Similar to fission products, neutron control material (like Boron) indeed does not absorb fast neutrons as much. Fortunately it still does absorb fast neutrons enough to control fast reactors (with which we have 430 reactor-years of experience or so).

Indeed fast reactors need higher fissile concentration to be critical. Some kinds of fast reactors (like the Bill Gates Traveling Wave Reactor idea, which, disclaimer: I have professional connections to) can breed up plutonium in spent fuel and then burn it down without reprocessing (to be fair, the spent fuel would have to be hot refabricated into metallic fuel first, but that's still not separations/reprocessing). This kind of reactor needs enrichment once (to start up) and then can just be fed natural, DU, or SNF and it will run on a stream of it happily "forever" (until the vessel life is reached, at which point you transfer the core to a new machine and keep on shuffling). Only fast neutron systems can do such a thing. The Fast Mixed-Spectrum Reactor idea of BNL in 1980 was similar. Many fast-neutron MSRs are the same.

Fast fission is not all that's at play in a fast reactor. It's all about Eta (neutrons released vs. neutrons absorbed in fuel). In a fast reactor, it skyrockets around 0.5 - 1 MeV because of three different physical facts: neutrons released per fission goes up with fast incident neutrons, fission cross section stays flat for all actinides around this range, and capture cross sections drop off towards zero around this range. This is discussed on: https://whatisnuclear.com/fast-reactor.html#havingmore

Having extra neutrons around means you can afford to invest more in breeding fissile material, and only fast neutrons can get you to the point where reactivity is flat or increasing instead of decreasing as fission products are produced. The only exception is Thorium fuel cycle, where U-233 releases a lot of neutrons even with thermal neutron absorption. In that case you have to be removing the fission products as they are created with separations, and this basically is only practical with fluid fuel. This was the idea of the Molten Salt Breeder Reactor project of Oak Ridge in the late 1960s.

With reprocessing, you can definitely re-concentrate the remaining fissile material and get a thermal reactor critical again to burn it. But you cannot meaningfully use the other actinides as fuel, which is the whole idea of "running on spent fuel".

Regarding that slide deck, I love the idea of molten fuel in tubes. I investigated it heavily once upon a time, being super excited about it. I ran into the problem that while salt fuel mass densities are low, separating them out even more physically makes the fissile density annoyingly low, to the point that I couldn't get the reactor performance I wanted for my then target market. I still think this concept is very interesting so I hope these folks do well. They won't be extracting any non-fissile energy from SNF though until they replace that moderating fluid with something like lead, gas, or sodium and change their fuel salt to a chloride.

CANDUs can't use straight-up SNF out of a LWR without first stripping out the fission products. They're an example of "burning down the fissile material even better" but not "extracting the 100x more energy from the fertile U238".

BTW the safety example in that slide deck is a bit disingenuous by suggesting that traditional reactors are not physically stable. If you know those guys you might want to tell them. Overmoderation and Doppler and the NRC ensure that they are inherently stable at power. These things aren't like the F-22 or whatever, requiring active systems to stay in the air. The engineered safety systems are simply for removing decay heat, which can be done passively in Gen III+ plants and indefinitely in any Gen-IV plant.

Fission products have larger cross section than enriched Boron-10. It is all about concentration; that is the number of B-10 atoms per number of atoms of fuel. Here is the cross sections plot I took yesterday: https://imgur.com/a/uWb0SAa

Robert Steinhaus' Question for fast reactor folks: "While the theoretical case for fast reactors being used to burn nuclear waste down to fission products with short half lives has been made for decades, there has in all that time not been a single demonstration of a fast reactor actually experimentally burning any significant quantity (kilograms) of separated Minor Actinides and Transuranics down to fission products to a batch completeness exceeding 90%. The proposal of using fast reactors to treat nuclear waste has been vigorously put forward for decades. Why do fast reactor proponents not demonstrate with one existing fast reactor the burn up of some kilograms of separated long half life Minor Actinide waste to prove the technical feasibility of fast reactors for the application of waste burning and waste treatment?"

