Simplifying Blanket salt processing

Threads relating to the design of liquid-chloride, fast-spectrum reactors as well as fuel reprocessing.
Jonathan Lee
Posts: 11
Joined: Jan 27, 2008 12:25 pm
Location: UK

Simplifying Blanket salt processing

Post by Jonathan Lee » Jan 28, 2008 2:13 pm

I'm new, so this may have been thought of and dismissed, but anyway.

If the blanket salt used a chloride mixture, rather than fluorides, then getting the Protactinium out is straightforward, as the PaCl4 bubbles out at 400C.

There would be issues with neutron economy, but if the reactor is using a blanket salt then most of the neutrons there will already be moderated, so it shouldn't be too much of a problem, and you might get away with using natural chlorine

Now somebody tell me why that was all wrong.

User avatar
Kirk Sorensen
Posts: 4152
Joined: Nov 30, 2006 3:30 pm
Location: Alabama
Contact:

Re: Simplifying Blanket salt processing

Post by Kirk Sorensen » Jan 28, 2008 3:18 pm

Jonathan Lee wrote:I'm new, so this may have been thought of and dismissed, but anyway.

If the blanket salt used a chloride mixture, rather than fluorides, then getting the Protactinium out is straightforward, as the PaCl4 bubbles out at 400C.

There would be issues with neutron economy, but if the reactor is using a blanket salt then most of the neutrons there will already be moderated, so it shouldn't be too much of a problem, and you might get away with using natural chlorine

Now somebody tell me why that was all wrong.


Hi Jonathan, that's a very interesting number about PaCl4. I had not known that. Does that still apply if the Pa is in a halide mixture of some sort (like a breeding blanket would be)?

I've gone ahead and created a new section of the forum because you're the first person to bring up chloride reprocessing, and it's a topic I think deserves attention.

Having a chloride blanket in a fluoride reactor (which may have been what you were originally talking about) could be interesting--I confess I've never thought about it. Certainly there would be little to no moderation from the constituents of the blanket (unlike fluorides) but in a fluoride reactor the flux would be pretty well-moderated by graphite (or maybe D2O) anyway.

User avatar
Kirk Sorensen
Posts: 4152
Joined: Nov 30, 2006 3:30 pm
Location: Alabama
Contact:

Post by Kirk Sorensen » Jan 28, 2008 5:03 pm

I'm not sure where this image came from, but I found it in some memos from Eric Ottewitte of INEL. I cleaned up the original image and here's the technique:

Image

Jonathan Lee
Posts: 11
Joined: Jan 27, 2008 12:25 pm
Location: UK

Post by Jonathan Lee » Jan 28, 2008 8:18 pm

Pure PaCl4 sublimes at 400C. I believe that the number holds in mixtures, but a molten ThCl4 mixture would be at significantly higher temperatures, so it shouldn't be an issue.

The idea would be best for designs like OwenT or David's, since they don't use the blanket for moderation, just shielding. Then this just reduces the blanket reprocessing needs to a gas pump and a fluoridation column to get UF6 for adding to the fuel stream.

Owen T
Posts: 191
Joined: Dec 26, 2007 11:45 am

Post by Owen T » Jan 29, 2008 1:34 pm

That's very interesting.

Wouldn't this type of reactor produce weapons-grade U233?

User avatar
Kirk Sorensen
Posts: 4152
Joined: Nov 30, 2006 3:30 pm
Location: Alabama
Contact:

Post by Kirk Sorensen » Jan 29, 2008 1:36 pm

Owen T wrote:That's very interesting.

Wouldn't this type of reactor produce weapons-grade U233?


I would recommend spiking the thorium feed with ionium (Th-230) to make sure the U233 discharge was very unattractive for weapons.

dezakin
Posts: 280
Joined: Dec 20, 2006 7:50 pm

Post by dezakin » Jan 29, 2008 4:42 pm

Play devils advocate here. Say you want to produce weapons from U233? Whats the best route?

Obviously a reactor with protactium partitioning ability. If you use a thorium breeder without partitioning (relying instead on high thorium concentration in the blanket) inevitably some will go on to become U234 and U232, but if we're relying on ionium spiking a foreign power will switch out the fuel load so its pure thorium, no?

Jonathan Lee
Posts: 11
Joined: Jan 27, 2008 12:25 pm
Location: UK

Post by Jonathan Lee » Jan 29, 2008 5:38 pm

Playing as a devils advocate, can't Th-230 spiking can be evaded by removing protactinium in batches and letting the Pa-232 decay out first?

If protactinium removal is going to be done for the neutron economy, then there will have to be some level of monitoring of movement of materials. Either you restrict Uranium coming out or you match Thorium going in to power generation.

At least in a MSR, there is no real excuse for moving fissiles out of the plant unless you're starting up a new one, as reprocessing is on-site.

User avatar
Kirk Sorensen
Posts: 4152
Joined: Nov 30, 2006 3:30 pm
Location: Alabama
Contact:

Post by Kirk Sorensen » Jan 29, 2008 6:41 pm

dezakin wrote:Play devils advocate here. Say you want to produce weapons from U233? Whats the best route?


You're already on the wrong route. You either enrich uranium or lightly toast natural uranium and remove bred plutonium. Two reasons why no one's ever built an operational U233 weapon.

User avatar
Kirk Sorensen
Posts: 4152
Joined: Nov 30, 2006 3:30 pm
Location: Alabama
Contact:

Post by Kirk Sorensen » Jan 29, 2008 7:07 pm

Jonathan Lee wrote:Playing as a devils advocate, can't Th-230 spiking can be evaded by removing protactinium in batches and letting the Pa-232 decay out first?

