Uranium and Diamonds

  • #1

Main Question or Discussion Point

 
Last edited by a moderator:

Answers and Replies

  • #2
Astronuc
Staff Emeritus
Science Advisor
18,753
1,807
It's not clear that this would be economical since each pellet would have to be coated (on the circumferential surface), which is done for certain fuel pellet designs. The bulk of the thermal gradient is within the ceramic pellet. UO2 simply has poor thermal conductivity compared to UN or UC, or some metal alloys. LWR fuel does require an enrichment of the U in U-235, and in the course of irradiation, some U-238 is converted to Pu-239/240/241, which also fission. The benefit of the oxide is that is traps a number of fission products.

Another consideration for a pellet coating is the pellet cracking due to differential thermal expansion within the ceramic pellet.
 
Last edited by a moderator:
  • #3
It's not clear that this would be economical since each pellet would have to be coated (on the circumferential surface), which is done for certain fuel pellet designs. The bulk of the thermal gradient is within the ceramic pellet. UO2 simply has poor thermal conductivity compared to UN or UC, or some metal alloys. LWR fuel does require an enrichment of the U in U-235, and in the course of irradiation, some U-238 is converted to Pu-239/240/241, which also fission. The benefit of the oxide is that is traps a number of fission products.

Another consideration for a pellet coating is the pellet cracking due to differential thermal expansion within the ceramic pellet.
Wouldn't a durable diamond surface be another way to trap fission products if the differential thermal expansion can be solved? If I understand you correctly I think the process of coating might be able to overcome the problem as the coating only has to be a few microns thick so tiny uranium particles could be coated and then merged and coated again into ever larger particles until you have the desired pellet size . . .

JDM
 
  • #4
The bulk of the thermal gradient is within the ceramic pellet. UO2 simply has poor thermal conductivity compared to UN or UC, or some metal alloys.
I wonder how a UO2/Zr cermet would fare in this regard. (IIRC in cermets used for cutting tools metal content is usually only about 10%. Since in this case we aren't going for strength, we can use even lower percentage of Zr metal.)
 
  • #5
Astronuc
Staff Emeritus
Science Advisor
18,753
1,807
Wouldn't a durable diamond surface be another way to trap fission products if the differential thermal expansion can be solved? If I understand you correctly I think the process of coating might be able to overcome the problem as the coating only has to be a few microns thick so tiny uranium particles could be coated and then merged and coated again into ever larger particles until you have the desired pellet size . . .
The differential thermal expansion is a consequence of the fission/energy distribution in the pellet, which at steady state establishes an approximately parabolic temperature distribution between the surface of the pellet and the center. The difference in temperature can be a few hundred °C to about 1000 °C depending on the pellet fission rate. The UO2 ceramic has grain sizes on the order of 10 microns to 20 microns (with some grains smaller and some larger). Adding inert material such as diamond or metal like Zr as nikkkom indicated with UO2/Zr cermet means that U would be displaced, so the enrichment would have to be increased to get to the necessary enrichment.

UO2 pellets are fabricated in a bulk/batch process where the power is pressed into pellet in a rotary press to a density of about 55-60% of theoretical density. The pressed (green) pellets are then collected in Mo boats or on sheets and are then sintered in a low humidity H2 (from cracked ammonia) environment at 1600-1800 °C for 2 to several hours depending on the temperature.

Modern PWR fuel cycles use enrichments up to 4.9 to 4.95% U-235 in U. The maximum enrichment is always just below 5% since 5% is the legal limit in commercial LWR fuel, and the 0.1 to 0.05% difference is there to allow for uncertainty in the manufacturing process.
 
Last edited:
  • #6
People have been looking at Molybdenum matrix cermet fuel:

http://www.nrg.eu/docs/nrglib/2004/2004_nucl_techn_146_3_bakker_klaassen.pdf [Broken]

Looks quite good, although they need to remove Mo-95.
 
Last edited by a moderator:
  • #7
mheslep
Gold Member
311
728
The metal fuel rod technology from your TR reference would be superior in heat conduction to any attempt to place a *coating* around some traditional oxide fuel, as the improvement of heat conduction from the center of the fuel to the edge of the fuel would greater than any aide offered by a coating alone.
 
Last edited by a moderator:
  • #8
mheslep
Gold Member
311
728
The differential thermal expansion is a consequence of the fission/energy distribution in the pellet, which at steady state establishes an approximately parabolic temperature distribution between the surface of the pellet and the center. The difference in temperature can be a few hundred °C to about 1000 °C depending on the pellet fission rate. The UO2 ceramic has grain sizes on the order of 10 microns to 20 microns (with some grains smaller and some larger). Adding inert material such as diamond or metal like Zr as nikkkom indicated with UO2/Zr cermet means that U would be displaced, so the enrichment would have to be increased to get to the necessary enrichment.

