Energy From Thorium Discussion Forum

It is currently Dec 14, 2017 3:59 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 34 posts ]  Go to page 1, 2, 3  Next
Author Message
PostPosted: Oct 31, 2013 4:57 pm 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
One of the challenges in making a reactor concept is to calculate the physical properties of a liquid salt mixture.

The most promising salt mixtures for a chloride salt reactor is a mixture of NaCL-KCl-UCl3-PuCl3 plus NdCl3 and BaCl2, CsCl (the last 3 of them are the most commen fission product chlorides).

There seems to be a calculation method to calculate the phase diagrams of multi-component salt mixtures with the Gibbs energies.

Does anyone know a method that a simple minded engineer can understand how to calculate the liquidus temperature of a multi-component salt mixture?

Does anyone know a method ...how to calculate the specific heat of a multi-component salt mixture?

Is there a reference where to find the decomposition temperature of SmCl3, PmCl3?

Holger


Top
 Profile  
 
PostPosted: Oct 31, 2013 5:44 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
Unfortunately there are no simple methods to calculate liquidus temperatures of multi component salt mixtures. There are theoretical models, but they often churn out nonsense. They can't be relied upon for liquidus testing. You'd have to do lab tests with small amounts to check the results to be certain or you're risking nasty surprises if your reactor ever gets beyond the paper phase.

In terms of theoretical modelling, Ondrej Benes is a bit of a pioneer here. Here's his thesis:

http://publications.jrc.ec.europa.eu/re ... _Benes.pdf

The basic approach is to determine the Gibbs energy of the solid and liquid of one component and then interpolate to find the largest difference, the idea is that this will be where the eutectic is. Fortunately Benes also did measurements on various interesting eutectics. The calculations on NaCl-UCl3-PuCl3 showed quite high melting points, though Benes should have checked NaCl-PuCl3-UCl4-UCl3 with a lot more UCl4 than UCl3.

ThCl4 eutectics may also interest you as they can get quite low in melting point without using KCl. A NaCl-ThCl4-UCl4-UCl3-PuCl3 eutectic is very attractive.

GdCl I'm pretty sure is out of the question because of its neutron capture, even in a fast spectrum you don't want to be having tens of tonnes of this stuff around.

For heat capacity, yes there's a simple method, known as the Dulong-Petit equasion.

https://en.wikipedia.org/wiki/Dulong%E2%80%93Petit_law

It isn't acurate at all for engineering systems design purposes; only around 80% accuracy. Worse, sometimes it's accuracy is better than 1%, at other times it will be 20% off. So really unpredictable this. I think you just can't use simple equasions with molar heat capacities because there are other effects relating to the composition (interaction effects) that affect heat capacity in some cases.


Top
 Profile  
 
PostPosted: Nov 02, 2013 5:17 am 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Cyrill...Thank you very much for your quick and very helpful answer!

Let me have 3 comments...

I read a chinese paper (in english) about the calculation of liquidus temperatures of liquid salt mixtures. I had a similar feeling.

UCL4 is usually not a preffered component in a MCFR. The published concepts usually intend to run such a reactor underchlorinated to avoid corrosion of the structure material.

GdCl3 is a fission product chloride in such a reactor. In my concept it is 16g of GdCl3 per Kg of fp in the reactor or roughly 1g/Kg of fuel.

Best regards

Holger


Top
 Profile  
 
PostPosted: Nov 02, 2013 7:39 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
HolgerNarrog wrote:
UCL4 is usually not a preffered component in a MCFR. The published concepts usually intend to run such a reactor underchlorinated to avoid corrosion of the structure material.


Certainly, but that doesn't mean you must use 100% UCl3. A 90% UCl4, 10% UCl3 ratio should be sufficient. Redox can be actively controlled to maintain this U4/3 ratio, either by adding reductant like uranium or sodium, or by electrodes. Uranium metal rods suspended in a cage and some lowering mechanism would allow very good control of the redox, and unlike the fluoride reactor version, you don't have to worry about +3 species solubility, so you get a lot more redox window in your design space.

Please keep in mind that overreducing is potentially debillitating to structural materials. Particularly anything with carbon in it (most alloys have some carbon), can go into a reaction with overreduced uranium. U metal reacts readily with carbon, forming uranium carbides somewhere you don't want them. Actually I presume you don't want uranium carbides anywhere.