If protactinium removal is going to be done for the neutron economy, then there will have to be some level of monitoring of movement of materials. Either you restrict Uranium coming out or you match Thorium going in to power generation.

At least in a MSR, there is no real excuse for moving fissiles out of the plant unless you're starting up a new one, as reprocessing is on-site.


I'm assuming that the U-233 made in chloride reactors as "start charges" for fluoride reactors would be secure. Assuming that, the only time the U233 would have to be moved would be from the chloride reactor to the fluoride reactor. To make the U233 unattractive to terrorists or other would-be divertors, you spike the thorium blanket of the chloride reactor with ionium, boosting the U-232 in the U-233 to unattractive levels. Once the U-233 is loaded in the fluoride reactor, it never needs to be removed from the site again.

drbuzz0
Posts: 73
Joined: Nov 18, 2007 11:52 pm
Contact:

Re:

Post by drbuzz0 » May 21, 2008 1:17 am

Kirk Sorensen wrote:
Owen T wrote:That's very interesting.

Wouldn't this type of reactor produce weapons-grade U233?
I would recommend spiking the thorium feed with ionium (Th-230) to make sure the U233 discharge was very unattractive for weapons.
I was thinking about this and I'm thinking that adding some natural or depleted uranium (or even slightly enriched uranium as a start charge) would be useful. If you run the reactor on a combination of uranium and thorium then you always have some U-238 in the mix so any uranium you remove will be no good for a weapon without enrichment. You'll generate some plutonium, but not much in a thermal reactor. That's no matter anyway, because the plutonium will fission just fine anyway and you'll never have enough plutonium to remove it for a weapon. Since it would take so long to breed any significant amounts of plutonium, you'll always end up with way too much Pu-240. And also, due to the thorium cycle you will have a small amount of Pu-238 which makes it even less attractive.

It would also have the advantage of making the reactor more flexible in fuel type if you could use a number of combination ratios of uranium and thorium, depending on what might be avaliable on hand and so on.

David
Posts: 909
Joined: Mar 07, 2007 11:02 am
Location: Ottawa

Re: Simplifying Blanket salt processing

Post by David » May 21, 2008 9:56 am

I did not notice this thread before, some interesting posts. To add some thoughts...

I don`t think a chloride based blanket for a fluoride core would be very good. Certainly not natural chlorine as this approach only is possible in very hard neutron spectrum because chlorines cross section is way too high at lower neutron energies. Even a core without graphite would still moderate neutrons a fair amount if it is a fluoride mix. The other big complication is that in order to transfer breed U233 from blanket to core you`d have to convert from chloride to fluoride (no idea how hard that would be).

In terms of using chloride based salts to get protactinium out, there are about 3 threads discussing the issue of protactinium removal in the other "Reprocessing" section. Many of us feel that it just adds too much proliferation risk and financial cost to justify employing it. If your idea of protactinium chloride bubbling out is correct that could change the financial cost side of the equation but I still think proliferation concerns win out.

In terms of Ionium (Th230) doping, I don`t think it has yet been shown that it would be possible to gather the needed amounts from uranium tailings to be practical. Bruce Hoglund and Ralph Moir were looking into that but I haven't heard them mention anything regarding it in quite some time.

In terms of adding U238 to denature (dilute) the U233 that is certainly an option but a quite complex one for a whole range of reasons. Please check out the discussion recently started in another thread in the fluoride Reactor Design section.

http://www.energyfromthorium.com/forum/ ... ?f=3&t=710

User avatar
honzik
Posts: 90
Joined: Apr 21, 2007 1:42 am
Location: Silicon Valley, CA

Re: Simplifying Blanket salt processing

Post by honzik » May 21, 2008 8:56 pm

Just for grins, can anyone post a plot of the cross-sections of the Chlorine isotopes as a function of neutron energy and for comparison, Fluorine?

User avatar
jaro
Posts: 1938
Joined: Nov 30, 2006 9:18 pm
Location: Montreal
Contact:

Re: Simplifying Blanket salt processing

Post by jaro » May 21, 2008 9:33 pm

First plot is F19 with Cl35 and Cl37. Second is F19 alone, for clarity....

GRLCowan
Posts: 97
Joined: Feb 21, 2007 7:06 pm

Re: Simplifying Blanket salt processing

Post by GRLCowan » Dec 29, 2008 12:50 pm

But the non-graphical summaries at http://www.matpack.de/Info/Nuclear/Nuclids/C/Cl37.html and
http://www.matpack.de/Info/Nuclear/Nuclids/C/Cl35.html show that the lighter isotope has a number of ways of, with significant probability, making a neutron useless besides just capturing it.

In fission-spectrum neutrons: (n, alpha) 56.45 mb, (n, p) 96.25 mb, ... OK, that's not "a number" of ways, it's two, unless two can be considered a number. Resonance integrals, for whatever they might mean here, 466.3 mb and 965.1 mb, respectively. That (n, alpha) process makes loads of sulphur.

I've been reading up on this because I think fast reactors may have been missing out on the advantage water reactors have of a transparent coolant with high volumetric heat capacity. If they could be cooled with some alloy of Na(37-Cl), Rb(37-Cl), and Pb(37-Cl)2, maybe they could gain that advantage.


--- G.R.L. Cowan (How fire can be domesticated)

Post Reply