|
Post by David B. Benson on May 8, 2020 8:59:47 GMT 9.5
|
|
|
Post by Roger Clifton on May 8, 2020 15:37:59 GMT 9.5
The idea is to either add some 5% LEU for a hybrid fuel-repo fuel mix, or to use the fuel for CANDUs. In the Russian REMIX solution, the extracted fuel of say, 1% remaining fissiles is topped up with 20% LEU to provide 3.5% recycled fuel to go back in the PWR. The increase in mass is here 15%. That is quite tolerable, especially since the removal of the fission products has already reduced it by 5% or so. An attractive feature of REMIX is that at end-of-life most of the U238 of raw U has been extracted during the relatively high enrichment, never gets irradiated and is destined to be stored as the relatively innocent depleted uranium (DU). Better to become DU than the familiar and nasty reprocessed uranium (RepU), which is high in U234, U232 etc and probably traces of Pu etc. In contrast, in classic CANDU fuel all of the U238 gets irradiated and almost all of it ends up as RepU.
|
|
|
Post by cyrilr on May 8, 2020 17:00:35 GMT 9.5
The idea is to either add some 5% LEU for a hybrid fuel-repo fuel mix, or to use the fuel for CANDUs. In the Russian REMIX solution, the extracted fuel of say, 1% enrichment is topped up with (maximal) 20% LEU to provide 3.5% recycled fuel to go back in the PWR. The increase in mass is here 15%. That is quite tolerable, especially since the removal of the fission products has already reduced it by 5% or so. An attractive feature of REMIX is that at end-of-life most of the excess U238 is destined to be stored as the relatively innocent depleted uranium (DU), instead of the familiar and nasty reprocessed uranium (RepU), high in U234, U232 etc and probably traces of Pu etc. In contrast, in classic CANDU fuel all of the U238 gets irradiated and almost all of it ends up as RepU. For sure a higher enrichment top-up would help, but it's getting more and more difficult to get 20% LEU, especially commercially for large scale power production. Many enrichment facilities aren't licensed over 5 (some 7%), a lot of the 20% LEU comes from downgraded weapons grade material which is a finite and small source for bulk power production levels. I think the CANDU approach is the best, as it avoids a higher fissile top-up and has other advantages that could lead to large uprates. It would also have the highest proliferation resistance, with ThO2 additions making U232, U233, U234, plus the U235 and U236 in the fuel it would mean impossible isotopics for weapons grade (even if the uranium would be separated).
|
|
|
Post by engineerpoet on May 9, 2020 5:02:31 GMT 9.5
Are there enough CANDUs to make that a viable re-use pathway going forward? Aren't they all planned to be replaced in a few decades anyway?
|
|
|
Post by engineerpoet on May 9, 2020 5:26:33 GMT 9.5
Are Diamonds The Answer To Our Nuclear Waste Problem? A serious boner in that piece: The entire US fleet makes about 2300 tons of HLW annually, not each plant.
|
|
|
Post by Roger Clifton on May 9, 2020 15:08:51 GMT 9.5
Candu... would also have the highest proliferation resistance, with ThO2 additions making U232, U233, U234, plus the U235 and U236 in the fuel it would mean impossible isotopics for [enriching to] weapons grade [uranium]... Having particularly low concentration of U235, Candu fuel has been more attractive to bomb makers for its Pu239 content in fuel easily extracted early in the burn. In other reactors, early burn fuel might be (perhaps less easily) extracted for the same purpose. Without affecting the chemistry of subsequent processing, high proliferation resistance for fresh fuel could be achieved by adding reprocessed fuel to otherwise pure uranium fuel. In the early burn of pure uranium fuel, furtive bomb makers (deprived of the more tractable U235) might be attracted by the growing concentration of Pu239. That is, before it is denatured later in the burn by subsequent conversion of some of it into Pu240. If more than say, 10% Pu240 in Pu is considered useless for bomb-making, surely the answer is to ensure that the fuel starts off with enough Pu240 that it is always greater than 10% of the evolving Pu throughout the burn. Let's say that a proliferation theft waits until the irradiation of some pure uranium fuel had accumulated 0.1% Pu239 but still had low Pu240, before it accelerates. (With no one noticing the premature shutdown and substitution of a fuel assembly, of course.) If when fresh the fuel had included 0.1% of RepPu, of which say(*) 20% is Pu240, the thieves would be pleased to find that they had nearly twice as much plutonium as they expected, then disappointed to find that all of it is more than 10% Pu240. If all fresh fuel contained at least this minimum denaturant, then none of it would attract undesirable attention and there would be no need for the expensive security, mystique and fear surrounding the traffic in fuel. (*) Okay, ex-PWR RepPu isn't 80:20 but contains around 53% Pu239, 25% Pu240, 5% Pu242 and 2% Pu238, and only experienced bomb makers would know what level of dilution makes a mixture with newly-created Pu239 useless, but my guess wouldn't be far off.
