neutrons are moving very fast and will damage the materials which form the vacuum chamber necessary to allow the fusing plasma. There are proposed schemes to line the walls of such a chamber with materials that the neutrons could irradiate to produce fissile materials that could be used as fuel in fission reactors or 6Li to breed tritium (which you would actually want to do because natural tritium is vanishingly rare) but they would also have to withstand the temperatures and radiation conditions in the vacuum chamber without interfering with the plasma, which is something that the ITER project is actually intended to develop and test.

Setting aside whether this is really an overunity event and the difficulty of converting energetic fusion yields into useful electrical or thermal energy, we often see claims that a successful nuclear fusion power generator will solve all energy scarcity and climate problems when a supposed breakthrough occurs. In practice, this is unlikely. For one, most plausible nuclear fusion power production systems are gigantic, multi-billion dollar construction projects that take years—in the case of ITER, decades—to construct and bring on line, and end up being very prone to operational problems such as loss of vacuum, misalignment of magnetic coils (for magnetic confinement), difficulty in handling materials, et cetera. If a proposed solution costs US$10B per operational facility and requires thousands of highly trained engineers and technicians to operate and maintain it, it isn’t very scaleable or practical for any but the most advanced nations.

Even if some method of nuclear fusion were developed that was highly scaleable, i.e. Tony Stark’s “ARC reactor (referred to in fusion circles as “fusion in a tuna can”) it doesn’t really solve the distribution and conversion problems in general. The general assumption is that nuclear fusion would be used to produce electricity, which is great for nations that have expandable and robust energy distribution infrastructure; however, few if any nations could switch their heating, manufacturing, and transportation systems to be fully electrified even setting aside the costs of retrofitting buildings, developing new manufacturing methods, and deploying enormous fleets of battery electric vehicles or producing synthetic fuels for existing vehicles derived from electricity. And fuel for fusion is not free; deuterium is readily available in normal seawater just by fractional distillation, but at some cost, while tritium would have to be bred from 6Li as noted above, but this isotope is pretty rare and would require extraction from clay deposits or seawater at enormous expense.

In short, even if the Financial Times article is completely accurate (unlikely; if this were a verified event one would expect notes in Physical Review Letters, Journal of Fusion Energy, Nature, Science, et cetera) it doesn’t mean that practical nuclear fusion power generation is around the corner. At best it is achieving a technical milestone that is one of many gates necessary to get to the point that fusion power production is a practical means of offsetting current power generation methods, and is certainly not something we should be relying upon to ‘fix’ climate change or provide surplus global energy.

And you don’t have to take my word on it; physicist and skeptical science communicator Sabine Hossenfelder has a very salient short lecture on the topic:

And yet, it never goes out of style. Which tells you something about how we perennially underestimate the difficulty of producing sustained nuclear fusion in terrestrial conditions.

Stranger

Nah, it just tells us that the research funding is never sustained for as long as it’d need to be.

From the article above

  • the people with knowledge of the results said, adding that the data was still being analysed".

  • “However, the exact yield is still being determined and we can’t confirm that it is over the threshold at this time. That analysis is in process, so publishing the information…before that process is complete would be inaccurate.”

So basically, it’s the same news story we’ve gotten since forever.

Thank you!

I fully appreciate that the engineering issues and remaining technical issues are gigantic, and that pre-pre-pre-release information is notoriously wrong, but I’ll sit here with my “Lottery’s up to a billion, might as well enjoy buying one ticket and enjoy the what if’s” stage of unjustified hope.

I’m curious about this. A uranium fission reactor produces fast neutrons (not this fast, but still fast), which are then moderated to sustain the reaction and to extract some heat. Are 14 MeV neutrons less amenable to moderation to thermal energies than 2 MeV neutrons? Even if you needed more moderator, I’d think it would be simpler since you don’t need to reflect and shape the neutron flux back into the reactor.

