Is a Molten Uranium Fast Breeder Reactor Feasible?

  • #1
2,075
400
I was reading lately about fast reactors, breeders, and desirable high-temp reactors; and it seemed most new reactor concepts being studied today are more difficult than existing reactors in at least one way.

Supercritical water reactors need to battle with ever more hot and high pressure water (for one, Zr cladding won't work);

Reduced moderation reactors have such narrow passages for water around fuel rods (about 1mm) that I imagine there are concerns about local overheating and rod damage if anything goes just a little bit not as planned.

Heavy water reactors seem to be easier than above, but they produce more tritium (a bit of PR nuisance) and also they need heavy water on the order of $1 billion per reactor. Also, same problems attempting to go to higher temps as with ordinary water.

Molten lead-cooled reactors seem to be nice, but (1) lead is corrosive, (2) breeding is hard to achieve (barely reaches 1.0), (3) lead is heavy (problems with seismic resistance).

Molten sodium is fire hazard (empirically proven by Japanese).

Etc...

Which made me thinking. Aren't there ways to dodge these problems, instead of attacking them head-on ("we will make even better cladding from super stable grade of stainless steel" etc)?

Also, won't KISS principle be good in nuclear engineering too? As an example, liquid fluoride thorium reactor, or aqueous homogeneous reactor.

On one hand, I'm far from thinking nuclear engineers are less clever than me, and didn't think about it before. So my ideas are likely to be already considered and deemed not viable.

OTOH, if I will not ask, I will not find out, right?

So, here is my (likely stupid) question:

Is it possible to construct a high temperature, fast breeder reactor where the fuel is in the form of molten uranium (possibly with plutonium)? Akin to molten lead fast reactor, but... without lead.

Pros:
* cladding problems are gone because there is no cladding :)
* excellent neutron economy :)
* near-atmospheric pressure
* naturally high-temperature (above 1405 K, wow)
* homogeneous active zone (no hot spots, local overheating damage hardly possible)

Cons:
* Needs refractory reactor vessel (Mo? W?)
* Is molten uranium corrosive?
* Is it safe (is temperature coefficient negative)?
* If it freezes, may be hard (impossible?) to restart
* Will it damage reactor vessel when melt freezes?
* Others?
 
Engineering news on Phys.org
  • #2
Here is a report on the compatibilty of U with W.
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19690010576_1969010576.pdf

There are a couple of relevant statements:

"Molten U and molten U alloys have been reported (refs. 1 and 2) to be extremely corrosive. Depending on the container material (type and purity) and reactor operating conditions (e. g., temperature and temperature gradients), corrosion can occur by chemical reaction, alloying, integranular attack, and/or mass-transfer attack." p. 2 (6 of 34 in pdf)

"A previous corrosion study on metals (ref. 1) indicated that tungsten (W) resists U attack in short term (<lo hr) iosthermal (800° to 1500° C) tests to a greater extent than many other metals (including chromium, manganese, iron, cobalt, nickel, titanium, zirconium, vanadium, niobium, tantalum, molybdenum, iridium, rhenium, 18: 10: 1 stainless steel, Nimonic 80, and Inconel). Tungsten apparently exhibited the best U corrosion resistance because no eutectic or chemical compound forms between W and U, and because U and W exhibit low mutal solubilities at elevated temperatures (refs. 3 and 4)." p. 2 (6 of 34 in pdf)

Corrosivity increases with temperature. Pb corrosion is mitigated with Bi.

Uranium alloy fuels have been considered (e.g., U-Mo, U-Zr).

Part of the challenge is the migration of fission products and the formation of gases (Xe, Kr) in the fission process, and low melting temperature products, e.g., alkali and halide elements.

Fuel materials are clad in order to contain the fission products and prevent their circulation in cooling system. With highly radioactive coolant, inspection and maintenance must be performed remotely (robotically).
 
  • #3
Astronuc said:
Here is a report on the compatibilty of U with W.
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19690010576_1969010576.pdf

There are a couple of relevant statements:

"Molten U and molten U alloys have been reported (refs. 1 and 2) to be extremely corrosive. Depending on the container material (type and purity) and reactor operating conditions (e. g., temperature and temperature gradients), corrosion can occur by chemical reaction, alloying, integranular attack, and/or mass-transfer attack." p. 2 (6 of 34 in pdf)

Thanks a lot, I read entire document, very informative!

