Energy From Thorium Discussion Forum

It is currently Apr 25, 2018 2:57 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 117 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8
Author Message
PostPosted: Jan 27, 2014 6:22 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 257
I like the idea of a pool type design and I would like to discuss the economics of a pool type MSR.

I looked at the quantity of coolant and vessel's dimensions of other pool type designs :

The SFR Superphenix has a thermal power of 3000 MW, 1240 MW electric, the vessel has a diameter of 21 m, 19.5 m height, 3300 tons of sodium in the primary loop and 1400 tons in the secondary loops.

The LFR reactor ELSY has a thermal power of 1500 MW, 600 MW electric, the vessel contains 9000 tons of lead and has an hemispheric form (maybe 12 meters for diameter).

I tried to imagine a pool type DMSR of 4000 MWth, 1800 MW electric ( big power for economy of scale ) with 9000 tons of buffer salt (like the 9000 tons of ELSY but ELSY has 3 times less electric power). With a low power density (5 MWth/m^3 like the ORNL design) the core and primary loops of the reactor will maybe weigh 4000 tons. This system is submerged in a tank of 9000 tons of buffer salt (like in Cyril's design). So the buffer salt vessel must bear 13 000 tons of matter (so 0.3 MWth per ton, compare to 0.16 MWth per ton for ELSY).

For the buffer salt, it seems to me that Flinak (with natural Li) is the best in a physical point of view (low density, high thermal capacity, high boiling point, low viscosity, better natural convection) but it is costly. 9000 tons is approximattely 4090 m^3 of Flinak. Consider that the mean temperature of the buffer salt in normal operation is 520 °C. If we consider all the decay heat generated by the reactor, 7 days after shutdown, that's 8.47 * 10^12 J so if all this heat is dumped into the buffer salt It causes a rise of only 500 °C (even without counting the thermal capacity of the core). The boiling point (1570 °C) is reached 22 days after the shutdown (of course everything has failed except maybe the ceramic insulation), so the thermal inertia is really amazing.

The main drawbacks of this system is cost, we are speaking of around 80 millions of dollars for the buffer salt (supposing 20 dollar per liter of Flinak) and a high cost for the buffer salt vessel (I cannot estimate, maybe 300 millions dollars). The reactor would cost more than 4 billions dollars so this will probably cost less than 10 % of the total cost. And there is also economic advantages on other places :

  • with Cyril's design we don't need DRACS or other passive or active safety grade system to remove decay heat, simplicity helps
  • if we use DRACS or other passive or active cooling system, we will maybe not need redundancy and safety grade because if everything fail we have plenty of time to mitigate the accident (compare this to the 3 days passive cooling requirements for AP1000 and ESBWR, we have much more time than 7 days), the four emergency building of the EPR cost more than 400 millions dollars I think ...
  • with these great safety advantages the licensing is easier and the acceptance of the public is higher
  • as discussed before, if we need to store batches of spent fuel, we can use the buffer salt pool rather than use an other building


So I have few questions :

Is this economically realistic to have so much buffer salt ?

Is Flinak too costly for the advantages it gives you ?

What do you think is the right size, right quantity and right salt ?



PS : if we use Flibe with enriched Li7 as a fuel salt, we can use the depleted lithium for the Flinak of the buffer salt, reducing the total cost for the 2 salts


Top
 Profile  
 
PostPosted: Jan 27, 2014 8:39 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3068
Earlier explorations of a pool buffer salt design show that moving the heat from the core to the blanket to the buffer salt is quite a challenge. Solve that one first, then we can address solving the total heat capacity.

Though it really looks like the pool can't simply absorb the heat for a long time - we have to dump heat to the atmosphere or an external body of water.


Top
 Profile  
 
PostPosted: Jan 27, 2014 9:49 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 257
Quote:
moving the heat from the core to the blanket to the buffer salt is quite a challenge


I am not sure to understand what you mean by "blanket" (the reflectors ?) I was speaking about a single fluid DMSR with a very low power density (no blanket salt).

I didn't thought about this issue, it's maybe possible to do some thermal engineering with bands of highly conductive material in the graphite reflectors, you will also maybe have natural convection in the loops with a lot of surface for heat exchange, unfortunately I do not have thermohydraulic softwares so I can't simulate the system.