UO2 pellets are fabricated in a bulk/batch process where the power is pressed into pellet in a rotary press to a density of about 55-60% of theoretical density. The pressed (green) pellets are then collected in Mo boats or on sheets and are then sintered in a low humidity H2 (from cracked ammonia) environment at 1600-1800 °C for 2 to several hours depending on the temperature.

Modern PWR fuel cycles use enrichments up to 4.9 to 4.95% U-235 in U. The maximum enrichment is always just below 5% since 5% is the legal limit in commercial LWR fuel, and the 0.1 to 0.05% difference is there to allow for uncertainty in the manufacturing process.
As I understand the metal rod concept from Lightbridge, the idea is to replace oxygen in UO2, not uranium.
 
  • #9
mheslep
Gold Member
311
728
I wonder how a UO2/Zr cermet would fare in this regard. (IIRC in cermets used for cutting tools metal content is usually only about 10%. Since in this case we aren't going for strength, we can use even lower percentage of Zr metal.)
The reference from the OP refers to a U/Zr alloy metal.
http://www.technologyreview.com/news/530981/new-nuclear-fuel-could-boost-reactors-but-also-safety-worries/

The Lightbridge fuel is instead made of zirconium/uranium alloy, with a cross configuration and spiral shape, which makes it look like a piece of Twizzlers candy. The metal composition means heat transfers far faster, and the shape increases the contact area between fuel and water by more than 35 percent. To cope with the increased intensity, water must move through the reactor core more quickly, but existing water pumps can handle this because the fuel provides less resistance to the flow.
...
Inserted in a conventional reactor, the new fuel could boost power 10 percent. Replacing equipment including turbines with larger-size ones would increase that to 17 percent, Lightbridge says.
 
Last edited:
  • #10
Astronuc
Staff Emeritus
Science Advisor
18,753
1,807
Lightbridge has an interesting concept that has yet to be proven in an LWR environment. Such fuel will have potential issues with fuel-cladding chemical interaction, depending on the fuel alloy. Metal fuel is beneficial from a fissile density and thermal conductivity perspective, but there are other disadvantages with respect to reduced melting temperature and fission product mobility.

The OP mentioned uranium pellets, and the bulk of commercial fuel is comprised of UO2 ceramic pellets clad in zirconium alloy tubing. Even most fast reactor fuel is UO2 or (U,Pu)O2, although some carbide, nitride and carboxide or carbonitride fuel has been tested.

In addition to normal operation, a designer must also consider the off-normal and failed fuel situations, as well as accident (RIA and LOCA) situations, and the consequences of the fuel form and it's behavior in the different situations. UO2 is more chemically stable in high temperature water coolant than UC/MC or UN/MN, where M = (U, Pu).

I'm not aware of actual fuel fabricated by LB unless it's in Russia.
 
Last edited:
  • #11
Carbide and nitride fuels have a problem of generating C-14, which bioaccumulates and has a very nasty half-life of 5700 years.
 
  • #12
mheslep
Gold Member
311
728
Lightbridge has an interesting concept that has yet to be proven in an LWR environment.
The reference suggests metal alloys have been tried in subs, thus military LWRs.


but there are other disadvantages with respect to reduced melting temperature and fission product mobility.
I would think the melting point of UO2 is not the primary accident event issue, post Fukushima, but rather the hydrogen production from Zirc metal when oxidized by water, which is well under way at 700C. Fission product mobility, as you suggest, seems like an intractable problem for a system that won't well tolerate those hot isotopes in the loop.
 
  • #13
Astronuc
Staff Emeritus
Science Advisor
18,753
1,807
The reference suggests metal alloys have been tried in subs, thus military LWRs.
or more generally naval reactors. I believe metal fuel has been used in reactors of Russian icebreakers, e.g., KLT-40 reactor.
http://scienceandglobalsecurity.org/archive/sgs14diakov.pdf
http://www.iaea.org/NuclearPower/Downloadable/aris/2013/25.KLT-40S.pdf

I would think the melting point of UO2 is not the primary accident event issue, post Fukushima, but rather the hydrogen production from Zirc metal when oxidized by water, which is well under way at 700C. Fission product mobility, as you suggest, seems like an intractable problem for a system that won't well tolerate those hot isotopes in the loop.
Certainly the rapid oxidation and disintegration of the cladding (Zr-alloy or otherwise) is a concern. There are other design considerations on the cladding as well. Not too many systems handle high temperature steam conditions, especially as the temperature exceeds 500°C for long periods, or possibly as may have been the case with Fukushima, > 1000°C.

