Use Fusion to Zap Nuclear Waste

[phyzmatix]

I was reading this "almost-a-real-article" in Popular Mechanics and came across the following extract

Researchers at the University of Texas recently unveiled a design for a hybrid fission–fusion reactor, a best-of-both-worlds device that would dispose of the deadliest waste from traditional nuclear power plants while generating power along the way. Most nuclear waste can be reprocessed for use as fuel in standard fission reactors, although that’s not currently the practice in the United States. The hybrid reactor would be a next step. It would employ fusion reactions to flood the remaining, highly dangerous transuranic waste with neutrons, allowing it to be burned in a fission process. One-third of the resulting energy would be fed back into the fusion process and the remaining 700 megawatts would be fed into the grid. According to senior research scientist Swadesh Mahajan, at the end of the process, about 99 percent of all nuclear waste could be eliminated. “What we really want to do is to tell the world, Please allow the expansion of nuclear energy, through standard light-water reactors,” Mahajan says. “It’s the only thing that can be ramped up quickly enough to replace coal. Do not worry about the waste. Because we’re going to give you the solution in 20 years. We will make it in time.”

http://www.popularmechanics.com/science/earth/4322757.html

Which, to my layman's ear, sounded absolutely fantastic until I read one of the comments

Fast breeder reactors were developed at Argonne National Labs in the 1980's, then de-funded. They are capable of turning 97% of our existing nuclear waste into fissionable fuel--enough fuel to meet all our electricity needs for 300 years. Why not take this tech of the shelf? Because a crisis solved is a crisis wasted. Better to dream up some bone-headed scheme like fusion-zapping, so that billions can be earmarked for a technology that will always be just around the corner.

Now, it seems that http://www.ph.utexas.edu/fogs/symposium/abstracts/mahajan.html should know what he's talking about. However, after Googling "fast breeder reactors" and in the process stumbling across Polywell fusion, all I've managed to do is confuse myself. I understand that this is an inherently complicated subject to start with, but the quagmire of information Google threw at me has left me floundering and completely out of my depth. Could someone dumb this down for me a bit so that I can make sense of it all?

Also, for those of you in the know, what are your opinions on the quoted article? Do you think Mahajan is overly-optimistic or do you think his views are realistic?

Thanks!
phyz

 

[mgb_phys]

Fast breeder reactors use the extra neutrons from the fission reaction to convert useless non-reactive isotopes of uranium into fissile isotopes that can be used in regular reactors. So you make more fuel than you burn (in theory). They have mostly fallen out of use since the cost of fissile uranium isn't high enough to really need to breed more fuel and they are very complicated and expensive to build and run.

Mixed oxide fuel (MOX) uses reactor waste, including elements like plutonium as fuel alongside it's normal uranium. So it destroys some of the waste while generating energy. MOX isn't very popular in some countries because it uses up plutonium which could be more profitably used in bombs, and is unpopular in other countries because it uses plutonium which since it is used in bombs is bad m-kay.

 

[vanesch]

Fast breeder reactors use the extra neutrons from the fission reaction to convert useless non-reactive isotopes of uranium into fissile isotopes that can be used in regular reactors. So you make more fuel than you burn (in theory). They have mostly fallen out of use since the cost of fissile uranium isn't high enough to really need to breed more fuel and they are very complicated and expensive to build and run.

Well, my reading of the history is that breeders fell out of political favor at a certain point. As you say, the nuclear industry didn't really need it at that time (still a lot of uranium to "waste"), and breeders have become the political target of anti-nuclear activists. This has been the case in many places: not only in the USA, but also in France and Germany (Superphenix and Kalkar). So the breeder programme has been abandoned, especially on political grounds (not enough pro-lobbying, and strong anti-lobbying). It is true that it was technologically somewhat more involved, but that was not the main reason. There ARE some breeders working. There's a small one in France (Phenix), and there are a few in Russia. The Japanese also have one.

Now, when I read the OP, I'm surprised that one seems to sell a fusion-driven fission reactor as SIMPLER than a breeder. I agree essentially with the comment quoted in the OP by "unknown".

That said, the original author has a point. In order to burn transuranic waste, you do need a subcritical reactor, which can be accelerator-driven, or fusion-driven. You can't use a breeder for that, because transuranic materials (especially americium) tend to render critical reactors unstable if there's too much of it. So to massively burn transuranics, you need such an installation.

 

[mgb_phys]

It is true that it was technologically somewhat more involved, but that was not the main reason. There ARE some breeders working. There's a small one in France (Phenix), and there are a few in Russia. The Japanese also have one.

