Liquid Fluoride Thorium Reactor

[mesa]

LFTR has a complex radiological path, and all of it is running at molten fluoride temperatures. Molten fluorides are NOT fun things to work with, they are very active. There are significant engineering hurdles for making a 700 C Material that can handle fluence for a reactor. Since there is no fuel loading - additional reactivity is inserted as needed from 233U-F4 salts in storage as needed, and fission products are removed in a chemical treatment of the main coolanant/fuiel salt, you're going to need materials which can handle 10^15 n/cm^2/s at 700 C for decades, not just a few years.

I can understand that but some of the Oak Ridge scientists who have worked with this type of reactor seem to think it really wouldn't be that big of a deal to figure out, here is a link:


Also, don't we have better suited materials today than these guys had 47 years ago?
Are the materials the biggest concern for building this type of reactor?

 

[mesa]

I can understand that but some of the Oak Ridge scientists who have worked with this type of reactor seem to think it really wouldn't be that big of a deal to figure out, here is a link:

I just realized that interview is kind of lengthy, Dick Engel gives his thoughts about these materials at timeframe 23:10 (although I found the interview as a whole really quite insightful).

 

[zapperzero]

I just realized that interview is kind of lengthy, Dick Engel gives his thoughts about these materials at timeframe 23:10 (although I found the interview as a whole really quite insightful).

TL;DL: "we kinda sort of thought we might be able to solve the corrosion problems at some unspecified point in the future because we did a few lab tests"

Yeah. Well.

 

[zapperzero]

Are the materials the biggest concern for building this type of reactor?

One of the. There is also a big concern with designing adequate valves and things of that nature.

 

[nikkkom]

Building a small reprocessing plant near every reactor also doesn't sound inspiring. Those things are complex, expensive, and deal with very nasty stuff.

 

[Rive]

Building a small reprocessing plant near every reactor also doesn't sound inspiring. Those things are complex, expensive, and deal with very nasty stuff.

As I get it the reprocessing would be part of the reactor (or at least the block), not a separated plant.

 

[mesa]

TL;DL: "we kinda sort of thought we might be able to solve the corrosion problems at some unspecified point in the future because we did a few lab tests"

Yeah. Well.

If you are going to quote someone it must be accurate:

Dick Engel during his interview by Kirk Sorenson was asked, “Did the people on the program, in particular the chemists and material scientists feel that corrosion was an insurmountable problem?”

Engel replied, “Uhh, no, I think the people that I dealt with, or spoke with, said ‘okay this is an issue, specifically the tellurium issue but we can get around that’. And some of the subsequent work, subsequent to the initial shutdown they did some experimental work that bode very favorably for an ability to solve that issue.”

Coming from an engineer that has actual experience with this type of reactor it would seem reasonable to assume the materials are not as big an issue as you have thought.

 

[mesa]

Building a small reprocessing plant near every reactor also doesn't sound inspiring. Those things are complex, expensive, and deal with very nasty stuff.

I thought for LFTR's this was part of the reactor and not some 'separate' re-processing plant. I understand that the liquid salts are dangerous to work with but are these systems by any means as complex (or dangerous for that matter) as the way our current nuclear reactors are run?

 

[mheslep]

LFTR has a complex radiological path, and all of it is running at molten fluoride temperatures. Molten fluorides are NOT fun things to work with, they are very active. There are significant engineering hurdles for making a 700 C Material that can handle fluence for a reactor. Since there is no fuel loading - additional reactivity is inserted as needed from 233U-F4 salts in storage as needed, and fission products are removed in a chemical treatment of the main coolanant/fuiel salt, you're going to need materials which can handle 10^15 n/cm^2/s at 700 C for decades, not just a few years.

I'm not sure why it must be so that material lasts the life of the reactor, when the design specifies the fluoride salt can be drained away from the fission core / moderator area at any time, allowing replacement of the core material (graphite?) at whatever schedule desired.

Yes there will need to be thorough certification process for material in contact with the salt (Hastelloy-N?), but then again that effort should be seen in the context of the conditions which the LFTR would replace: a PWR with 153 atm water at 300C and fuel reaching 600C in zircalloy, also w/ 10^15 n/cm^2/s.

 

[mheslep]

I thought for LFTR's this was part of the reactor and not some 'separate' re-processing plant. I understand that the liquid salts are dangerous to work with but are these systems by any means as complex (or dangerous for that matter) as the way our current nuclear reactors are run?

I think the point was, built-in or loosely coupled, the reprocessing step is required for LFTR which adds significant complexity that existing PWR/BWRs don't require.

On the other hand, the advantage of LFTR over PWR/BWR is that i) the fuel enrichment / production step is greatly simplified or goes away entirely, ii) waste is greatly reduced and the waste that is produced has a much shorter half life, iii) no 150 atm water/steam to contain.

 

[mesa]

I think the point was, built-in or loosely coupled, the reprocessing step is required for LFTR which adds significant complexity that existing PWR/BWRs don't require.

On the other hand, the advantage of LFTR over PWR/BWR is that i) the fuel enrichment / production step is greatly simplified or goes away entirely, ii) waste is greatly reduced and the waste that is produced has a much shorter half life, iii) no 150 atm water/steam to contain.

I would imagine given a choice of chemical separation vs. isotopic, chemical will always be the easier path so long as rates of reaction are good. The idea that we shouldn't develop LFTR 'cause we haven't done it yet' seems absurd.

If the advocates are correct and the LFTR is capable of doing what they say I am on board, but getting the rest of the public and the political will in Washington will likely become the biggest challenge.

 

[mheslep]

I would imagine given a choice of chemical separation vs. isotopic, chemical will always be the easier path so long as rates of reaction are good. ...

Different problems. Unlike enrichment of uranium, the chemicals in a LFTR will have strong gamma and beta emitters, and the process will necessarily be in close proximity to an operational reactor.