With regard to diamond coating of fuel or fuel particles, one has to look at the engineering feasibility and cost, and one must be familiar with the technology, much of which is proprietary or otherwise sensitive.

There is also the neutron damage to the diamond structure, which would also change the composition through transmutation, in addition to displaced atoms.
 
  • #14
jim hardy
Science Advisor
Gold Member
2019 Award
Dearly Missed
9,839
4,868
I can only speculate here..

Interesting thought. Mix diamond right in with the fuel pellet.

What would be effect of having so many carbon atoms in such proximity to the fuel ?
Seems to me moderator temperature feedback would be really quick.

But would such energetic neutrons break up the diamond crystal lattice?
 
  • #15
I can only speculate here..

Interesting thought. Mix diamond right in with the fuel pellet.
In the limit, you end up with just having uranium carbide fuel. Which may be not that bad, although I don't like that neutron activation will produce lots of C-14.
 
  • #16
QuantumPion
Science Advisor
Gold Member
902
42
In the limit, you end up with just having uranium carbide fuel. Which may be not that bad, although I don't like that neutron activation will produce lots of C-14.
No it won't. Carbon does not absorb neutrons. That is why it is a good moderator. C-14 in the environment is produced from nitrogen.
 
  • #17
jim hardy
Science Advisor
Gold Member
2019 Award
Dearly Missed
9,839
4,868
The absorbtion cross section for C13 is small but finite, around a millibarn.
Nitrogen's is a bit shy of 2 whole barns.
http://www.isis.stfc.ac.uk/learning/neutron-training-course/downloads/general/sears---neutron-cross-section-table10655.pdf

useless and boring trivia -
That's why we purged our incore neutron detector tubes with CO2 not air. Flux in there was around 10^14nv.
We lubricated the steel drive cables with graphite. As best i recall the prominent activation product on the drive cables was cobalt. Of course the little fission chambers themselves (Oralloy) got the hottest.

old jim
 
  • #18
201
10
So I was wondering given that its been shown that a more thermally conductive material around uranium pellets improves the efficiency of a power plant and that diamond films are cheap http://www.sciencedaily.com/releases/2013/06/130628102929.htm does coating uranium pellets with diamond make sense?
I know a Czech researcher working with a group studying diamond coatings who is studying coating the outside of the cladding material to reduce oxidation during accidents. Another research group is looking at using diamond coating on the inside of the cladding to mitigate some chemical interactions between fission products and the Zr. Neither of these are trying to take advantage of the thermal conductivity but instead the chemical and mechanical stability. The coating in question is a mix of diamond and graphite (to give flexibility and toughness).

The biggest issue with UO2 fuel is that it performs best at low temperatures (thermal expansion, fission product retention, margin to melting ect) but it has a poor thermal conductivity. To improve this property people have suggested doping UO2 to produce heterogeneous materials with better macroscopic heat transfer properties. A number of materials are being looked at mostly BeO, SiC, diamond and nano-tubes. In these cases it isn't a coating, but instead bulk material added to the pellet. You could imagine a pellet with many strings of high thermal conductivity material carrying heat from the center of the fuel to the surface.
 
  • #19
No it won't. Carbon does not absorb neutrons. That is why it is a good moderator. C-14 in the environment is produced from nitrogen.
At high neutron flux figures typical for power reactors at 100% power, some carbon in carbide _will_ absorb neutrons. Fuel is exposed to this flux for at least one year, typically more.
 
  • #20
In these cases it isn't a coating, but instead bulk material added to the pellet. You could imagine a pellet with many strings of high thermal conductivity material carrying heat from the center of the fuel to the surface.
Yes, that's why cermet (a ceramic sintered with about 10% of metal powder) immediately pops into my mind. Its thermal conductivity is higher than of pure ceramic.

It is a well-known class of materials, used for cutting tools and other high tech applications. There are tons of research on it - for example, how different sizes of initial powders affect the resulting material, what happens if you add a powder (possibly of a different metal) which consists of tiny needle-like particles, not spheres (some tests show that it tends to reinforce material, like microscopic rebar)...

As a starting point, I'd try a cermet with uranium metal (added benefit of more uranium in the fuel, by weight). Also l'd try depositing a thin layer of Zircalloy on the cermet pellet's surface - there's hope this can eliminate chemical interaction with cladding tube.
 
