3 ms·

Good writeup. As a tech nerd, it's easy to get carried away with fancy things (breeders! MSR's! pyroprocessing!), but what nuclear really needs to succeed is dr

by jabl 7y ago

Good writeup. As a tech nerd, it's easy to get carried away with fancy things (breeders! MSR's! pyroprocessing!), but what nuclear really needs to succeed is driving down costs and improving public acceptance. > Many advanced reactor companies talk about using high-assay low-enriched uranium (20% enriched rather than 5%), which is economic bonkers. But enrichment and uranium are dirt cheap at the moment due to overcapacity from the Japanese and German nuclear shutdowns. This won't last though, and the HALEU reactors will very likely be too expensive to operate if they do actually get built. Hmm, sure, enrichment cost doesn't increase linearly as a function of enrichment, but it shouldn't be prohibitive either, particularly as fuel cost is such as small part of the total? Of course, to get correspondingly higher burnup you probably need a new fuel design, and probably also burnable poisons, which will increase costs. > Put simply, this means we must build reactors that have only a tiny number of valves, cables, motors, etc. that are needed to safely shut down and cool the reactor. Isn't this what Nuscale is basically proposing (in addition to riding the SMR bandwagon, which may or may not work out)? Their design is cooled by convection, and can be indefinitely passively cooled after shutdown. > In 1957, Walter Zinn (director of Argonne National Lab) suggested that liquid metal cooled (low pressure) natural uranium fueled (like a CANDU) reactors were most likely to succeed economically. Interesting, got a reference? I couldn't find anything via some quick searching. Though 1957 was also before centrifuges drove down enrichment cost, so the benefit from low enrichment isn't as great today as it was back then.

acidburnNSA 7y ago
> particularly as fuel cost is such as small part of the total? The current nuclear industry says things like: "Fuel is free!" all the time, and it nearly is. At EDF's fleet, fuel is something like 5% of the total cost. But this is in a world where capital costs have soared. If we're not careful, we can easily make fuel not free. Many advanced reactor companies are running too far with the fuel-is-free mentality and making choices that will inevitably challenge their long-term goals from a fuel cost point of view. Burning HALEU is a perfect example of this, as if you do that, especially in a fast reactor where you need high fissile concentration, you can make fuel more than 50% of total costs. If nuclear scales 10x or more as we need it to for climate change mitigation, the fuel-is-free mantra will break down. > Isn't this what Nuscale is basically proposing Pretty much. They're onto something for sure, as far as light-water reactors go. It will be interesting to see how much safety-grade equipment they end up with as the licensing continues. > Interesting, got a reference? I couldn't find anything via some quick searching. Page 413 of Hewlett and Holl's Atoms for Peace and War: https://www.energy.gov/management/downloads/hewlett-and-holl-atoms-peace-and-war https://www.energy.gov/management/downloads/hewlett-and-holl...
- jabl 7y ago
  > Burning HALEU is a perfect example of this, as if you do that, especially in a fast reactor where you need high fissile concentration, you can make fuel more than 50% of total costs. The way I think breeder might make sense today for a new commercial design (without a government funded reprocessing program) is if you have a design with a breeding ratio slightly over 1 enabling very high burnups without reactivity swing (iso-breeding, breed-and-burn, or whatever you want to call it). Of course, that requires a fuel design that won't fall apart at high burnups. IIRC EBR-II metallic fuel reached up to 200 GWd/tHM, and of course fast spectrum molten salt designs like Elysium have no such limits at all. But yeah, for thermal reactors with no plans for reprocessing, much higher enrichments than the current LWR standard is most likely not justified, in my non-expert opinion. > Page 413 of Hewlett and Holl's Atoms for Peace and War: https://www.energy.gov/management/downloads/hewlett-and-holl.. https://www.energy.gov/management/downloads/hewlett-and-holl.... Thanks, looks like an interesting book (though not really much details why Zinn thought the way he did..).

by jabl 7y ago

acidburnNSA 7y ago

> particularly as fuel cost is such as small part of the total? The current nuclear industry says things like: "Fuel is free!" all the time, and it nearly is. At EDF's fleet, fuel is something like 5% of the total cost. But this is in a world where capital costs have soared. If we're not careful, we can easily make fuel not free. Many advanced reactor companies are running too far with the fuel-is-free mentality and making choices that will inevitably challenge their long-term goals from a fuel cost point of view. Burning HALEU is a perfect example of this, as if you do that, especially in a fast reactor where you need high fissile concentration, you can make fuel more than 50% of total costs. If nuclear scales 10x or more as we need it to for climate change mitigation, the fuel-is-free mantra will break down. > Isn't this what Nuscale is basically proposing Pretty much. They're onto something for sure, as far as light-water reactors go. It will be interesting to see how much safety-grade equipment they end up with as the licensing continues. > Interesting, got a reference? I couldn't find anything via some quick searching. Page 413 of Hewlett and Holl's Atoms for Peace and War: https://www.energy.gov/management/downloads/hewlett-and-holl-atoms-peace-and-war https://www.energy.gov/management/downloads/hewlett-and-holl...

jabl 7y ago
> Burning HALEU is a perfect example of this, as if you do that, especially in a fast reactor where you need high fissile concentration, you can make fuel more than 50% of total costs. The way I think breeder might make sense today for a new commercial design (without a government funded reprocessing program) is if you have a design with a breeding ratio slightly over 1 enabling very high burnups without reactivity swing (iso-breeding, breed-and-burn, or whatever you want to call it). Of course, that requires a fuel design that won't fall apart at high burnups. IIRC EBR-II metallic fuel reached up to 200 GWd/tHM, and of course fast spectrum molten salt designs like Elysium have no such limits at all. But yeah, for thermal reactors with no plans for reprocessing, much higher enrichments than the current LWR standard is most likely not justified, in my non-expert opinion. > Page 413 of Hewlett and Holl's Atoms for Peace and War: https://www.energy.gov/management/downloads/hewlett-and-holl.. https://www.energy.gov/management/downloads/hewlett-and-holl.... Thanks, looks like an interesting book (though not really much details why Zinn thought the way he did..).