4 ms·

One thing I don’t understand about fusion is the mechanics of gain factors. If we can achieve a fusion reactor with a very small gain factor of say 1.01, is tha

One thing I don’t understand about fusion is the mechanics of gain factors. If we can achieve a fusion reactor with a very small gain factor of say 1.01, is that sufficient to kick off an energy revolution, or do we need something more extreme like 10x or 100x? I suppose it boils down to what the “saturation threshold” of nuclear reactors is, where you can’t pump more energy in without breaking the thing. In any case, what are the benchmarks that engineers are shooting for?

XorNot 4y ago
Q=10 is believed to be necessary to make it economically viable.
- raydiatian 4y ago
  Is there any insight into how long a road from 1 to 10 it is?
  - XorNot 4y ago
    So the assumption is that once you exceed 1 with some reactor architecture, you've more or less got it provided you can build a bigger containment vessel. ITER is as large as it is because within reasonable magnetic field strengths, you still need a certain amount of distance to curve the charged particles back towards the centre - and probability means you have a distribution, some but not all will make it. Bigger vacuum vessel, the more particles you can loop back into your reactor. So, if we can provably get over 1, and validate our plasma behaviour models, we can then derive the correct engineering equations to target a specific Q factor when building the production type reactors.
    - xupybd 4y ago
      So we might have to go much bigger than ITER?
      - XorNot 4y ago
        Yes. The ITER successor is DEMO which is larger and meant to be the production power plant design. The idea is ITER proves out the plasma control physics and scaling laws for fusion, so you can then build DEMO, get the predicted results and say "right, this is how a fusion plant looks". ITER will do Q=10 if we're right anyway, but the scaling laws are all volume based: the real trouble with fusion is that building gigantic vacuum vessels is very hard to do, but we're at the limits of magnets (unless MIT can get HTSCs working, but even then, you still benefit from volume). [1] https://physics.stackexchange.com/questions/175830/nuclear-fusion-scaling-with-reactor-size https://physics.stackexchange.com/questions/175830/nuclear-f...
        DennisP 4y ago
        Tokamaks scale with the square of plasma volume but the fourth power of magnetic field strength. The REBCO used by CFS should let them get ITER-level output from a reactor a tenth as large.
  - willis936 4y ago
    I like to show this graphic. Q>1 is in the "we don't know but we have a good idea" territory. Self-heating (burning) plasma is an untested regime. That will change soon. https://twitter.com/jb_fusion/status/1506964692627034118 https://twitter.com/jb_fusion/status/1506964692627034118
  - DennisP 4y ago
    Not too long, given scaling laws. In tokamaks for example, fusion output is proportional to the square of plasma volume, and the fourth power of magnetic field strength.
nukemandan 4y ago
The lifetime cost being on par will. No power factor in itself will do that.

XorNot 4y ago

Q=10 is believed to be necessary to make it economically viable.

raydiatian 4y ago
Is there any insight into how long a road from 1 to 10 it is?
- XorNot 4y ago
  So the assumption is that once you exceed 1 with some reactor architecture, you've more or less got it provided you can build a bigger containment vessel. ITER is as large as it is because within reasonable magnetic field strengths, you still need a certain amount of distance to curve the charged particles back towards the centre - and probability means you have a distribution, some but not all will make it. Bigger vacuum vessel, the more particles you can loop back into your reactor. So, if we can provably get over 1, and validate our plasma behaviour models, we can then derive the correct engineering equations to target a specific Q factor when building the production type reactors.
  - xupybd 4y ago
    So we might have to go much bigger than ITER?
    - XorNot 4y ago
      Yes. The ITER successor is DEMO which is larger and meant to be the production power plant design. The idea is ITER proves out the plasma control physics and scaling laws for fusion, so you can then build DEMO, get the predicted results and say "right, this is how a fusion plant looks". ITER will do Q=10 if we're right anyway, but the scaling laws are all volume based: the real trouble with fusion is that building gigantic vacuum vessels is very hard to do, but we're at the limits of magnets (unless MIT can get HTSCs working, but even then, you still benefit from volume). [1] https://physics.stackexchange.com/questions/175830/nuclear-fusion-scaling-with-reactor-size https://physics.stackexchange.com/questions/175830/nuclear-f...
      - DennisP 4y ago
        Tokamaks scale with the square of plasma volume but the fourth power of magnetic field strength. The REBCO used by CFS should let them get ITER-level output from a reactor a tenth as large.
- willis936 4y ago
  I like to show this graphic. Q>1 is in the "we don't know but we have a good idea" territory. Self-heating (burning) plasma is an untested regime. That will change soon. https://twitter.com/jb_fusion/status/1506964692627034118 https://twitter.com/jb_fusion/status/1506964692627034118
- DennisP 4y ago
  Not too long, given scaling laws. In tokamaks for example, fusion output is proportional to the square of plasma volume, and the fourth power of magnetic field strength.

nukemandan 4y ago

The lifetime cost being on par will. No power factor in itself will do that.