Anyways you can use drain tanks in the buffer salt, or an heat exchanger in the buffer salt with natural circulation. And next use the Cyril's design for moving heat into the atmosphere. Extract heat from the core is a requirement for all system, buffer salt or not, so it's not a specific problem of the buffer salt design ( except if you absolutely don't want to use drain tanks, so yes moving heat by conduction and convection in the core and the loops may be impossible). The buffer salt is a liquid so I guess you have better natural convection cooling than with air or other gases. If there are leaks of the fuel salt, the buffer salt will dilute and cool it, if the fuel salt boils, the buffer salt can maybe condense it and catch the non volatil radioactive products. Mitigation is also easier, it is preferable to put water on a non radioactive buffer salt (depending on the salt) than directly on the fuel salt, idem for air cooling (considering accidents mitigation with atmospheric air directly in contact of the salt).

Quote:
Though it really looks like the pool can't simply absorb the heat for a long time - we have to dump heat to the atmosphere or an external body of water.


Of course but I meant that you have maybe a week to mitigate an accident (and even more to mitigate a severe accident).


Top
 Profile  
 
PostPosted: Jan 27, 2014 11:36 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3068
fab wrote:
Quote:
moving the heat from the core to the blanket to the buffer salt is quite a challenge


I am not sure to understand what you mean by "blanket" (the reflectors ?) I was speaking about a single fluid DMSR with a very low power density (no blanket salt).

No blanket will make the problem slightly easier. We still have the challenge of moving the heat out of the core to the next layer of salt. It doesn't look like we can do it with just the natural surface area of the vessel - and much less so if it is surrounded by a graphite reflector or neutron absorber as I presume these will add substantially to the thermal resistance. Putting an HX inside the core is tough on the neutronics. Putting one outside the core other than the primary HX is tough on the out of core fraction. So, in all, we may be pushed toward a drain tank to move the fuel into a position where it is optimized for heat transfer rather than for neutronics.
Quote:
And next use the Cyril's design for moving heat into the atmosphere. Extract heat from the core is a requirement for all system, buffer salt or not, so it's not a specific problem of the buffer salt design ( except if you absolutely don't want to use drain tanks, so yes moving heat by conduction and convection in the core and the loops may be impossible). The buffer salt is a liquid so I guess you have better natural convection cooling than with air or other gases. If there are leaks of the fuel salt, the buffer salt will dilute and cool it, if the fuel salt boils, the buffer salt can maybe condense it and catch the non volatile radioactive products. Mitigation is also easier, it is preferable to put water on a non radioactive buffer salt (depending on the salt) than directly on the fuel salt, idem for air cooling (considering accidents mitigation with atmospheric air directly in contact of the salt).

Of course but I meant that you have maybe a week to mitigate an accident (and even more to mitigate a severe accident).


I started with the hope of simply putting in enough buffer to absorb the heat even without passive cooling. Got my education that there is an enormous amount of heat to deal with. So I'm thinking that we have to passive cooling.


Top
 Profile  
 
PostPosted: Jan 28, 2014 12:13 am 
Offline

Joined: Apr 19, 2008 1:06 am
Posts: 2229
For buffer salt, you could use cost effective NaCl-MgCl2.


Top
 Profile  
 
PostPosted: Jan 28, 2014 10:33 am 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 257
Quote:
We still have the challenge of moving the heat out of the core to the next layer of salt.

You speak about the thin layer of salt between the core graphite block and the reflectors (the radial gap and the plenums) is that it ? (I believed this was considered as a part of the core)


In fact I didn't thought about the problem of transfering heat to the buffer salt, I thought this was not a big issue because this was not mention before in this thread (or I missed it), well I apologize. But let us talk about that if you want. Forget Flinak and take an other buffer salt with less activation issues. If we can use some drain tanks or some heat exchangers outside the core that's good, if we can not, ok I give up. Now it would maybe be better to not use drain tanks and use an integral design in a pool type design. If we take something like the IMSR of David Leblanc, we have natural convection in the core with DRACS above the core. It's maybe possible to modify this design in order to have enough natural circulation in the core (and maybe go for an higher density in order to have less fuel salt). Something like replacing the DRACS by some heat exchangers with the buffer salt, or more simply just using the surface of the vessel ( just have the vessel, as thin as possible, between the fuel salt and the buffer salt ) for heat exchange is maybe sufficient to cause enough natural circulation in the core. Sadly I don't have software to see if these solutions are realistic.


Quote:
got my education that there is an enormous amount of heat to deal with. So I'm thinking that we have to passive cooling.


Of course I was not thinking of not have passive cooling ( i wanted to use something like the Cyril's design) and there are a lot of components that will not support a rise of 500 °C, so you will maybe lose the power plant. But if you just want to impede a release of Cesium, Strontium and actinides from the salts, the salts must be below something like 1400 °C no ? And you will have a lot of time before that happens.