|
|
|
Post by Roger Clifton on May 9, 2020 15:36:45 GMT 9.5
A serious boner in that piece: “The typical nuclear power plant creates about 2,300 tons of waste annually,” reports Big Think.The entire US fleet makes about 2300 tons of HLW annually, not each plant. Indeed. The US nuclear fleet generates 92 GW, implying an essential waste stream of only 92 tons per annum of fission products. Of course, bean counting would have the industry minimise the cost of reducing used fuel down to its fission products, but 92 t/a is at least a very low minimum. It could be buried in a single borehole a year, as deep as the fearful require.
|
|
|
Post by cyrilr on May 9, 2020 19:37:46 GMT 9.5
A serious boner in that piece: “The typical nuclear power plant creates about 2,300 tons of waste annually,” reports Big Think.The entire US fleet makes about 2300 tons of HLW annually, not each plant. Indeed. The US nuclear fleet generates 92 GW, implying an essential waste stream of only 92 tons per annum of fission products. Of course, bean counting would have the industry minimise the cost of reducing used fuel down to its fission products, but 92 t/a is at least a very low minimum. It could be buried in a single borehole a year, as deep as the fearful require. The article from Oilprice is crap. Just more scaremongering of "ooh 2 million year half life! WHOPPING!". Gee what about the BILLIONS of TONNES of URANIUM that occur NATURALLY with BILLION YEAR HALF LIVES. WE ARE ALL DOOMED. Oh no wait. We're not. Granite rocks aren't a threat to humanity after all. Even though granite rocks will still be radioactive when our sun runs out of hydrogen to fuse. Then there is the nonsense about how the waste is vulnerable to terrorism. Really? Armored dry casks holding ceramic pellets in metal tubes in the middle of nowhere? What scenario is postulated here that would lead to a large number of casualties? How does one get a lethal dose of radiation to the nearest house? It's simply not credible, and alluding or suggesting that there is a big risk is absurd hyperbole and bad journalism. As usual articles like these don't even bother to elucidate. There is a mention of "a lot of WASTE". Oh scary. No attempt to explain the trivial amounts of actual fission products, or even what a fission product is, or the difference between high level and low level waste. Or indeed anything that would be useful to comprehend in order to understand risks. Just another article worthy of no better storage cabinet than the garbage bin. Then again do we really expect anything else from an oil website?
|
|
|
Post by cyrilr on May 9, 2020 19:49:12 GMT 9.5
Are there enough CANDUs to make that a viable re-use pathway going forward? Aren't they all planned to be replaced in a few decades anyway? There's potential to uprate and extend the life of the CANDUs but obviously there aren't enough. Then again if we want to replace fossil fuels globally we'd need at least 10x the current nuclear capacity. A lot of that - probably most - would be Gen IV. There's tons more options there for re-use. Fast reactors being obvious but there are some not immediately obvious options too. Fluorination to remove the uranium and then simply using the bulk fluorinator bottoms as fuel feed for a fluoride MSR is a great way to go. The mix is mainly TRU/lanthanide/noble metals and has a high fuel worth despite the high concentration of fission products. I suspect though that we will continue building Gen III and Gen III+ for a long while, and simply re-baking the fuel for re-use in a CANDU seems like a useful part of the solution. As long as we are still using billions of tons of coal a year I'm not picky on the flavor of nuke. All hands on deck.
|
|
|
Post by Roger Clifton on May 10, 2020 21:22:05 GMT 9.5
things like vacuum baking, vacuum distillation and zone refining (latter two more for metal fuels) are better avenues of approach CyrilR, I'd be interested to hear you develop your concept of zone refining of metal fuel. It would have to draw a solid alloy containing all the actinides but none of the lanthanides out of a melt containing both. We are most familiar with zone refining of a almost pure compound, where the solid being drawn up out of the melt is a crystal, in which the components find a lower energy state than the impurities to be excluded.
|
|
|
Post by cyrilr on May 10, 2020 22:28:33 GMT 9.5
In theory zone refining would have mixed results. The actinides should have good solubility for each other. The lanthanides likely have good solubility too though so likely less effective. Things like cesium are not well soluble so are removed but then any baking would largely remove that already. Strontium could be somewhat effectively removed. Zirconium and molybdenum would not be effectively removed as they have good solubility, but Zr is already part of the metal fuel alloy and Mo is not detrimental.