They always seemed to spend most of their time shutdown fixing leaks.
Although this might be because they are mostly research/prototypes rather than mainstream reactors.

You can't use a breeder for that, because transuranic materials (especially americium) tend to render critical reactors unstable if there's too much of it. So to massively burn transuranics, you need such an installation.

Is there much of a problem with transuranics in waste?
Most of them are rather useful if you can extract and reprocess the.

 

[phyzmatix]

Most of them are rather useful if you can extract and reprocess the.

The what...the what? You're keeping us in suspense!

 

[vanesch]

Is there much of a problem with transuranics in waste?

Well, they make for the long activities (the "thousands of generations"). It's true that the waste management would be simpler if they weren't there. Good geological disposal can deal with it (we hope), but if the times were shorter so that human made containment could handle the whole period, that would be better.

Most of them are rather useful if you can extract and reprocess the.

Americium and curium ? I guess, in small amounts, they can have uses. I would expect the offer to surpass the demand

 

[oldsloguy]

That said, the original author has a point. In order to burn transuranic waste, you do need a subcritical reactor, which can be accelerator-driven, or fusion-driven. You can't use a breeder for that, because transuranic materials (especially americium) tend to render critical reactors unstable if there's too much of it. So to massively burn transuranics, you need such an installation.

Interesting subject, I assume that you mean minor actinides. When you say “critical reactor”, in the context, I’m assuming that includes breeders as well. What levels would be too much or what fraction of the minors could be fed back into a breeder fuel stream without causing operational stability problems?

Which minors would be the main proplem, i.e. would the higher Pu isotopes be in that group? What physics differences (from U235 and Pu239) cause the problem operational stability with these minor isotopes?

Thanks, hope I'm not a pest!

 

[vanesch]

Interesting subject, I assume that you mean minor actinides. When you say “critical reactor”, in the context, I’m assuming that includes breeders as well. What levels would be too much or what fraction of the minors could be fed back into a breeder fuel stream without causing operational stability problems?

Which minors would be the main proplem, i.e. would the higher Pu isotopes be in that group? What physics differences (from U235 and Pu239) cause the problem operational stability with these minor isotopes?

The main problem, in as much as I remember from a course on it, is Am-243, which cross section covers the best U-238 peaks that give us a negative Doppler coefficient ; in other words, too much Am-243 decreases the stabilizing effect of the Doppler effect. In a fast reactor, to get a negative temperature coefficient, you can use two effects: Doppler, and geometry. Usually the void coefficient of the coolant is positive (in contrast to a light water reactor), so you hope to compensate for that with Doppler and geometry. If your doppler effect diminishes (or even becomes positive, I don't know - it's not impossible) you usually cannot get a guaranteed negative temperature coefficient.

 

[oldsloguy]

The main problem, in as much as I remember from a course on it, is Am-243, which cross section covers the best U-238 peaks that give us a negative Doppler coefficient ; in other words, too much Am-243 decreases the stabilizing effect of the Doppler effect. In a fast reactor, to get a negative temperature coefficient, you can use two effects: Doppler, and geometry. Usually the void coefficient of the coolant is positive (in contrast to a light water reactor), so you hope to compensate for that with Doppler and geometry. If your doppler effect diminishes (or even becomes positive, I don't know - it's not impossible) you usually cannot get a guaranteed negative temperature coefficient.

Thanks for the reply.

Let me try and restate what you said a little differently to see if I fully understand it.

With no Am243 present, as the core heats, the neutron energy increases. At higher neutron energies fewer U235 fissions occur and more neutrons are absorbed by the U238, both of which reduce the reaction rate thereby cooling the core.

When Am243 is present, the fission cross-section of the Am243 increases with increasing neutron energy and is larger than the absorption cross section of the U238. Thus the presence of the Am243 causes additional fissions that would not have occurred as the core warms.

Now, if I’ve understood you about that correctly, another couple of questions come to mind.

Is the delayed neutron fraction in Am and the other minor actinides for that matter, similar to that in U253 and Pu239?

Is destroying the minor actinides the main reason that people are proposing these accelerator based sub-critical reactors?

Thanks again

 

[fatra2]

Hi there,

Sounds like what already exist. Waste from fission products can be "recycled" into ready to be used nuclear fuel. The only difference that I see, is that now the recycling of waste product is done in a different facility than the reactor core.

By the way, recycling of nuclear fuel "costs" energy. But it's feasible.

Cheers

 

[vanesch]

Thanks for the reply.

Let me try and restate what you said a little differently to see if I fully understand it.

With no Am243 present, as the core heats, the neutron energy increases. At higher neutron energies fewer U235 fissions occur and more neutrons are absorbed by the U238, both of which reduce the reaction rate thereby cooling the core.

