Energy From Thorium Discussion Forum

It is currently Oct 23, 2017 3:38 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 15 posts ] 
Author Message
PostPosted: Mar 20, 2013 12:02 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
In reading some of Dostal's famous work on supercritical CO2 turbines, something struck me as a potential problem.

http://stuff.mit.edu/afs/athena/course/ ... dostal.pdf

The CO2 inlet into the heat exchanger coming from the main compressor is 61 degrees Celsius, am I reading that correctly? And recompression entering around the 158 degrees Celsius point. See figure 10.1 for a temperature-entropy diagram.

This seems like a freezing problem, even compared to a modern regenerative supercritical steam cycle, where the feedwater is heated to >270 degrees Celsius. Even though dense CO2 has only a fraction of the volumetric heat capacity of pressurized water, it is much higher than ideal Brayton cycle gasses.

550 degrees Celsius for the basic design could work with molten nitrate salt, but NaNO3-KNO3 has a freezing point of >220 Celsius.


Top
 Profile  
 
PostPosted: Mar 21, 2013 6:49 am 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 462
Location: Teesside, UK
Regeneration has multiple benefits!

Fig 10.1 is a repeat of Fig 6.2, on P115 of the thesis or P134 of the pdf, (in my version - there are several near-identical documents on the net). Fig 6.1 just above this gives the key to the state point labels. Point 3 is the merge point for the recompressed and heat rejection streams. Point 4 at 397C is the entry point to the reactor/heat exchanger. The heating from 3 to 4 is from the regenerator, cooling the expanded stream from 6 to 7 as it goes. So you need a salt with freezing point <390C, prefferably <350C to give a bit of margin.


Top
 Profile  
 
PostPosted: Mar 21, 2013 8:02 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Thanks Luke. I missed that regeneration step, good news for salt-heated S-CO2 cycles. 397 degree C is fine, this would work well a fluoride salt such as NaF-KF-ZrF4 (m.p. 385 C).

The advanced and high performance designs, with point 4 (reactor inlet) being at 489 and 531 degree C respectively, would be more suitable towards fluoride salts wrt freezing.

But the nitrate salt NaNO3-KNO3 is still attractive for the basic 550 C design, with this supposedly being a first commercial build operating temperature.

As a top notch chemist, Luke could probably tell me if zirconium fluoride reacts with CO2 to produce zirconium oxyfluoride? Or even the oxide?


Top
 Profile  
 
PostPosted: Mar 21, 2013 9:21 am 
Offline

Joined: Jun 05, 2011 6:59 pm
Posts: 1337
Location: NoOPWA
If there is an issue with freezing, can't you just use a less efficient HEX and get a larger deltaT?

_________________
DRJ : Engineer - NAVSEA : (Retired)


Top
 Profile  
 
PostPosted: Mar 21, 2013 11:17 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
KitemanSA wrote:
If there is an issue with freezing, can't you just use a less efficient HEX and get a larger deltaT?


Normal operation won't be a problem, given good design. During transients there are potential problems. One of the most important is a pump error where one loop keeps pumping (or trips too late) but a salt loop does not. Then the salt loop could freeze if the minimum temperature in the loop that keeps pumping is below the salt freezing temp. Station blackout is another one that is difficult to protect freezing if the salt melt point is high.


Top
 Profile  
 
PostPosted: Mar 21, 2013 1:14 pm 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 462
Location: Teesside, UK
Cyril R wrote:
.........if zirconium fluoride reacts with CO2 to produce zirconium oxyfluoride? Or even the oxide?

Sorry, couldn't find thermodynamic data on the oxyfluoride. The oxide won't form, as exchanging CO2 and ZrF4 for CF4 and ZrO2 is both endothermic (+270.6 kJ/mol) and exentropic (6.4 J/mol-K entropy lost). I only found data for the solid, but ZrF4 melts and solutions must be lower Gibbs energy than the pure solid, or they would drop ZrF4 solid out, so the conclusion is conservative. The oxyfluoride would have to be highly favoured over ZrF4 + ZrO2 for there to be a problem

Cyril R wrote:
........ One of the most important is a pump error where one loop keeps pumping (or trips too late) but a salt loop does not. Then the salt loop could freeze if the minimum temperature in the loop that keeps pumping is below the salt freezing temp. Station blackout is another one that is difficult to protect freezing if the salt melt point is high.


Does having a high temperature for the working fluid at HX inlet even help much for this? The turbine/compressor will keep running so long as there is a temperature difference between hot and cold ends of the loop, although the efficiency soon becomes very poor. If the salt loop pump fails, the gas loop must trip to bypass the heat exchanger - and fast! Suppose the volume of salt in the exchangers is 20m^3, and it's volumetric heat capacity is a bit better than water, say 5 J/ml/K = 5MJ/M^3/K. The heat engine is normally moving 2GW (thermal), so the salt temp drops at (2000MJ/s)/(5*20 MJ/K) = 20K/s. This rate will fall quickly, but there is not going to be very much time available. For printed circuit heat exchangers with low salt inventory, there is even less time.


