Get rid of transuranians in Liquid Fluoride Thorium Reactors?

[kiskrof]

Liquid Fluoride Thorium Reactors might be a fantastic source of energy for the future. It looks like the world is divided into two clans and the moment: those who have heard of LFTR and are enthusiastic, and those who have hardly ever heard of it because it is hardly mentioned even in the most recent textbooks.
At some moments I even wonder if LFTR is not a hoax along with perpetual motions or Martian technology. But the more I read, the more I am convinced about it. Everyone who does not know about LFTR should google it immediately or look for information on this very forum.
LFTR would produce only fission products radioactive a few centuries and no transuranians radioactive for thousands of years. Among the often quoted advantages of LFTR is the possibility of eliminating transuranian waste from "archaic" (I mean "present") nuclear technology by feeding some of it into the LFTR. I have no reason to doubt that but I wonder how long it might take.
It is claimed that about two kilograms of Thorium a day would be enough to make an average LFTR (1 GW) work. It is absolutely awesome. But if we imagine that the quantity of transuranians fed into the reactors represents a 10% fraction of the Thorium, it would be 200g a day. I read there are about 20000 tons of Plutonium on Earth at the moment. It means 100 LFTRs would need about 250 years to "burn" all this plutonium. Am I in the right time scale or not at all?

 

[QuantumPion]

The properties of LFTR's are somewhat exaggerated and misunderstood. Most of the advantages come from the reprocessing of the fuel, but this can be done with any reactor type. Transuranics are only a problem in spent fuel that is not reprocessed because of their long half-lives. They are alpha emitters and not particularly hazardous by themselves.

 

[mfb]

Many tons of a material that is highly radioactive, has to stay hot to be liquid all the time, directly reacts with air moisture on contact, contains radioactive gases that have to be separated in a controlled way and safely stored for months to decades, contains various corrosive fluorine compounds of other elements and needs extensive chemical/physical reprocessing steps with radioactive materials, to name just a few of the disadvantages.

It is certainly an interesting concept, but it has its own challenges.

But if we imagine that the quantity of transuranians fed into the reactors represents a 10% fraction of the Thorium

That looks quite optimistic, as getting a breeding ratio of 1 is a challenge on its own, so a reactor would probably be just a bit above 1.
It is not just plutonium, we have some 100,000 tonnes of highly radioactive waste.

 

[nikkkom]

Among the often quoted advantages of LFTR is the possibility of eliminating transuranian waste from "archaic" (I mean "present") nuclear technology by feeding some of it into the LFTR. I have no reason to doubt that but I wonder how long it might take.

Reprocessing is a quite expensive process.
Separating transuranics from fission products would require adding a number of steps to it, making it even more expensive.

Currently, no one is doing that. Not even French, the world leaders in reprocessing. It might end up being uneconomical.

 

[mesa]

The properties of LFTR's are somewhat exaggerated and misunderstood. Most of the advantages come from the reprocessing of the fuel, but this can be done with any reactor type. Transuranics are only a problem in spent fuel that is not reprocessed because of their long half-lives. They are alpha emitters and not particularly hazardous by themselves.

The difference being the fuel is already liquid and ready to be processed in LFTR's. If we want 'melt down proof' reactors (which the post Fukishima public demands) then removing things like the transuranics while in operation is required. MSR's are the only reactors I am aware of that can accomplish this.

 

[nikkkom]

The difference being the fuel is already liquid and ready to be processed in LFTR's. If we want 'melt down proof' reactors (which the post Fukishima public demands) then removing things like the transuranics while in operation is required.

How removing transuranics help with that?
Almost all decay heat load comes from fission products, not transuranics.

 

[mesa]

How removing transuranics help with that?
Almost all decay heat load comes from fission products, not transuranics.

I was trying to stay on topic hence the wording 'like transuranics'. I was going to edit but didn't get around to it.

Either way you are completely correct of course, the transuranics in an MSR would stay in place and be fissioned while the fission products of these elements and U233 would be removed which in turn would create the 'melt down proof' design aspect of a LFTR.

