Sources for nuclear fuel pellet fabrication methods?

[Phys pilot]

TL;DR Summary

Looking for sources on nuclear fuel fabrication, specifically the creation of fuel pellets (mchanical and ceramic methods)

Hello,
I'm looking for some sources and bibliography focused on nuclear fuel fabrication.
In this context, I am referring to the creation of fuel pellets, excluding the uranium enrichment procedures. I am aware of at least two methods, the mechanical and the ceramic one.

With fabrication I don't mean the procedure of enriching the uranium but the fabrication of the pellets, as far as I know there are at least two methods, the mechanical and the ceramic one, I'm specially interested in the ceramic one.

I'm particularly interested in the ceramic method of nuclear fuel pellet fabrication. I would appreciate any recommendations or resources that address this specific aspect.
Thank you!

 

[sbrothy]

Forgive me if this is daft question, but you're referring to classical nuclear reactors as opposed to fusion, right?

EDIT: BTW, I wouldn't be surprised if much of this information is classified.

 

[Astronuc]

TL;DR Summary: Looking for sources on nuclear fuel fabrication, specifically the creation of fuel pellets (mchanical and ceramic methods)

Hello,
I'm looking for some sources and bibliography focused on nuclear fuel fabrication.
In this context, I am referring to the creation of fuel pellets, excluding the uranium enrichment procedures. I am aware of at least two methods, the mechanical and the ceramic one.

With fabrication I don't mean the procedure of enriching the uranium but the fabrication of the pellets, as far as I know there are at least two methods, the mechanical and the ceramic one, I'm specially interested in the ceramic one.

I'm particularly interested in the ceramic method of nuclear fuel pellet fabrication. I would appreciate any recommendations or resources that address this specific aspect.
Thank you!

One could look through US and EU patents on various processes over the past 70 years. One might find some discussion on manufacturers websites, or in older transactions of conferences on LWR fuel performance.

Light water reactors typically use UO₂, since it has a high melting point and is more stable than metal fuel in contact with water.

There are two routes (wet or dry) one could take starting with UF₆, which is converted to uranyl fluoride, UO₂F₂. In the wet process, it would be converted to uranyl nitrate hexahydrate (UNH), then precipitated, or ammonium diuranate (ADU) or ammonium uranyl carbonate (AUC). In some cases, ADU or AUC may be co-precipitated with UNH. The precipitate is then thermally reduced to a UO₂ powder. The powder is then processed with additives by milling, slugging and granulating to get a coarse powder that will 'flow' into a die. The powder is blended with a binder that serves as a pore former and die lubricant, and ultimate poured into a die (usually on a rotary press) where two die punches compress the powder to cylindral pellets to a density around 50 to 60% of theoretical density (green ceramic). Some 'addback' (recycled powder and pellet grains) may be added to the powder before preparation for pressing, and in some cases the recycled powder/pellets may be oxidized to U₃O₈.

The dry process is more recent in which the UF₆ is reacted with steam to form UO₂F₂, which is then reacted with more steam and hydrogen to reduced the uranyl fluoride to UO₂. The powder is very fine and must be processed to make it active, i.e., 'flowable', then it is processed as described above for powder to pellet.

The green ceramic pellet is sintered, initially heated at lower temperature to volatize the binder (pore former and die lubricant), then at high temperature 1650-1850°C for hours (the lower the temperature the longer the time) in a reducing environment (hydrogen from cracked ammonia) to the desired density of about 95 to 98% of theoretical density.

 

[Phys pilot]

Forgive me if this is daft question, but you're referring to classical nuclear reactors as opposed to fusion, right?

EDIT: BTW, I wouldn't be surprised if much of this information is classified.

I'm referring to classical nuclear reactors. Basically uranium pellets. There is not much info so I guess that manufacturers have their own methods and they don't public them.

One could look through US and EU patents on various processes over the past 70 years. One might find some discussion on manufacturers websites, or in older transactions of conferences on LWR fuel performance.

Light water reactors typically use UO₂, since it has a high melting point and is more stable than metal fuel in contact with water.

There are two routes (wet or dry) one could take starting with UF₆, which is converted to uranyl fluoride, UO₂F₂. In the wet process, it would be converted to uranyl nitrate hexahydrate (UNH), then precipitated, or ammonium diuranate (ADU) or ammonium uranyl carbonate (AUC). In some cases, ADU or AUC may be co-precipitated with UNH. The precipitate is then thermally reduced to a UO₂ powder. The powder is then processed with additives by milling, slugging and granulating to get a coarse powder that will 'flow' into a die. The powder is blended with a binder that serves as a pore former and die lubricant, and ultimate poured into a die (usually on a rotary press) where two die punches compress the powder to cylindral pellets to a density around 50 to 60% of theoretical density (green ceramic). Some 'addback' (recycled powder and pellet grains) may be added to the powder before preparation for pressing, and in some cases the recycled powder/pellets may be oxidized to U₃O₈.