Looks like tungsten won't work as a material for such reactor.
Maybe even there is no metal/alloy which would.
There may be some hope with coating RPV's inner side with ceramics/composites. IIRC there is interesting advances in SiC in this area.

Uranium alloy fuels have been considered (e.g., U-Mo, U-Zr).

Part of the challenge is the migration of fission products and the formation of gases (Xe, Kr) in the fission process, and low melting temperature products, e.g., alkali and halide elements.

I know that solid metal fuel was considered for power reactors, and that it is used today in some research and small scale reactors.

I was thinking about molten one.

With molten fuel, fission products can't be retained in it and will need to be collected from RPV.
 
  • #4
nikkkom said:
Thanks a lot, I read entire document, very informative!

Looks like tungsten won't work as a material for such reactor.
Maybe even there is no metal/alloy which would.
There may be some hope with coating RPV's inner side with ceramics/composites. IIRC there is interesting advances in SiC in this area.
Yes - there is current interest in SiC cladding in LWRs - to replace Zr-alloys - which oxidize at high temperature. SiC and pyrolytic carbon (PyC) have been proposed for TRSIO fuel in gas-cooled reactors.

There are obvious challenges with SiC tubing - mainly the porosity and sealing the ends.

I know that solid metal fuel was considered for power reactors, and that it is used today in some research and small scale reactors.

I was thinking about molten one.

With molten fuel, fission products can't be retained in it and will need to be collected from RPV.
Two challenges: 1) the removal of the heat to a heat exchanger and ultimately to a power generation cycle, and 2) collection and processing of fission products, to ensure the fission products are immobilized and prevented from entering the environment. This is the same technical challenge for molten-salt reactors.

Current US law requires that the core of a nuclear reactor is coolable and controllable (the reactor can be made subcritical upon demand) in order to prevent the movement of fission products from core to the environment, or if there is a release, that the dose to plant and offsite be below regulatory limits. For current commerical plants, there is an additional requirement there be two independent reactivity control systems to ensure that a reactor remains subcritical once it is shutdown, i.e., that there is no uncontrolled or unintended recriticality. The requirement for coolability means that the core geometry must be maintained to ensured controllability or prevent uncontrolled release of fission products.
 
  • #5
I try to access Astronuc reference with no success.

I find two references related to this topic:

1-CONSTITUTION OF URANIUM AND THORIUM ALLOYS
http://www.orau.org/ptp/PTP Library/library/Subject/Thorium/thorium6.pdf

2-Elemental Solubility Tendency for the Phases of Uranium by Classical Models Used to Predict Alloy Behavior
http://www.inl.gov/technicalpublications/Documents/5517238.pdf

3-Elemental neutronic propierties plots: http://atom.kaeri.re.kr/cgi-bin/endfplot.pl

On the first reference U6Mn melts at 716C at eutectic 80%U-20%Mn.
On the second reference, Fe solubility is <1 on U6Fe compound.
Using the third reference, Mn has better fast spectrum neutronic propierties than Fe.

I think it will be great to study the corrosion propierties of molten U6Mn on Fe alloys.
 
Last edited by a moderator:
  • #6
A tungsten hyper speed coating can also be researched.

The hyper speed W atoms literally fuse into the reactor metal alloy and may provide a flexible corrosion layer.

I do not remember which company does this coating process.
 
  • #7
why do people complain about the generation of tritium, don't they use it for fuel in both MCF and ICF??
 
  • #9
I agree with you. 233U produced from Thorium can run with high neutron yield on the thermal or fast neutron spectrum and can perform very good on a fast breeder reactor.
The main reason for using Plutonium and Uranium is the spent fuel legacy from light and heavy water reactors. The second reason is due to proliferation concerns.
 
  • #10
Kidphysics said:
why do people complain about the generation of tritium, don't they use it for fuel in both MCF and ICF??

For fusion experiments, tritium can be bred from lithium.
Tritium inadvertently generated during normal operation of power reactors is problematic because separating it is hard.

For example, reprocessing plants' processing stages give off various gases, including water vapor. These gases are not just vented to the atmosphere - they are filtered, removing entrained radioactive fission products, particles, etc. But no filter can filter out tritiated water molecules from ordinary ones!
 
  • Like
Likes 1 person
  • #11
Kidphysics said:
the report seems to be offline is there any way you can find it again?
It seems NASA and the government have removed such reports from public access. One would have to request a copy from NTRS.
 

Similar threads

Back
Top