***EDIT*** On a completely different note, I just noticed this is my 555th post, hooray!

 

[Hologram0110]

I believe another advantage of the thorium cycle with regards to waste management is that Th-232 requires many more neutron activations and beta decays before becoming Pu or Am. This part has nothing to do with the molten salt part though, just the starting fuel composition.

As a side note, some MSR designs do not call for reprocessing, or reprocessing in batches. Thus fission product loading can still be significant in these designs. The real advantage of MSR is the ability to passively drain into high surface area tanks for passive cooling.

 

[mesa]

I believe another advantage of the thorium cycle with regards to waste management is that Th-232 requires many more neutron activations and beta decays before becoming Pu or Am. This part has nothing to do with the molten salt part though, just the starting fuel composition.

As a side note, some MSR designs do not call for reprocessing, or reprocessing in batches. Thus fission product loading can still be significant in these designs. The real advantage of MSR is the ability to passively drain into high surface area tanks for passive cooling.

The former Scientists at ORNL felt confident they were close to getting the engineering right on the Hastelloy (for a reactor that removes these products). If left in they likely will introduce some new challenges on the materials side of design. Then again leaving them in simplifies reactor design as well.

 

[Astronuc]

Either way you are completely correct of course, the transuranics in an MSR would stay in place and be fissioned while the fission products of these elements and U233 would be removed which in turn would create the 'melt down proof' design aspect of a LFTR.

It is rather humorous to refer to a 'Molten' Salt Reactor as 'melt down' proof, because the fuel/core is in a molten state.

The benefit of an MSR, assuming that the chemical processing and reprocessing are part of it, is that the equilibrium inventory is rather lower as compared to a conventional system in which the fission products and TU nuclides accumulate in the fuel.

The U-233 is recycled into the core, ostensibly with some U-234, U-235 and perhaps U-236, which should be at lower levels than a U-235 fueled core.

The fission products still need to be accumulated, calcined and vitrified in order to be stabilized.

 

[mesa]

It is rather humorous to refer to a 'Molten' Salt Reactor as 'melt down' proof, because the fuel/core is in a molten state.

Agreed, I am open to suggestions on terminology :)

The benefit of an MSR, assuming that the chemical processing and reprocessing are part of it, is that the equilibrium inventory is rather lower as compared to a conventional system in which the fission products and TU nuclides accumulate in the fuel.

The U-233 is recycled into the core, ostensibly with some U-234, U-235 and perhaps U-236, which should be at lower levels than a U-235 fueled core.

It's unfortunate ORNL never had a chance to get this far.

The fission products still need to be accumulated, calcined and vitrified in order to be stabilized.

Certainly the best option for long term containment.

 

[Stanley514]

Many tons of a material that is highly radioactive, has to stay hot to be liquid all the time, directly reacts with air moisture on contact, contains radioactive gases that have to be separated in a controlled way and safely stored for months to decades, contains various corrosive fluorine compounds of other elements and needs extensive chemical/physical reprocessing steps with radioactive materials, to name just a few of the disadvantages.

That thing potentially wastes tons of lithium and perhaps beryllium (both are rare and valuable metals).

 

[mheslep]

The properties of LFTR's are somewhat exaggerated and misunderstood. Most of the advantages come from the reprocessing of the fuel, but this can be done with any reactor type. ...

Major advantages obtain via:
1. Molten fuel.
i) Enables online reprocessing which means no build up of neutron absorbing fission products which in turn means a) no periodic shut down for refueling and b) high burn up.
ii) Allows a natural fail safe by draining the fuel off the moderator.
iii) Low pressure design, reducing reliance on containment, perhaps eliminating it.
2. Thorium fertile fuel, which means a drastic reduction in the scale of the uranium fuel enrichment cycle.
3. Near zero plutonium production.

These can not be accomplished by starting with existing solid fuel PWR/BWRs and simply shipping their removed waste to current reprocessing centers.

