3 ms·

I'll just mention there are a bunch of privately-funded fusion reactor projects, including TAE, General Fusion, Commonwealth, Tokamak Energy, LPP, Helion (funde

I'll just mention there are a bunch of privately-funded fusion reactor projects, including TAE, General Fusion, Commonwealth, Tokamak Energy, LPP, Helion (funded in part by YCombinator), and others. There are also lots of companies working on molten salt reactors. In theory that's easier than fusion, but governments are more prickly about letting you work on fission.

pfdietz 7y ago
TAE: it was explained to them 20 years ago why their idea didn't work, even theoretically. General Fusion: acoustic compression was shown to be unworkable, so they pivoted to a slower scheme with a central conductor down the middle of the chamber. Good luck getting that conductor to survive 100T magnetic fields and orders of magnitude higher neutron loading than in conventional reactor concepts. Commonwealth and TE: overall volumetric power density of their concepts will be very poor compared to fission reactors, even if every other problem is avoided. LPP: there is a serious disconnect between accomplishment and hype. Also, pulsed approaches with very short bursts of fusion output make wall engineering much harder, and it's already very hard in continuous fusion reactor concepts. The only one I cannot immediately dismiss is Helion, although that may be because I know less about them. The goal of direct conversion is admirable if they want to have a chance of being cheaper than fission, but the demands on that seem formidable. The high current, high voltage, billion reps over the lifetime of the reactor switching requirements seem extreme.
- DennisP 7y ago
  I'm not trying to claim that these particular projects will succeed. I'm just saying that some VCs are in fact funding the sort of high-risk breakthrough physical technologies that the article claims are not being funded. But a nitpick on LPP: wall engineering isn't really an issue for them since they're attempting boron fusion. Like Helion they're also planning direct conversion, since the output is a pulsed beam of charged particles. I once asked Lerner why they don't start with D-T and he had an argument that their concept actually works better with boron (don't remember why).
  - pfdietz 7y ago
    Of course wall engineering is going to be a problem with boron fusion. Any very hot plasma will emit photons, and H-11B plasmas will emit them even more than DT plasmas (because the rate of emission increases rapidly with the charge of the nuclei). In fact, a H-11B plasma will normally emit photons faster than it can reheat itself from fusion. LPP makes noise about making tiny knots of plasma that are so dense and ultramagnetized that this doesn't happen -- but as those knots expand, the supposed effects that suppress radiation would die away, and out would pour the photons.
    - DennisP 7y ago
      The reactor would only be about 5MW, and the wall of the reactor core can be however far away it needs to be to not get melted by x-rays. It's not like it's surrounded by superconductors or breeding blanket, it's just vacuum containment. Generally it's neutron radiation that's considered difficult to deal with, and there'd be little of that. And the reactor won't produce net power at all unless they're right that bremsstrahlung will be suppressed long enough by the plasma's magnetic field. If it is, then most of the energy will go into the alpha particle beam.
      - pfdietz 7y ago
        The reactor is described to be 5 MW average output. But if it's very bursty the instantaneous power will be far higher. Damage to the wall will be determined by heating of the surface by these flashes (and possibly mechanical damage from shocks induced by vaporization in that surface layer). Keeping the instantaneous power low enough to avoid damage could make the reactor unacceptably large.
        DennisP 7y ago
        So your criticism is that maybe, possibly, the reactor will be too big. I'll take it.
        7y ago

pfdietz 7y ago

TAE: it was explained to them 20 years ago why their idea didn't work, even theoretically. General Fusion: acoustic compression was shown to be unworkable, so they pivoted to a slower scheme with a central conductor down the middle of the chamber. Good luck getting that conductor to survive 100T magnetic fields and orders of magnitude higher neutron loading than in conventional reactor concepts. Commonwealth and TE: overall volumetric power density of their concepts will be very poor compared to fission reactors, even if every other problem is avoided. LPP: there is a serious disconnect between accomplishment and hype. Also, pulsed approaches with very short bursts of fusion output make wall engineering much harder, and it's already very hard in continuous fusion reactor concepts. The only one I cannot immediately dismiss is Helion, although that may be because I know less about them. The goal of direct conversion is admirable if they want to have a chance of being cheaper than fission, but the demands on that seem formidable. The high current, high voltage, billion reps over the lifetime of the reactor switching requirements seem extreme.

DennisP 7y ago
I'm not trying to claim that these particular projects will succeed. I'm just saying that some VCs are in fact funding the sort of high-risk breakthrough physical technologies that the article claims are not being funded. But a nitpick on LPP: wall engineering isn't really an issue for them since they're attempting boron fusion. Like Helion they're also planning direct conversion, since the output is a pulsed beam of charged particles. I once asked Lerner why they don't start with D-T and he had an argument that their concept actually works better with boron (don't remember why).
- pfdietz 7y ago
  Of course wall engineering is going to be a problem with boron fusion. Any very hot plasma will emit photons, and H-11B plasmas will emit them even more than DT plasmas (because the rate of emission increases rapidly with the charge of the nuclei). In fact, a H-11B plasma will normally emit photons faster than it can reheat itself from fusion. LPP makes noise about making tiny knots of plasma that are so dense and ultramagnetized that this doesn't happen -- but as those knots expand, the supposed effects that suppress radiation would die away, and out would pour the photons.
  - DennisP 7y ago
    The reactor would only be about 5MW, and the wall of the reactor core can be however far away it needs to be to not get melted by x-rays. It's not like it's surrounded by superconductors or breeding blanket, it's just vacuum containment. Generally it's neutron radiation that's considered difficult to deal with, and there'd be little of that. And the reactor won't produce net power at all unless they're right that bremsstrahlung will be suppressed long enough by the plasma's magnetic field. If it is, then most of the energy will go into the alpha particle beam.
    - pfdietz 7y ago
      The reactor is described to be 5 MW average output. But if it's very bursty the instantaneous power will be far higher. Damage to the wall will be determined by heating of the surface by these flashes (and possibly mechanical damage from shocks induced by vaporization in that surface layer). Keeping the instantaneous power low enough to avoid damage could make the reactor unacceptably large.
      - DennisP 7y ago
        So your criticism is that maybe, possibly, the reactor will be too big. I'll take it.
        7y ago