Top
 Profile  
 
PostPosted: Nov 02, 2013 6:32 pm 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Dear Cyrill,

having a look on the MCFR a potential structure material has to widthstand an 800°C salt mixture consisting of stable salts as NaCl, less stable salts as UCL3, SmCl3, some free chlorine, a few ppm oxygen, 1 ppm tellurium, < 1ppm sulphur and a high neutron flux for decades.

The only material group known to me that might widthstand these challenges are the molybdenum alloys. Tests with fluoride salts did show titanium leaching. Mo-tzm consist of 0.01 - 0.04. It is not a major alloying component. Hence I do not see decarbonization as a major corrosion mechanism.

Do you know any corrosion tests with liquid salts and molybdenum alloy structure materials?
Do you know any other potential structure material for an MCFR*?

The fission of 1 Kg PuCl3 or 1440g PuCl3 will create a chlorine surplus of 45g. An U-Pu Alloy with 20% Pu (melting point!) could be used in the fission zone to balance the chlorine content.

Best regards

Holger

*Graphite performs very well concerning corrosion but does not have sufficient mechanical properties (100.000h tensile strength above 100MPa).


Top
 Profile  
 
PostPosted: Nov 03, 2013 3:40 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
I'm not aware of any chloride fuel salt tests with TZM.

TZM tubing and plates are easy to get, but joining thick sections for large vessels is by no means an easy feat. ORNL figured out how to weld molybdenum with unconventional welding techniques, but it's tricky business. High moly alloys are very structure sensitive in their weld performance, and radiation embrittles it causing an increase in the ductile-to-brittle-transition temperature (DBTT). That last bit is likely not a problem for a molten salt reactor though - it never gets to cold shutdown when fuel salt is loaded.

Still, the properties of TZM are very unsuitable for pressure vessel construction. Usually a softer, weaker, but more ductile material is preferred for safety.

I think that high end stainless steel should be looked into. They will corrode more, but this can be accomodated by increasing thicknesses and maybe a nickel coating. Corrosion of stainless steel in molten salts is only a surface phenomenon after all. Stainless steels are readily available and weldable and already certified for a number of nuclear and nuclear high temperature uses. They are particularly resistant to neutron damage so you can make a vessel out of it. Fast spectrum has more issues with vessel irradiation.


Top
 Profile  
 
PostPosted: Nov 03, 2013 12:32 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1494
What about just 3D printing the vessel in one piece using one of those fancy new stainless steel printers?


Top
 Profile  
 
PostPosted: Nov 03, 2013 12:38 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
E Ireland wrote:
What about just 3D printing the vessel in one piece using one of those fancy new stainless steel printers?


If that works, then it should also work for a molybdenum printer, where it would be far more attractive (stainless is easy to weld).


Top
 Profile  
 
PostPosted: Nov 03, 2013 5:02 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
You can do some cool things with Mo alloys using powder metallurgy techniques. From what little I've been able to find on welding, Mo is extremely sensitive to oxygen inclusion, so welding needs to take place in full argon atmosphere completely free from atmospheric oxygen, with special pretreatment to remove oxides and trace moisture, so I don't think that it will lend itself to 3D printing due to the effects of oxygen and moisture weakening the final material.

I know someone who used to use TZM in plasma experiments and they used to rivet it as simple practical way joining things at a physics lab for the experiments that they were doing. While far from ideal and not leak free, riveting is a very low tech and practical way of joining materials.


Top
 Profile  
 
PostPosted: Nov 03, 2013 5:49 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
Here's a short study of some welding of molybdenum, the welds produced were of good quality. Authors mentioned the low ductility as a greater problem than oxygen affinity as inert gas environment seems easy to maintain.

http://www.daaam.info/Downloads/Pdfs/pr ... vaatal.pdf

The authors recommend electron beam welding to weld moly.

Probably the low ductility will make it a major regulatory problem to use this material for a reactor vessel.

Why is it hard to have inerted 3d printing? Printers don't need oxygen, unlike human welders.

Riveting is out of the question for molten salt reactor vessels. I cannot think of any application that is more demanding towards leak tightness. Anything that isn't leak tight like riveting is a no-no. We will need a welded vessel or a vessel produced out of one part.


Top
 Profile  
 
PostPosted: Nov 03, 2013 8:07 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
Cyril R wrote:
Why is it hard to have inerted 3d printing? Printers don't need oxygen, unlike human welders.