 

[Hologram0110]

Major advantages obtain via:
1. Molten fuel and thus enabled online reprocessing means no build up of neutron absorbing fission products which in turn means i) no periodic shut down for refueling and ii) high burn up.
2. Thorium fission fuel, which means a virtual end to the uranium fuel enrichment cycle.
3. Near zero plutonium production.

These can not be accomplished by starting with existing solid fuel PWR/BWRs and shipping their removed waste to current reprocessing centers.

1 and 2 are possible with conventional heavy water or graphite moderated designs.

3 is primarily a political problem if one allows reprocessing. Pu is very good fuel (higher actinides not so much).

 

[mheslep]

1 and 2 are possible with conventional heavy water or graphite moderated designs.

Forgot about CANDU so I grant no enrichment there. But how does one remove fission product alone from *any* solid fueled reactor?

3 is primarily a political problem if one allows reprocessing. Pu is very good fuel (higher actinides not so much).

No just reprocessing, else the growing stockpile of Pu would be not be growing. The appropriate reactor is required. Politics enters there, but so does engineering of (largely) Pu fueled reactors.

 

[Hologram0110]

Sure there isn't online removal of the fission products from solid fuel. However, if you have a short irradiation cycle and reprocessing you can do it. Many MSR don't plan on online reprocessing because of how difficult it is to deal with radioactivity of short-lived fission products.

There isn't a significant technical problem to using Pu in many existing reactors. MOX fuel is a very established technology. The problem is simply economics. Right now virgin uranium and manufacturing new fuel is cheap enough that reprocessing doesn't make economic sense. Things may eventually change if there is a break through in reprocessing technology or shortage of uranium supply. Until then, reprocessing is mostly a political decision to reduce 'waste' or preserve natural resources.

 

[mesa]

That thing potentially wastes tons of lithium and perhaps beryllium (both are rare and valuable metals).

Last time I checked lithium and beryllium are easier to come by than U235.

 

[mheslep]

...Many MSR don't plan on online reprocessing because of how difficult it is to deal with radioactivity of short-lived fission products.

For a thorium fueled MSR, as per the OP, online reprocessing is unavoidable.

 

[mheslep]

That thing potentially wastes tons of lithium and perhaps beryllium (both are rare and valuable metals).

Last time I checked lithium and beryllium are easier to come by than U235.

Forget the isotope, Li is a couple orders of magnitude more abundant than natural U; Be a couple multiples more abundant than natural U.
http://en.wikipedia.org/wiki/Abunda...ust#mediaviewer/File:Elemental_abundances.svg

 

[mheslep]

Sure there isn't online removal of the fission products from solid fuel. However, if you have a short irradiation cycle and reprocessing you can do it. ...

The point is not solely what is technically possible, but what also what is economically plausible. The major advantage of high fuel burnup *in place* is cost reduction. Adding a loop where, the fuel is removed from containment, shipped off for dis-assembly, is reprocessed, reassembled and reinserted into the reactor can not be remotely as cost effective.

 

[Stanley514]

Last time I checked lithium and beryllium are easier to come by than U235.

Even more so.

 

[Hologram0110]

The point is not solely what is technically possible, but what also what is economically plausible. The major advantage of high fuel burnup *in place* is cost reduction. Adding a loop where, the fuel is removed from containment, shipped off for dis-assembly, is reprocessed, reassembled and reinserted into the reactor can not be remotely as cost effective.

I agree that reprocessing may never be cost effective. In this discussion there is the implication that online reprocessing of MSR fuel WILL be economical because the fuel is liquid. Personally, I think that remains to be seen. I think that many of the proponents of MSR overstate the advantages while ignoring the challenges. I just wanted to point out that other reactor systems could make similar claims provided we ignore their problems.

I think the earliest MSR will feature a much simpler fuel cycle, likely uranium not thorium. The reactor can be batch refueled and if any reprocessing is done, it will likely remove a small subset of fission rather than a complete cleaning of the fuel. More likely more fissile isotopes will simply be added to the fuel to maintain reactivity.

