TAE Technologies Targets Commercial Nuclear Fusion by 2030

TAE has produced a stable plasma at 50,000,000°C within a compact reactor and repeated this over hundreds of testing cycles. They have had over 25,000 fully integrated fusion reactor core experiments. The experiments were optimized with machine learning from a collaboration with Google. The current goal is a commercial fusion power plant by 2030.

They recently raised an additional $280 million investment. Total funding is over $880 million.

A larger-scale new reactor called ‘Copernicus’

The approximate cost of the Copernicus reactor is $200 million. TAE started construction and test runs start in 2023.

The Da Vinci device is a proposed successor device to Copernicus. It is a prototype commercially scalable fusion energy reactor designed to bridge between D-T and a p-11B fuel. Conditional on the success of Copernicus, it will be developed in the second half of the 2020s and will be designed to achieve a plasma of 3 billion°C, keep a proton-boron fuel stable and produce fusion energy.

Norman reactor is about 100 feet long. Copernicus is about 200 feet long and will have criss-cross like an asterisk (*). TAE has had schedule slippage but this is expected for ambitious new technology that has to solve many science problems and raise a lot of funding.

There are several companies working towards commercializing nuclear fusion. TAE Technologies is the best-funded of the non-government projects.

SOURCES – TAE, Engineering and technology
Written By Brian Wang, Nextbigfuture.com

46 thoughts on “TAE Technologies Targets Commercial Nuclear Fusion by 2030”

  1. I'm cool with making the thing 90 meters longer so that it doesn't accumulate neutron radiation.

  2. Hey great news everyone! Instead of being perpetually 20 years away, commercial fusion is now perpetually only 10 years away.

  3. They've been pretty clear about the fact that they don't want to do any neutronic fusion testing, because as soon as they do, the equipment becomes radioactive, which tremendously complicates their research.

    That said, I wouldn't be shocked if they could achieve engineering breakeven fairly quickly, by using the focus fusion device to produce neutrons that cause fission in a U238 liner. Every 1MEV of neutrons the fusion produced would cause about 200MEV of fission in the liner. 

    That sort of fusion/fission reactor could produce usable power, while breeding fuel for fission reactors, and it would be much more fail safe than a reactor using U235, as well as not requiring isotopic separation to run.

    But that's not their gig, obviously.

  4. Is there a figure of merit for this approach so that we could compare it to other approaches? A product of temperature, density and confinement time, or similar?

  5. Exactly. And he also seems to forget capex to reach said power. If you get the heat for nothing and the turbine is really cheap, then the whole system would be cheap. Which is counter to his claim.

  6. I think they're holding out hope of aneutronic fusion as bait for investors who don't know the physics. Then if they actually do achieve engineering breakeven using D-T, they'll be all, "Hey, let's commercialize on this, at least it works!"

    I find that more plausible than them not knowing the physics as well as a mechanical engineer in his 60's who just reads a lot.

    Alternatively they could have some crazy clever scheme for suppressing the Brem that I've never heard of, and are just keeping it a secret. But I'll be shocked if that's it, Brem is a pretty basic physics problem.

  7. I don't get it. So coal is uneconomical, why does that prove that turbines with 30% efficiency is uneconomical by themselves? There could be many reasons why coal would be uneconomical that are not related to the turbine efficiency.. Say, cost of transportation, cost of cleaning the exhaust, taking care of the ash.. and many more factors that I have not considered.

    Here is a source that estimates large turbines + generators at ~400 USD per kW of power, i.e. dirt cheap [1].

    Could you clarify your argument?


  8. No they use ion beams external to the plasma to hold the plasma. That's what the eight attachments are along the long shaft. TAE uses a FRC to produce plasma and then holds it in a stable configuration for tens of seconds while fusion occurs.

    I'm not clear how they harvest energy. LPP is relatively straightforward in that they harvest fusion energy as He ions move away from the anode and x-rays via the photoelectric effect.

    If you crunch the numbers there is also some reason to harvest waste heat. I recall that they produce ~2MJ/s of waste heat.

  9. Worth pointing out that LPP is a "mini bomb". Lots of energy in a very small plasmoid for a very short period of time.

  10. D-Li6 works best in thermonuclear "devices". Its small cross section is a killer. I do suspect that LPP's DPF could be a more generic fusion machine and I wish they had the budget to pursue D-D as a side project.

    He3 is definitely easier for Toks. Breeding it costs money, not that Tritium breeding is free just a convenient necessity with all the neutrons Toks make.

    pB is probably only workable as pulsed power (LPP, Z Machine, or laser fusion). Or unless you work at TAE and know something we don't know.

    Also we never hear anything about the Z machine. Or at least I don't. Then again none of this is remotely near my day job.

  11. Reply to myself: as far as I am aware TAW has a magnetic field in the same ballpark as Toks. Which is to say Brem is an issue. I think that they expect to have lower temperatures but longer confinement and that is how they get a usable Lawson criteria.