Top
 Profile  
 
PostPosted: Mar 21, 2013 2:17 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Well, we'd have to check what the regenerator outlet temperature would change to with a choked heat source upstream.

Perhaps the scenario isn't realistic if some component in the CO2 cycle will trip on process parameters. And having multiple redundant trip logic for the CO2 cycle would probably be a good idea anyway that is easy to do.

Station blackout is a bit more tricky, but that's also where having normal min CO2 temp above salt freezing point helps a lot. There'd be a flow coastdown and then the min salt temp creeps down to the min CO2 temp. It would be interesting to see the effect of natural circulation on the salt loops and the gas loop.


Top
 Profile  
 
PostPosted: Mar 21, 2013 2:44 pm 
Offline

Joined: Apr 28, 2011 10:44 am
Posts: 247
The problem isnt so much the freeze, but the subsequent thaw.
If the higher melting point fluid is on the shellside of a tube and shell,
you can set up to thaw from the top down.

Compact HX or high melting fluid in tubes, I think yr screwed.


Top
 Profile  
 
PostPosted: Mar 22, 2013 4:23 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
djw1 wrote:
The problem isnt so much the freeze, but the subsequent thaw.
If the higher melting point fluid is on the shellside of a tube and shell,
you can set up to thaw from the top down.

Compact HX or high melting fluid in tubes, I think yr screwed.


For a tube-and-shell HX, the high pressure fluid/gas will always be in the tubes, with the salt on the shell side, so that's good.

For compact HXs, recent short term rupture testing by Heatric showed that a HX can take 10x the design pressure without rupture. That's amazing, and it makes me wonder if thaw expansion would cause such stresses (which would be >250 MPa for a 25 MPa CO2 cycle).

But freezing complicates the safety or at least the operational/plant protection system. It is very difficult to predict the hydraulic effects (liquid hammer, etc.) with local freezing in a complicated geometry HX. Making freezing a non-plausible event would help in simplifying development of the systems.

Does anyone have information of the different salts expansion/shrinkage on freezing/thawing? It appears FLiBe doesn't expand much at all, but I can't find data on other salts.


Top
 Profile  
 
PostPosted: Mar 22, 2013 5:56 am 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 462
Location: Teesside, UK
ORNL-TM-4308, Stanley Cantor, 1973 says

FLiBe +12% ThF4 has density decrease of about 7% on melting. NaF/NaBF4 has 5%-8% depending upon assumptions, but a 12.7% change in density between 2 solid forms at 243C.

NIST has data for some pure salts, but I've found nothing for mixtures. The behaviour for nitrite/nitrate solasr salt mixture must be known, but is not readily accessible.

For comparison, the water/ice density difference is about 9%, though obviously with the solid less dense, unlike most materials.


Top
 Profile  
 
PostPosted: Mar 22, 2013 6:14 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
That's with thorium added. But I recall some progress report or such mentioning FLiBe has little expansion and that this is one of the little mentioned attractions towards this salt for a coolant salt.

Nitrate salt, googling finds one mention of 3% expansion upon thawing, not sure if it's the Hitec or solar salt/draw salt. Apparently some work has been done with CSP receivers and piping, with some bellows device invented even to accomodate freeze/thaw. Perhaps this is suitable for a fluoride salt too, though high temperature bellows in fluoride salt are not available and would likely require a good deal of testing, and maybe it won't solve the HX freezing issue (just large pipes).


Top
 Profile  
 
PostPosted: Mar 22, 2013 10:41 am 
Offline

Joined: Apr 28, 2011 10:44 am
Posts: 247
These salts are nearly incompressible.

Dont have numbers for flibe or the like
but the bulk modulus of NaCl is about 25 GPa.

You dont want to go up against these forces.


Top
 Profile  
 
PostPosted: Mar 22, 2013 12:44 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
djw1 wrote:
These salts are nearly incompressible.

Dont have numbers for flibe or the like
but the bulk modulus of NaCl is about 25 GPa.

You dont want to go up against these forces.


I don't think we have full axial restraint in HXs, just full radial restraint at worst.


Top
 Profile  
 
PostPosted: Mar 22, 2013 1:33 pm 
Offline

Joined: Apr 28, 2011 10:44 am
Posts: 247
True, but with 1000's of 10 mm OD tubes maybe 5 meters long,
you are looking at a very difficult, if not impossible, situation.


Top
 Profile  
 
PostPosted: Mar 22, 2013 1:52 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
djw1 wrote:
True, but with 1000's of 10 mm OD tubes maybe 5 meters long,
you are looking at a very difficult, if not impossible, situation.


PCHEs would be typically some 1-1.5 meters long modules. I'd feel the same about the difficulties and complexities, and would prefer to avoid freezing as design basis for those reasons, but it'd make for a fun and useful experiment, for sure. This would be a quite cheap experiment, at the cost of a PCHE module and some coolant salt. Would be a useful experiment for AHTRs at least.


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:  
cron
Powered by phpBB® Forum Software © phpBB Group