I would also argue that a fast reactor can achieve very high burnup between each reprocessing. This is because fission product poisoning is less of an issue at fast energies. Uranium from sea water may put an upper limit on the cost of natural uranium, limiting recycling.

Also, while lithium might be more prevalent than uranium, lithium has more uses than uranium mainly batteries.

 

[mfb]

Also, while lithium might be more prevalent than uranium, lithium has more uses than uranium mainly batteries.

Sure, but the amount needed for nuclear reactors would be small. That is true for most things fusion or fission reactors need / would need, with cooling water as a huge exception.

 

[Hologram0110]

Sure, but the amount needed for nuclear reactors would be small. That is true for most things fusion or fission reactors need / would need, with cooling water as a huge exception.

Looks like your right. Wikipedia says global production is about 600 000 tonnes a year. I can't find any core inventory numbers on how much is in a MSR, but I'd guess it is relatively small (maybe 10 tonnes?)

 

[Astronuc]

Looks like your right. Wikipedia says global production is about 600 000 tonnes a year. I can't find any core inventory numbers on how much is in a MSR, but I'd guess it is relatively small (maybe 10 tonnes?)

I think it depends on the size of the core and power density.

A ~3400 MWt (~1100 MWe) PWR has about 100 MT of fuel.

Mostly likely modern MSRs would be smaller, as in modular.

I would like to see how they would distribute the enrichment, or moderation, radially, as well as axially. And I'm curious about the delayed neutron fraction.

 

[mheslep]

Typical burnup for a PWR or BWR is supposedly ~45 GWdays/ton, i.e. ~5%. By "high burnup" in the case of thorium, I mean the like of 170 GWd/t which has already been attained from experience at Fort St Vrain, and upwards of 96% per theory.
http://www.world-nuclear.org/info/Current-and-Future-Generation/Thorium/

 

[Astronuc]

Typical burnup for a PWR or BWR is supposedly ~45 GWdays/ton, i.e. ~5%. By "high burnup" in the case of thorium, I mean the like of 170 GWd/t which has already been attained from experience at Fort St Vrain, and upwards of 96% per theory.
http://www.world-nuclear.org/info/Current-and-Future-Generation/Thorium/

PWRs fuel typically leads BWR fuel in burnup by about 5 GWd/tU. We consider batch, assembly and rod average, as well as peak pellet. The higher the average burnups in the batch and assembly, the greater the utilization. Some PWR fuel approaches about 50 GWd/tU batch average, with the peak assembly a bit higher, and peak rod approaching 59/60 GWd/tU. Some core may be higher, but LWR fuel is limited on a peak rod average burnup of about 62 GWd/t (current US NRC approved license limit).

In a fixed lattice system, the burnup is limited due to cladding integrity issues, as well as fission product accumulation, and technical limits on rod internal pressure due to fission gas production and release. Fast reactor fuel usually has a very large plenum volume to accommodate a large volume of fission gas.

Burnup in a LFTR is a bit different since the fuel is reprocessed online. It could conceivably approach 90+% FIMA (with 1% FIMA ~ 9.7 GWd/tHM). The relationship between FIMA depends on the energy per fission.

LFTRs with reprocessing have the benefit of removing fission products, including fission gases.

Some fast reactor fuel can approach 200 GWd/tU, and higher.

 

[Moniz_not_Ernie]

For a thorium fueled MSR, as per the OP, online reprocessing is unavoidable.