 

[mesa]

...Unlike enrichment of uranium, the chemicals in a LFTR will have strong gamma and beta emitters, and the process will necessarily be in close proximity to an operational reactor.

Okay, is it that the gamma and beta radiation will interfere with the chemical reactions? or are there problems with shielding for personel? both? or something else I am completely missing?

 

[wizwom]

I'm not sure why it must be so that material lasts the life of the reactor, when the design specifies the fluoride salt can be drained away from the fission core / moderator area at any time, allowing replacement of the core material (graphite?) at whatever schedule desired.

Yes there will need to be thorough certification process for material in contact with the salt (Hastelloy-N?), but then again that effort should be seen in the context of the conditions which the LFTR would replace: a PWR with 153 atm water at 300C and fuel reaching 600C in zircalloy, also w/ 10^15 n/cm^2/s.

I suppose one could do a tubesleeve system, although swelling issues are significant, and then replace the tubes when they seem to have lost cohesion. The lack of significant pressure will also alleviate the material concerns; you can be more brittle when your hoop stresses are lower.

Since ZrF4 was used as fluoride salt component in various MSRs, I'd expect Zircaloy is right out as a tubing material; The temperature range puts us into SiC or ZrC ranges; but they are non-ductile. I expect ODS alloys will be the likely tubing.

Of course, this is assuming we can't separate the corrosion resistance and ductility under radiation problems; if we can work out a reasonable method for SiC coating parts that stands up to radiation and thermal changes then almost all corrosion difficulties can be ignored, and the structural material can be chosen on retention of ductility alone.

 

[mesa]

I suppose one could do a tubesleeve system, although swelling issues are significant, and then replace the tubes when they seem to have lost cohesion. The lack of significant pressure will also alleviate the material concerns; you can be more brittle when your hoop stresses are lower.

Since ZrF4 was used as fluoride salt component in various MSRs, I'd expect Zircaloy is right out as a tubing material; The temperature range puts us into SiC or ZrC ranges; but they are non-ductile. I expect ODS alloys will be the likely tubing.

Of course, this is assuming we can't separate the corrosion resistance and ductility under radiation problems; if we can work out a reasonable method for SiC coating parts that stands up to radiation and thermal changes then almost all corrosion difficulties can be ignored, and the structural material can be chosen on retention of ductility alone.

Are there issues with using Hastelloy N that were not adressed during the running of the research reactor at ORNL or are these simply better choices (cost, durability, etc.) by comparison due to material advancement in the last half century?

 

[mheslep]

I suppose one could do a tubesleeve system, although swelling issues are significant, and then replace the tubes when they seem to have lost cohesion. The lack of significant pressure will also alleviate the material concerns; you can be more brittle when your hoop stresses are lower.

Since ZrF4 was used as fluoride salt component in various MSRs, I'd expect Zircaloy is right out as a tubing material; The temperature range puts us into SiC or ZrC ranges; but they are non-ductile. I expect ODS alloys will be the likely tubing.

Of course, this is assuming we can't separate the corrosion resistance and ductility under radiation problems; if we can work out a reasonable method for SiC coating parts that stands up to radiation and thermal changes then almost all corrosion difficulties can be ignored, and the structural material can be chosen on retention of ductility alone.

We may be talking about two different things.

In the case of a liquid molten salt reactor, it seems to me there are two primary materials to select. The first is the moderator, which will suffer the neutron flux, but has little structural support responsibility. The ONR experiment used a graphite block w/ channels through which the salt was pumped. I assume that's still the first choice for a moderator. The second material is for structural containment. It receives relatively small neutron flux, high radiation, and must structurally contain the ~700C salt. ONR used Hasteloy N.

 

[mheslep]

Are there issues with using Hastelloy N that were not adressed during the running of the research reactor at ORNL or are these simply better choices (cost, durability, etc.) by comparison due to material advancement in the last half century?

Yes there's a problem that was recognized but not yet addressed (AFAIK) in an operation. It is mentioned in the video interview link you provided and on the MSR experiment wiki page. They found that Tellurium, a fission product, causes cracking presence of radioactivity in the alloy ONR used. This would not be trivial thing to test.

 

[mesa]

Yes there's a problem that was recognized but not yet addressed (AFAIK) in an operation. It is mentioned in the video interview link you provided and on the MSR experiment wiki page. They found that Tellurium, a fission product, causes cracking presence of radioactivity in the alloy ONR used. This would not be trivial thing to test.

Agreed, but certainly possible.

Here is another interview with Dick Engel where he discusses this exact problem, it rings deeper than just the Tellurium (which it seems the material scientists had a solution for)

http://www.youtube.com/watch?v=ENH-jd6NhRc&feature=player_embedded

The question is raised at 17:25 and goes to 20:56, although (once again) I really found the discussion as a whole very interesting.

I like Dick Engals take on how to test materials for future reactors, same link but starting at time frame 19:41.

 

[mheslep]

...

The question is raised at 17:25 and goes to 20:56, although (once again) I really found the discussion as a whole very interesting.

Yes I'd seen it previously. I just watched it again and Engel raised an obvious point that I missed before. He points out that if the Te corrosion problem is solved, the overall corrosion problem may not be solved because another element might cause trouble. The larger point being that fission of course means a large chunk of the periodic table would be present, everything from gallium to hafnium, including the very reactive alkali and halogen groups. Does this mean a chemical analysis the interaction of most of the elements in the periodic table against Hasteloy N must be done under LFTR conditions?

One of the advantages of LFTR is supposed to be that high burnup and low waste is possible in part because fission poisons, esp. xenon, can be chemically removed from a liquid fueled reactor, unlike a solid fueled reactor which must have the fuel replaced every couple years. But while targeting the removal of some elements is surely feasible, I doubt it is so easy to remove most of the periodic table.

