4 ms·

People want fusion because fission reactors leave long-lived radioactive waste, can melt down dangerously, and use the same fissile materials as nuclear weapons

People want fusion because fission reactors leave long-lived radioactive waste, can melt down dangerously, and use the same fissile materials as nuclear weapons, making combating nuclear proliferation difficult. That's why there are extremely high regulatory costs associated with fission reactors. (I'm not saying we should wait for fusion reactors, but there's a lot of good reasons to develop them, and once fusion reactors are available there's a lot of good reasons to stop building fission reactors at that point.)

dmitrygr 6y ago
The issue is that the current fusion choices (deuterium) produce neutrons during fusion, that hit your containment vessels, and ... create long-lived radioactive isotopes while slowly embrittling the containment vessel. When you replace it, it is in fact, a long lived radioactive husk that you need to safely contain for a long time. There are other fusion options that do not, but they are MUCH harder to fuse than deuterium (what we are still trying to achieve)
- reissbaker 6y ago
  While true, it's still better than fission in terms of waste. (And much less liable to Fukushima-style meltdowns, and less likely to lead to nuclear weapon proliferation.)
  - dmitrygr 6y ago
    Well, actually, a nice source of a large number of neutrons is a HUGE proliferation thread: weaponizable Pu-239 is produced out of easily-available U-238 by neutron capture. Only generation 2 (or 3) fusion, when we move past deuterium, will be proliferation-safe.
    - reissbaker 6y ago
      It's a threat comparable to fission reactors only if fusion reactor containment vessels are built out of uranium, which they aren't. To put it another way: operating a nuclear reactor today is expensive due to regulatory constraints meant to prevent nuclear weapon proliferation. If the fuel for your reactor can't be mistaken for nuclear bomb parts, and the components of your reactor can't be mistaken for nuclear bomb parts, it's a lot cheaper to build and operate. And it's a lot safer for someone to sign off on "Yep that's a whole bunch of lithium for a fusion reactor" than looking at a bunch of uranium and being like... Well...
      - dmitrygr 6y ago
        > operating a nuclear reactor today is expensive due to regulatory constraints meant to prevent nuclear weapon proliferation The way I read that, you seem to imply that this cost dominates all others. If you do mean that, i'd like to see a citation please.
        reissbaker 6y ago
        Here you go: https://www.americanactionforum.org/research/putting-nuclear-regulatory-costs-context/ https://www.americanactionforum.org/research/putting-nuclear... "These figures have profound implications for the industry’s bottom-line. Based on a review of per-plant profitability, there are at least six plants nationwide where regulatory burdens exceed profit margins." Regardless, as I mentioned, the entire process is safer from a proliferation perspective.
        pfdietz 6y ago
        Those regulatory burdens described there exceed profit margins on operations. Those margins ignore the sunk capital cost of the plants. For new nuclear construction, these regulatory costs would be a small compared to the cost of actually building the plants. Of course, new nuclear plants would be outrageously unprofitable.
        reissbaker 6y ago
        You can't ignore operating costs because of sunk capital costs. That's the sunk cost fallacy! You can sometimes ignore sunk costs if operating revenue is good. But if it's negative, that's the problem.
        pfdietz 6y ago
        I wasn't ignoring operating costs. I was pointing out you can't ignore capital and financing costs.
      - deleted 6y ago
  - nickik 6y ago
    This is exactly my point, a properly designed molten salt thorium fission reactor is really terrible to build nuclear weapons (no nation would ever do it) and a Fukushima-style meltdown is literally impossible. The waste from a closed cycle breeder reactor only has to be stored less then 300 years and you can put it back into a mine before that if you want.
pfdietz 6y ago
People want fusion because of mental inertia. Fusion was long sold as the future, so the idea that's it's desirable has become a social default. But it's a fossil belief, perhaps true decades ago, but with little to justify it now.
- elihu 6y ago
  Fusion has long seemed unattainable, but now we seem to be approaching the point where we can actually build a useful reactor. Why wouldn't we use fusion if we could? It's a lot better than the coal and gas we're still using. Wind and solar will be a big part of the power grid going forward, but the installation costs are substantial and we'll need a lot more battery storage to fully transition to renewables. Even assuming SPARC (or one of its competitors) works, it'll be awhile before the technology becomes mature and we can assess whether it's actually cheaper/better than fission or wind/solar/batteries. But from where we stand now it looks promising. Why the pessimism?
  - pfdietz 6y ago
    > now we seem to be approaching the point where we can actually build a useful reactor. I think not. Reaching ignition in a fusion reactor would be akin to what fission achieved in 1942. There would then be enormous engineering obstacles to overcome, particularly to produce a design that could be competitive with other sources of energy. Fusion has grave disadvantages (low power density, complexity, reliability, difficulty of testing components) that must be overcome. I see nothing from existing efforts that suggest they will be able to surmount these obstacles.

