Energy From Thorium Discussion Forum

It is currently May 21, 2018 10:10 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 70 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next
Author Message
PostPosted: Jan 28, 2014 12:56 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
KitemanSA wrote:
jagdish wrote:
... or with BeO moderator.
How about a BeF2 moderator? Could we effectively eliminate the graphite swelling issue by replacing the solid C log with a solid BeF2 core + thin graphite or CCComposite skin?
At that point, you have a phase change heat dump inside the core in the form of meltable BeF2.
Hmmm.


Unfortunately BeF2 is a lot more volatile than BeO and the F2 eats surprisingly large amounts of neutrons. BeF2 has a low melting point and is ultra viscous to the point of gummy, it is going to do weird things when heated by all that neutron radiation, like bloat/deform and blow up.

Graphite would be pretty good, like E Ireland says it does not require the precautions that beryllium does.

For beryllium there are superior neutronics and superior lifetime over graphite. It likely does not need replacement. A surprising problem is the ultra high tritium emissions from that much beryllium. It makes a lot more tritium than even a CANDU. Coupled to beryllium toxicity it would result in more containment issues/cost.


Top
 Profile  
 
PostPosted: Jan 29, 2014 4:34 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1512
Supercritical water does prevent on load refueling, which increases fuel costs measurably due to the reduced discharge burnup but I suppose that cannot be avoided.....

Ideally we should be aiming for the $1/W threshold but I doubt we can reach that no matter how many high cost components are eliminated with all the regulatory issues.


Top
 Profile  
 
PostPosted: Jan 29, 2014 4:48 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
E Ireland wrote:
Supercritical water does prevent on load refueling, which increases fuel costs measurably due to the reduced discharge burnup but I suppose that cannot be avoided.....


On the bright side, not having the online refuelling machine avoids many complications and safety scenarios (refuelling machine cooling breakdown or LOCA @ fresh decay heat outputs). It could mean you can go for the more interesting vertical configuration, meaning a tighter pitch becomes possible, fewer issues with sagging and the like. The refuelling could be done like PWRs do. Much cheaper and simpler refuelling crane in stead of a machine with full active cooling and whatnot.


Top
 Profile  
 
PostPosted: Jan 29, 2014 9:14 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1512
As I understand it the main problem with graphite irradiation is the graphite starts swelling which is not good when your reactor has a precisely tuned set of dimensions?
However, since our graphite is just a pile that is not under pressure wouldn't it be possible to design the system so that the zirconium pressure tubes (and presumably the other channels as well, such as control rods and block cooling gas and what not) could shift with the graphite?
This would require a fiendishly hard to engineer set of expansion joints on the tube bottoms and headers, but if the graphite is not constrained surely the tubes would simply move further apart, with the graphite heat conduction split-rings fitted around them keeping them with contact with the blocks even as they shifted?

Although I imagine this would be complicated because neutron flux is not constant across the core and thus there would be differential expansion, but it might be worth considering.

I would rather like to avoid the idea of discharging graphite during operation because it complicates the design of the reactor beyond it simply being a concrete encased pile of graphite.


Top
 Profile  
 
PostPosted: Jan 29, 2014 9:40 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Most nuclear grade graphites first shrink to a compacted state and then swell endlessly. Modern nuclear grade graphites could easily last 4 years at CANDU sized calandria neutron flux levels.

If the graphite blocks are just stacked next to each other with coolant passage in between (just helium hot cell gas on natural circulation) then they can be easily removed online or offline. At the bottom you can have central pins holding down the graphite blocks in axial position but free to expand wherever they want. When the swelling starts the coolant passage starts to narrow which is bad so you remove the graphite blocks and lift in or push in new ones. If you can do refuelling then you can certainly do this. An advantage of pressure tube design is the moderator is not part of the pressure boundary.

You could even have just a pebble bed of graphite spheres that is continuously replenished. It would make online graphite replacement easy, at the cost of higher void fraction and the need for a calandria to hold the spheres (but at least tank leaks are never a concern, compared to heavy water calandria).