It may be that in the case of long term corrosion the issue turns in favor short turn fuel supplies, as while fission also generates products in the solid fuel Zirc alloy rods, they're pulled out of service while LFTR is intended to keep going for 30 years or so.

 

[mheslep]

I've just read some of the critical final report on the Molten Salt Reactor Experiment written by the AEC in 70's after the shutdown. I knew of its existence but had avoided it given the politics of the time heavily favoring light water reactors and liquid metal breeders, and I thought it biased.

The structural material discussion starts on page 30. The argument seems valid, if overly absolute ("not suitable").

http://www.energyfromthorium.com/pdf/WASH-1222.pdf

 

[Astronuc]

I've just read some of the critical final report on the Molten Salt Reactor Experiment written by the AEC in 70's after the shutdown. I knew of its existence but had avoided it given the politics of the time heavily favoring light water reactors and liquid metal breeders, and I thought it biased.

The structural material discussion starts on page 30. The argument seems valid, if overly absolute ("not suitable").

http://www.energyfromthorium.com/pdf/WASH-1222.pdf

On page 32 of WASH-1222 is the statement:
"In addition to the intergranular corrosion problem, the standard Hastelloy-N used in the MSRE is not suitable for use in the MSBR because its mechanical properties deteriorate to an unacceptable level when subjected to the higher neutron doses which would occur in the higher power density, longer-life MSBR."

Note that MSRE was only 8 MW (without the electrical generation system, and with off-line batch processing), while the MSBR was planned for 2250 MWt (1000 MWe) and use of on-line continuous processing. See Table III, p. 21 of WASH-1222.

Nickel is problematic in any neutron environment. It absorbs neutrons and becomes active (producing Co-58 and some Co-60) and suffers from an (n,α) reaction. An alloy with lower Ni content would be preferable, something more along the lines of more Cr-Mo (Hastelloys are Ni-Cr-Mo).

Several other technical issues are mentioned. The MSBR concept proposes high temerpature steam cycle, and that presents a challenge, particularly with respect to the heat exchanger, which basically can't be allowed to fail (leak), and then there is the materials compatibility issues between the steam and salt loop. The chemical separation part of the plant would also be challenging. Storage of Xe, Kr, I would be challenging, as well as ultimate disposition of the other fission products (ostensibly they would be converted to oxides and vitrified).

A large scale LFTR would not be a trivial undertaking.

 

[mheslep]

On page 32 of WASH-1222 is the statement:
"In addition to the intergranular corrosion problem, the standard Hastelloy-N used in the MSRE is not suitable for use in the MSBR because its mechanical properties deteriorate to an unacceptable level when subjected to the higher neutron doses which would occur in the higher power density, longer-life MSBR."

Note that MSRE was only 8 MW (without the electrical generation system, and with off-line batch processing), while the MSBR was planned for 2250 MWt (1000 MWe) and use of on-line continuous processing. See Table III, p. 21 of WASH-1222.

Nickel is problematic in any neutron environment. It absorbs neutrons and becomes active (producing Co-58 and some Co-60) and suffers from an (n,α) reaction. An alloy with lower Ni content would be preferable, something more along the lines of more Cr-Mo (Hastelloys are Ni-Cr-Mo).

Yes, apparently He production inside the hastelloy is also a concern. However, it seems to me the neutron flux can be held to some arbitrarily low limit for the outer, structural support holding the salt, where no fission need occur, and with an arbitrary amount of salt or other neutron stops been the graphite-salt-core and the containment. So before WASH stated the material was "unsuitable" without caveat they might have demonstrated how a large neutron flux on the containment was unavoidable.

Several other technical issues are mentioned. The MSBR concept proposes high temerpature steam cycle, and that presents a challenge, particularly with respect to the heat exchanger, which basically can't be allowed to fail (leak),

The high temperature (~700C) is different from a PWR, but I don't know that such temperatures are more challenging than those encountered by any existing Brayton cycle system (e.g. jet engine)

and then there is the materials compatibility issues between the steam and salt loop. The chemical separation part of the plant would also be challenging. Storage of Xe, Kr, I would be challenging, as well as ultimate disposition of the other fission products (ostensibly they would be converted to oxides and vitrified)...

Yes, removal and storage of most every element below U would be required over the lifetime of reactor. There must be some kind of overall chemical architecture to address that issue, as working piecemeal against each and every fission product and their daughter products seems intractable.

 

[mesa]

Yes I'd seen it previously. I just watched it again and Engel raised an obvious point that I missed before. He points out that if the Te corrosion problem is solved, the overall corrosion problem may not be solved because another element might cause trouble. The larger point being that fission of course means a large chunk of the periodic table would be present, everything from gallium to hafnium, including the very reactive alkali and halogen groups. Does this mean a chemical analysis the interaction of most of the elements in the periodic table against Hasteloy N must be done under LFTR conditions?

One of the advantages of LFTR is supposed to be that high burnup and low waste is possible in part because fission poisons, esp. xenon, can be chemically removed from a liquid fueled reactor, unlike a solid fueled reactor which must have the fuel replaced every couple years. But while targeting the removal of some elements is surely feasible, I doubt it is so easy to remove most of the periodic table.

It may be that in the case of long term corrosion the issue turns in favor short turn fuel supplies, as while fission also generates products in the solid fuel Zirc alloy rods, they're pulled out of service while LFTR is intended to keep going for 30 years or so.

What really matters is to what degree these elements formed, your statement is too broad and your link is for fissioning U235 not U233 but let's take a closer look anyway. You mentioned Gallium, pretty nasty stuff when in contact with most metals however it isn't even showing on your distribution scale for the graph you linked.

You had asked:
"Does this mean a chemical analysis the interaction of most of the elements in the periodic table against Hasteloy N must be done under LFTR conditions?"

