Energy From Thorium Discussion Forum

ORNL 2771 doesn't seem to be related ("Emulsion stabilization by silicic acid") - perhaps the number is wrong?

I see plenty of concern about the graphite in the core but not so much about the shell. I see a report of less than 0.2mils/year erosion of the shell. There was concern about chemical attack - but there seems to be confidence that this problem can be solved both by the Russians and French. There is concern about the temperature and that definitely sets an upper limit on our operating temperature. But I don't see anything that suggests a plan to periodically replace the shell. Is it neutronic damage that you are concerned about? Wouldn't this scale with neutron flux?

Try this one:

ORNL-2751

Lars,

Yes, sorry. It is ORNL 2751. I notice Kirk just added a link above (thanks..), it had a broken link the depository before, it might be back up now.

Using simple graphite as a barrier between a core salt and blanket salt might be a bit hard to ensure proper structural integrity. If we could use it I am pretty sure we'd still have a quite limited lifetime due to the fact that unless we have a really low power density the graphite will start to swell beyond its original size after just a few years or less. To me at least this would probably bring the danger of cracking.

You might have found more recent comments from the French than I've seen but I don't think any group is talking about using graphite for longer than the conventional limit given by the total fluence of fast neutrons before it starts to swell beyond its original volume (it shrinks a bit first, then swells). If you can run the graphite a bit cooler you get a longer time, hotter and less. From what I've read of the French work (often written in French which I don't read well) they don't seem to be taking on the engineering challenges of ensuring a stable barrier between their core salt and radial blanket. Perhaps they don't think it will be too difficult or perhaps they just view it as a problem to be later dealt with. Often they've shown their blanket as being individual graphite hexagon tubes that overall form a cylinder around the core. That might be a way to deal with things but again with a certain lifetime and probably won't work for their current fast spectrum designs.

To sum things up, you mention you "don't see a plan to periodically replace the shell". Just because they don't mention a plan, doesn't mean they might not need one. Hopefully we can find a suitable fix that allows at least a 40 year lifetime for any design needing such a barrier. If not then we have to make sure things are reasonably easily replaceable. If we fail on both counts, then we still have many other ways to run things without barriers (i.e. not 2 Fluid designs).

As you pointed out though, in 1 1/2 Fluid designs the barrier can be at a lower flux so barriers can last longer. This is the case in some earlier French studies at least but I think their current fast spectrum designs will certainly have similar challenges of making a barrier last. As I mentioned though, it might be balance between an expensive barrier that lasts 20 years versus something simple and easy to replace that might only last a couple.

In the Russian work, at least the recent Transuranic burner design of MOSART, they don't have a barrier between two salts. However, they do look to use graphite simply to line the vessel wall because the Hastelloy N of the vessel wall might have a hard time lasting a full design lifetime if they don't cut down the flux. This graphite has cooling channels (1% volume) and is meant to be periodically changed (can't recall how often but less than 10 year lifetime I believe).

David LeBlanc

I'm not sure I understand a no barrier two salt design. What does this mean?

Sorry I wasn`t clear, I just edited things above in italics. I meant we have lots of other possible design choices that do not have 2 separated fluids. The obvious being Oak Ridge`s basic Single Fluid design, the MSBR, but there are many possibilities.

David L.

"The connections between pipes seen in figure 7 were carefully studied since it seemed an important aspect for assembling and dismantling the components in the context of a 4 year replacement schedule."

Excerpt from the Proceedings of the COMSOL Users Conference 2007 Grenoble
J. P. Caire*, A. Roure,
LEPMI-ENSEE, 1130 Rue de la Piscine, 38402 Saint Martin d’Hères, France
*Corresponding author: Jean-Pierre.Caire@lepmi.inpg.fr

This shows a graphite based heat exchanger concept. There appears to be a presumption of a 4 year replacement cycle. No mention of why replacement is needed.

I've been digging around. I did not find a lot of concern about neutron damage (though that might become a concern for a true 2 fluid design). There seems to be some concern about corrosion but the general sense I get is that folks know what they need to do to solve this. I haven't seen anything about deposition of the noble metals. The latest French reports say that the He purge is needed to keep the noble metals from causing problems. But the ORNL report on fission products (4865) indicates that the noble metals do not come off with the off-gas system in significant quantities but prefer to plate out in regions of modest flow and lots of bubbles (the pump well).

It seems they are assuming periodic replacement of the heat exchangers. I hope this can be designed out.