  12. Thanks for reminding me of that paper.

    On page 139, he relates calculating the size of plasma at plausible densities necessary to render the plasma optically 'thick', so that the em radiation could be retained. 6*10^20 meters!

    You might be able to build a p-B bomb, but not a magnetic confinement reactor.

    LLP gets around this by using a quantum effect to keep the electrons much lower in temperature than the ions. (Ions are where the temperature is useful!) But it requires a magnetic field so strong you'd normally only find it on a neutron star.

  13. Fast moving electrons hit slower boron nuclei and emit x-Ray radiation which cools the plasma fast. It’s why you can’t just make the plasma hotter to get more fusion.

    LPP relies on a quantum mechanical effect that happens in their gigagauss magnetic field to force a minimum electron speed. If it works great, if it fails no fusion for LPP but they get a Nobel. Win-win?

    Look up Bremstrallung radiation on Wikipedia.

  14. TAE doesn’t seem to discuss their approach much. They use neutral ion beams to confine and hold a hot plasma for long duration (order of a minute).

    But Brem losses should haunt them. Also not entirely clear how they harness energy from the fusion ions.

    Many aspects of their design elude me and I’ve been trying to follow them for years.

    I wish them all the best and their investors like them so what can I say.

  15. You seem to be forgetting energy density. It is evident that a turbine, even a fairly inefficient one, connected to a functioning fusion reactor would be economically viable. Besides that there are many industrial, agricultural, desalination and housing uses for waste heat.

  16. Todd Rider's thesis is generally thought to disprove the possibility of aneutronic fusion from plasmas in equilibrium, due to radiation loss.

    But Rider's appendix E goes through a bunch of methods that might get around his main result. Suppressing bremsstrahlung with magnetic fields was one of them. Some of the others look like they might apply to HB11's petawatt laser fusion. It doesn't seem like any would apply to TAE.

    (I'm no physicist though.)

  17. I feel like by the time we have fusion ready for power plants, we'll also have supercritical CO2 turbines.

  18. People like the P-B reaction because neither of the fuels can react in neutronic fusion, so you only have to worry about a rare side chain.  And you use the most common isotope of both elements.

    But the ignition temperature is almost 10 times higher than D-T, and the reaction cross section several hundred times smaller. And half the energy per reaction.

    D-He3 has an ignition temperature 4 times higher, and the reaction cross section is 55 times smaller. He3 is pretty rare, but easy enough to breed if there were a demand for it. And it gives you slightly more energy than D-T. But you'd be getting a substantial amount of D-D fusion going on.

    D-Li6 ignition temperature is 5 times higher, and the cross section is 850 times smaller. And Li6 isn't the common isotope, but light elements are relatively easy to separate, and you get half again as much energy as D-T. But, again, D-D fusion, too.

    P-B is not the first aneutronic reaction I'd go for, that's for sure. In fact, it's close to the last.

  19. I generally agree and all I can say is that TAE seems to fall back on the idea that they can maintain their plasma conditions for long durations.

    It would be great if TAE would do a bit of explaining how this all works out. They generally propose scaling up to pB despite pB being a different beast than DT.

  20. Comparing neutron flux of D-T to neutron flux of pB is like saying both Chicago and Singapore have a murder problem.

  21. True also comparing He side reaction neutrons to D-T which must produce neutrons to make power is apples to oranges. Neutron flux between the two is off by many orders of magnitude.

  22. Beryllium doesn't neutron activate worth a darn. It mostly reflects neutrons, but when it does absorb a neutron, it then emits TWO neutrons, and promptly fissions into a couple of alpha particles. There's a rare side chain that can happen if the neutron is very low energy, that gets you a bit of tritium, though.

    Any neutrons hitting the electrode are going to eventually end up elsewhere.

  23. Don’t disagree that Tok fusion is uneconomical.

    Reliability is nice though, not that I expect Toks to be as reliable as gen 2-4 fission.

  24. LPP’s neutrons happen as a side reaction which TAE would also have. At the brief moment when He fuses and makes a neutron it is surrounded by a dense plasma of H, B and trace He. All things that slow neutrons.

    Does this work to adequately reduce neutron activation of the Beryllium anode? I don’t know.

  25. And LLP has the problem of being so small, that neutrons on the reactor wall is in the same range as a D-T reactor. So they have a neutronicity problem. 20x closer=400x more neutrons per area.

  26. LPP is relying on using/abusing/hacking QM to reduce Brem.

    Which lets be honest is how physics should work. 

    Let us not question The Maker for giving us convenient game exploits.

  27. Thinking by not-analogy would point out that pB needs much higher temps and Brem is a real issue.

    TAE seems very eager to congratulate themselves on solving these and I genuinely hope they are right.