It is more complicated than this. In fact, U-238 has many resonant absorption peaks, and when the core heats, due to the Doppler effect, these effective peaks broaden. When these peaks broaden, the absorption rate in U-238 increases, which leaves less of the neutrons for the fission, and hence the reaction rate decreases. So it is not mostly because the neutrons go faster (in a fast reactor, they are not much moderated) that the U-238 absorps more (although one has to take this into account also). It is because the resonant peaks in U-238 broaden. However, the absorption+fission cross section of Am-243 is stronger than that of U-238 and hence masks most of the peaks of the U-238. This diminishes the negative feedback of the doppler effect in U-238.

When Am243 is present, the fission cross-section of the Am243 increases with increasing neutron energy and is larger than the absorption cross section of the U238. Thus the presence of the Am243 causes additional fissions that would not have occurred as the core warms.

Yes, but it is not only the neutron energy that plays a role (although it does play some role). It is mainly the "good" Doppler effect in U-238 which is masked.

Is the delayed neutron fraction in Am and the other minor actinides for that matter, similar to that in U253 and Pu239?

I think it is much smaller, in fact, but I don't know from the top of my head.

Is destroying the minor actinides the main reason that people are proposing these accelerator based sub-critical reactors?

I think so.

 

[Astronuc]

Fast breeder reactors use the extra neutrons from the fission reaction to convert useless non-reactive isotopes of uranium into fissile isotopes that can be used in regular reactors. So you make more fuel than you burn (in theory). They have mostly fallen out of use since the cost of fissile uranium isn't high enough to really need to breed more fuel and they are very complicated and expensive to build and run.

Well, my reading of the history is that breeders fell out of political favor at a certain point. As you say, the nuclear industry didn't really need it at that time (still a lot of uranium to "waste"), and breeders have become the political target of anti-nuclear activists. This has been the case in many places: not only in the USA, but also in France and Germany (Superphenix and Kalkar). So the breeder programme has been abandoned, especially on political grounds (not enough pro-lobbying, and strong anti-lobbying). It is true that it was technologically somewhat more involved, but that was not the main reason. There ARE some breeders working. There's a small one in France (Phenix), and there are a few in Russia. The Japanese also have one.

The issue with fast reactors was proliferation. During the 70's, the LWR operation was based on 3 annual cycles with reloads of about 1/3 core and enrichment about 3-4%. The spent fuel would have been reprocessed and the Pu-239/240/241 used. The fast reactor would have been complimentary - it was supposed to burn Pu-239/240/241 - and the driver fuel would have been MOX (0.80 U, 0.20 Pu). However, Carter was concerned about setting a precedent for producing Pu which could be diverted by other countries to make nuclear weapons. The countries that want to make Pu will do so regardless of US policy or practice (e.g., India, Pakistan, N. Korea). The Russians have several FRs, France has Phenix and Japan has Monju. The US has decommissioned it's two FRs, FFTF and EBR-II. The UK shutdown is FR program, and decommissioned Dounreay and Prototype FR have been closed since 1994.


I'm not sure fusion neutrons eliminate radioactive waste, unless they intend to convert all long lived isotopes to much faster decaying radionuclides. The shorter-lived isotopes would still need to be isolated for about a 1000 years or so, assuming half-lives of 30 years or less.

 

[oldsloguy]

It is more complicated than this. In fact, U-238 has many resonant absorption peaks, and when the core heats, due to the Doppler effect, these effective peaks broaden. When these peaks broaden, the absorption rate in U-238 increases, which leaves less of the neutrons for the fission, and hence the reaction rate decreases. So it is not mostly because the neutrons go faster (in a fast reactor, they are not much moderated) that the U-238 absorps more (although one has to take this into account also). It is because the resonant peaks in U-238 broaden. However, the absorption+fission cross section of Am-243 is stronger than that of U-238 and hence masks most of the peaks of the U-238. This diminishes the negative feedback of the doppler effect in U-238.

This may be a dumb question, but you’ve got me doing this imaginary experiment in my head about the reason for increased neutron absorbtion caused by the Doppler broadening.

Suppose I had an imaginary neutron gun with which I could produce beams of mono-energetic neutrons and by turning a dial to adjust the neutron energy as desired. Further imagine that I have a U238 nucleus sitting perfectly stationary. Start by firing groups of mono-energetic low energy neutrons at the stationary nucleus and measuring the absorption probability, then, differentially increase the energy of the neutron group and take another measurement. In theory, I think, I would have mapped out the resonance peak shapes.