Yes, more neutrons are released for Pu239 fast fission. In CANDU, U238 fast fission is more because 99% of fuel is U238. If we increase Pu concentration Pu fast fission happens as well. U238 fast fission is ~3% of CANDU's power. This is because all neutrons are born fast. A FFR (MSR) can adjust heterogeneity as required, because heat transfer takes place outside the core. A MSR can have thicker fuel regions and higher lattice pitch than a CANDU and exploit fast fission further.

CANDU can use straight-up SNF. Check this or search it: http://www.iaea.org/inis/collection/NCLCollectionStore/_Publ...

"These things aren't like the F-22 or whatever, requiring active systems to stay in the air." That is exactly how pressuriser in a PWR works. A control system loop with Temp & pressure sensors with heater and water injection. Almost all reactors are directly synced with grid, reactors will participate in tiny load adjustments and there is also cooling water temperature variations in a day.

MSRs can be isolated from the power conversion using a thermal salt reserve. That is called inherent safety. Water temperature variations or frequency control wont hit the reactor.

Even if MSR is connected to grid directly, there is no DNB control, the boiling point margin is high. If Xenon is removed with >90% efficiency, -ve reactivity can take care of leftover xenon as well, no control rod circus for MSR till a xenon equilibrium is attained.

Whoops: the dominant control reaction in B-10 is (n,alpha), not (n,gamma)! That's a common mistake. As you'll see, it's larger than most fission products across the board See https://imgur.com/6ODUcRx. There are a few Hafnium and Europium nuclides that can beat it but they're pretty expensive so most people stick to Boron control.

Regarding that paper on CANDUs, I'd like to see some lattice physics calcs supporting that. Has anyone run CASMO on it? I would be absolutely shocked if you took a core of 55 MWd/kg (avg) LWR spent fuel, put it straight in a CANDU with no separations (all FP inventory accounted for) and saw it push through an additional normal amount of burnup.

Today, most CANDU's use ~2% enriched feed, I believe.

PWRs are not unstable. They are in their most critical configuration during operation. If they heat up, the water density goes down. Neutrons fail to be moderated and the source of thermal neutrons back into the fuel goes down, and the chain reaction shuts down. This is a regulatory requirement (GDC 11 of the NRC). Unstable cores run away in power excursions when poked (see Chernobyl). Modern cores don't do that. They have negative MTC and Doppler.

I'm well aware of what inherent safety is. EBR-II, a sodium-cooled fast reactor, was the first to actually demonstrate passive shutdown and passive decay heat following unprotected loss of heat sink and unprotected loss of flow.

BTW I don't advocate for fast reactors to burn SNF. It's much more economical to just dispose of it and mine new uranium. I just was pointing out that some people messed up their attempts to burn SNF in thermal reactors.

Yeah, I missed it the first time. You commented when I started editing it. I took (N,TOT) this time. Fission product cross sections is still higher. https://imgur.com/a/pjoNv1p

CANDU: DUPIC fuel cycle is not demonstrated. Canadians don't run PWRs. India, China and South Korea run PWR and CANDU, not much advanced R&D in India. Chinese CANDU in Qinshan may be testing this. US is most advanced in nuclear R&D. Why not build a CANDU in the US and demonstrate this?

I was talking about pressuriser and DNB. If there is no pressuriser (PWR) and constant adjustment of recirculation flow (BWR), clad may get damaged and fuel may melt because of DNB or denucleate boiling. Not inherently safe, engineered systems are needed. (In BWR fuel itself is engineered top-to-bottom with different enrichment levels and different burnable poison concentrations.)

Control rod adjustments are needed till a xenon equilibrium is attained. Also flux flattening circus, boron dilution circus etc. Operators are needed to babysit a LWR or CANDU. This is not like diesel generator or PV panel. Passive shutdown does not mean passive operation like a diesel.

I don't think first generation of MSRs can be like diesel generators. Operators are needed. But, there is potential for nuclear reactors to be like diesel generators with fluid-fuel reactors.