dmitrygr 6y ago

The issue is that the current fusion choices (deuterium) produce neutrons during fusion, that hit your containment vessels, and ... create long-lived radioactive isotopes while slowly embrittling the containment vessel. When you replace it, it is in fact, a long lived radioactive husk that you need to safely contain for a long time. There are other fusion options that do not, but they are MUCH harder to fuse than deuterium (what we are still trying to achieve)

reissbaker 6y ago
While true, it's still better than fission in terms of waste. (And much less liable to Fukushima-style meltdowns, and less likely to lead to nuclear weapon proliferation.)
- dmitrygr 6y ago
  Well, actually, a nice source of a large number of neutrons is a HUGE proliferation thread: weaponizable Pu-239 is produced out of easily-available U-238 by neutron capture. Only generation 2 (or 3) fusion, when we move past deuterium, will be proliferation-safe.
  - reissbaker 6y ago
    It's a threat comparable to fission reactors only if fusion reactor containment vessels are built out of uranium, which they aren't. To put it another way: operating a nuclear reactor today is expensive due to regulatory constraints meant to prevent nuclear weapon proliferation. If the fuel for your reactor can't be mistaken for nuclear bomb parts, and the components of your reactor can't be mistaken for nuclear bomb parts, it's a lot cheaper to build and operate. And it's a lot safer for someone to sign off on "Yep that's a whole bunch of lithium for a fusion reactor" than looking at a bunch of uranium and being like... Well...
    - dmitrygr 6y ago
      > operating a nuclear reactor today is expensive due to regulatory constraints meant to prevent nuclear weapon proliferation The way I read that, you seem to imply that this cost dominates all others. If you do mean that, i'd like to see a citation please.
      - reissbaker 6y ago
        Here you go: https://www.americanactionforum.org/research/putting-nuclear-regulatory-costs-context/ https://www.americanactionforum.org/research/putting-nuclear... "These figures have profound implications for the industry’s bottom-line. Based on a review of per-plant profitability, there are at least six plants nationwide where regulatory burdens exceed profit margins." Regardless, as I mentioned, the entire process is safer from a proliferation perspective.
        pfdietz 6y ago
        Those regulatory burdens described there exceed profit margins on operations. Those margins ignore the sunk capital cost of the plants. For new nuclear construction, these regulatory costs would be a small compared to the cost of actually building the plants. Of course, new nuclear plants would be outrageously unprofitable.
        reissbaker 6y ago
        You can't ignore operating costs because of sunk capital costs. That's the sunk cost fallacy! You can sometimes ignore sunk costs if operating revenue is good. But if it's negative, that's the problem.
        pfdietz 6y ago
        I wasn't ignoring operating costs. I was pointing out you can't ignore capital and financing costs.
    - deleted 6y ago
- nickik 6y ago
  This is exactly my point, a properly designed molten salt thorium fission reactor is really terrible to build nuclear weapons (no nation would ever do it) and a Fukushima-style meltdown is literally impossible. The waste from a closed cycle breeder reactor only has to be stored less then 300 years and you can put it back into a mine before that if you want.

pfdietz 6y ago

People want fusion because of mental inertia. Fusion was long sold as the future, so the idea that's it's desirable has become a social default. But it's a fossil belief, perhaps true decades ago, but with little to justify it now.

elihu 6y ago
Fusion has long seemed unattainable, but now we seem to be approaching the point where we can actually build a useful reactor. Why wouldn't we use fusion if we could? It's a lot better than the coal and gas we're still using. Wind and solar will be a big part of the power grid going forward, but the installation costs are substantial and we'll need a lot more battery storage to fully transition to renewables. Even assuming SPARC (or one of its competitors) works, it'll be awhile before the technology becomes mature and we can assess whether it's actually cheaper/better than fission or wind/solar/batteries. But from where we stand now it looks promising. Why the pessimism?
- pfdietz 6y ago
  > now we seem to be approaching the point where we can actually build a useful reactor. I think not. Reaching ignition in a fusion reactor would be akin to what fission achieved in 1942. There would then be enormous engineering obstacles to overcome, particularly to produce a design that could be competitive with other sources of energy. Fusion has grave disadvantages (low power density, complexity, reliability, difficulty of testing components) that must be overcome. I see nothing from existing efforts that suggest they will be able to surmount these obstacles.