No, its imperative to be more careful about allowing opinion to get in the way before looking at the facts.

 

[mesa]

I've just read some of the critical final report on the Molten Salt Reactor Experiment written by the AEC in 70's after the shutdown. I knew of its existence but had avoided it given the politics of the time heavily favoring light water reactors and liquid metal breeders, and I thought it biased.

The structural material discussion starts on page 30. The argument seems valid, if overly absolute ("not suitable").

http://www.energyfromthorium.com/pdf/WASH-1222.pdf

I have heard about this report, seems to be fairly well known amongst the LFTR community; their opinions of it are less than favorable but a closer look should be done before forming an opinion on this matter.

On page 30 I noticed the report starts by talking about how the ORN scientists wanted to 'freeze' the salt along the walls to prevent corrosion in the flourinator, seems like a pretty clever idea, but is it feasable? The author thinks no but this report was not written by the scientists actually working on the project who had experience working with similar techniques; Time frame 8:16:

http://www.youtube.com/watch?v=ENH-jd6NhRc&feature=player_embedded

So now come the questions, how hard would it be? how much energy does it use? what is the cost of a system like this? From my own experience in refrigeration I don't think this would be difficult to add on. What is your opinion?

 

[mesa]

A large scale LFTR would not be a trivial undertaking.

No doubt, but does it look promising enough to justify further develop?
Do you see the materials for the reactor as being the largest obstacle?

Yes, removal and storage of most every element below U would be required over the lifetime of reactor. There must be some kind of overall chemical architecture to address that issue, as working piecemeal against each and every fission product and their daughter products seems intractable.

I thought removal of these elements is part of the design with much talk (amongst advocates, so it should be investigated more thoroughly) about how they have high value for the industrial and research markets.

On another note, I don't see why we have to remove 'most every element below U' if their concentration is almost undectectable and there is little to no effect on the reactor itself. It would seem more important to concentrate on the elements that effect the lifecycle of the reactor i.e. materials longevity, efficiency, waste, etc.

 

[mheslep]

What really matters is to what degree these elements formed,

Yes, though operation for 30 years means everything has time to accumulate, unlike in solid fuel reactors.

your statement is too broad and your link is for fissioning U235 not U233

Why is the slight difference between 233 and 235 products relevant to the point, which is that a broad swath of periodic table is dumped into the salt over time via fission products?

You had asked:
"Does this mean a chemical analysis the interaction of most of the elements in the periodic table against Hasteloy N must be done under LFTR conditions?"

No, ..

Why not? In addition swath of fission products, there are other paths for the introduction of elements in elemental form, including the elements from the salt itself - lithium, beryllium, fluorine - then higher Z elements formed from neutron capture of those elements, the decay daughter products of the fission products, carbon from the moderator, and so on.

 

[mheslep]

I thought removal of these elements is part of the design with much talk (amongst advocates, so it should be investigated more thoroughly) about how they have high value for the industrial and research markets.

Yes, that is my understanding, that in fluid fueled reactors it is feasible to remove fission products so that, by removing Xenon via chemical processing, poisoning can be stopped allowing high burn-up. This capability is not feasible in solid fuel designs.

I'm suggesting that along with the advantage comes a problem. While the dispersal of fission products throughout the reactor makes them removable, if the chemical means are put in place, it also means the reactor structural containment must accommodate contact with all of those products which accumulate over long periods.

...
On another note, I don't see why we have to remove 'most every element below U' if their concentration is almost undectectable and there is little to no effect on the reactor itself. It would seem more important to concentrate on the elements that effect the lifecycle of the reactor i.e. materials longevity, efficiency, waste, etc.

Again, a light water reactor w/ solid fuels would have very similar fission products in the short term. The difference with MSRs is that the fuel salt stays in the reactor for the life of the reactor, as I understand it. So that in a solid fuel reactor the minor products might only accumulated at trace levels, while in the MSR they have 30 years to accumulate. After that much time would minor products still be "undetectable"? I don't know that to be the case.

PS One speculative idea that comes to mind: After a high fuel burnup, dump the salt, say, every ten years. The MSR is designed for this for safety reasons in any case. Give it some decay time (short because of the low concentration of actinides in a Thorium cycle), then bury/dispose?

The idea might be a step in the wrong direction, i.e. away from passive, walk away safety. As it implies a design that it a *dump* maintenance is neglected the structural containment is at threat of failure.

 

[mesa]

Yes, though operation for 30 years means everything has time to accumulate, unlike in solid fuel reactors.



Why is the slight difference between 233 and 235 products relevant to the point, which is that a broad swath of periodic table is dumped into the salt over time via fission products?


Why not? In addition swath of fission products, there are other paths for the introduction of elements in elemental form, including the elements from the salt itself - lithium, beryllium, fluorine, then higher Z elements formed from neutron capture of those elements, the decay daughter products of the fission products and so on.

That it does but for most of the elements it looks like that accumulation is still trivial even after 30 years. Do you know of a good source of data on fission byproducts that we could use to make actual calculations? Otherwise this is just a circular arguement.

As far as your graph link I was simply pointing out it was for the wrong fissile material, your new link is much better, thanks for posting it.

 

[mheslep]

That it does but for most of the elements it looks like that accumulation is still trivial even after 30 years. Do you know of a good source of data on fission byproducts that we could use to make actual calculations? Otherwise this is just a circular arguement...

I think the information is roughly available from that products graph.

For instance, for every mole of U233 consumed, 2% of a mole of some fission product (with atomic weight 85) is produced, 7% Zr, 6% Cs and so on. Burn another mole of U233, get another 2%, 7%, 6%, ... which the remains in the reactor, unless it has a fast decay path thus becoming something else, or unless it happens to have a high neutron capture cross section thus becoming something else, ...