Really needs an experiment - which is not that hard to do as one can start with depleted uranium and non radioactive noble metals, lanthanides, strontium etc. Transuranics aren't really necessary to add I think as there seems no reason to suggest they have low solubility in molten uranium so that will simplify the experimental setup.
For fast reactor metal fuel we'd obviously not care much about some lanthanides and Zr, Mo being left. If excessive amounts remain then it starts to become an issue at some point just due to physical high concentration in the fuel after many recycles.
Silver, cesium, strontium, could all be removed by distillation. Unfortunately most of the lanthanides could not as their volatility is often close to the TRUs (though this is a method to separate TRUs/lanthanides together from uranium).
Unfortunately I do not have the time or experimental resources to develop this technology. It would be a fun national lab project.
|
|
|
Post by Roger Clifton on May 11, 2020 13:22:33 GMT 9.5
...removed by distillation Distillation would remove many of the lighter fission products, until it had boiled off the pervasive strontium at 1382° C. However that is a lot of bother just to make the fuel easier to handle. The lanthanides are more refractory still. Even samarium would not come off until 1900° C. Considering that the distillates are quite radioactive, it seems a hard ask to paint them on the walls of a vacuum furnace, hot pump and condensation chamber, inside a hot cell that may not be opened for decades. However the lanthanides do form a group of similar electrochemical potentials, just as the actinides form a group of similar but lower electrochemical potentials. The chloride electrolysis in the IFR successfully extracted the actinides group clear of the lanthanides, which were left in the chloride melt. So electrolysis can separate out the lanthanides.
|
|
|
Post by cyrilr on May 11, 2020 15:28:54 GMT 9.5
Distillation should not prove very difficult. If you can operate, and run electric current through, a molten chloride electrolyte with molten cadmium, and scrape the actinides off it - and likely have to use distillation to separate salts and metals - then you can do a much simpler vacuum still. In any case we are talking full blown hot cells and high activity gas management. Throwing cesium-ridden spent fuel in a hot molten chloride bath... pretty nasty offgas regardless. Sr is not as big a contamination issue as Cs.
In any case the chloride electrolysis process has issues of its own, you end up with at least 3 additional waste streams: cathode, anode and electrolyte, with due processing and waste storage complications. Anode processing looks pretty hairy to me, and anything that processes the electrolyte, the anode or the cathode will itself become high level or intermediate level waste, and is hands-off full-blown-hotcell territory.
If we are going to bother with salts then zone refining would be an interesting option there as well, to remove some fission products.
For a still, 1400C would not be required, 800C should be enough to remove strontium and most of the barium. This is a vacuum still not atmospheric. Though a higher temperature design, using a refractory composite still (and possibly a separate, cooled, stainless steel vacuum vessel as pressure containment) would be interesting. In reality we'd likely want to run just above the fuel melting point at least anyway, it'll simplify the process to have a liquid actinide still bottoms that can be gravity drained to form the ingot for extrusion into new fuel elements.
Keep in mind too that zone refining involves melting - so high temperatures (and possible offgas issues) are there in any case. Given the small quantities of offgas and distillate it does not seem difficult to manage and immobilize. Cs can be immobilized by gettering to form a stable compound (CsF possibly, using CuF2 getter).
A combo of vacuum distillation and zone refining would be decent enough I think. Maybe not all lanthanides are removed, but in a fast reactor, that is not that important anyway.
A still of sufficient capacity for a 1000 MWe reactor would fit in a car's trunk. Tabletop equipment.
|
|
|
Post by cyrilr on May 11, 2020 19:21:36 GMT 9.5
Found a great reference. "Fission product separation from thorium-uranium alloy by arc-zone melting" by R.D. Burch, C.T. Young. catalog.hathitrust.org/Record/100902743As expected Cs and Sr are volatilized out effectively. Zr stays with U, also expected. Lanthanides (R.E.) removal rate better than I expected. Pretty good actually, a limited number of passes would suffice (perhaps only 1 for IFR fuel). Now all that is left is to figure out TRUs behavior. If they stick with the lanthanides then that end can be cut off and processed differently, with a greatly limited flow into that second process.
