13 ms·

Research into fusion went down a blind alley; a means of escape may be at hand

39 points by JDDunn9 11y ago

guimarin 11y ago
The problem with Fusion has always been that R&D associated with it has no bearing on any first order weapons technology. Sure you might be able to put a reactor on an aircraft carrier, but studying powergeneration from fusion reactions does not yield higher yield!
- jerf 11y ago
  The advantages from getting it to work are sufficient on their own merits to not require massive military application. Practical fusion is worth trillions of dollars in the decades after its development. If we get it, all the billions we've poured into it, even down a blind alley, will seem like an underinvestment.
  - guimarin 11y ago
    and yet the US has defunded and underfunded new Fusion research related to power generation at every turn for the past 20 years. Pinning our hopes on the ITER boondoggle.
- ChuckMcM 11y ago
  Ok, and I thought I was cynical :-) I'd challenge your assessment that there isn't a weapon here. Look no further than the rail gun which consumes copious amounts of electrical charge, which a fusion reactor could replace rapidly. There are also the laser weapons which have demonstrated that chemical lasers are too fragile for battle. Large arrays of semiconductor lasers though, collimated into a single beam get you there. Also a consumer of lots and lots of electricity. So one day your cruiser class naval ship starts boiling the ocean around it while it fires of its lasers and rapid fire rail gun.
  - tlb 11y ago
    Since these fusion reactors generate heat, you still need the steam turbine and generator. So the capital cost ($/watt) can't be noticeably less than oil burners. For weapons, fuel cost hardly matters because it's operating for such short periods. So thermal fusion reactors don't enable rail guns.
    - ChuckMcM 11y ago
      Not necessarily. See [1] although there are better sources out there. But one of the things that makes fusion much more interesting is that it is theoretically practical to do direct energy conversion. This is much much harder in fission systems as the only thing fission generates is massive amounts of heat, whereas fusion can generate massive amounts of magnetic flux. [1] https://en.wikipedia.org/wiki/Direct_energy_conversion https://en.wikipedia.org/wiki/Direct_energy_conversion
- trhway 11y ago
  >The problem with Fusion has always been that R&D associated with it has no bearing on any first order weapons technology. actually i think it is quite opposite - the inertial confinement approaches have from the start been the most promising, and they are relatively easy to miniaturize - thus paving the way for the neutron weapons without fission primary. No first-rate power - who already has nuclear weapons and thus has no real need for new types of neutron weapons - is interested in second-grade players obtaining such weapons bypassing the need for fission primary, as all the hurdles related to fission primary is what provides for non-proliferation. Thus all the government funded research has been only in non-miniaturizable (as seen decades back) Tokamak and laser confinement. Sandia Z-machine (great inertial confinement approach) could have been burning DT targets at the end of 199x, yet they started to do it only 15 years later (at least officially) - somebody definitely wasn't in a rush :) With tremendous progress in solid state lasers i kind of curious what fate (ie. government funding) is waiting for NIF as upgrading its warehouse size 3% efficient lasers to container size 20-50% efficient solid state would obviously be in order. Yet it doesn't seems it is going to happen - the last monthly NIF's status update was in May 2014 and officially it has stopped ignition experiments and back to material research (simulating plutonium compression in the nukes)
  - Florin_Andrei 11y ago
    > and they are relatively easy to miniaturize Wait, really? How?
    - trhway 11y ago
      while it is obviously a sarcasm, it suggests that there is need to point obvious difference between energy and power and its density. Inertial confinement schemas don't require large energy, they require huge power and power density though. I.e. a few joules in a femtosecond impulse concentrated in a few cubic millimeters volume. Reaching that power and density is just a pure engineering task, there is no hard physical limits preventing it as there are for example on the minimal size of tokamak.
  - cstross 11y ago
    Nit-pick: the problem with the Z-machine fusing D-T is that the D-T reaction kicks out a ton of neutrons. Once you turn a plasma research experiment over to actual fusion, you end up with secondary activation of the machine's core components, thereby turning them into high-level radioactive waste. This stops you tearing it down and poking around inside and reconfiguring it. This was the fate of the UK's JET experiment at Culham (the Joint European Torus, predecessor to ITER): once they finally bit the bullet and went to actual fusion containment experiments, the "hot" reactor became a lot less accessible for tinkering and reconfiguration. So you don't take that step until you're real sure you don't want to play with your plasma containment experiment any more.
- Sanddancer 11y ago
  The Navy already is investing in fusion research through the Polywell design. Nuclear reactors on ships are big, heavy maintenance intensive, relatively fragile, and require a rather involved port call for refuelling. Any sort of fusion-powered craft would be quite literally floating on its own fuel supply, making it cheaper to operate, safer, and cleaner, which makes the end of life costs for a ship lower.