 

[mesa]

I think the information is roughly available from that products graph.

For instance, for every mole of U233 consumed, 2% of a mole of some fission product (with atomic weight 85) is produced, 7% Zr, 6% Cs and so on. Burn another mole of U233, get another 2%, 7%, 6%, ... which the remains in the reactor, unless it has a fast decay path thus becoming something else, or unless it happens to have a high neutron capture cross section thus becoming something else, ...

Well let's look back at your first link (similar enough to U233) since it shows a bit more of the dropoff at atomic masses of less than 75 at a rate of .0001% of all fissions and falling off drastically from there. Gallium (like you had mentioned earlier) is 69.723amu so what percentage of fission products produce this element? We need to be careful as well and take into account all isotopes.

With this data we can simply calculate the accumulation of this element of the course of say a 30 year life cycle based off of anticipated (MWt energy of a reactor)/(energy per fission)*time for a rough estimate.

Astronuc, can you point us in the direction of a source with more detail of the fission products from U233?
Nevermind, found it, here is a link for anyone interested in running some calculations:
http://www-nds.iaea.org/relnsd/vchart/

 

[mheslep]

...

With this data we can simply calculate the accumulation of this element of the course of say a 30 year life cycle based off of anticipated (MWt energy of a reactor)/(energy per fission)*time for a rough estimate.

...

PWR typical burnup is around 50 GWdays/ton, or 5% of the fuel. Up to 500 GWdays/ton is expected in an experimental reactor, says the wiki. LFTR supposedly will have very high burnup, so optimistically assume 500 GWdays/ton, or ~120GWdays per 1000 moles of U, or given a 33% efficient reactor, 40GWe-days/1000 moles, or ~11GWe-years/1e5 moles U.

So for every 11 years of operation, and again following the fission products curve, a 1GWe reactor produces 7e3 moles of Zr, 6e3 moles of Cs, etc, for the high probability products. Or, all products with amu's from 82 to 105, and 127 to 150 would accumulate 5e2 moles, or higher, in 11 years. Those concentrations will change through decay or neutron capture.

The consequence of the result would depend on chemistry of the particular element in contact with the alloy which is beyond me.

 

[mesa]

Note that MSRE was only 8 MW (without the electrical generation system, and with off-line batch processing), while the MSBR was planned for 2250 MWt (1000 MWe) and use of on-line continuous processing. See Table III, p. 21 of WASH-1222.

PWR typical burnup is around 50 GWdays/ton, or 5% of the fuel. Up to 500 GWdays/ton is expected in an experimental reactor, says the wiki. LFTR supposedly will have very high burnup, so optimistically assume 500 GWdays/ton, or ~120GWdays per 1000 moles of U, or given a 33% efficient reactor, 40GWe-days/1000 moles, or ~11GWe-years/1e5 moles U.

So for every 11 years of operation, and again following the fission products curve, a 1GWe reactor produces 7e3 moles of Zr, 6e3 moles of Cs, etc, for the high probability products. Or, all products with amu's from 82 to 105, and 127 to 150 would accumulate 5e2 moles, or higher, in 11 years. Those concentrations will change through decay or neutron capture.

The consequence of the result would depend on chemistry of the particular element in contact with the alloy which is beyond me.

Interesting approach, I did it this way using Astronuc's thermal value above for a commercial generating facility of 2250MWt:
2250MWtx24hoursx365daysx30years/((MeV per fission)x(4.4504902416667x10^(-17))) = total number of fissions for the life cycle of the reactor. From here we can just multiply by the Cumulative Fission Yields to get:

4.9x10^21 Ga atoms produced, or .0081mols
Using your method I get .0025mols Ga for the same time frame.

If we are correct Gallium will not be an issue. Granted we could also account for Ga production from U235 since small amounts will also appear in this reactor but that lowers our values since they are an order of magnitude less in production of Ga in the thermal spectrum. Also, as Astronuc pointed out in the other thread, 8-10% of fission in LFTR will be fast neutrons, however this value is comparitively insignificant as well for this particular case.

 

[mesa]

PWR typical burnup is around 50 GWdays/ton, or 5% of the fuel. Up to 500 GWdays/ton is expected in an experimental reactor, says the wiki. LFTR supposedly will have very high burnup, so optimistically assume 500 GWdays/ton, or ~120GWdays per 1000 moles of U, or given a 33% efficient reactor, 40GWe-days/1000 moles, or ~11GWe-years/1e5 moles U.

So for every 11 years of operation, and again following the fission products curve, a 1GWe reactor produces 7e3 moles of Zr, 6e3 moles of Cs, etc, for the high probability products. Or, all products with amu's from 82 to 105, and 127 to 150 would accumulate 5e2 moles, or higher, in 11 years. Those concentrations will change through decay or neutron capture.

The consequence of the result would depend on chemistry of the particular element in contact with the alloy which is beyond me.

We should go visit Borek in the Chemistry section and see what his thoughts are on this.

As for the remainder, calculations for the rest of the elements produced along with their constituent isotopes (and variations) would be helpful but improvement is needed on how calculations are performed to get decent sig figs.

Any thoughts?

 

[Astronuc]

* Independent fission yield (%): number of atoms of a specific nuclide produced directly (not
via radioactive decay of precursors) in 100 fission reactions
* Cumulative fission yield (%): total number of atoms of a specific nuclide produced
(directly and via decay of precursors) in 100 fission reactions

From http://www-nds.iaea.org/publications/tecdocs/iaea-tecdoc-1168.pdf

These may not include activation (n-capture).

--------------------------------------------------
Fission product pairs for U (Z, 92-Z; A, 234-A for U235 or 232-A for U233), assuming 2 neutrons released per fission. The neutrons affect A, not Z.