|
|
|
Post by Roger Clifton on May 12, 2020 16:21:15 GMT 9.5
CyrilR, thank you for providing a link to the paper. However it doesn't really support your ideas. Had you actually read it through yourself? Using a tungsten arc electrode, the researchers drew a bead of liquid along a small bar of thorium containing 3% uranium, which had been irradiated in a reactor. Considering that the melting point of thorium is 2023 K, it is no surprise that cesium and strontium boiled off. More refractory fission products including rare earths were removed with various efficiencies, having been either boiled off or enriched into the liquid bead. Uranium, the main fissile source in the fuel sample was redistributed between the bead and the solid. The minor actinide protactinium gave uncertain behaviour. The authors claim they successfully removed tellurium and ruthenium by zone refinement, and cesium and strontium by volatilisation. However their sample only contained 3% of the more fissile element, uranium. Another sample, thorium with 12% uranium, was less successful. Although the authors use the term "zone melting" to explain the behaviour of elements in the bead, they did not claim to establish a method of zone refinement to clean fuel. The fact that their process separated out some of the essential uranium from the thorium alloy sample is unimpressive. Far from "refining" the fuel, they destroyed it. Considering that the melting point of plutonium is only 912 K and the eutectics it would form with uranium are lower still, it is hard to see that a zone melting process acting on a realistic plutonium/uranium fuel could mobilise the much more refractory rare earths into the liquid while leaving all the Pu behind in a purified fuel solid.
|
|
|
Post by Roger Clifton on May 12, 2020 16:46:54 GMT 9.5
Throwing cesium-ridden spent fuel in a hot molten chloride bath... pretty nasty offgas regardless. You have a point there. The melting point of potassium chloride, the electrolyte in the IFR process, is 1043 K, whereas the boiling point of cesium metal is 963 K. Unless metal fuel had been baked beforehand, it would sputter when immersed in the chloride. However baking beforehand is probably routine in any reprocessing, because there are other volatiles to be released including xenon, krypton and possibly some of the iodine and bromine. The latter two could be trapped in alkali and returned to the electrolyte. Unless there is a market for any of the fission products, once the actinides have been sufficiently removed, the KCl electrolyte and its burden of fission products would be easily cast into a shape ready for deep disposal.
|
|
|
Post by cyrilr on May 12, 2020 16:52:27 GMT 9.5
Results are sensitive to arc speed and specimen size, so these need to be optimized. Don't see big challenges here.
Take a look at Table II. Distribution coefficient for uranium is about unity, for almost all FPs is well below unity. Then Table III, the uranium has clearly moved to the back, along with Te, Ru, Zr. But R.E. are removed from the back. Volatilization does explain some (and in case of Cs and Sr all) of the loss. Don't see why that matters very much - I am actually thinking of an integrated vacuum distillation unit + zone refiner, basically where the zone refiner is in a vacuum vessel fitted with offgas filter and distillate trap. Whether zone refining or volatilization removes it is not that relevant. I don't expect Pu to volatilize at all, Pu is demonstrated to be able to be zone refined for bombs.
Agree on mixed results for Pa. Ruthenium is clearly zone refined, concentrating in the end, it goes with uranium but not a big deal.
as to "far from refining the fuel they destroyed it" now you're the one making claims not supported by the reference. But Th fuel definately is different from U/Pu fuel. I think the results are encouraging but need a U/Pu fuel experiment. Even if Pu goes with R.E. though it would still be a useful processing method (or did you not actually read my previous comment).
|
|
|
Post by cyrilr on May 12, 2020 17:11:16 GMT 9.5
Also. In Table IV you can clearly see by the ratios that U, Ce, Te, Pa, Zr, and Ru are zone refined - concentrating toward the back. R.E. seem to have a small zone refining effect but there does appear some and in the right direction (the average of the first two increments is greater than the average of the last two increments for R.E.). But clearly volatilization explains the large loss of R.E. Curious that Ce goes to the back though, not that this matters, it's innocuous and can go into the recovered fuel.
Even a quite small distribution rate would be acceptable as zone refining is a simple process, so more passes can be added.
As long as Pu does not volatilize I don't see major issues. If it does volatilize we would have to recover from the distillate using some different process.
Zone refining of salts (fluorides or chlorides) is also something I'm interested in, but hard to find good references.
|
|
|
Post by David B. Benson on Sept 20, 2022 10:58:14 GMT 9.5
|
|
|
Post by David B. Benson on Sept 20, 2022 11:12:09 GMT 9.5
|
|