Riveting is out of the question for molten salt reactor vessels. I cannot think of any application that is more demanding towards leak tightness. Anything that isn't leak tight like riveting is a no-no. We will need a welded vessel or a vessel produced out of one part.
Good point you could inert the 3D printer, I would still worry that the combination of impurities and poor grain structure would severely affect the mechanical properties of printed Mo alloys. Actually I wonder if the real benefit would be in printing accurate cans to permit the making of one-piece structural elements from Mo powders and powder metallurgy techniques

Interesting article, thanks for posting it, clearly Mo alloys can be welded under the right conditions, but I would like to know more and in particular would be possible to restore the fine grain structure of the material by post weld heat treatment of the whole item?


Top
 Profile  
 
PostPosted: Nov 04, 2013 6:43 am 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Hi Cyrill, Hi All of You!

According to tests done with some chloride salts the corrosion rate for nickel based materials in chlorides salts at such temperatures is about 1.1mm/yr. Stainless steel is worse. Nickel based material could work for a reactor vessel. It could be made similar to a PWR vessel of 250mm thick material. It could work as well for the primary pipes. The heat exchangers have wall thicknesses of about 1 - 3mm. A nickel based material would corrode within a year. Further the mechanical properties of nickel based materials reduce significantly above 700°C. The heat transition of nickel based materials is low compared to other metals. Nickel forms helium under neutron irradiation. Helium makes nickel based material brittle. Neither nickel based materials nor stainless steel are suitable materials for a heat exchanger for an MCFR. It is the preferred material for chloride salt storage tanks and pipes.

A potential method of joining molybdenum TZC is to manufacture the piece by sintering and put some Re at the edges that has to be welded. A electron beam welding in vacuum with a MoRe filler and some MoRe at the edges to be welded provide a ductile welding. This is for sure not the least expensive way of welding some pieces but as the testing, certification and documentation makes up for a large fraction of the costs of a nuke it might not blow-up the cost calculation.

Other options for joining molybdenum pieces are friction welding, diffusion welding, brazing...

I do not think that laser sintering will be useful for a reactor in the next decades. The density of laser sintered materials is according to my assumptions lower than the current sintering applying pressure and temperature. The surface is big and hence the corrosion properties below that of arc casted molybdenum alloys. The mechanical properties as well. (This is an assumption!)

Holger


Top
 Profile  
 
PostPosted: Nov 04, 2013 9:02 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
Most of the corrosion tests I've seen did not employ sufficient cleanup of the chloride (H2+HCl sparging + mechanical filtering) and did not employ effective continuous redox control. So these high corrosion rates are not a surprise.

Any molten salt reactor needs to employ a great degree of purification prior to startup, and have additional purification and redox control during operation.


Top
 Profile  
 
PostPosted: Nov 05, 2013 10:19 am 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Hy Cyrill,

you are right the purity of the salts is vital for the corrosion properties. I read a russian paper about a corrosion test with chloride salts in a molybdenum container. 2/3 of the paper were dedicated to their efforts to clean the salt.

In a technical reality the purity of salt has some limitations. Fission will bring in sulphur, tellurium...and other corrosive elements.

I put some thoughts about it on my homepage....


http://kernkraftwerkderzukunft.npage.de ... -mcfr.html

Holger


Top
 Profile  
 
PostPosted: Nov 05, 2013 11:20 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5056
Tellurium corrosion is not expected at all with very reducing environments. About 15% UCl3, 85% UCl4 as fuel should not have any Te corrosion. Te would not be present as the chloride in such an environment. It would be elemental or as chromium telluride. If no Cr is available in the loop it will be completely elemental. It can be removed by filtering and de-misting (has significant vapor pressure @ 700C) so equilibrium concentrations in the salt would be tiny as well.

I'm not sure if tellurium corrosion is any issue at all with molybdenum. Nickel was the element conductive to the attack, if I recall. It may be completely absent with molybdenum, but you'd have to check. Although if moly forms a very low melting eutectic with Te or any other fission product then there might be other issues.

Be careful with the Zinkle paper. It's for a fusion reactor application where you have much faster neutrons, even much faster than a molten chloride fast reactor would have. The operating limits in that paper are also open to discussion (some are too arbitrary).


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 34 posts ]  Go to page 1, 2, 3  Next

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group