Z    A        92-Z 234-A for U-235; 232-A for U-233
63   Eu        29   Cu
62   Sm        30   Zn
61   Pm        31   Ga
60   Nd        32   Ge
59   Pr        33   As
58   Ce        34   Se
57   La        35   Br
56   Ba        36   Kr
55   Cs        37   Rb
54   Xe        38   Sr
53   I         39   Y
52   Te        40   Zr
51   Sb        41   Nb
50   Sn        42   Mo
49   In        43   Tc
48   Cd        44   Ru
47   Ag        45   Rh
46   Pd        46   Pd

--------------------------------------------------
Another factor to consider is the delayed neutron precusors that leave the core. Delayed neutrons are important with respect to control the reactor as well as irradiating the structure and piping outside the core.

Reactivity control is another consideration, so a large MSBR may require use of control elements.

The graphite must be supported, so there is a core support plate (not graphite), which will receive a neutron flux.Differences in thermal expansion between graphite and the structural alloy will have to be investigated. Hideout of the molten salt could be an issue. Note the MSRE operated 4 years and surface defects of 7 mils were found. Larger defects may propagate. Also, a 40 to 60 year lifetime is preferable.

The numerous technical issues should be listed and discussed separately.

 

[mesa]

* Independent fission yield (%): number of atoms of a specific nuclide produced directly (not
via radioactive decay of precursors) in 100 fission reactions
* Cumulative fission yield (%): total number of atoms of a specific nuclide produced
(directly and via decay of precursors) in 100 fission reactions

From http://www-nds.iaea.org/publications/tecdocs/iaea-tecdoc-1168.pdf

These may not include activation (n-capture).

--------------------------------------------------

Another factor to consider is the delayed neutron precusors that leave the core. Delayed neutrons are important with respect to control the reactor as well as irradiating the structure and piping outside the core.

Reactivity control is another consideration, so a large MSBR may require use of control elements.

The graphite must be supported, so there is a core support plate (not graphite), which will receive a neutron flux.Differences in thermal expansion between graphite and the structural alloy will have to be investigated. Hideout of the molten salt could be an issue. Note the MSRE operated 4 years and surface defects of 7 mils were found. Larger defects may propagate. Also, a 40 to 60 year lifetime is preferable.

The numerous technical issues should be listed and discussed separately.

Agreed.

I received an email from FliBe energy giving a link to the pdf files of the ORNL research program on the MSR. There is a substantial amount of information:

http://energyfromthorium.com/pdf/

This should be helpful.

 

[mheslep]

Reactivity control is another consideration, so a large MSBR may require use of control elements.

The MSRe had a *negative* temperature reactivity coefficient. The salt expands with temperature, density falls, reactivity falls. Is there some reason that control method must change with large reactor?

 

[mesa]

The MSRe had a *negative* temperature reactivity coefficient. The salt expands with temperature, density falls, reactivity falls. Is there some reason that control method must change with large reactor?

Here is Chris Holdens reason for it @6:16 in his presentation for his reactor design, here is a link:



Calculating for if they are necessary would be good, however there are many things Astronuc suggested that seem like viable avenues to look at. This is already a proven technology and it would seem the question is whether it is needed or not; it is reasonable to assume regulatory agencies could insist on such measures as they are a standard today even if shown to be unneccesary for LFTR.

 

[mesa]

The graphite must be supported, so there is a core support plate (not graphite), which will receive a neutron flux.Differences in thermal expansion between graphite and the structural alloy will have to be investigated. Hideout of the molten salt could be an issue. Note the MSRE operated 4 years and surface defects of 7 mils were found. Larger defects may propagate. Also, a 40 to 60 year lifetime is preferable.

The numerous technical issues should be listed and discussed separately.

Rusty Holden had an interesting idea about a different moderator @ 3:12:


What is 'hideout'? Are you referring to areas in the reactor where flow rates of the salt drop significantly?

"Also, a 40 to 60 year lifetime is preferable."
That would seem reasonable.

 

[mheslep]

Here is Chris Holdens reason for it @6:16 in his presentation for his reactor design, here is a link:



... This is already a proven technology and it would seem the question is whether it is needed or not; it is reasonable to assume regulatory agencies could insist on such measures as they are a standard today even if shown to be unneccesary for LFTR.

Regulatory agencies could insist on anything they like, just because that's the way it has been done. But that's not technically relevant. No MSR is going to see approval in the US by the NRC for decades to come. The design will have to be built abroad, so I don't see tailoring a design to NRC inertia without valid technical reasons, driving up cost, as particularly wise.

 

[zapperzero]

So now come the questions, how hard would it be? how much energy does it use? what is the cost of a system like this? From my own experience in refrigeration I don't think this would be difficult to add on. What is your opinion?

It sounds stupid, wasteful in terms of energy on the one hand and on the other I do not see how you could control the quality of the salt layer. The interface would surely see a lot of stress and cracks and whatnot. Would they propagate to the walls? How could you tell if they did? And so on.

 

[Astronuc]

The MSRe had a *negative* temperature reactivity coefficient. The salt expands with temperature, density falls, reactivity falls. Is there some reason that control method must change with large reactor?

The negative temperature and void coefficients are useful for limiting a reactivity excursion, which is the case in LWRs. However, they are not suitable for power maneuvering a reactor. The delayed neutrons determine the period or rate at which power increases for a given insertion of positive reactivity (e.g., increase in fuel enrichment or removal of a neutron poison). The objective is to maintain control of the power level, and to avoid a rapid increase in reactor power.

Another matter to consider is the guide structure in the core. Control rods are positioned at the edge of the core for rapid insertion. The control rod and guide structure materials must be able to resist the high fluence and fluoride salt interaction.

A lot of the issues mentioned in this thread are also being explored in the Gen-IV MSR program.