(Edit: Reply to "Today, most CANDU's use ~2% enriched feed, I believe." No. ALL heavy water reactors run with natural uranium in most tubes, depleted uranium or thorium in few pressure tubes for flux flattening circus. No enriched uranium in any tubes.)

Total includes scattering which isn't a loss mechanism. Best bet is to do a spectrum weighted 1-group macroscopic absorption XS (including all neutron loss mechanisms) in various reactors for a meaningful comparison. Most nuclear textbooks have these in the appendix. All I'm telling you is that boron is used for control in fast reactors. I think you asked how it was done a while back.

I kept reading about all these advantages of "slightly enriched uranium" in candus so I'm surprised they aren't using it anywhere! [1]

Autonomous control can theoretically be done with many kinds of reactors and fuels. Sodium reactors have vast pools of liquid metal with extraordinary thermal conductivity and heat capacity. Lead-cooled reactors are similar. So are salt-cooled FHRs, and pebble-bed gas reactors like x-energy. There are a few decades of regulatory catchup before that can happen. Arpa-e has a current project into operator assistance which is aiming to move to more autonomous control.

Don't forget about maintenance costs. Equipment and chemistry will cause maintenance of any reactor to be higher than an average diesel. Fluid fuel will have 50% of the periodic table in thermal gradients, plating out in heat exchangers and other cold surfaces. Sodium reactors have sodium fires. Lead reactors have corrosion. Gas reactors have power cycle leaks. It's a worthy goal but we have lots of work before we get there. We don't have enough experience with fluid fuel yet to judge how maintenance of a commercial unit will pan out. It might be great, or it might prove difficult. Very worthy goal though!

[1] https://inis.iaea.org/search/search.aspx?orig_q=RN:20038082

RE "This kind of reactor needs enrichment once (to start up) and then can just be fed natural, DU, or SNF and it will run on a stream of it happily "forever" (until the vessel life is reached, at which point you transfer the core to a new machine and keep on shuffling)" As fission product concentration increases, even fast reactors have to remove fuel and reprocess it. What works for boron also works for fission products. I don't think any reactor (fast or thermal) can close fuel cycle without the help of reprocessing. May be fast reactors need less frequent reprocessing, that's it.

(LWR-SNF cannot go fast critical. May be fission products is not removed, but reprocessing is needed to concentrate Pu in LWR-SNF.) Even if LWR-SNF has to be used as-is, it has to be toasted in blanket region of fast reactor before it is used as fuel. Some designs like TWR claim no reprocessing needed between toast step and burn step. I agree with that. But, for complete fission of SNF, fission-product removal is needed at some stage when fission product concentration exceeds a particular level in the fuel rod. No exception for travelling wave burnup.

For a critical reactor core some minimum length of fuel rod is always required to be in use. If fission product starts accumulating from bottom of the fuel-rod towards top, somewhere at the midway core can't go critical because a minimum length of rod is needed for criticality.

4% of 1GW-LWR fuel runs for 4.5 years. 100% of this fuel has potential for 112.5 years. A solid-fuel rod can't maintain integrity for 100 years. The crystal structure damage can't be reversed by any amount of maintenance. This is the same phenomenon how solar panels get degraded in sunlight. A fuel rod has crystal structure damage by radiation as well as by fission products.

Fluid-fuel can be used indefinite time and it is easy to reprocess liquid fuel by any method. There is no crystal structure to get damaged. ORNL was working on physical separation of fission products like vacuum distillation when the project was cancelled. Chemical separation and isotopic separation may raise proliferation concerns and they may not be environmentally friendly.

There is a chemical separation of fuel from fission products called fluoride volatility, which is used in industrial scale for enrichment of fuel. But, it works only for fluoride salts. ORNL has demonstrated this process when they switched from U235 to U233. They irreversibly damaged the drain tanks and hastelloy plumbing by doing this. MSRE would have lasted longer if they had not done this fluorination experiment.

RE "Fluid fuel will have 50% of the periodic table in thermal gradients, plating out in heat exchangers and other cold surfaces."

