Energy From Thorium Discussion Forum

It is currently Sep 22, 2018 6:34 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 12 posts ] 
Author Message
PostPosted: Oct 03, 2011 8:19 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Can anyone tell me or point me to references that model the temperature rise in a single fluid core during complete loss of forced cooling without scram? Particularly what happens long-term if we 'walk away' from the reactor with no power and no scram. Passive cooling removes the decay heat, but Pa decay adds reactivity which increases the stable temp over months.


Top
 Profile  
 
PostPosted: Oct 04, 2011 3:00 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
You might be mixing up a couple concepts. First, I guess you are assuming the salt remains in the core and doesn't drain? (i..e. the freeze plug either jams or wasn't in the design). Second, yes, Pa turning into U233 with time will gradually raise the temp the reactor would be critical at (i.e. more reactive so critical at a higher temp). Not sure how fission products transmuting with time might also change reactivity (beyond the obvious Iodine to Xenon effect). I've seen estimates of Pa effect but can't quite recall (I think the French group has an estimate somewhere but of course it will depend on the reactor type and spectrum). I think maybe Luke or someone else here might have also made a rough estimate (50C jumps to mind but don't quote me). However the missing issue is that if forced cooling fails and all the decay heat (or 2/3rds anyhow) is going into the reactor core the temperature will rise on its own since only for a quite small reactor could decay heat leave by the walls of the reactor (perhaps that is what you are getting at though, decay heat removal in situ). So yes, if decay heat can be removed long term, it becomes an interesting calculation to the stable temp the reactor will stay at since reactivity will build up a little as the weeks and months pass (less so in DMSR types since there is less Pa). In reality you might get some weird power and temperature fluctuations if it is just sitting there on its own.

David L.


Top
 Profile  
 
PostPosted: Oct 04, 2011 3:28 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Not mixing things up, just not mentioning everything explicitly :lol:

The question came from the discussion we had the other day about not having dump tanks (where I also failed to mention that explicitly, sorry!), using something like the PB-AHTR pool containment design, and that PRACS heat removal thing. The AHTR had modelling on this and found a quite high temperature rise intially since only doppler is available for fission shutdown. Interestingly they also found that too high a PRACS decay heat cooling capacity actually led to overcooling that caused the reactor to want to start fissioning again.

http://www.nuc.berkeley.edu/pb-ahtr/pap ... riveau.pdf

Now the AHTR modelling assumed no thorium. So that Pa decay would have to be accomodated for a thorium design.

For the MSR it will be better with such large negative dilatation coefficients and no excess fuel stored energy (those pebbles get HOT inside and that heat must be absorbed by the coolant upon LOFC). But I'd like to know the exact amount for say an ORNL DMSR design.

If uranium only is used this Pa decay reactivity insertion wouldn't be present. So likely not relevant for a first design.

For the record I don't see any means to make the control rods not insert. They are heavier than the fuel salt and are pushed up by the flow so if flow stops they sink in. Seems hard to prevent this. But lets hypothesize a limit analysis eartquake that deforms the graphite so the control rod can't insert and the power goes out. What is the temperature rise?

We can use silver-copper alloy cans filled with gadolinium fluoride and place them in various locations in the vessel, that melt upon overheating of the fuel salt, to prevent any criticality in this event, but perhaps it won't even be necessary if the temp rise is modest enough.


Top
 Profile  
 
PostPosted: Oct 04, 2011 4:02 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3063
A pure 238U/239Pu cycle DMSR will also have a similar effect as Pa decay from 239Np decay. Since the half-life is 2.3 days instead of 27 days there will only be about 1/10th as much inventory to decay. The French did an estimate of the reactivity change assuming an accident where instead of adding a daily amount of thorium (roughly 1000kg/365 days) they added the same mass of 233U. You could take that same analysis scale it by the half-life in days to get the reactivity change due to the extra fuel, then divide that by the thermal coefficient to get the temperature rise. I suspect it will also roughly scale inversely with the fissile inventory but I'm not sure of that.

I don't see why we should get lots of temperature fluctuations in a LFTR where the salt is not flowing. The fission process seems so much faster than the other time constants for cooling and decay that I would expect the overall loop to be as stable as the reactivity versus temperature.

