Energy From Thorium Discussion Forum

It is currently Oct 23, 2017 3:32 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 22 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: Apr 17, 2008 11:13 am 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 914
Location: Ottawa
I was hoping to get some opinion and/or educated guesses regarding the concept of lowering the peak salt temperature by a modest amount. There are a variety of benefits to be had by a peak reduction. Hastelloy N suffers from helium embrittlement from (n,alpha) reactions if we try to use it in the higher flux regions of the reactor. I have learned that this would not be a problem if the peak temperature was lowered somewhat. Also, common stainless steels such as 304 and 316 might be possible to use throughout the circuit if the peak temperature was a bit lower, certainly if we went as low as 500 C and hopefullly even by 600 C. Finally, graphite can have a much longer lifetime if the peak temperature is lowered.

A drop in thermodynamic efficiency on its own is not really an issue is terms of fuel consumption as thorium itself is so cheap and plentiful. The balance of plant can be more expensive though with a lower efficiency. However, even if pumps, heat exchangers and turbines are larger, if they are using far less expensive materials then perhaps overall the effect might actually be a net gain (ex Hastelloy N is 6 to 8 times the cost of stainless steel).

Here are the basic MSBR parameters that everyone seems to start with...

Primary Salt Inlet/Outlet 566 C/704 C
Intermediate Salt Inlet/Outlet 454 C/621 C
Efficiency on Rankine Steam cycle = 44%
Efficiency on Brayton Gas cycle = 48%

So my first question is, what would the projected efficiency be if we dropped all those values by 50 C, 100 C or 200 C. Please try to include estimates for the standard MSBR case for comparison (as your method may not match the above)

A related question is that the temperture drop across the core is quite large at 138 C. We could certainly drop that by increasing the pumping and heat exchanger size. So the next question is how efficiency might change if we lower the delta T across the core, say by half or 69 C (so 635 inlet and 704 outlet with the related change for the intermediate salt). How would that effect the standard MSBR, and 50C, 100 C or 200 C lower peak temperature.

Thanks in advance for any input...


Top
 Profile  
 
PostPosted: Apr 17, 2008 11:22 am 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3306
Location: Alabama
David, comparing to the "standard MSBR" is rather difficult, since the standard MSBR was coupled to a complicated supercritical steam Rankine power conversion system. Because of the nature of the steam-Rankine system, you typically add heat to the working fluid (water) isothermally, since the temperature doesn't change during the boiling process. Therefore, it is to your advantage when coupling a fluid-fueled reactor like the MSBR to a Rankine cycle to keep the enthalpy (essentially temperature) rise across the core down, since a large temperature differential just complicates your salt-steam heat exchanger.

The fact that the MSBR was intended to be coupled to a supercritical steam cycle muddied the water even more.

I've shown in another thread and given the derivation to show that the efficiency of the Brayton cycle will be constant in the ideal case, and rather flat in the real case, for a constant turbine outlet temperature and a variable turbine inlet temperature. Net work will decrease for lower turbine inlet temperatures, but efficiency won't change much.


Attachments:
File comment: 1972 MSBR steam-system diagram
ca1972_MSBRsteamSystem.gif
ca1972_MSBRsteamSystem.gif [ 33.82 KiB | Viewed 3631 times ]
Top
 Profile  
 
PostPosted: Apr 17, 2008 7:27 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 914
Location: Ottawa
Kirk,

I mention the standard MSBR conditions since almost all molten salt work after 1965 starts from the same assumption of an inlet of 566 C (1050F) and outlet of 704 (1300F) even if people are discussing the Brayton gas cycle. I am simply looking to find an approximate answer to how the efficiency will lower if the peak temperature is dropped a small to moderate amount. I know it is extremely complex to set up the entire chain from primary to intermediate to gas or steam but I would hope some approximations wouldn't be too hard.

David L.


Top
 Profile  
 
PostPosted: Apr 18, 2008 9:26 am 
Offline

Joined: Jan 08, 2007 5:31 pm
Posts: 57
The diagram that Kirk posted (1972 MSBR steam-system diagram) is what I was thinking would be great to have as a standard to discuss the efficiency effect in the Constant Efficiency for all Turbine Inlet Temperatures thread.

Is it possible to have hyperlinks links between threads? Someone mentioning a particular system could have a link that took you to that description and its thread discussion. Just a thought.