Just throw away the hot leg plumbing and heat exchangers every 2.8 years initially (MSRE salt loop was circulated for 2.8 years) and increase this incrementally. Fact: In solid fuel reactors fuel itself acts as a primary heat exchanger. In 4.5 years LWRs throw away a heat exchanger worth of Zr tubes/cladding. Earlier LWRs threw away solid-fuel every 3 years. Why point fingers at MSR folks when solid-fuel reactors throw away stuff? Note the difference between Inconel 625 (similar to hastelloy-N) and Zr. https://www.tricormetals.com/cost-comparison.html

> But, for complete fission of SNF, fission-product removal is needed at some stage when fission product concentration exceeds a particular level in the fuel rod. No exception for travelling wave burnup.

Of course. This kind of thing is very well established things like [1]. TWR is not interested in burning SNF, that would require reprocessing to convert it from oxide in the first place. Reprocessing is expensive and has historical proliferation concerns. TWR's entire purpose is to approach closed fuel cycle advantages (Gen-IV safety, sustainability, cost) without needing any reprocessing whatsoever. It's a natural step after the US CRBRP mega-boondoggle. Don't reprocess SNF, bury it in geologic repositories and/or boreholes. Burn U-238 at global scale. That's the idea there. It's called a Modified Open Cycle, or Deep-burn once-through fuel cycle.

> Fluid-fuel can be used indefinite time and it is easy to reprocess liquid fuel by any method. There is no crystal structure to get damaged.

Fluid fuel trades solid fuel performance challenges for chemical/corrosion challenges and radionuclide containment challenges. At MSRE, ORNL has yet to account for about 50% of the radioiodine produced. No one knows where it went. that's a huge but not impossible challenge. I agree with the "Easy to reprocess liquid fuel" advantage, but recognize that it is also a proliferation disadvantage.

> Just throw away the hot leg plumbing and heat exchangers every 2.8 years initially (MSRE salt loop was circulated for 2.8 years) and increase this incrementally. These are rad waste remote operations, which has step influence on cost. What's the operational cost of this at FOAK? What's the operational cost at NOAK? It may be very cheap (Thorcon is well on their way to this) or it may be prohibitively expensive. The only way to really find out is to build and operate such a reactor commercially. To this end we will all benefit.

> Why point fingers at MSR folks when solid-fuel reactors throw away stuff? Who's pointing fingers? I mentioned downsides to all forms of reactors in the parent comment. MSR is wonderful and exciting and we as a community need to build many more of them and shake them down.

[1] https://www.iaea.org/publications/7112/implications-of-parti...

Both metallic fuel and salt fuel are not suitable for geological repository. Salt fuel may be stored in salt mines, but unproven. More R&D is needed for geological disposal; but, the point in having metallic fuel or salt fuel is the ease of separating fission products and reusing fuel, not geological storage.

Iodine: Not 50%. It is a probability range. "Thus, of the order of one-fourth to one-third of the iodine has not been adequately accounted for."

RE: "but recognize that it is also a proliferation disadvantage"

MSR without reprocessing: MSR can be as hardened as anyone wants. Because of liquid fuel-form, MSR can be sealed tamper-proof. Fuel goes in and nothing comes out during operation.

Safeguard scenario in case of maintenance: By design the fuel will move to drain tank when shutdown for maintenance. IAEA can have the keys for the drain tank. Fuel can't be pumped back without the presence of IAEA inspector. If someone tampers this, license will be cancelled and the country will have no electricity.

MSBR with reprocessing: How easy separation of fission products from denatured liquid-fuel can be a proliferation disadvantage? It is actually a advantage because fuel always stays inside a tamper-proof system. ORNL attached a "add-on" apparatus for fluoride volatility, and fuel always stayed inside MSRE building. This is actually an advantage.

As material science improves, MSRs last longer and longer without need for maintenance. Safeguards will also become less frequent and less expensive.

I strongly agree that MSR needs a chance. MSR graduated with excellent results with MSRE. MSRE has answered most of the questions and raised very few new questions. MSR should be given a suitable employment like process heat or medical isotope production as soon as possible. We can scale it up for electricity later on.