  28. From the article "This approach may allow for scaling to the conditions necessary for an economically viable fusion plant."
    That's not true. ANY steam powered turbine is effectively uneconomic now except in some corners of the planet (so uneconomic in like 89% of the market). So until you see a viable high temperature supercritical CO2 turbine based system to convert heat into power, it isn't economically viable. You could have a magic wand that provides the heat and just an old 33% efficient turbine island and it still wouldn't be economically viable in most places. Water price and cooling towers are now too expensive for that, you need to be at least 50% thermodynamically efficient. Otherwise it's like hydro, it's too particular to climate and water price (aka geography) to be a general solution. In 2017 the number of coal plants peaked in the world and have by net been dropping since. That'll continue. If coal is uneconomic even without a carbon price then fusion certainly is not.
    World coal consumption peaked around 2013

  29. Cute.

    Typical power system scale up is 3-5x per generation of equipment. If everything goes perfectly, each generation of equipment take 3-5 years.

    They've never made energy, but let's pretend they're at 1 MW by the end of 2021.

    Wildly optimistically, they could be 5MW by 2024, 25MW by 2027, 125MW by 2030, 500MW commercial scale demonstration by 2033, first actually commercial plant in 2036. Realistically, probably a decade slower than this.

    Lots of companies have tried to skip scale-up steps. They typically result in big expensive heaps of garbage that never works. And sometimes bankrupt companies.

  30. People aren't usually dealing with thermal radiation from optically thin plasmas. You're talking about radiation that's produced in the bulk, but not appreciably absorbed, so scale doesn't matter for it.

    Now, it still matters for a lot of other loss modes, so those can scale well. But for a high Z fuel at extremely high temperatures, thermal radiation is going to be the dominant loss mechanism.

    If you recall the articles about proton-boron fusion being attempted by LLP, the reason they thought they could make it work is that quantum magnetic phenomena would suppress cyclotron radiation, and their plasma would become optically thick due to high density. And even so they needed to have very efficient power conversion to have any hope of hitting engineering breakeven.

    Normal magnetic confinement doesn't have either advantage.

  31. "losses scale exactly with the amount of plasma present"

    My bullshiit detector is going off on that one. It defies all my common sense. I can't think of any other situation where you would get less extractable energy by increasing the size of the thing.

  32. With the D-T reaction you actually need the neutrons, to breed more Tritium.

    Like I said, my concern with PB, is that the losses are primarily in the form of T^4 thermal radiation. Where T is absurdly high, (10 times higher than DT, which means 10,000 times the thermal radiation even ignoring Boron's higher Z) and the higher atomic number of the boron enhances radiation. The plasma in this reactor is going to be optically thin, meaning that the losses scale exactly with the amount of plasma present, and enlarging the reactor doesn't help.

    In fact, enlarging the reactor just increases the thermal radiation load on the inner wall. Which is going to be terrifying.

    And P-B isn't even a particularly energetic reaction.

    I don't believe PB is a feasible reaction in reactors with optically thin plasma.

    I understand the attraction of mostly aneutronic reactions, but PB is really, really hard. It's not the first aneutronic reaction you should go for.

    D-H3 is far easier, and H3 could be bred if we were running reactors.

  33. Isn't this almost exactly like a space launch vehicle.

    It's enormously more useful to have a reusable rocket than one that you have to throw away each time.

    BUT, it's easier and cheaper to get a throw-away one working first.

  34. I'm pretty sure that for reducing em losses you'd need an optically thick plasma, which you don't see outside of inertial confinement or LLP's pinches.

  35. There's different types of hard. PB just needs more confinement. That's almost just a matter of how big you make the reactor.

  36. Setting aside that PB doesn't mean no neutrons, just few, you have to be able to reach breakeven before caring about neutrons. PB fusion is REALLY hard.

  37. From the ET article.
    "The TAE Technologies design does not confine plasma using external magnetic field; instead, it confines the plasma on closed magnetic field lines without central penetration. The plasma is held in a rotating, self-stable doughnut shape."

    So, maybe that's a way of saying plasmoid? Sort of like plasmoid formed by a DPF machine? If negative particles exit one side, and positive the other, like with DPF direct energy conversion with P-B11, or P-He3 reactions is a piece of cake. Take a hike you mechanical engineers, no heat engines for you!

  38. ! agree. With D-T reaction you'd have an amazing amount of neutron activation, and heat gain for many meters radius from the reaction chamber.

  39. Not really. You could just as easily argue that D-T in a tokamak is a non-starter because it irradiates the chamber. That seems like a bigger problem than needing to increase the triple product by a couple orders of magnitude.

    To explain further, imagine you had this choice:

    A reactor that's 10 meters long but you have to deal with irradiated chamber walls, OR a reactor that's 100 meters long but no neutron radiation at all.

    Looking at it that way, increasing the triple product is actually much easier in the real world, than dealing with neutron radiation.

  40. "It is a prototype commercially scalable fusion energy reactor designed to bridge between D-T and a p-11B fuel. "

    I was under the impression that PB fusion is basically a non-starter for conventional reactors, due to the high temperature, low gain, and high Z of the fuel.

Comments are closed.