 

[mheslep]

As I recall the ONR MSR ~7MWth experiment mainly used load following to control the reactor. Increase the load which removes heat faster, the salt cools, reactivity increases to meet the load.

 

[mesa]

It sounds stupid, wasteful in terms of energy on the one hand and on the other I do not see how you could control the quality of the salt layer. The interface would surely see a lot of stress and cracks and whatnot. Would they propagate to the walls? How could you tell if they did? And so on.

One of the big issues with this type of reactor is the materials reacting with the salt and byproducts of fission; keep in mind that rates of reaction go up drastically with temperature and solids are no where near as reactive as liquids so this idea, (that came from the scientists at ORNL/MSR), seems to have some validity.

Either way we should look through the documents first to see what their proposed approach was before attempting to invalidate/validate this idea with arguement. Here is the link if you missed it:

http://energyfromthorium.com/pdf/

 

[mesa]

* Independent fission yield (%): number of atoms of a specific nuclide produced directly (not
via radioactive decay of precursors) in 100 fission reactions
* Cumulative fission yield (%): total number of atoms of a specific nuclide produced
(directly and via decay of precursors) in 100 fission reactions

Okay, thank you.

These may not include activation (n-capture).

--------------------------------------------------
Fission product pairs for U (Z, 92-Z; A, 234-A for U235 or 232-A for U233), assuming 2 neutrons released per fission. The neutrons affect A, not Z.

Z    A        92-Z 234-A for U-235; 232-A for U-233
63   Eu        29   Cu
62   Sm        30   Zn
61   Pm        31   Ga
60   Nd        32   Ge
59   Pr        33   As
58   Ce        34   Se
57   La        35   Br
56   Ba        36   Kr
55   Cs        37   Rb
54   Xe        38   Sr
53   I         39   Y
52   Te        40   Zr
51   Sb        41   Nb
50   Sn        42   Mo
49   In        43   Tc
48   Cd        44   Ru
47   Ag        45   Rh
46   Pd        46   Pd

--------------------------------------------------

This information is very useful but just for clarification what column is Z and which is A, or are the columns just not lined up?

 

[Astronuc]

This information is very useful but just for clarification what column is Z and which is A, or are the columns just not lined up?

The Z is over the atomic number (number of protons in the nucleus). The A and 234-A are over the letters designating the element (nuclide) corresponding to the Z.

If one fission produces Eu (Z=63, A=158) then the other fission product is necessarily Cu (Z=29, A = 234-158 = 76) + 2 neutrons. If Eu-159 was the fission product, then Cu-75 would be the other fission product + 2 neutrons. If 3 neutrons are released during fission, then the pair would be Eu-158, Cu-75 or Eu-159, Cu-74.

When U-233/U-235 absorbs a neutron and becomes an excited U-234/U-236 nucleus and fissions, the atomic numbers of the fission products, Z1 and Z2 must sum to 92 (or Z, 92-Z). The atomic numbers, A1 and A2, sum to 232/234 if 2 fission (prompt) neutrons are released (or A2 = 232-A1, or 234-A1), or 231/233 if 3 fission (prompt) neutrons are released. Some fission products release 'delayed' neutrons as well - usually fractions of a second up to 60 to 80 seconds later. The fraction of delayed neutrons with U-233 is less than for U-235.

 

[mesa]

The Z is over the atomic number (number of protons in the nucleus). The A and 234-A are over the letters designating the element (nuclide) corresponding to the Z.

If one fission produces Eu (Z=63, A=158) then the other fission product is necessarily Cu (Z=29, A = 234-158 = 76) + 2 neutrons. If Eu-159 was the fission product, then Cu-75 would be the other fission product + 2 neutrons. If 3 neutrons are released during fission, then the pair would be Eu-158, Cu-75 or Eu-159, Cu-74.

When U-233/U-235 absorbs a neutron and becomes an excited U-234/U-236 nucleus and fissions, the atomic numbers of the fission products, Z1 and Z2 must sum to 92 (or Z, 92-Z). The atomic numbers, A1 and A2, sum to 232/234 if 2 fission (prompt) neutrons are released (or A2 = 232-A1, or 234-A1), or 231/233 if 3 fission (prompt) neutrons are released. Some fission products release 'delayed' neutrons as well - usually fractions of a second up to 60 to 80 seconds later. The fraction of delayed neutrons with U-233 is less than for U-235.

Okay, I understand; I thought your chart represented something else, but it is still good for quick reference.

I would like to put together a data table on fission products that have high cross sectional areas for capturing thermal neutrons in the Th/U233 breeder cycle and see which are of biggest concern (like zenon 135).

It would also be good to run through the fission products and see which will have a high likelyhood for rate of reactivity/concentration (like tellurium) with the Hastelloy N. This part will likely prove difficult to compute without experimentation; hopefully there is sufficient information in the ORNL documents.

 

[Astronuc]

One would have to do some calculations based on flux and fuel composition, or find detailed tables that list specific nuclides and their decay chains, for example -

Ba¹⁴⁷ -> La¹⁴⁷ -> Ce¹⁴⁷ -> Pr¹⁴⁷ -> Nd¹⁴⁷ -> Pm¹⁴⁷ -> Sm¹⁴⁷ (stable), but each nuclide can absorb a neutron (but with different cross sections). Sm is a moderate neutron poison. And there are heavier nuclides, e.g., Pm¹⁵⁵ -> Sm¹⁵⁵ -> Eu¹⁵⁵ -> Gd¹⁵⁵, where Eu and Gd are stronger neutron poisons, but their fractional yields are quite low.