I’ve seen this comment before but it isn’t just an issue of funding; every time a new innovation comes along that ‘promises’ to achieve a viable threshold, physicists discover that there are even more complex problems in sustaining confinement. In the ‘Fifties, the twisted toroidal ‘Stellarator’ would allow for confinement and minimize Bremsstrahlung losses but suffered from massive diffusion losses. In the late ‘Fifties and ‘Sixties, tokamak designs were thought to be a solution to the diffusion problem. ‘Seventies to early ‘Eighties, it was assumed that microwave heating of the plasma and the use of magnetic mirrors (what became known as the “Bumpy Torus”) would be able to reinforce the plasma, but that demonstrated scaling deficiencies long before it got anywhere near even a tiny fraction of unity. “High temperature” superconductors and very high power isomer lasers in development in the late ‘Eighties and ‘Nineties were thought to be means to magnetic and inertial confinement, and in fact the NIF was an outgrowth of the latter (although the NIF ended up using a more conventional glass fiber laser as the pump supply to a large array of focusing lasers). The current path of fusion tech startups is magnetic confinement high tesla near-room temperature superconductors controlled by a machine intelligence system that will somehow find the right combination of factors (Commonwealth Fusion) along with various inertial and electrostatic confinement systems of varying degrees of implausibility. (I think pretty much everyone has given up on muon-catalyzed fusion because of the seemingly intractable “alpha sticking problem”, which admittedly would require a retooling of electrodynamics but is my favorite approach from a standpoint of practical implementation.)

All of the funding in the world doesn’t make the fundamental physics issues go away, and we are only now getting to the point of having sufficient computational capability of being able to more or less directly simulate the complex electrodynamics of a charged plasma to a sufficient degree of fidelity to even understand how difficult the problem is. And I suspect the simulation tools and machine learning is just going to expose even more difficulties because that has been the path of fusion development to date.

To be clear, it is entirely possible to create a working fusion reactor. By scaling up a system you reduce the issues of Bremsstrahlung and thermal losses to the point that fusion is almost trivial, and in fact if you bring together ~1029 kg of hydrogen fusion will start to happen whether you want it to or not. But on the scale of practical fusion in terrestrial facilities, the difficulty of sustaining fusion conditions in a plasma are extremely difficult notwithstanding how you would actually extract useful energy from such a system.

With nuclear fission you simply suspend your ‘fuel’ (fissile elements) in water or some other medium which slows the neutrons to ‘thermal’ speeds; without the water the neutrons would actually be moving too fast, on average, to achieve a criticality (result in more than one fission reaction for every neutron released). Although the interactions between the neutrons and the water does produce some heating, most of the heat occurs within the fuel elements themselves and is conducted to the water through the cladding. Other nuclear fission systems such as molten salt or high temperature gas reactors work somewhat differently but on the same basic principle of carrying away heat from the fuel elements into a coolant loop where it does work by driving a turbine.

With fusion, the plasma has to exist in an otherwise evacuated chamber, and you actually don’t want to carry away the heat that sustains the plasma. The recover of energy has to be from the momentum of the products converted to electricity or thermal energy by…some process. There are designs that use the neutron yield to drive fission reactions (hybrid fission-fusion), essentially using the fusion as a neutron generator, but they haven’t proven to be very viable in an overall energy balance, and pure fusion systems would require having some kind of system to convert the 14.06 MeV of the neutrons into some useable energy or otherwise accept a baseline efficiency of 20% even before any losses. One of the biggest problems with fusion, though, is just the fact that the plasma is constantly trying to reject energy to come to an equilibrium, and the requirement to maintain the plasma at the necessary pressure and temperature conditions for even tiny fractions of a second is just physically problematic outside of the conditions of the core of a star. Even if you could capture all of the energy coming out of the reactor, converting it back into power to sustain the plasma at fusion conditions itself is a daunting challenge, and doing so on a practical scale that doesn’t require building a multi-billion dollar facility just may not be feasible.

Stranger

I thought this was announced over a year ago. Or is that something different?

Thank you. That clears things up for me.

They’re also not as much in a hurry to get a story out without proper research and vetting of sources. So their lack of immediate coverage isn’t much of a red flag in and of itself.

I think this is a different event. The NIF and any other high energy facility should be demonstrating regular advances on the way to its planned capability.

Actually, there isn’t that much of a ‘story’ to present. This is clearly popsci outlets (Financial Times isn’t specifically popsci but they do have a lot of focus on technical innovations) grasping onto a headline-making story even though it isn’t all that meaningful. You see such outlets printing statements such as “It’s believed fusion energy, the same reaction which powers the sun, could one day provide a limitless and cleaner alternative to fossil fuels,” which compacts so many misapprehensions into a single statement that it is hard to know where to start correcting it. There are legitimate reasons to be ‘excited’ about this from a technical standpoint albeit more from the simulation of nuclear weapon designs or thermonuclear processes in stellar physics than any practical achievements toward fusion power production.

Stranger

Gonna file this in my bulging “Norwegian Doctors Find Cancer Cure & Other News” folder until further notice…