- tlb 11y ago
  Or some say the opposite: that the practical development of fusion energy at the NIF and Sandia played a secondary role to weapons research. So they made engineering decisions to probe the limits of fusion physics, rather than making something that works.
  - guimarin 11y ago
    I think it's a stretch to ever think there was practical development for energy occurring at those facilities. When you think for 10s about the efficacy of dropping a deuterium pellet every 100ms into a laser chamber that literally needs the lasers to fire 3X orders of magnitude faster than currently possible to sustain a fusion reaction you come to the realization that this was always and only about weapons research.
wmil 11y ago
I'm curious what HN thinks of that last line in the article, about solar. "More likely, cheap photovoltaic and energy-storage technology will mean that much of humanity’s energy comes from a different fusion reactor—one 150m kilometres away, called the sun." 5 billion years of evolution haven't given us solar powered animals that can run around. It seems like there are some significant hard limits on solar technology. Energy storage technology actually makes nuclear much more practical... the big problem it faces is that it generates very consistent power, but power demand fluctuates by time of day.
- chrischen 11y ago
  > 5 billion years of evolution haven't given us solar powered animals that can run around. It seems like there are some significant hard limits on solar technology. What do you think animals run on? Gasoline? Plants are like batteries, storing solar energy. Animals, including humans, charge up by consuming these plants. And for those who don't eat plants, they eat charge up by eating the animals who charge up on the plants.
  - dnautics 11y ago
    There are some organisms that get energy off of chemical gradients deep underground.
  - JoeAltmaier 11y ago
    Again, we can call oil 'solar powered' by that means. Of course 'solar powered' should mean 'directly powered by sunlight captured by the creature itself'. No animal can do this, because the light X capture efficiency is far too low.
    - mikeash 11y ago
      Solar power for electricity more closely resembles animals getting energy from plants which get energy from the sun, rather than animals which directly get energy from the sun. Aside from calculators and a couple other items, we don't put solar panels on our machines to power them, we connect solar panels to the electrical grid and connect machines to the grid too. Fossil oil qualifies too, but the fact it's all ancient stored solar energy rather than solar energy captured now makes it vastly different. Oil produced by capturing solar power now should qualify as "solar power" just as much as photovoltaics.
jjtheblunt 11y ago
Where, by using the subjunctive verb "may", they mean "is not".
astazangasta 11y ago
Can anyone who knows about fusion comment on non-plasma containment methods for producing fusion reactions? In a plasma only a small minority of particles will have enough energy to produce fusion; this means the vast majority of collisions will just send your ions pinging out of the magnetic bottle they are in. This seems a futile, difficult way to get what you want. Surely there are strategies where only high-energy collisions are likely; what are they?
- mng2 11y ago
  If there were, then fusion would be much easier. Consider a reactor that consists of two colliding beams. Should work great, right? Well, the reaction rate is minuscule, so you're wasting a whole bunch of energy in the ones that don't collide. It's a numbers game; in the sun there's a process that has a reaction rate of something like 1 in billions of years, but it's okay because there are orders of magnitude more protons in the sun. The point of tokamaks and stellarators is that the field lines are closed, and you won't have leakage out of cusps like in mirror machines. You prop up the energy distribution with techniques like neutral beam injection and electron cyclotron resonance heating, since the ions tend to dump their energy to the electrons. In other machines, mostly inertial confinement devices, the claim is that a pulsed mode of operation will be effective. However this remains to be shown, and some work (see Todd Rider's PhD thesis) claims to prove that such non-equilibrium schemes cannot reach breakeven.
- drjesusphd 11y ago
  Well, there's "beam" fusion, where you send only high energy particles into target that they fuse with. The issue here is getting the beam up to speed in the first place. What's especially nice about thermonuclear fusion is that it's feasible that the fusion reaction itself keeps the plasma hot enough. > this means the vast majority of collisions will just send your ions pinging out of the magnetic bottle they are in Here's the thing: collisions get less frequent the higher in energy you go. This seems counter-intuitive, but consider that the vast majority of "collisions" are actually weak Coulomb interactions at a large distance. If you're approaching an ion at very high speed, your trajectory is not deflected very much.
- drjesusphd 11y ago
asgard1024 11y ago
Although I think the fusion will eventually happen (and ITER will help), I wish it wouldn't. I think having fusion is a danger for our civilization. People will use this energy for stupid things, like moving mountains; eventually, it will either heat up the planet or produce huge amount of toxic waste or use up all the water we will be able to find. It will be like with carbon fuels but order of magnitude worse. I would much rather see our civilization to learn to live with the plentiful, yet on decent timescales sustainable, solar energy.