  • #21
201
10
Yes, that's why cermet (a ceramic sintered with about 10% of metal powder) immediately pops into my mind. Its thermal conductivity is higher than of pure ceramic. As a starting point, I'd try a cermet with uranium metal (added benefit of more uranium in the fuel, by weight). Also l'd try depositing a thin layer of Zircalloy on the cermet pellet's surface - there's hope this can eliminate chemical interaction with cladding tube.
U metal in an LWR/PHWR is pretty though sell in terms of safety. From a reactor physics U metal is great (density, thermal conductivity) but it is far too chemically reactive during accidents and won't contain fission products. It would significantly improve reactor performance in a number of ways. It would be great if you could rely on no cladding failures to eliminate the need for fuel/coolant chemical compatibility, but historical evidence doesn't support that. There are too many failure modes for fuel cladding (core damage, debris fretting, corrosion, over pressure).

In reactors cooled with something other than water you may be on to something since you don't need to worry about fuel oxidation.
 
  • #22
U metal in an LWR/PHWR is pretty though sell in terms of safety. From a reactor physics U metal is great (density, thermal conductivity) but it is far too chemically reactive during accidents and won't contain fission products.
I know about this concern. That's why I'd look into adding an additional barrier - a thin coating of Zr over the cermet pellet. Even if Zr tube fails and water touches pellets, it won't come into contact with uranium metal - it will touch zirconium.

Historical evidence is not very nice towards current breed of purely ceramic pellets either. Since specific powers have gone up, and burnup gone up too, even ceramic gets damaged: IIUC pellets crack and swell, and if Zr tube fails and water touches pellets, significant fraction of more volatile fission products, namely krypton and xenon, iodine, caesium, strontium get washed out.

Cermet pellets may fare better because they would be under far lesser thermal stress.
 
  • #23
201
10
I know about this concern. That's why I'd look into adding an additional barrier - a thin coating of Zr over the cermet pellet. Even if Zr tube fails and water touches pellets, it won't come into contact with uranium metal - it will touch zirconium.

Historical evidence is not very nice towards current breed of purely ceramic pellets either. Since specific powers have gone up, and burnup gone up too, even ceramic gets damaged: IIUC pellets crack and swell, and if Zr tube fails and water touches pellets, significant fraction of more volatile fission products, namely krypton and xenon, iodine, caesium, strontium get washed out.

Cermet pellets may fare better because they would be under far lesser thermal stress.
To prevent fuel coolant interaction your fuel would have to be resistant to cracking due to the temperature gradient. This is a significant issue for most fuels. For solid-metal fuels swelling is a big problem which would also destroy the integrity of any coating on the surface. The idea of diamond coating on the outer surface of fuel (or inner surface of the cladding) is certainly helpful, but doesn't fix the core of the problem steam interactions with fuel can still happen.

I've read about a fuel design which was briefly tested which used liquid metal fuel in a stainless steel cladding. The idea is that you design the fuel to melt by design, thus you do not have complicated fuel-cladding interaction, virtually no fuel swelling and good thermal contact. Unfortunately the problem was chemical reactions with the cladding (due to fission products and radiation). Such a design might benefit from a diamond coating.

Personally I think the simpler answer to increase the surface area to volume ratio on fuel by moving away from cylindrical fuel pins (like the lightbridge design). This lowers the temperate improving all the other properties.
 
  • #24
Here's a link where cermet properties with different metals, ceramics and grain sizes are looked at. They look at it from cutting tool POV, specifically woodcutting. It turns out some grades of wood are somewhat acidic, they literally dissolve the tool as it cuts:

http://www.carbideprocessors.com/pages/carbide-parts/making-cermet-material.html

Other attempts at reducing fuel temperature gradients are looked at by Russians, they experiment with fuel in a form of small ceramic grains, almost powder, vibro-packed into a Zr tube.
 
  • #25
QuantumPion
Science Advisor
Gold Member
902
42
At high neutron flux figures typical for power reactors at 100% power, some carbon in carbide _will_ absorb neutrons. Fuel is exposed to this flux for at least one year, typically more.
You said "lots". Specifically, you inferred that uranium carbide fuel would be undesirable merely due to C-13 activation. The amount of C-14 produced would be totally insignificant compared to activation products from any other impurities in the fuel, cladding, or coolant.
 

Related Threads on Uranium and Diamonds

  • Last Post
Replies
6
Views
5K
  • Last Post
6
Replies
127
Views
21K
  • Last Post
Replies
11
Views
7K
  • Last Post
Replies
7
Views
4K
  • Last Post
Replies
1
Views
2K
  • Last Post
Replies
7
Views
8K
  • Last Post
Replies
15
Views
4K
  • Last Post
Replies
9
Views
27K
  • Last Post
Replies
14
Views
16K
Top