Top
 Profile  
 
PostPosted: Jan 29, 2014 9:45 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
A graphite pebble bed option also allows you to dump the pebbles via a big hatch with electromagnet switch, for the secondary shutdown system.

Its strange I cannot find any mention of pebble bed calandrias on the web.


Top
 Profile  
 
PostPosted: Jan 29, 2014 10:20 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1512
Unlike previous reactors (RBMK) we probably have to keep the pressure tubes cold, on order of ~80 celsius to allow us to keep the zirconium pressure tubes cold enough to prevent us having to use enormously thick tubes with all the neutron loss problems that causes.

This means either compressed gas or probably some sort of secondary boiling [atmospheric pressure] heavy water loop to cool the tubes since several percent of reactor power will be distributed into the moderator. (The pressure tubes will have to insulated from the blocks which must be kept hot).

So you would propose either pebbles (which I think is problematic because of all the extra systems required to handle them and distribute them properly) or some sort of prismatic moderator blocks?
If we have to have a removable moderator I think the latter is preferable, although I would much prefer to not have to dismantle the reactor at all, it requires a far larger crane be provided than to lift out individual fuel assemblies during reloading and it also complicates the structural design of the building and the placement of the upper biological shield.

I am not convinced the extra money on foundations (since you will need comparable quantities of graphite either way) for a larger core is going to be greater than all the extra costs involved in removing highly radioactive material (not as radioactive as Spent Fuel admittedly, but as we know spent fuel is absurdly radioactive) and storing the graphite in intermediate storage because you can't use the reactor structure itself as a dry store as you can with the permanent core concept.

Having more graphite in the core also adds more thermal mass to the system, especially if we keep the graphite at around the annealing temperature. We will also need a cooling system to cool the blocks independently of the pressure tubes.


Top
 Profile  
 
PostPosted: Jan 29, 2014 10:32 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Graphite can't be run cold because it will build up Wigner energy. The British CO2 cooled reactors found that out when the plants were built already, so they had to do high temperature annealing runs that ended up with the metal fuel and clad setting fire (Windscale).

Graphite is hardly radioactive at all if prepared such that no activated chlorine and such are used. The material can be handled with gloves and a standard rad suit instantly after it comes out of the reactor. For MSRs the issue is fission product going into the graphite surface making it nasty, that is not the case with solid fuel reactors especially not with the moderator outside the primary circuit altogether.

Actually I wouldn't be that interested in graphite moderator at all, my choice would be a light water cooled, light water moderated (calandria) arrangement. That keeps the pressure tubes cool so we can use the internally insulated pressure tube concept, and we can put the entire fuel channel reactor submerged in a pool of water for emergency heat sink. It's the lowest capital cost arrangement, and no tritium (saves surpringly large equipment costs). It is also inherent pressure suppression so we get a cheap pressure suppression containment without the complication of a wetwell/drywell containment.

Biggest downside is higher enrichment need - similar to PWR. But we get higher burnup in return so it is not bad. These days with uranium coming out of efficient centrifuges it is not a major issue anymore.


Top
 Profile  
 
PostPosted: Jan 29, 2014 10:46 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1512
Better neutron economy would still give an economic advantage in fuel terms because the maximum allowable LEU enrichments would permit higher burnups in the graphite moderated reactor.
I hear stories of ceramic fuel being able to handle 100GWd/t or more.

I will work a bit more on some ideas I have for block cooling and get back to the thread when I have some numbers worked out.
(Insulating the pressure tubes on the outside as well as inside to reduce the cooling load on the secondary loop to almost nothing and so on).

(Also for reference: the Windscale fire was not in a carbon dioxide cooled pile but in one of the air cooled plutonium production reactors)


Top
 Profile  
 
PostPosted: Feb 01, 2014 7:39 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1512
So my proposal is thus:
A slightly more complicated pressure tube design to enable the insulation to be used despite the high temperature of the moderator and retaining the low neutron cross section of the tube design.