We need some definitions here. Processing is too broad a term. Some folks will think immediately of PUREX, and that is way off.
There is one on-line process common to all MSR designs. That is called an off-gas system. The 135-Xenon, in particular, will bubble out of the liquid with a little encouragement. The Krypton comes out, too, and traces of other volatiles. This is stored separately, which is, of course, a possible path to failure. Balanced against that risk, the volatiles are already "in the can" if a spill occurs.
A second system is used if Thorium is the fuel, one example of which is the LFTR. It adds Fluorine gas to the blanket salt, which gasifies the 233-U. This is then pumped into the core as makeup fuel. There is no plan to separate the 233-Pa. That was an ORNL notion from a time when getting breeding ratios well over one was considered essential. Given the discovery of plenty of uranium ore, that is no longer a priority. (Just making the blanket volume bigger reduces most of the 232-Pa absorption losses.)
The Denatured MSR (ORNL's "fourth" design) had only an off-gas system. It was supposed to work for twenty years with no processing other than removing the inert gases.
I have heard of three other forms of online processing. One uses Cerium to shove out other, neutron-hungry rare Earth's. Another uses molten Bismuth, I forget why. The most promising is a simple vacuum distillation pot. Hopefully, the neutron-hungry and low soluble (jewelry) elements will stick to the bottom of the pot as the expensive fuel and salt boil off and get put back in the core.
In my opinion, the off-gas system and distillation are nowhere near as complicated, expensive, dirty, and dangerous as PUREX or any other process associated with solid fuels.

 

[mheslep]

A second system is used if Thorium is the fuel, one example of which is the LFTR... Given the discovery of plenty of uranium ore, that is no longer a priority...

The latter implies, what, for a thorium machine? That the design would continuously have its neutron count boosted by addition of uranium (not just at start up)? If so it once again becomes dependent on an enriched fuel cycle, discarding a major advantage of a thorium design.

There is no plan to separate the 233-Pa. That was an ORNL notion from a time when getting breeding ratios well over one was considered essential.

So three neutrons per fission event can be thrown away without concern?
232-Th+n ->233-Pa -> 233-Pa +n -> 234-Pa -> beta + 234-U -> 234-U+n -> 235-U

 

[mheslep]

Some fast reactor fuel can approach 200 GWd/tU, and higher.

Demonstrated? Even if a fast reactor enables consumption of other actinides, how does a solid fueled fast reactor dispense with poisons, or cladding degradation, or any of the other problems that are likely to demand fuel removal before completion of a high burn up?

 

[Moniz_not_Ernie]

The latter implies, what, for a thorium machine? That the design would continuously have its neutron count boosted by addition of uranium (not just at start up)? If so it once again becomes dependent on an enriched fuel cycle, discarding a major advantage of a thorium design.

So three neutrons per fission event can be thrown away without concern?
232-Th+n ->233-Pa -> 233-Pa +n -> 234-Pa -> beta + 234-U -> 234-U+n -> 235-U

The first neutron is required to produce 233-U anyway.
233-Pa has a capture cross-section of 39.5, and a half-life of a month. It is a bit of a race, but I think far more will decay before it absorbs. I don't know the math needed to prove that. I just figure a fuel atom "lives" in the core for maybe a year without fissioning. I get that from the following logic: If new fuel is 5% enriched, and (say four years later) old fuel is 1%, an individual atom has 20% chance of surviving four years.
The fuel has a fission cross-section of 530. Throw in the thinner neutron flux in the blanket, compared to the core. Might that come to only 1% losses to 234-U? (Granted, two neutrons are lost by that route.)

 

[Astronuc]

Demonstrated? Even if a fast reactor enables consumption of other actinides, how does a solid fueled fast reactor dispense with poisons, or cladding degradation, or any of the other problems that are likely to demand fuel removal before completion of a high burn up?

Fast reactor fuel uses stainless steel cladding in a liquid metal coolant. The end of life issues are different that those of a Zr alloy in a water-cooled environment. Fast reactor fuel is also enriched to about 20% fissile, which is usually Pu-239/240/241 and U in (U,Pu)O₂ or (U,Pu)C or (U,Pu)N. One can return a fuel element to the core and drive it with fresh fuel nearby, so the burnups achieved can be quite high. Xe-135/135m, which is a strong poison in an LWR thermal neutron flux is not so strong in a fast flux.