Top
 Profile  
 
PostPosted: Apr 21, 2008 8:11 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 914
Location: Ottawa
I should not have been so fast to ask for help, in doing things yourself you learn so much more. With Kirk's urging I went back to his Brayton simulator and sorted out a great deal for myself. The first very important thing I learned was we should not be so quick to repeat the value of 48% for the standard 705 C MSBR if employing the Brayton cycle that is often mentioned.

This value comes from work on the molten salt "cooled" designs (AHRT, changed names several times since...) where solid fuel is used but cooled by the same LiF-BeF2 molten salt. As materials have been already rated for use up to 705 C (1050F) that was the starting point of their work and then they also examined higher temperatures.

However the assumptions they made are pretty optimistic. I should double check but the data I have on the MSBR had an inlet/outlet of 565/705 (130 delta T) and a drop of 83 C between primary and intermediate coolant. The AHRT supposed 670/705 (35 deltaT) and a final turbine inlet temp of 675 C. Thus they are only losing 30 C from primary to intermediate to gas (those are some pretty impressive heat exchangers!). Does anyone know off hand if the fast breeder with liquid sodium to liquid sodium to steam can come anywhere close to this small a temperature drop between loops?

Note added later: Metal cooled reactors do seem to have a similar small temperature drop from primary coolant (sodium, lead etc) to steam inlet. However, when the primary loop has fuel, I don't think we will want as so much surface area (and thus salt volume) in the heat exchanger in order to have very small temperature differences from the primary to intermediate loop. If any have opinions, please chime in. My assumption of 100 C drop mentioned below is maybe something that could be lowered a bit. I'll look into that further...Anyhow, getting back to my own original question about dropping the temperature a bit. Here are some values

By Kirk's simulator set to the AHRT LT (low temp) setting, it shows the turbine inlet at 675 C (948K) from the AHRT work. This gives 47.97% efficiency but as mentioned above this seems very optimistic if peak salt temp is only 705 C.

If we assume a much more conservative drop of 100 C across the two heat loops then the turbine inlet would be 605 C (878K) the efficiency comes out as 44.55% which is quite close to the values ORNL published on steam cycles (44.4%).

So starting from that value of 44.55% for a 705 C molten salt (liquid fluoride) reactor, and a "conservative" 100 C drop before the turbine, what if we lowered the peak temp a bit.

For 705C peak and 605 C turbine inlet gives 44.55% (Conservative MSBR settings)

For 650 C peak and 550 C turbine inlet gives 41.44%

For 600 C peak and 500 C turbine inlet gives 38.2%

For 500 C peak and 400 C turbine inlet gives 30.24%

Not sure if the last number makes as much sense. Are gas cycles not very good at modest temperatures? A PWR gets well above that with a primary coolant temperature of 310 C.

Anyhow food for thought, dropping all the way from 705 to 500 C certainly will be a big loss of efficiency but a drop down to 600 C might be a good middle ground to get perhaps have the ability to switch to inexpensive stainless steels for pumps, heat exchangers etc. (slighlty larger, but much cheaper...)


Top
 Profile  
 
PostPosted: Jul 11, 2008 9:04 am 
Offline

Joined: Jan 11, 2007 9:43 am
Posts: 2
Location: Lynchburg, VA
Can someone put the URL or a link here to Kirks Brayton simulator? I want to play around with it also. Thanks.

_________________
William A. Casino Jr
Graduate Student: Nuclear Engineering
University of Tennessee at Knoxville
E-mail: wcasino@utk.edu

Nuclear Engineer II
AREVA Federal Services
Email: william.casino@areva.com


Top
 Profile  
 
PostPosted: Jul 11, 2008 9:56 am 
Offline

Joined: Apr 24, 2008 4:54 am
Posts: 491
Location: Columbia, SC
bcasino wrote:
Can someone put the URL or a link here to Kirks Brayton simulator? I want to play around with it also. Thanks.


It's in this thread:

viewtopic.php?f=30&t=15


Top
 Profile  
 
PostPosted: Jul 15, 2008 5:39 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
I also believe that lowering peak salt temperature is desirable from an economics, complexity and durability viewpoint. Supercritical carbon dioxide is under development as an alternative to helium cycles. With the right design features, it's possible to create a cycle with comparable efficiency to the helium cycle but operating under much lower temperature regimes. I've found one article on a system that got very high efficiency, 45.3 percent @ just 550 degrees C turbine inlet (probably around 650 degrees C reactor temp or lower), claimed to be similar to a Helium cycle @ 850 degrees C turbine inlet. The key lies in optimizing the compressor, using higher pressure (20 MPa). That's higher than the helium compression but a very workable pressure in terms of engineering. The cycle is supposed to be cheaper than He cycles and have much higher power density. More info:

http://dspace.mit.edu/handle/1721.1/17746

Cheers, Cyril


Top
 Profile  
 
PostPosted: Jul 15, 2008 8:20 am 
Offline
User avatar

Joined: Jun 24, 2007 10:43 am
Posts: 253
Location: Dallas, TX
David, I commend you for thinking creatively about practical issues. You are asking good questions.


