Lars wrote:In other words with uranium you have 3% UF3 or 1 ton of UF3 but with plutonium you have 5 tonnes. Using FLiBe salt you have some trouble here. Changing salt to another will do better for trifluoride solubility but the neutronics get worse unless you are talking about a fast reactor so you won't be able to start with TRU and feed it only thorium. You can continue to feed it TRUs and burn a lot of them but at the end the neutronics will bite you and you will have to remove the remaining plutonium that will be much poorer fuel than the original LWR SNF. This problem will be worse if you start out with more fission products than TRUs.Cyril R wrote:
Yes, but the uranium from LEU is much bigger in quantity. You need about 5 tonnes of plutonium to start up a thermal reactor based on the DMSR. But ORNL's DMSR had roughly 30 tonnes of LEU (almost 29 tonnes U238) at the end of life inventory. You want at least 2% UF3 and probably more like 3% UF3 to keep all the tellurium in elemental form. So you need to have as much as 1 ton of uranium as UF3 for this design. Much more if you want to run on uranium only converter.
5 tonnes of plutonium is about a cubic meter, which is less than 1% of the fuel salt. No problem for NaF-BeF2 eutectic solvent. You also have 0.9% trifluoride fission products, so at the end of the 30 year life @ 75% capacity you'd be close to the solubility limit. But that's good enough I think, it may be attractive to run somewhat shorter even.
Almost forgot, the 5 tonnes plutonium will decrease over time as the thorium chain takes over (later on it is the majority of fissions), so it could be that the amount of trifluoride stays fairly constant, with fission product trifluorides burning in as plutonium is increasingly phased out.
Yes you have 5 tonnes of SNF Pu but that is only 1 cubic meter or about 1% of the fuel salt of the DMSR (a large reactor). You keep feeding Pu in, maybe 300 kg/year (?) to make up for subunity breeding. Then the U233 increasingly takes over since more than half of the fissions are from the thorium chain. Probably the trifluoride fission products that burn in are roughly similar to the burnout rate of plutonium, so you should stay close to that 1% for many years. At some point you get into trouble so you shut down the reactor and fluorinate out the U233. Then you have a lot of U233 to startup new LFTRs with online processing for isobreeding, and you have burnt up most of the plutonium. All without online processing. Not bad I think. Sure you’ve still got a sizeable TRU inventory left, but it is much smaller than what you’ve put in (maybe as much as 10x smaller). You could still process it in the future with some liquid metal exchange process and put the TRUs in LFTRs, this option is still open, and with the stuff all in fluoride form you’re in a better position for future processing.
ORNL said that the trifluoride solubility for FLiBe at the lowest operating temperature is about 1.3%. It is higher for FNaBe, at the cost of only a small decrease in breeding (this would be a converter so not that big a deal anyway).
If we don’t remove the trifluorides in the fluorinator bottoms, how much trifluoride will come with the Pu?