39 points by JDDunn9 11y ago

The problem with Fusion has always been that R&D associated with it has no bearing on any first order weapons technology. Sure you might be able to put a reactor on an aircraft carrier, but studying powergeneration from fusion reactions does not yield higher yield!

jerf 11y ago
The advantages from getting it to work are sufficient on their own merits to not require massive military application. Practical fusion is worth trillions of dollars in the decades after its development. If we get it, all the billions we've poured into it, even down a blind alley, will seem like an underinvestment.
- guimarin 11y ago
  and yet the US has defunded and underfunded new Fusion research related to power generation at every turn for the past 20 years. Pinning our hopes on the ITER boondoggle.
ChuckMcM 11y ago
Ok, and I thought I was cynical :-) I'd challenge your assessment that there isn't a weapon here. Look no further than the rail gun which consumes copious amounts of electrical charge, which a fusion reactor could replace rapidly. There are also the laser weapons which have demonstrated that chemical lasers are too fragile for battle. Large arrays of semiconductor lasers though, collimated into a single beam get you there. Also a consumer of lots and lots of electricity. So one day your cruiser class naval ship starts boiling the ocean around it while it fires of its lasers and rapid fire rail gun.
- tlb 11y ago
  Since these fusion reactors generate heat, you still need the steam turbine and generator. So the capital cost ($/watt) can't be noticeably less than oil burners. For weapons, fuel cost hardly matters because it's operating for such short periods. So thermal fusion reactors don't enable rail guns.
  - ChuckMcM 11y ago
    Not necessarily. See [1] although there are better sources out there. But one of the things that makes fusion much more interesting is that it is theoretically practical to do direct energy conversion. This is much much harder in fission systems as the only thing fission generates is massive amounts of heat, whereas fusion can generate massive amounts of magnetic flux. [1] https://en.wikipedia.org/wiki/Direct_energy_conversion https://en.wikipedia.org/wiki/Direct_energy_conversion
trhway 11y ago
>The problem with Fusion has always been that R&D associated with it has no bearing on any first order weapons technology. actually i think it is quite opposite - the inertial confinement approaches have from the start been the most promising, and they are relatively easy to miniaturize - thus paving the way for the neutron weapons without fission primary. No first-rate power - who already has nuclear weapons and thus has no real need for new types of neutron weapons - is interested in second-grade players obtaining such weapons bypassing the need for fission primary, as all the hurdles related to fission primary is what provides for non-proliferation. Thus all the government funded research has been only in non-miniaturizable (as seen decades back) Tokamak and laser confinement. Sandia Z-machine (great inertial confinement approach) could have been burning DT targets at the end of 199x, yet they started to do it only 15 years later (at least officially) - somebody definitely wasn't in a rush :) With tremendous progress in solid state lasers i kind of curious what fate (ie. government funding) is waiting for NIF as upgrading its warehouse size 3% efficient lasers to container size 20-50% efficient solid state would obviously be in order. Yet it doesn't seems it is going to happen - the last monthly NIF's status update was in May 2014 and officially it has stopped ignition experiments and back to material research (simulating plutonium compression in the nukes)
- Florin_Andrei 11y ago
  > and they are relatively easy to miniaturize Wait, really? How?
  - trhway 11y ago
    while it is obviously a sarcasm, it suggests that there is need to point obvious difference between energy and power and its density. Inertial confinement schemas don't require large energy, they require huge power and power density though. I.e. a few joules in a femtosecond impulse concentrated in a few cubic millimeters volume. Reaching that power and density is just a pure engineering task, there is no hard physical limits preventing it as there are for example on the minimal size of tokamak.
- cstross 11y ago
  Nit-pick: the problem with the Z-machine fusing D-T is that the D-T reaction kicks out a ton of neutrons. Once you turn a plasma research experiment over to actual fusion, you end up with secondary activation of the machine's core components, thereby turning them into high-level radioactive waste. This stops you tearing it down and poking around inside and reconfiguring it. This was the fate of the UK's JET experiment at Culham (the Joint European Torus, predecessor to ITER): once they finally bit the bullet and went to actual fusion containment experiments, the "hot" reactor became a lot less accessible for tinkering and reconfiguration. So you don't take that step until you're real sure you don't want to play with your plasma containment experiment any more.
Sanddancer 11y ago
The Navy already is investing in fusion research through the Polywell design. Nuclear reactors on ships are big, heavy maintenance intensive, relatively fragile, and require a rather involved port call for refuelling. Any sort of fusion-powered craft would be quite literally floating on its own fuel supply, making it cheaper to operate, safer, and cleaner, which makes the end of life costs for a ship lower.
tlb 11y ago
Or some say the opposite: that the practical development of fusion energy at the NIF and Sandia played a secondary role to weapons research. So they made engineering decisions to probe the limits of fusion physics, rather than making something that works.
- guimarin 11y ago
  I think it's a stretch to ever think there was practical development for energy occurring at those facilities. When you think for 10s about the efficacy of dropping a deuterium pellet every 100ms into a laser chamber that literally needs the lasers to fire 3X orders of magnitude faster than currently possible to sustain a fusion reaction you come to the realization that this was always and only about weapons research.