Now set the nucleus in oscillating motion and repeat the experiment. My gun would register the same neutron energies, but U238 nucleus would feel the impact velocity as the vector summation of the neutron velocity and the nucleus velocity. Thus, I can see how the moving nucleus could absorb neutrons from energy groups that were once outside the resonance region of the stationary nucleus.

What bothers me is that while I would gain neutrons by this process, I would also lose neutrons, that were once within the peak, by the same process, negating the effect.

So, is the advantage of the wider peak just that there are more energy bins under the peak for neutrons to scatter into, thus making more neutrons available for absorption?

 

[vanesch]

What bothers me is that while I would gain neutrons by this process, I would also lose neutrons, that were once within the peak, by the same process, negating the effect.

So, is the advantage of the wider peak just that there are more energy bins under the peak for neutrons to scatter into, thus making more neutrons available for absorption?

Yes, you are right, the "integral" under the resonance peak remains the same. However, the thing that "saves" us, and makes Doppler-broadening effective, is the following. If you have a certain neutron population, exposed to a certain absorbant, then the neutron density tends to have a strong dip where there is an absorbant. This is sometimes called the "self-protection" effect: it is as if neutrons FLEE those regions in the spectrum where they tend to get absorbed most. Of course that's just a matter of speaking. So, when you have a very strong absorption peak, you almost completely "kill" the neutron density in this area. What happens is that neutrons that "dare come in this region" are immediately absorbed, period. So what gets absorbed in a strong resonance peak is determined by how many neutrons get scattered into this energy region, and the specific height of the peak doesn't really matter, as long as it is "high". The absorption is not proportional anymore to the height of the peak, but to the WIDTH of the part of the spectrum, because that determines the "production rate" of neutrons in this region, which will "end up absorbed in any case".

And now you see why broadening has a net effect: by broadening, as you point out, the peak gets larger, but less high. But as long as it is still "high enough" you now just cover a larger part of the spectrum where "essentially everything gets absorbed". As the "production" of neutrons in this larger part of the spectrum is (proportionally) larger, you get more neutrons absorbed. Of course, they get "slightly" less quickly absorbed, but as long as it is quickly enough, that doesn't matter.

To make the analogy with your beam, imagine you have a beam that SCANS energy, and imagine your absorber is thick enough. At low temperature, with a high, and narrow peak, if the energy is not tuned, your beam gets through, and if you are "on top" of the peak, 99.99% of the beam gets absorbed. Now, next imagine that you increase temperature, and the peak gets 10 times wider, and 10 times less high. As you scan, the region where absorption takes place is now 10 times larger. However, you won't absorb 99.99% anymore, but, say, only 99%. You see, that doesn't matter. You still essentially absorb everything. But on a 10 times larger band.

 

[oldsloguy]

Yes, you are right, the "integral" under the resonance peak remains the same. However, the thing that "saves" us, and makes Doppler-broadening effective, is the following. If you have a certain neutron population, exposed to a certain absorbant, then the neutron density tends to have a strong dip where there is an absorbant. This is sometimes called the "self-protection" effect: it is as if neutrons FLEE those regions in the spectrum where they tend to get absorbed most. Of course that's just a matter of speaking. So, when you have a very strong absorption peak, you almost completely "kill" the neutron density in this area. What happens is that neutrons that "dare come in this region" are immediately absorbed, period. So what gets absorbed in a strong resonance peak is determined by how many neutrons get scattered into this energy region, and the specific height of the peak doesn't really matter, as long as it is "high". The absorption is not proportional anymore to the height of the peak, but to the WIDTH of the part of the spectrum, because that determines the "production rate" of neutrons in this region, which will "end up absorbed in any case".

And now you see why broadening has a net effect: by broadening, as you point out, the peak gets larger, but less high. But as long as it is still "high enough" you now just cover a larger part of the spectrum where "essentially everything gets absorbed". As the "production" of neutrons in this larger part of the spectrum is (proportionally) larger, you get more neutrons absorbed. Of course, they get "slightly" less quickly absorbed, but as long as it is quickly enough, that doesn't matter.

To make the analogy with your beam, imagine you have a beam that SCANS energy, and imagine your absorber is thick enough. At low temperature, with a high, and narrow peak, if the energy is not tuned, your beam gets through, and if you are "on top" of the peak, 99.99% of the beam gets absorbed. Now, next imagine that you increase temperature, and the peak gets 10 times wider, and 10 times less high. As you scan, the region where absorption takes place is now 10 times larger. However, you won't absorb 99.99% anymore, but, say, only 99%. You see, that doesn't matter. You still essentially absorb everything. But on a 10 times larger band.

Thanks, really interesting