In LWRs, used fuel assemblies can be 'driven' to higher burnups by face adjacent fresh fuel. In addition, high burnup fuel rods can be placed into locations in fresh fuel assemblies and those older fuel rods can be 'pushed' to higher burnups. However, several key issues for high burnup LWR fuel include fission gas release and rod internal pressure, oxidation of the cladding (leading to metal wall loss), hydriding of the cladding (leading to brittle cladding or enhanced oxidation, and growth of the fuel rod in an assembly not designed for increased growth.

Some LWR test fuel has gone above 60 GWd/tU up to ~100 GWd/tU, but that was done under special test conditions not in a commercial reactor. Commercial reactors have much more severe limits due to safety concerns.

 

[Astronuc]

The first neutron is required to produce 233-U anyway.
233-Pa has a capture cross-section of 39.5, and a half-life of a month. It is a bit of a race, but I think far more will decay before it absorbs. I don't know the math needed to prove that. I just figure a fuel atom "lives" in the core for maybe a year without fissioning. I get that from the following logic: If new fuel is 5% enriched, and (say four years later) old fuel is 1%, an individual atom has 20% chance of surviving four years.
The fuel has a fission cross-section of 530. Throw in the thinner neutron flux in the blanket, compared to the core. Might that come to only 1% losses to 234-U? (Granted, two neutrons are lost by that route.)

This is only part of the story. In an LWR, with 5% enrichment, i.e., 5% U-235 and 95% U-238, some of that U-238 is converted to Pu-239, -240, -241, and other transuranic radionuclides that accumulate with time. So even though U-235 is depleted to 1% or so, at high burnup, the Pu-239 is being fissioned, and at some point, more so than the U-235, which is partially shielded by the Pu-239.

LWRs have a fast as well as thermal population of neutrons. Some fissions (~8-10%) are fast fissions, while the rest occur from thermal-neutron induced fissions. Pu-239 (and other transuranics) accumulates on the periphery of the fuel pellets, and during the course of a lifetime, the U-235 in the interior of the pellet is shielded by the Pu-239 on the exterior. If a fuel pellet has a burnup of 60 GWd/tU, then the rim (outer 10%) has a very high burnup of ~150-180 GWd/tU. This induces the so-called 'rim effect' seen in high burnup LWR fuel.

 

[Moniz_not_Ernie]

This is only part of the story. In an LWR, with 5% enrichment, i.e., 5% U-235 and 95% U-238, some of that U-238 is converted to Pu-239, -240, -241, and other transuranic radionuclides that accumulate with time. So even though U-235 is depleted to 1% or so, at high burnup, the Pu-239 is being fissioned, and at some point, more so than the U-235, which is partially shielded by the Pu-239.

LWRs have a fast as well as thermal population of neutrons. Some fissions (~8-10%) are fast fissions, while the rest occur from thermal-neutron induced fissions. Pu-239 (and other transuranics) accumulates on the periphery of the fuel pellets, and during the course of a lifetime, the U-235 in the interior of the pellet is shielded by the Pu-239 on the exterior. If a fuel pellet has a burnup of 60 GWd/tU, then the rim (outer 10%) has a very high burnup of ~150-180 GWd/tU. This induces the so-called 'rim effect' seen in high burnup LWR fuel.

Interesting geometry there. And I didn't know the fissions from fast neutrons was so high. But that depends on your definition of "fast". I have seen fission yield data for three speed bins, labeled 0.0253, 500K, and. 14 million eV. Do you include the middle energy bin in you 8-10%? What is your upper limit for slow neutrons?
I ask because I'm designing an MSR simulation. One basic question is whether we want fast or slow reactors. However, if the data comes in three speeds, So I may allow the user to choose reactor specs a little precisely. It will probably require aggregating ENDSF data into two or three bins. Wish I could contract that task out, but I'm on a tight budget ($=0).
I'm not a nuclear engineer. I'm a simulation designer, who got interested in this stuff through my hobbies; first astronomy (your username is intriguing), then cosmology and stellar evolution. I'll be asking some newbie questions here.

 

[Stanley514]

This one article states that Lithium or Beryllium are not consumed by LFTR.
http://www.2112design.com/blog/lftr/
Could somebody verify it?