The dry process is more recent in which the UF₆ is reacted with steam to form UO₂F₂, which is then reacted with more steam and hydrogen to reduced the uranyl fluoride to UO₂. The powder is very fine and must be processed to make it active, i.e., 'flowable', then it is processed as described above for powder to pellet.

The green ceramic pellet is sintered, initially heated at lower temperature to volatize the binder (pore former and die lubricant), then at high temperature 1650-1850°C for hours (the lower the temperature the longer the time) in a reducing environment (hydrogen from cracked ammonia) to the desired density of about 95 to 98% of theoretical density.

Hi Astronuc, I appreciate your response, your knowledge shown in this forum always impresses me, especially considering the limited information I could find these methods. Thank you.

I'll explore public patents to get a better understanding of the process. As you mentioned, I did come across some information on manufacturer websites, but it wasn't very comprehensive. Thanks once again

 

[sbrothy]

I am nowhere near as knowledgeable as either Astronuc or indeed the great majority here to put it mildly, but I know how to search. I'm guessing though that you know too so these are probably not new to you:

Preparation of powdered uranium oxides by microwave heating of substandard ceramic pellets of oxide nuclear fuel.

Ceramic processing of uranium–plutonium mixed oxide fuels (U1−yPuy)O2 with high plutonium content.

Ceramics as nuclear reactor fuels.

They're probably behind paywalls though.

 

[Astronuc]

I'm referring to classical nuclear reactors. Basically uranium pellets. There is not much info so I guess that manufacturers have their own methods and they don't public them.

Correct, manufacturer processes are indeed proprietary, since the IP provides some advantage in theory, although practically, that is not nearly so clear. Similarly, the cladding design and materials (primarily materials in terms of composition and microstructure) are proprietary (trade secret).

Hi Astronuc, I appreciate your response, your knowledge

I've had about 40 years experience doing fuel performance modeling and simulation, and more recently material/component design, in addition to nuclear reactor design. I was fortunate to spend about 10 years out of graduate school doing technical surveillance and QA/QC audits of nuclear fuel fabrication from the raw material to the finish product (fuel rods/assemblies) at most of the US, European and Asian manufacturers. I learned a first hand.

During my undergraduate and graduate programs, I took courses in materials engineering and science, and corrosion. I figured I would need to learn/know about materials' behavior in a variety of service environments.

I missed an opportunity to take a foundry course (Engineering Technology), as I was unaware until too late. I also which I had taken some classes in welding, which itself is a serious art.

The hot area now is Additive Manufacturing (AM). How one adapts that to producing bulk materials in the tens of tons range is a challenge - otherwise, one still has to do large casting and forgings, or welded plate.

 

[Astronuc]

Tens of of tons range?!

For LWRs, the reactor pressure vessels are large, ~350 to 500 t, depending on vessel diameter and height.

ENS indicates a 1300 MWe (~3600 MWt) reactor vessel is about 530 t, excluding the internals.
https://www.euronuclear.org/glossary/reactor-pressure-vessel/

A SONGS decommissioning site mentions the unit 2 reactor head (cap of the RPV) weighed 77 t, and the RPV from Unit 1 weighed 770 t, which seems 2x heavy from what I remember. I suspect it was filled with internals and concrete, but I'd have to check. Unit 1 was an older Westinghouse design, while Units 2 and 3 were Combustion Engineering designs, which were much larger than Unit 1.
https://www.songscommunity.com/deco...loads-part-of-decommissioning-a-nuclear-plant

However, let us stay on topic, which is the manufacturing of nuclear fuel (ceramic) pellets for LWR fuel.

 

[Astronuc]

I'll explore public patents to get a better understanding of the process. As you mentioned, I did come across some information on manufacturer websites, but it wasn't very comprehensive. Thanks once again

This link provides a paper that gives some process outlines.
https://www.osti.gov/etdeweb/servlets/purl/20269232

Regarding dopants for increasing grain size, chromia, is widely accepted now. To my knowledge, neither titania, nor nobia, worked as hoped. Alumina and alumina-silica have also been used. There are patents on dopants as well.

For burnable poisons, gadolinia and erbia, have been added to UO₂, and doping with a little (< ~1%) gadolinia (and probably erbia) seems also to reduce fission gas release. One also havs to be concerned with the reduction of pellet (ceramic) thermal conductivity with increasing dopant levels. As I recall, a rough ballpark estimate is that 1% gadolinia reduces thermal conductivity equivalent to a burnup of 4 GWd/tU (burnup also reduces thermal conductivity of UO₂ (or MOX) ceramic fuel).

The processes for MOX (mixed oxide, e.g., (U,Pu)O₂, (U,Th)O₂, (Th,Pu)O₂) are much the same as for UO₂.