4 ms·

This is a good point, thank you for bringing it up. I watched a talk by one of the MIT people on this, and their case is tremendously compelling. This is parti

by AlanSE 9y ago

This is a good point, thank you for bringing it up. I watched a talk by one of the MIT people on this, and their case is tremendously compelling. This is particularly true when you compare it to the pitfalls of the ITER project. At some point, your megaproject is just too flawed and too out of date to make sense on its own. ITER's consumption of Tritium may do more to damage the future of fusion power than what it offers in terms of understanding of the plasma. Sure, the newer designs are risky, but how is ITER not risky? Any decent risk analysis needs to look at things in terms of alternatives. I think ITER will give important knowledge in terms of plasma physics, I just cannot begin to fathom how that is worth the 10s of Billions and waiting 2 decades to get it. The tragedy of megaprojects comes in the form of opportunity cost.

tim333 9y ago
I had similar thoughts after watching the MIT video (https://www.youtube.com/watch?v=KkpqA8yG9T4 https://www.youtube.com/watch?v=KkpqA8yG9T4). It would seem to make sense to scrap the $20bn ITER abd build MITs proposed $300m SPARC prototype instead https://www.nextbigfuture.com/2016/10/mit-set-nuclear-fusion-plasma-record.html https://www.nextbigfuture.com/2016/10/mit-set-nuclear-fusion... Quicker to make, cheaper, more modern tech and much closer to a possible commercial design.
- majewsky 9y ago
  Sunk cost fallacy.
Retric 9y ago
ITER is about farm more than just plasma physics. Tritium breeding from lithium is critical for any D-T reactor design and requires taking the vessel apart remotely and processing the blanket because of it's half life. That's just one of the many things involved and why you want a 'safe' design.
- DennisP 9y ago
  ARC handles breeding by using FLiBe, a molten salt coolant containing lithium. No disassembly required to harvest the tritium.
  - Retric 9y ago
    You can't really use liquids around plasma. You can try to have liquid lithium behind inside a solid container, but wall thickness directly relates to wall lifetime AND you want the thinnest walls possible to maximize tritium production. Worldwide Tritium production is on the order of 1 lb per year, that's not going to cut it when you want to scale up.
    - DennisP 9y ago
      There's a solid inner reactor wall. It's 3-D printed and replaced once a year; the reactor has a modular design that makes it easy to do that.
      - Retric 9y ago
        Exactly, replaced once a year that's easy to write but very hard to actually accomplish. Just like ITER it's extremely radioactive during that process, so you need to handle all of that remotely and your reactor design must be capable of both opening and closing regularly. Sure, it does not seem as sexy as superconductors, plasma, etc but it's a surprisingly hard problem. PS: My point about not scaling up Tritium production just means reactor designs need to be net possessive Tritium producers, which is why the walls need to be thin and swapped out regularly.
        DennisP 9y ago
        One thing that makes it relatively easy to open the reactor is that the superconducting tapes can have joints without significant resistance, instead of being a continuous wrapping. They've tested the joints to make sure of this. Here's a rendering showing how the reactor is put together: https://www.youtube.com/watch?v=efOlmF3wjJE https://www.youtube.com/watch?v=efOlmF3wjJE
- AlanSE 9y ago
  That's an excellent point, and this topic was specifically dealt with in the article. > Just 2 percent of the neutrons will be intercepted by test modules for investigating tritium production in lithium, but 98 percent of the neutron streams will simply smash into the reactor walls or into devices in port openings. It kind of sounds like they're not sure if breeding will work in this kind of reactor design. If not, I guess D-T fusion is just off the table as an option generally? If that's true, this technology track simply doesn't work for the purpose of energy production. Maybe I'm being too unfair? Maybe these test modules are exactly what's needed, maybe a larger area of the modules would have costed too much. That just seems like a strange argument. I don't feel like I can make heads or tails of this. It just feels suspicious.

by AlanSE 9y ago

tim333 9y ago

I had similar thoughts after watching the MIT video (https://www.youtube.com/watch?v=KkpqA8yG9T4 https://www.youtube.com/watch?v=KkpqA8yG9T4). It would seem to make sense to scrap the $20bn ITER abd build MITs proposed $300m SPARC prototype instead https://www.nextbigfuture.com/2016/10/mit-set-nuclear-fusion-plasma-record.html https://www.nextbigfuture.com/2016/10/mit-set-nuclear-fusion... Quicker to make, cheaper, more modern tech and much closer to a possible commercial design.

majewsky 9y ago
Sunk cost fallacy.

Retric 9y ago

ITER is about farm more than just plasma physics. Tritium breeding from lithium is critical for any D-T reactor design and requires taking the vessel apart remotely and processing the blanket because of it's half life. That's just one of the many things involved and why you want a 'safe' design.

DennisP 9y ago
ARC handles breeding by using FLiBe, a molten salt coolant containing lithium. No disassembly required to harvest the tritium.
- Retric 9y ago
  You can't really use liquids around plasma. You can try to have liquid lithium behind inside a solid container, but wall thickness directly relates to wall lifetime AND you want the thinnest walls possible to maximize tritium production. Worldwide Tritium production is on the order of 1 lb per year, that's not going to cut it when you want to scale up.
  - DennisP 9y ago
    There's a solid inner reactor wall. It's 3-D printed and replaced once a year; the reactor has a modular design that makes it easy to do that.
    - Retric 9y ago
      Exactly, replaced once a year that's easy to write but very hard to actually accomplish. Just like ITER it's extremely radioactive during that process, so you need to handle all of that remotely and your reactor design must be capable of both opening and closing regularly. Sure, it does not seem as sexy as superconductors, plasma, etc but it's a surprisingly hard problem. PS: My point about not scaling up Tritium production just means reactor designs need to be net possessive Tritium producers, which is why the walls need to be thin and swapped out regularly.
      - DennisP 9y ago
        One thing that makes it relatively easy to open the reactor is that the superconducting tapes can have joints without significant resistance, instead of being a continuous wrapping. They've tested the joints to make sure of this. Here's a rendering showing how the reactor is put together: https://www.youtube.com/watch?v=efOlmF3wjJE https://www.youtube.com/watch?v=efOlmF3wjJE
AlanSE 9y ago
That's an excellent point, and this topic was specifically dealt with in the article. > Just 2 percent of the neutrons will be intercepted by test modules for investigating tritium production in lithium, but 98 percent of the neutron streams will simply smash into the reactor walls or into devices in port openings. It kind of sounds like they're not sure if breeding will work in this kind of reactor design. If not, I guess D-T fusion is just off the table as an option generally? If that's true, this technology track simply doesn't work for the purpose of energy production. Maybe I'm being too unfair? Maybe these test modules are exactly what's needed, maybe a larger area of the modules would have costed too much. That just seems like a strange argument. I don't feel like I can make heads or tails of this. It just feels suspicious.