Top
 Profile  
 
PostPosted: Jan 28, 2014 11:12 am 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3068
fab wrote:
Of course I was not thinking of not have passive cooling ( i wanted to use something like the Cyril's design) and there are a lot of components that will not support a rise of 500 °C, so you will maybe lose the power plant. But if you just want to impede a release of Cesium, Strontium and actinides from the salts, the salts must be below something like 1400 °C no ? And you will have a lot of time before that happens.

Impeding the release of strontium fluoride I think will be quite effective since it is non-volatile. Cesium fluoride though is more volatile. In normal vacuum distillation it comes out before the LiF at 1000C and low pressure. So one would need to check how volatile it is at atmospheric pressure and a higher temperature. I presume your temperature will drop pretty rapidly as you move away from the core. Never thought about the stack filters before but it seems reasonable that by the time any CsF gets to the stack filters it will be at a much lower temperature and able to be easily captured. So, for CsF I think we need to be a bit more careful but that it won't prove to be a real problem in the end.


Top
 Profile  
 
PostPosted: Jan 28, 2014 11:34 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5061
The emission of cesium will likely be inversely proportional to the amount of buffer salt it gets dissolved into. This is because of an extension of Raoult's Law. In any case it will be tiny because the amount of cesium is tiny compared to the much less volatile buffer salt (Raoult's Law).

It's easy to lose a lot of heat through not having insulation or having very little of it. 0.25% is very easy. So you don't need weeks of thermal store, I expect that the economic optimal is 1-3 days. Longer is only good to protect against long term freezing transients, shorter is ok if you don't care about long term freezing (if you can accomodate it for example) and you want a smaller reactor building. A buffer salt vessel could operate near freezing at normal operation to insulate things. Then melt the crust again during an emergency. This is how aluminium is manufactured with NaF-AlF3 peritectic. They operate the electrolytic cell with a thick crust of fluoride salt, it insulates well and prevents the inside of the cell from freezing. It will work better for us with eutectics. It is better that I do not talk too much about this subject, sorry. Big secret.


Top
 Profile  
 
PostPosted: Jan 28, 2014 11:42 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5061
I should add that thermal resistance in the reactor vessel is indeed a major problem with these schemes. Having a low velocity viscous fluid on the buffer salt side results in a bigger reactor vessel peak temperature than you'd initially think. Iain pointed this out earlier.

The PB-AHTR earlier work used a PRACS to increase the surface area for heat transfer of the reactor vessel in order to have decay heat removal to the buffer salt with a bigger reactor power. Its a lot of thin tubes so this solves the problem elegantly. Not as elegant as cool-through-vessel of course. In a MSR application a leak is a big concern. Since the PRACS use thin tubing and lots of little welds that can have pinhole leaks. An MSR could use PRACS but you would probably need a closed salt loop PRACS to prevent leak problems. That is acceptable.


Top
 Profile  
 
PostPosted: Jan 28, 2014 1:16 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 257
Thanks all for your responses guys.

Quote:
For buffer salt, you could use cost effective NaCl-MgCl2.


Yes Flinak is probably too much costly for this application. NaCl-MgCl2 is a good cheap salt, maybe some issues with corosion in chloride salts and activation with 36Cl. There are other cheap fluoride salts.

Quote:
I expect that the economic optimal is 1-3 days.


So 1-3 days of thermal storage to prevent failure (or great damages) of major components (vessels, ...).

Quote:
Having a low velocity viscous fluid on the buffer salt side results in a bigger reactor vessel peak temperature than you'd initially think


So you think that even with a big IMSR type design ( low power density, big surface of the vessel and just have the vessel, as thin as possible, between the fuel salt and the buffer salt ) the natural convection in the core and the buffer salt is insufficient ?
So we absolutely need PRACS or drain tanks ? ( or maybe there is too much fuel salt with the integral design and PRACS with higher power density is more cost effective).
That is unfortunate, I like cool-through-vessel and integral designs.


Top
 Profile  
 
PostPosted: Jan 28, 2014 1:37 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5061
For small high power density reactors or large very low power density reactors, the vessel itself is sufficient for heat transfer but it could be better (we get a significant heatup on loss of flow). A big high power density machine like ORNL MSBR or even worse the TMSR is not going to work, not even close, at least not if we want to use things like Hastelloy vessels.

ATWS events are something to watch out for since it increases the power to be transmitted through the vessel wall.


Top
 Profile  
 
PostPosted: Jan 28, 2014 2:57 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 257
OK thanks guys.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 117 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8

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:  
cron
Powered by phpBB® Forum Software © phpBB Group