Energy From Thorium Discussion Forum

It is currently Oct 16, 2018 5:46 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 15 posts ] 
Author Message
PostPosted: Nov 30, 2010 3:07 pm 
Offline

Joined: Dec 05, 2008 8:50 am
Posts: 336
I can't find the original data, can you refresh my memory : which was DMSR power density originally considered at the time by ORNL scientists ?
What does it change considering the option of pebble beds as moderator ?

I ask this because I noted in a David L.' s paper (Molten salt reactors: A new beginning for an old idea)
download/file.php?id=727

a huge difference in power density (and thus very likely in complexity, possibility of factory build and costs) between a two fluids, graphite free, tube within tube thorium isobreeder (Case 1 : 200 MW per cubic meter !) and other DMSR options (Case 2 and 3), both traditional ORNL one and pebble beds moderated (in this latter case, how much can power density be improved ?).

Can you comment it ?


Top
 Profile  
 
PostPosted: Nov 30, 2010 4:04 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Typical PWR is about 100 kWe/liter of active core.
Typical BWR is 50 kWe/liter of active core.
Original ORNL DMSR is also about 100 kWe/liter of active core.

If you take the entire vessel the power density is an order of magnitude lower. ORNLs DMSR was 8.3 meters by 8.3 meters for 1000 MWe.

David's tube in shell maybe 200 kWe/liter of active core.

Higher power densities may seem appealing but its kind of scary with so little thermal inertia. And the neutron damage thing. My guess is, with a low pressure vessel costs with size won't scale so much (unless the higher radiation incurs quality control costs etc.)


Top
 Profile  
 
PostPosted: Nov 30, 2010 4:17 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
The DMSR power density was an average of about 5 kw/liter of core (10 kw peak). This is about 4 times lower than the 1970s MSBR and much lower than what you can do with graphite free cores (Two Fluid etc.). The main reason to go so low (i.e. big core) was to avoid needing to change any graphite for the proposed 30 year life of the plant. (There are neutronic reasons as well, namely not losing as many neutrons to Pa233). I know first instincts say a 10 meter wide and tall core must be hugely more expensive than the small size you can do with Tube with Tube Two Fluid or the French MSFR (partial blanket). However, even a big core doesn't need to add much cost as it is mainly just a Hastelloy N vessel (which can welded together from pieces on site) and cheap graphite that is loaded in on site. It is true though that the bigger cores can make your building structure costs go up.

It is a tough call whether going to higher power density in the DMSR and just replacing graphite might be better (at the minor expense of a few extra neutrons lost to Pa). If one uses pebbles it makes replacement much easier and simpler to go to higher power density but I think I am a little less attracted to pebbles these days as I always seek to minimize any new technical uncertainty. Yes, many pebble bed designs have been built and/or proposed but no functional experience with molten salts. Perhaps if Per Peterson's group at Berkley gets one built in their molten salt cooled work (with solid fuel) then maybe it will push things in that direction.

Another comment is that no matter your power density, if you use graphite, you'll end up using about the same amount over the life of your plant. The DMSR design just needed to buy and install it all from day one. Another advantage of the BIG core route is you have a wonderful amount of thermal inertia which buys you plenty of time in terms of how quickly things heat up from decay heat or a power surge.

David LeBlanc

P.S. Not sure where you are getting the 100 kwe/liter number Cyril. Maybe you saw something showing power density in the salt itself which is 12.9% of the core volume. As well, I think your PWR and BWR numbers are for kw thermal not kw electrical. A PWR or BWR pressure vessel starts to look about the same size as the big DMSR but their cores inside are only a small fraction of that big volume.


Top
 Profile  
 
PostPosted: Dec 01, 2010 10:09 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Hello David, you’re right, the LWR figures are thermal, I messed that up, as well as miscalculating the DMSR power density, darn (the ORNL design used two active core sections, one central, one around that). That’s what I get for not looking stuff up before posting…

Ah well. I just looked up the AP1000 and it has 150 cubic meters vessel for 1117 MWe which is about 7 kWe/liter total vessel. Similar to a DMSR. Big difference is the order of magnitude thinner vessel for the DMSR. That’s important not just for costs, but also to scale quickly. Welding 30 cm nuclear grade steel is a very specialized business.


Top
 Profile  
 
PostPosted: Dec 01, 2010 11:11 am 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
Cyril R wrote:
Hello David, you’re right, the LWR figures are thermal, I messed that up, as well as miscalculating the DMSR power density, darn (the ORNL design used two active core sections, one central, one around that). That’s what I get for not looking stuff up before posting…

Ah well. I just looked up the AP1000 and it has 150 cubic meters vessel for 1117 MWe which is about 7 kWe/liter total vessel. Similar to a DMSR. Big difference is the order of magnitude thinner vessel for the DMSR. That’s important not just for costs, but also to scale quickly. Welding 30 cm nuclear grade steel is a very specialized business.