If the cooling is sufficient to overcome the decay heat then in the first 12 hours you will have some temperature drop due to 135I decaying into 135Xe reducing the reactivity until the 135Xe decays away and restores the reactivity. But likely we will absorb enough of the initial decay heat that we are over the critical temperature for the first day regardless of any Xe effect.


Top
 Profile  
 
PostPosted: Oct 04, 2011 5:12 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1946
Location: Montreal
Lars wrote:
A pure 238U/239Pu cycle DMSR will also have a similar effect as Pa decay from 239Np decay.

Thanks Lars - I wanted to point out this as well.

Cyril R wrote:
Can anyone tell me or point me to references that model the temperature rise in a single fluid core during complete loss of forced cooling without scram? Particularly what happens long-term if we 'walk away' from the reactor with no power and no scram. Passive cooling removes the decay heat, but Pa decay adds reactivity which increases the stable temp over months.

Cyril R wrote:
The question came from the discussion we had the other day about not having dump tanks (where I also failed to mention that explicitly, sorry!)

If I may digress slightly (?) to the HW-MSR concept, the complete loss of forced cooling without scram event is automatically taken care of by the resulting temperature rise, which causes boiling and un-priming of the fuel channel siphon inlets.
As a result, the fluid in the fuel channels dumps into the recirc pan (subcritical configuration), where the rise in level submerges the passive decay heat cooling coils (natural thermosiphon circulation).
If the forced cooling pumps are still running (without heat sink in the HX), the level rise in the upper manifold will cause the flow to short-circuit back to the recirc pan via overflow lines.
That pretty much terminates the event.
Next problem ?


Top
 Profile  
 
PostPosted: Oct 05, 2011 6:12 am 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 461
Location: Teesside, UK
Lars wrote:
.......The French did an estimate of the reactivity change assuming an accident where instead of adding a daily amount of thorium (roughly 1000kg/365 days) they added the same mass of 233U. You could take that same analysis scale it by the half-life in days to get the reactivity change due to the extra fuel, then divide that by the thermal coefficient to get the temperature rise......

That's what I did for the rough estmate David was refering to. Iain then corrected an error in my version. The estimate is 260C rise, of which half is in the first month. Taking the thermal coefficient from the French work means it only applies to their fastish spectrum design. As Lars says, for a highly thermalised reactor with less U233 inventory, the effect will be greater, partly from the increased reactivity worth of a given quantity of fissile, and partly from the positive thermal reactivity contribution from the graphite.


Top
 Profile  
 
PostPosted: Oct 05, 2011 10:35 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Np239 isn't an issue due to its much smaller concentration. But for a thorium reactor, Pa's 260 C rise is a bit much! LOFC without scram is considered beyond design basis (even Fukushima didn't have it). So if we want to operate without a dump tank we can add temperature fusible cans of gadolinium fluoride in the top plenum and solve this problem.

Luke, does your calculation include salt dilatation?

Jaro, thin pipes with high freezing point coolant or reactive coolants in contact with the ultimate heat sink (air or water) and that are susceptable to breakage by earthquakes, plus failure mode of the hot cell cooling system that you need for low temp hot cell operations, are what I'm trying to avoid here. If your cooling pipes fail then you rely on active cooling for the hot cell to keep containment. This is not walk-away safe.


Top
 Profile  
 
PostPosted: Oct 05, 2011 1:36 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3063
We need a different answer than accepting a 260C rise for something as slow as Pa decay.

I like the idea of control rods held up by the pressure of the flowing salt. Stop the flow and the control rods drop into place by gravity.

A freeze plug with liquid poison sounds like a plausible solution. With this approach:
1) how do we ensure mixing of the poison and the fuel if the pumps have stopped?
2) is the poison heavier or lighter than the fuel salt
3) is the poison soluble in the fuel salt or just a slurry

Another possibility would be deliberate mixing of fuel and blanket salts on signs of serious trouble.
It takes only ten seconds or so to mix them using the big pumps we got.
It would take some time (a couple of weeks?) to separate them again and restore the reactor afterword.

Another idea. Suppose we have a 1.5 or 2 fluid design and have vents between them that are shut because of the flowing salt. When the salt stops flowing the vents open. Given the temperature and density differences between the blanket and fuel can we arrange things so they will mix?