From the inside we have:

1. The Fuel element
2. The Supercritical Coolant space
3. The liner tube
4. The first layer of insulation
5. The Zircalloy pressure tube
Up to here the design is essentially identical to the Supercritical CANDU.

6. Onto the outside of the pressure tube is etched a spiral pattern (spirals are probably better than straight lines for this) of coolant channels in a manner similar to a PCHE.
7. A thin zircalloy sheath tube is then shrunk onto the primary tube (or whatever method is deemed appropriate - it might be possible to diffusion weld it) to contain the channels which will then be used to cool the pressure tube.
8. To protect the pressure tube from the moderator a layer of silica aerogel will be wrapped around the sheath tube, providing a large amount of insulation and allowing the pressure tube to be kept cooler than the moderator.
9. Depending on the issues with radiative heat transfer a zirconium foil might be wrapped around this assembly although this may proove to be unnecessary.

While this is far more complicated than the CANDU design it avoids adding large amounts of mass (or more specifically - neutron cross section) to the tubes, so should perform fairly well in my (admittedly only half educated) opinion.
The 'tube plenum' cooling fluid might be heavy water [pressurised or possibly even boiling - although at 80C the latter would put the circuit below atmospheric pressure which would be.... interesting] or (depending on the amount of heat that needs removing) might simply be an argon/helium gas circuit operating independantly of the primary graphite containment gas circuit.

As the density of the supercritical water in the cooling channels and potentially the plenum cooling circuit will vary significantly over the length of the tube it might be neccessary to machine cavities into the graphite blocks at the base of the stack to account for the loss of moderation at the top and balance power over the core although this might be unnecessary.

The graphite moderator could be cooled by supercritical water cooling through zircalloy tubes that can tolerate the high temperatures on account of being narrower in diameter than the fuel channels by virtue of not having fuel in them.

EDIT:

Some calculations - assuming an outer assembly diameter approaching 15cm (it has to include the channels, sheath and so on so I will be conservative) that implies that a surface area of something like 0.02 square metres per metre of tube length.
Which means a 6m long pressure tube would only have 0.12 square metres of surface area.

A 1 cm thick layer of aerogel and a ~600 degree celsius (at most - it could easily be a third of that) temperature difference across the insulation would lead to a heat transfer of something like 216W per tube.
Which is essentially nothing if a heavy water cooling circuit is used - even gas might be able to handle that.
The heat intrusion from inside the pressure tube wall will dominate but figures for that are hard to calculate - Supercritical CANDU estimates something like 1% of channel power - which is a lot considering (8MW channel power would lead to something like 80kW).

But even 80kW would be nothing if you can cool it with atmospheric pressure heavy water. Something like 300mL per channel per second if you use something like a 35K delta-T.


Top
 Profile  
 
PostPosted: Feb 01, 2014 10:19 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Interesting proposal. It gets really complicated with that many nested cooling jackets around right in the middle of a reactor core. You also have to cool the graphite somehow, I checked some numbers from CANDU and it appears that most energy deposited is actually neutron and gamma radiation (not conduction). It's a lot, like 5% of the thermal output of the reactor. So it would be 150 MWth for a 3000 MWth core if you're thinking a big reactor (1400 MWe). This is too much for natural circulation so you now have another nested cooling system for the graphite.

One of the things not to like about all this is the incompatibility of materials. Graphite does not burn in air, however, steam (H2O or D2O) reacts with graphite, H20 + C = CO + H2, that is then flammable and a noncondensable gas - noncondensables are annoying for containments because they cannot be pressure suppressed (like PWR deluge spray or BWR suppression pools). A lot of cooling jackets mean more potential for leaks. In a radiation environment, weird things happen in terms of growth of materials and the like, which in addition to the usual thermal expansion problems would further complicate the design, maybe having expansion pieces that are then prone to leakage and such.

It also remains to be seen that the neutronic efficiency is that good. Graphite is a fair deal worse than D2O in its captures, and if you have cooling jackets you have more Zr as well, and the YSZ insulator adds a lot of neutron poison as well.