Again though that is 7 kwe/liter whereas the DMSR is 5 kwth/L and closer to 4 if you quote for the whole vessel (10m by 10m holding a core 8.3 m by 8.3 m with reflectors). If you want to talk huge pressure vessels though, look at GEs new ESBWR which is a little more narrow than the DMSR vessel (7.1 m) but a massive 27.5 meters high! This comes out to way more vessel volume than the DMSR and the fun job of welding together 20 to 30 cm thick walls!


Top
 Profile  
 
PostPosted: Dec 01, 2010 11:36 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Yeah, the ESBWR is massive! There's gotta be like, a MILLION liters of water in that thing! Though that gets you the advantage of lots of thermal mass. Lots of water to boil away when a pipe breaks. Plus ESBWR is fully passive in operation. That means the core will never be uncovered. Its also much bigger than DMSR in power, about 1500-1600 MWe. The BWR's 75 atmosphere also sounds easier than 150+ that PWRs get. I think the AP1000 is designed for over 180 atmospheres for 60 years!

Bigger pressure vessels are also disproportionally harder to make than smaller tubes. A DMSR of 5000 MWe wouldn't be unthinkable. A bit impractical for sure (5x1000 MWe is much easier in construction and operatioin), but not absurdly hard compared to a 5000 MWe PWR.

How thick would a DMSR vessel be? Less than 1 cm?


Top
 Profile  
 
PostPosted: Dec 01, 2010 7:34 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3063
I've seen 5cm for MSBR and the French designs - though I don't know where that number comes from - seems much thicker than necessary.


Top
 Profile  
 
PostPosted: Dec 01, 2010 8:07 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1946
Location: Montreal
Lars wrote:
I've seen 5cm for MSBR and the French designs - though I don't know where that number comes from - seems much thicker than necessary.

The larger you make the reactor vessel, the more difficult it becomes to control the thermal expansion/contraction stresses during transients.

Thin walls may concentrate stresses to such a degree, that the vessel will tear itself apart.

Increased thickness may be able to handle the forces by distributing them over a greater amount of material, thereby reducing the stress per unit area of wall x-section.

Geometry also plays a big role -- with cylindrical vessels having more even stresses than cubes.

I get the impression that people don't appreciate the trade-offs between structural design and thermal inertia considerations: I would much rather deal with very low thermal inertia of a small fuel mass, than with the headaches of huge thermally-stressed structures !


Top
 Profile  
 
PostPosted: Dec 02, 2010 4:50 am 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
jaro wrote:
Lars wrote:
I've seen 5cm for MSBR and the French designs - though I don't know where that number comes from - seems much thicker than necessary.

The larger you make the reactor vessel, the more difficult it becomes to control the thermal expansion/contraction stresses during transients.

Thin walls may concentrate stresses to such a degree, that the vessel will tear itself apart.

Increased thickness may be able to handle the forces by distributing them over a greater amount of material, thereby reducing the stress per unit area of wall x-section.

Geometry also plays a big role -- with cylindrical vessels having more even stresses than cubes.

I get the impression that people don't appreciate the trade-offs between structural design and thermal inertia considerations: I would much rather deal with very low thermal inertia of a small fuel mass, than with the headaches of huge thermally-stressed structures !

5 cm or 50 mm is reasonably easy to work with, some cores look like they are supported from the top and if you think about supporting the total weight of the core and all of that very heavy fuel salt, the support systems do need to have reasonable amount of strength. Add on top that the additional forces that have to be designed for in terms of the seismic performance of the system, send it all adds up. Another aspect might be wanting to have healthy strength margins to compensate for neutron induced damage and the risk of corrosion, should it occur. An extra 5 mm of material at the time of construction, would be a lot cheaper than attempting a repair later on.

Re thick and thin sections, generally thin sections suffer lower thermally induced stresses than thick sections. With thick sections exposed to a rapid change in temperature, the inside surface and the outside surface are at different temperatures, this difference in temperature can create very large stresses internally within the heavy walled equipment due to differential expansion. Repeated deep thermal cycles on thick sections like high pressure steam drums can cause mechanical failure through low cycle fatigue (LCF).

A good practical example of thick vs thin and thermal cycling, large frame gas turbines used for power generation are usually heavy machines with heavy components, not dissimilar to steam turbines, and they don't respond well to repeated starts on a daily basis. Jet engines that might start and stop 5 - 15 times a day use thin-wall components that handle that repeated thermally cycling very well.