wmil 11y ago

I'm curious what HN thinks of that last line in the article, about solar. "More likely, cheap photovoltaic and energy-storage technology will mean that much of humanity’s energy comes from a different fusion reactor—one 150m kilometres away, called the sun." 5 billion years of evolution haven't given us solar powered animals that can run around. It seems like there are some significant hard limits on solar technology. Energy storage technology actually makes nuclear much more practical... the big problem it faces is that it generates very consistent power, but power demand fluctuates by time of day.

chrischen 11y ago
> 5 billion years of evolution haven't given us solar powered animals that can run around. It seems like there are some significant hard limits on solar technology. What do you think animals run on? Gasoline? Plants are like batteries, storing solar energy. Animals, including humans, charge up by consuming these plants. And for those who don't eat plants, they eat charge up by eating the animals who charge up on the plants.
- dnautics 11y ago
  There are some organisms that get energy off of chemical gradients deep underground.
- JoeAltmaier 11y ago
  Again, we can call oil 'solar powered' by that means. Of course 'solar powered' should mean 'directly powered by sunlight captured by the creature itself'. No animal can do this, because the light X capture efficiency is far too low.
  - mikeash 11y ago
    Solar power for electricity more closely resembles animals getting energy from plants which get energy from the sun, rather than animals which directly get energy from the sun. Aside from calculators and a couple other items, we don't put solar panels on our machines to power them, we connect solar panels to the electrical grid and connect machines to the grid too. Fossil oil qualifies too, but the fact it's all ancient stored solar energy rather than solar energy captured now makes it vastly different. Oil produced by capturing solar power now should qualify as "solar power" just as much as photovoltaics.

jjtheblunt 11y ago

Where, by using the subjunctive verb "may", they mean "is not".

astazangasta 11y ago

Can anyone who knows about fusion comment on non-plasma containment methods for producing fusion reactions? In a plasma only a small minority of particles will have enough energy to produce fusion; this means the vast majority of collisions will just send your ions pinging out of the magnetic bottle they are in. This seems a futile, difficult way to get what you want. Surely there are strategies where only high-energy collisions are likely; what are they?

mng2 11y ago
If there were, then fusion would be much easier. Consider a reactor that consists of two colliding beams. Should work great, right? Well, the reaction rate is minuscule, so you're wasting a whole bunch of energy in the ones that don't collide. It's a numbers game; in the sun there's a process that has a reaction rate of something like 1 in billions of years, but it's okay because there are orders of magnitude more protons in the sun. The point of tokamaks and stellarators is that the field lines are closed, and you won't have leakage out of cusps like in mirror machines. You prop up the energy distribution with techniques like neutral beam injection and electron cyclotron resonance heating, since the ions tend to dump their energy to the electrons. In other machines, mostly inertial confinement devices, the claim is that a pulsed mode of operation will be effective. However this remains to be shown, and some work (see Todd Rider's PhD thesis) claims to prove that such non-equilibrium schemes cannot reach breakeven.
drjesusphd 11y ago
Well, there's "beam" fusion, where you send only high energy particles into target that they fuse with. The issue here is getting the beam up to speed in the first place. What's especially nice about thermonuclear fusion is that it's feasible that the fusion reaction itself keeps the plasma hot enough. > this means the vast majority of collisions will just send your ions pinging out of the magnetic bottle they are in Here's the thing: collisions get less frequent the higher in energy you go. This seems counter-intuitive, but consider that the vast majority of "collisions" are actually weak Coulomb interactions at a large distance. If you're approaching an ion at very high speed, your trajectory is not deflected very much.
drjesusphd 11y ago

asgard1024 11y ago

Although I think the fusion will eventually happen (and ITER will help), I wish it wouldn't. I think having fusion is a danger for our civilization. People will use this energy for stupid things, like moving mountains; eventually, it will either heat up the planet or produce huge amount of toxic waste or use up all the water we will be able to find. It will be like with carbon fuels but order of magnitude worse. I would much rather see our civilization to learn to live with the plentiful, yet on decent timescales sustainable, solar energy.