Safety of Americium in Fast Reactors

[vanesch]

Hi all,

I have a question. In fast reactors, it is generally considered unsafe to have too high an amount of americium (mainly am-241) in the core, because it diminishes all the safety parameters of the core:
- it has too high an absorption cross section, so that it "shadows" the U-238 resonances which give the Doppler effect
- it has a much lower retarded neutron fraction, hence diminishing the safety margin to prompt criticality
- it has a rising fission cross section which overshadows that of U-238, hence increasing fission when the spectrum hardens, which is bad for the void factor.

At least, that's what I understand.

But now, my question is: when I look at the Pu-240 cross sections, they seem to have similar problems. So why isn't Pu-240 then such a problem ?

cheers,
Patrick.

 

[Astronuc]

Looking at Pu-241's SF rate and half-life, one notes that the half-life is ~14.29 yrs vs 432.6 yrs for Am-241, so the equilibrium concentration of Pu-241 will be much lower than Am-241.

I need to review the fast fission cross-sections before further comment.

 

[vanesch]

Looking at Pu-241's SF rate and half-life, one notes that the half-life is ~14.29 yrs vs 432.6 yrs for Am-241, so the equilibrium concentration of Pu-241 will be much lower than Am-241.

I need to review the fast fission cross-sections before further comment.

That wasn't really my question, I was talking about the Pu-240 (not 241) and I'm not talking about radioactive decay, but about their use in a fast reactor. One sometimes says that a fast reactor as a minor actinide incinerator is limited in performance because of the small amount of americium (mainly Am-241) that one can allow in most designs in the core fuel (at most a few percent), and the arguments that are put forward are those that I listed (and a few others). However, the recuperated plutonium from PWR with high burnup has a rather high amount of Pu-240 in it (I think that it is something like 23% for a burnup of ~50 GW-day/ ton or so) from the capture on Pu-239. And apparently that's not a problem (MOX fuel can be used in fast reactors). But when I read the qualitative arguments against the use of Am-241, and I look at the capture and fission cross sections of Pu-240, then I find that the arguments also apply (qualitatively) to Pu-240. Now, I didn't do any detailled calculations, it's just the general aspect of the cross sections I'm talking about.
So how come that 10% of Am-241 is a safety problem, but not 23% of Pu-240 ?

However, in order to illustrate my point, I made a new plot of the 3 capture cross sections of U-238, Am-241 and Pu-240, and when one looks more carefully it is true that Am-241 covers more of the low-lying resonances than Pu-240 does... maybe that's the reason... In the attached figure, the capture cross sections for Am-241 (blue), U-238 (green) and Pu-240 (red)...

 

[Astronuc]

I was talking about the Pu-240 (not 241)

Oops, sorry, my mistake.

With MOX (U, Pu)O₂, one does not have much choice with regard to isotopic vector. Pu, which originates from neutron transmutation of U-238, will be Pu-239, Pu-240, Pu-241, and trace of Pu-242. Pu-241 decays to Am-241, and Pu-242 decays to Am-242m, if they don't fission or undergo alpha decay.

Low burnup U (U235, U238) contains mostly Pu-239, Pu-240 which is used for nuclear weapons. Higher burnup U contains Pu-241, Pu-242, Am-241, and traces of others. There is a burnup limit on recycled LWR fuel because of the Am-241 build up, which I think is more related to the dose (fuel handling) issue.

I don't see the resonance absorption of Am-241 being necessarily an issue for Doppler - resonance absorption is resonance absorption - unless Am-241 resonances result in fission rather the n,gamma absorption. One concern might be production of Am-242m, but n-capture in Am-241.

The absorption cross-section above the resonance region would mean that resonance absorption has lower weighting, so that would be a problem for reactivity control.

The change (reduction) in delayed neutron fraction would certainly be a concern with respect of reactivity control.

This might be of interest.
http://www.ead.anl.gov/pub/doc/Americium.pdf
http://www.ead.anl.gov/pub/doc/Curium.pdf

 

[vanesch]

I don't see the resonance absorption of Am-241 being necessarily an issue for Doppler - resonance absorption is resonance absorption - unless Am-241 resonances result in fission rather the n,gamma absorption. One concern might be production of Am-242m, but n-capture in Am-241.

The absorption cross-section above the resonance region would mean that resonance absorption has lower weighting, so that would be a problem for reactivity control.

I got that from a course which I recently followed on nuclear engineering, and it wasn't clear to me either (hence my question). According to the lecturer (Janne Wallenius), the Doppler constant (Kd in k(0) - Kd ln(T)) drops for a fast reactor of a certain design from 530 pcm to 200 pcm and then to 60 pcm when we have respectively a pure U/Pu fuel, when we have 10% Am-241, and when we have 20% Am-241.
This was qualitatively explained by indeed 1) stronger absorption at higher energies (so that fewer neutrons get down to the resolved resonance domain) 2) suppression by autoprotection of a few important low-lying resonances in U-238 by a higher absorption cross section. Point is, I could understand this during the lecture without a problem... it is when I looked at the Pu-240 cross sections, which have kind of the same features (though less pronounced) that I got puzzled.

Next he showed that due to the rising of the fission cross section towards higher energies, a hardening of the neutron spectrum gave a strong rise of the coolant temperature coefficient with increasing Am-241 fraction together with a strong positive void coefficient increase, also by a detailled calculation result.

I guess that the qualitative arguments were in fact after-the-fact arguments once the numbers were given by a detailled calculation, and that one shouldn't use this backwards.