Yesterday I stumbled upon a report written by General Atomics ( GA-AI4211-UC77) in 1977 detailing the consequences of long term irradiation of various sorts of commercial graphite. Irradiation was done in ORNL'S ORR reactor & the specimens were exposed to a sufficient total "fast" (>118 kev) neutron flux to begin the swelling process.

The bottom line is that GA's report indicates that it's both reasonable & conservative to expect that an all-graphite barrier (reactor tube) made with a "modern" graphite would be absolutely safe up to a total flux of 10^22 neutrons/cm^2 (and probably 2-3 times as much).

The roughly 23m^2 barrier of David's original 1 GWe tube-in-shell 2-salt reactor would have to transmit a bit under 10^22 neutrons/cm^2/year of all energies (not just the "hot" ones that damage graphite) to achieve break even.

At relevant temperatures (about 700C) graphite possesses about one tenth of the tensile strength of high-nickel SSs but only absorbs about 1/1000 as many neutrons per unit thickness. This plus the fact that graphite actually gets stronger upon irradiation suggests that it would also be reasonable/conservative to propose that an annually replaced tube-in-shell breeder utilizing a 10 cm thick walled graphite reactor/barrier tube would be absolutely "safe".

The fact that such a reactor tube would only weigh about 4 tonnes & and could readily be broken up into chunks that would fit into a single "TRUPAC" (WIPP waste container) suggests that we might want to shift our attention from trying to make the barrier last longer by discovering/testing/utilizing some yet-to-be-"demonstrated" stuff to designing our reactor so that its barrier could be replaced in a single work shift. This in turn suggests that we should be looking into "novel" ways of connecting our graphite to the rest of the reactor via something like spring loaded flanges & graphite gaskets, not with the exotic graphite-to-metal "welds" that ORNL apparently assumed.

Any ideas?

Agree 100%.

So far, I haven't seen anything more easily adaptable than the Grayloc-type flange joint.
That's for simple pipe end connections.
No idea how one could simplify tube-in-shell connections.

That sounds awfully good - for graphite.
Can you cite the exact numbers, please? .....maybe I need to update my materials table at the Engineering Materials post.....

Interesting. What would you do with the used graphite tube-in-shell assemblies? Would it be possible to just refresh them by annealing at very high temperature under inert atmosphere, cool down, re-use?

I don't have "exact" figures Jaro. What I do have are the numbers quoted for irradiated graphite in the GA report (typ. about 2200 psi tensile strength - presumably measured at room temp), the fact that unlike metals, graphite gets stronger at elevated temps, & a figure describing the "hot strength characteristics" of various metal alloys that I found on the web yesterday - the best behaving ("austenitic") stainless steels exhibit about 25 kpsi tensile strength at 700C.

In the fire protection world they have grove lock fittings. Commonly call Victaulic fittings.

http://www.victaulic.com/content/groove ... nology.htm

Before they had these fittings they used to cut treads in large diameter pipe( very costly). Now they roll a grove in the steel pipe and clamp them together.

For this application you would need to cask or cut a groove in the graphite and the other metal pipe. You would also need to develop higher temperature gasket to go under the clamp. The clamp may also need to be wider or a different shape.

The size and thickness of graphite blocks are limited in their use in reactors in order to avoid cracking due to differential thermal expansion stresses (initially), and later on, due to differential neutron flux induced creep.

How do you propose to construct a reactor vessel out of such limited size blocks? How big? How thick?

Moly and Tungsten are used every day in industry all over the world. Don't knock yourself out over graphite.

http://elmettechnologies.com/products/Molybdenum/

<snip>

Wrought Molybdenum metal possesses unique properties which make it desirable for use in a number of manufacturing industries. Its most useful characteristic is its high melting point, 2610° C. In practical operation, for example, this high performance material can work as a load-bearing material in industrial furnaces at temperatures ranging up to 1900° C.



Segment of a forty-two inch diameter, twelve foot long, hot isostatic press heat shield.



A twelve inch conduit for molten glass, spun and machine construction.

There are two questions that come to mind:-
As Carbon is so important as a moderator and the crystalline structure is such a problem, why has someone not tried out refractory Carbides like Magnesium, Silicon or Aluminum Carbides?
Two fluid design amounting (in my understanding) to a liquid blanket and Thorium Carbide blanket have been discussed at length. However there has been no reaction to reusable thorium pellets (as part of thorium cookies) in a blanket. Is a fast reactor producing fuel for a thermal reactor such a no-no? Or is it reusing irradiated material without reprocessing?