I'd still prefer a simple light water cooled light water moderated reactor. The entire reactor module, up to and including the isolation valves or PCHEs for indirect design, could be submerged in a bunker full of water, which is also the calandria (except you've eliminated the calandria vessel for better inservice inspection, emergency cooling, etc.).


Top
 Profile  
 
PostPosted: Feb 01, 2014 10:28 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1512
You would probably cool the graphite with seperate small diameter zirconium pressure tubes that would get over the temperature weakening issue by simply being of far smaller diameter.
The Graphite direct heat could be removed at the operating temperature which is a marginal (but still measurable) increase in overall reactor efficiency. Only conduction heat flux has to be removed at the low temperature.

The heat removal from the graphite is also in many ways less strenous because the heat flux essentially stops immediately upon Reactor SCRAM.

I agree that you would have to do a lot of research to ensure the plenum 'headers' did not leak but an all zirconium tube design should prevent nasty differential expansion from causing too much trouble, and since they would be at atmospheric pressure a leak doesn't result in a jet of superheated water that is madly boiling as it slices things to pieces.

Would Carbon-14 release from steam in the containment be catastrophic immediately?
You could douse the stack with a ridiculous amount of water if necessary to stop any major steam-graphite reactions from occuring (Squib operated valves in the reactor top plate, presumably with the plate submerged in a water tank).
While the pressure tube walls remain intact there should be no fission products in the graphite to be released by such reactions - the pressure tubes could be rapidly depressurised if necessary to reduce stress on the walls and begin cooling the fuel.

Do we have any figures on how much worse Graphite is than Heavy Water?
While Yittrium does have a significant neutron capture cross section it is only present in relatively small amounts as a relatively minor component in a relatively thin layer of a relatively porous material.

The additional material over a standard SCANDU 'High Efficiency Tube' design is merely a thin plate of zirconium, some aerogel that weighs essentially nothing and maybe another thin layer of zirconium foil.


Top
 Profile  
 
PostPosted: Feb 01, 2014 11:19 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
The graphite has a different expansion coefficient than zircalloy tubing, so you either get bad thermal contact (insulating gaps) or thermal stress cracking your tubes. You'd want some sort of flexible thermally conducive binder in between (possibly graphite powder) to prevent this problem but it further complicates things. Bad contact will result in overheating.

There's a far more serious problem in the buildup of Wigner energy in graphite. It is negligible at molten salt reactor temperatures (>600C graphite) but if you're cooling with water tubes you get a lot of the graphite very cold.

Inspection is another issue. With good thermal contact, the cooling jacket for the pressure tube and cooling tubing for the graphite, are essentially not inspectable inservice, and would be tough to inspect out of service. Big graphite cores are opaque across the EM spectrum, and not practical for ultrasound inspections.

To retain neutron economy advantage for graphite you want heavy water in the graphite cooling tubing. So you will have tritium. Why not just have heavy water all over the place.

If we have just light water we eliminate all the tritium equipment and even the calandria itself. Visual inspection inservice becomes possible just like a research pool type reactor. We also get fully compatible materials and inherent pressure suppression from a completely submerged reactor.


Top
 Profile  
 
PostPosted: Feb 01, 2014 11:46 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1950
Location: Montreal
E Ireland wrote:
Do we have any figures on how much worse Graphite is than Heavy Water?
We do now....


Attachments:
H1,2_C12_O16_F19_Na23_Pb208_(n,g).gif
H1,2_C12_O16_F19_Na23_Pb208_(n,g).gif [ 28.66 KiB | Viewed 1717 times ]
Top
 Profile  
 
PostPosted: Feb 01, 2014 12:48 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
jaro wrote:
E Ireland wrote:
Do we have any figures on how much worse Graphite is than Heavy Water?
We do now....


That is just the cross section. Compensating for moderating power |(ie moderating ratio) makes light water look much better, and graphite much worse...


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 70 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 4 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group