Top
 Profile  
 
PostPosted: Oct 05, 2011 1:53 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
For a reactor with online vacuum distillation, the easiest thing is to use solid GdF3 packed in temperature fusible cans of a low melting noble metal alloy. Serious overheating melts the cans, dissolving GdF3 (just as soluble as any trifluoride lanthanide fission product). These cans would be positioned near the flow path of the PRACS so that natural circulation quickly disperses it. Even without natural circulation, dissolving GdF3 in the salt causes it to diffuse through the salt. After the accident, the vacuum still can be easily used as normally, to remove the GdF3.

We'd only need this in the extreme beyond design basis accident where everyone walks away from the reactor after an earthquake and power outage have occured. The earthquake has caused the control rod to jam in the cavity or graphite internals or whatever.

Basically two very implausible (perhaps impossible) scenarios. Looks like we can design against it easily which shows how robust the MSR technology is. Even without the poison cans, any failure of the primary vessel will push fuel salt into the buffer salt which will preclude any criticality (it won't challenge the containment).


Top
 Profile  
 
PostPosted: Oct 05, 2011 4:34 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1946
Location: Montreal
From another thread,
Lindsay wrote:
thermal engineering and monitoring will be required as it applies to the ground around the hot cell to ensure that the thermal performance of the system overall is adequate at all times.
Of course: that's why we have engineers !

Cyril R wrote:
Jaro, thin pipes with high freezing point coolant or reactive coolants in contact with the ultimate heat sink (air or water) and that are susceptable to breakage by earthquakes, plus failure mode of the hot cell cooling system that you need for low temp hot cell operations, are what I'm trying to avoid here. If your cooling pipes fail then you rely on active cooling for the hot cell to keep containment. This is not walk-away safe.
Similarly, cooling circuit piping can readily be designed (engineered) to AVOID "breakage by earthquakes", by using the appropriate seismic design basis.

Cyril R wrote:
thin pipes .....are susceptable to breakage by earthquakes
Are you suggesting that its easier to design big/thick pipes for earthquake resistance, than "thin pipes" ? ....because the opposite is in fact true.

I might add that, in this regard, seismically qualified design is facilitated by having SMALL reactor cores (including any thermal buffer tanks), to begin with.


Top
 Profile  
 
PostPosted: Oct 06, 2011 3:58 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Jaro, the insulation will be so thick as to not require cooling and avoid heating up the ground too much. And then we'll add a comfortable safety margin of even more insulation.

Rigid calcium silicate or more high load bearing ceramic insulation is quite cheap, and can get the heat flux to the ground very low.

I'm not familiar with how big an issue small pipes breaking in earthquakes would be. But the operating temperature is high, and ceramic or refractory metals tend to be more vulnerable to accellerating forces.

In the pool type design, the containment cooling is passive. No active cooling require. No pipes required, no freezing of high melting point fluids in thin pipes. If the PRACS pipes break, the fuel salt mixes with the pool buffer salt which doesn't challenge the passively cooled containment (heat balance stays the same).

In your HW-MSR, the small fuel salt quantity would cause it to rapidly boil in the event of cooling coil failure and then you have to deal with complex two phase calculations and issues to prove the safety case. The heavy water just like the hot cell cooling is also actively cooled so that will boil off during a station blackout. This is a quite complicated accident with heavy water and fluoride salt in different phases going everywhere in the hot cell, condensing on the hot cell wall surfaces with hydrofluoric acid etc.

I'm not saying one design is better than the other, but trade-offs will have to be made. My idea is to keep everything single phase at all times for regulation simplicity.


Top
 Profile  
 
PostPosted: Oct 08, 2011 6:36 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Ok, I found a reference of the FUJI-MSR reactivity accident. They use moderating rods as reverse control rods, and looked at a scram failure reactivity accident, at both zero power and full power, but apparently no loss of forced cooling.

https://smr.inl.gov/Document.ashx?path= ... d_p575.pdf

The temp rise of the biggest reactivity insertion is not acceptable (roughly 250 degrees Celcius it appears). Clearly using moderating control rods is not a good way to make a MSR inherently safe. Why use such rods for a fluid fuelled reactor that already has reactivity finetuning in the fuel itself?


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 12 posts ] 

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


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:  
Powered by phpBB® Forum Software © phpBB Group