One of the benefits of MSR's operating at low pressures is that many components are quite thin compared to other systems operating at much higher pressures. This in turn allows them to handle thermal transients better than thick-walled systems.


Top
 Profile  
 
PostPosted: Dec 02, 2010 6:18 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
When solar thermal powerplants store hot nitrate eutectics in megaliter quantities with 1-2 thermal cycles per day at low cost, it is clear that structural considerations are only a minor factor. Having thermal inertia however is a big benefit. Jaro wants to use UF4 UF3 in CANDU MSR with fast online processing. There are other advantages to lower fuel volume, in that case. The DMSR is different. No fuel processing to care about processing volumes. But lots of graphite to worry about. So you go big, and get thermal inertia as added bonus.

I'm quite surprised that DMSR only gets 2 kWe/liter of vessel. Modern PWR seems to get roughly 7 kWe/liter of vessel.


Top
 Profile  
 
PostPosted: Dec 02, 2010 6:30 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1946
Location: Montreal
Lindsay wrote:
Jet engines that might start and stop 5 - 15 times a day use thin-wall components that handle that repeated thermally cycling very well.

Excellent example !

How many tons of hot fuel does a jet engine have to support ? .....something like 0.000001 maybe ?

Certainly, in such circumstances thin-wall components are ideal -- which is precisely why I prefer systems with low mass = low thermal inertia.


Top
 Profile  
 
PostPosted: Dec 02, 2010 7:56 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
jaro wrote:
Lindsay wrote:
Jet engines that might start and stop 5 - 15 times a day use thin-wall components that handle that repeated thermally cycling very well.

Excellent example !

How many tons of hot fuel does a jet engine have to support ? .....something like 0.000001 maybe ?

Certainly, in such circumstances thin-wall components are ideal -- which is precisely why I prefer systems with low mass = low thermal inertia.


Jet engines don't have to deal with fission product decay heat.


Top
 Profile  
 
PostPosted: Dec 02, 2010 10:06 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1946
Location: Montreal
Cyril R wrote:
Jet engines don't have to deal with fission product decay heat.

Right !

.....I thought that with MSRs we separate that function from reactor operation in two ways:
1) a fraction of the FPs is continually or periodically removed from the core (typically at least the volatile FPs) to a vessel designed for cooling these FPs;
2) the remaining (larger) fraction of FP decay heat is also dealt with in a separate vessel designed for that purpose, and to which the fuel salt is drained following shutdown.

I either case, one may consider a choice between high-volume-high-thermal-inertia and low-volume-low-thermal-inertia.
My personal preference would be for the latter, because small volumes drain faster and are easier to expose to a large heat transfer area (transfering heat to the environment).
And of also for reasons mentioned previously -- namely easier design/construction of reactor vessels with thinner walls, that only need to accommodate a relatively small/light core mass.

Interestingly, one could argue that dealing with the volatile FPs is in some ways analogous to the jet engine: very low mass with extremely high heating rate per unit mass....


Top
 Profile  
 
PostPosted: Dec 02, 2010 5:49 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Cyril R wrote:
jaro wrote:
Excellent example !

How many tons of hot fuel does a jet engine have to support ? .....something like 0.000001 maybe ?

Certainly, in such circumstances thin-wall components are ideal -- which is precisely why I prefer systems with low mass = low thermal inertia.


Jet engines don't have to deal with fission product decay heat.

In a perfect world a jet engine supports 0 fuel and 0 fission product decay heat, but what we were talking about was stresses inside materials induced by thermal transients, and in relation to those circumstances thin walled systems perform better than thick walled systems.

If decay heat creates a temperature transient, then that would have to be addressed in the design. But any increase in wall thickness would be there as a design response to a lower allowable design stress at the higher temperature, not as a means of increasing the thermal inertia of the system. If we are talking about reactor vessels, the thermal inertia of the vessel will be low compared to the fuel salt and graphite if present.


Top
 Profile  
 
PostPosted: Dec 02, 2010 6:01 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3063
The issue with decay heat isn't the transient. It is that in an accident where the cooling system fails how do you deal with the resulting heat. The heat production decreases fairly rapidly so that the normal plans are to have an emergency cooling system for passively removing a certain amount of heat (say 1% of the full power level). Decay heat starts around 7% (in the case of LFTR perhaps more like 5% since we export the gases and noble metals so their heat is in a different location). The heat produced in excess of 1% then gets deposited in the fuel salt and causes a temperature rise. The higher the heat capacity the lower the temperature rise.

Another way to deal with this is to move the salt to a different cooling location (perhaps with more thermal inertia). Then we would need to see how fast we can move the salt and how fail safe is that motion.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 15 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