Meanwhile, these can provide some idea of the FP vector.

http://www.doitpoms.ac.uk/tlplib/nuclear_materials/nuclear_processes.php

http://en.wikipedia.org/wiki/File:ThermalFissionYield.svg

http://en.wikipedia.org/wiki/Fission_products_(by_element )

http://en.wikipedia.org/wiki/Fission_products_(by_element)#Tellurium-125.2C_127_to_132

http://en.wikipedia.org/wiki/Fissio...7.2C_and_samarium-149.2C_151.2C_152.2C_154.29


http://upload.wikimedia.org/wikiped...onYield.svg/750px-ThermalFissionYield.svg.png


There are preferred nuclides, i.e., those with high yield fractions.

Also of interest - http://en.wikipedia.org/wiki/Fluoride_volatility

 

[mesa]

One would have to do some calculations based on flux and fuel composition, or find detailed tables that list specific nuclides and their decay chains, for example -

Ba¹⁴⁷ -> La¹⁴⁷ -> Ce¹⁴⁷ -> Pr¹⁴⁷ -> Nd¹⁴⁷ -> Pm¹⁴⁷ -> Sm¹⁴⁷ (stable), but each nuclide can absorb a neutron (but with different cross sections). Sm is a moderate neutron poison. And there are heavier nuclides, e.g., Pm¹⁵⁵ -> Sm¹⁵⁵ -> Eu¹⁵⁵ -> Gd¹⁵⁵, where Eu and Gd are stronger neutron poisons, but their fractional yields are quite low.

Meanwhile, these can provide some idea of the FP vector.

http://www.doitpoms.ac.uk/tlplib/nuclear_materials/nuclear_processes.php

http://en.wikipedia.org/wiki/File:ThermalFissionYield.svg

http://en.wikipedia.org/wiki/Fission_products_(by_element )

http://en.wikipedia.org/wiki/Fission_products_(by_element)#Tellurium-125.2C_127_to_132

http://en.wikipedia.org/wiki/Fissio...7.2C_and_samarium-149.2C_151.2C_152.2C_154.29


http://upload.wikimedia.org/wikiped...onYield.svg/750px-ThermalFissionYield.svg.png


There are preferred nuclides, i.e., those with high yield fractions.

Also of interest - http://en.wikipedia.org/wiki/Fluoride_volatility

Yes, this will take some time.

It shold be fairly straightforward to find the products that need the most attention, we need to set up a formula to account for concentration (based on fission products/decay chains) and 'poisoning/absorbance' via cross secions, pretty straight forward.

To keep things simple a strictly Th232/U233 breeder cycle should be considered including U235 and other fissile isotopes formed in meaningful concenrations for calculations.
Any thoughts?

*Here is a link that may also be helpful:
http://www-nds.iaea.org/relnsd/vchart/
This interactice chart has a comprehiensive list of the nucleotide products and their decay chains, although the data has some minor conflicts with other sources (like we saw with Ga) and so there will have to be discussion before number crunching.

 

[Astronuc]

To keep things simple a strictly Th232/U233 breeder cycle should be considered including U235 and other fissile isotopes formed in meaningful concenrations for calculations.
Any thoughts?

*Here is a link that may also be helpful:
http://www-nds.iaea.org/relnsd/vchart/
This interactive chart has a comprehensive list of the nuclide products and their decay chains, although the data has some minor conflicts with other sources (like we saw with Ga) and so there will have to be discussion before number crunching.

I believe the approach is to start MSR (MSBR) with U-235 in Th-232 until sufficient U-233 is available - then perhaps wean the system from U-235 to U-233.

There is also the consideration of neutron spectrum, e.g., thermal, epi-thermal or even fast. One current MSR concept is for a graphite free core, which might imply more moderation from Be. In addition, the Li in the LiF should be depleted in Li-6 to minimize tritium production.

Here is a somewhat relevant report - www.princeton.edu/sgs/publications/sgs/pdf/9_1kang.pdf

See also - https://www.physicsforums.com/showpost.php?p=2546513

 

[mesa]

I believe the approach is to start MSR (MSBR) with U-235 in Th-232 until sufficient U-233 is available - then perhaps wean the system from U-235 to U-233.

Should we be so concerned with the initial injection of fissile U235 or concentrate on the Th232/U233 breeder cycle as the majority of operational time will go to that? On another note, perhaps the easiest way to set this up would be based on MWt generated since we can directly calculate fissions per U233 (and small amounts of U235 created from the breeder cycle)

There is also the consideration of neutron spectrum, e.g., thermal, epi-thermal or even fast. One current MSR concept is for a graphite free core, which might imply more moderation from Be.

That sounds reasonable, calculations will have to include all neutron energies that bring significant concentration of daughter nuclie(s) of interest (neutron poison).

On the graphite free core; very interesting idea employing beryllium in the salt as the moderator although the design requires Hastelloy 'tubes' for these salts, that could be technically difficult as the materials are one of the largest obstacles and this system would require a vast increase in surface area while being at minimum thickness for optimal heat transfer.

Perhaps we should just pick up where ORNL left off and assume for a graphite core in the interim. Other considerations can be taken into account after getting these initial values.

Also if we just base the calculations strictly off of MWt then they could be 'adjusted' to any of these systems estimated MWt.

In addition, the Li in the LiF should be depleted in Li-6 to minimize tritium production.

I have seen interviews of the scientists from ORN suggesting that removal of tritium is not an issue, also considering the difficulty in isotopic separation of Li₆ could add a great deal of expense to a reactor on commercial scale when considering the large quantities of salt required.

Also current worldwide production of tritium is remarkably small:
"According to the Institute for Energy and Environmental Research report in 1996 about the U.S. Department of Energy, only 225 kg (500 lb) of tritium has been produced in the United States since 1955. Since it continually decays into helium-3, the total amount remaining was about 75 kg (170 lb) at the time of the report" link here:
http://en.wikipedia.org/wiki/Tritium
At current market value of almost $30,000/g I would assume this is an asset, not a liability.

We are covering an aweful lot of ground here, what are your thoughts as far as where we should focus our energy for now?