Top
 Profile  
 
PostPosted: Feb 22, 2014 3:05 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 256
I understand that a stainless steel vessel running at 500°C is much cheaper than an Hastelloy vessel running at 600 °C and the conversion system will maybe cheaper too because of lower pressures and temperatures. The entire system will be cheaper if the reactor has a low power. But with a big reactor is that still right ?

If you run at 600°C rather than 500°C you will maybe have 4 more points in efficiency in your conversion system. With a big reactor (say more than 3500 MW of thermal power) that's more than 140 MW electric. If the power plant runs more than 40 years your hastelloy vessel becomes maybe profitable no ?

Thanks in advance for any response.


Top
 Profile  
 
PostPosted: Feb 22, 2014 5:59 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Stainless will always be cheaper than high Ni superalloy. In chemical industry its conventional to try to use mild steels first, if that can't be done then carbon or stainless steel, if that is also not possible (like very corrosive fluids) then go for the more expensive alloys. The national lab approach is often different, they often look for the alloy that has the best properties reckless of cost, then select that alloy as a baseline design material. ORNL appears to have done this for (what seems to me) no good reason at all, for the MSBR.

To tell you the truth, I'm not really sure where this idea of "stainless must run cooler than hastelloy" comes from, either. ASME allowable stresses are really not that different at 700C. In fact, code cases for alloy N are only up to 704C, stainless goes well above 800C.

We can't really operate too cold, because of salt freezing (depends on salt choice), and trifluoride solubility problems (depends on salt choice and reactor design). And the salts are pretty poor in terms of viscosity at the really low end of things.

Turns out it really is a compromise between lots of different factors in different fields from engineering to salt chemistry and physics, and it depends a lot on the reactor design. For your massively large reactor of 3500 MWth you'd want to stick with steam turbines, you can get over 45% net efficiency with a 550C steam turbine of that size. Really not much reason to go for the gas turbine unless you want to do what Dr. Peterson wants to do with the PB-FHR MK1.


Top
 Profile  
 
PostPosted: Feb 22, 2014 7:15 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 256
Thanks for your response.

If I understand well you think that ORNL should have done more work on stainless steel.

So you think that it's possible to use some stainless steel for the vessel and running at 650 °C for a long period of time without too much problems with corrosion ?


Top
 Profile  
 
PostPosted: Feb 22, 2014 7:47 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Sure, you can operate at that temp for a long time. Just increase the vessel thickness according to the worst case expected corrosion. This is the conventional industry approach, you look at the product you want to contain, how much does it corrode per year, how long you want it to last, the multiply and assume zero strength for the corroded zone.

It is still attractive to try and operate the vessel at a lower temperature if you can. Creep is the limiting factor, which is reduced greatly with temperature.

In some ways, stainless steel is more corrosion resistant than Hastelloy. Hastelloy has little chromium and a lot of nickel, this produces issues with grain boundary attack from sulphur-like fission products or mutants (Te, possibly Se). Stainless steel, with only some nickel, is also generally more rad resistant than higher nickel alloys.


Top
 Profile  
 
PostPosted: Feb 22, 2014 8:25 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 256
Quote:
Sure, you can operate at that temp for a long time. Just increase the vessel thickness according to the worst case expected corrosion. This is the conventional industry approach, you look at the product you want to contain, how much does it corrode per year, how long you want it to last, the multiply and assume zero strength for the corroded zone.


I was wondering about issues caused by the contamination of the fuel salt by the presence of materials coming from the corroded vessel but I guess there are not big problems since the salt already contains a lot of different chemical elements with the fission products and the actinides.

So you think that the MSR's designers should try to qualify stainless steel for operation rather than stick with Hastelloy ?

Quote:
It is still attractive to try and operate the vessel at a lower temperature if you can. Creep is the limiting factor, which is reduced greatly with temperature.


What do you propose for peak salt temperature ? 600°C ? I guess it's difficult to go under 560°C with Flibe and its high melting point, you will have to use other salts.


Top
 Profile  
 
PostPosted: Feb 23, 2014 3:17 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
Austenitic stainless steels are already qualified for nuclear pressure vessels at lower temperatures, and non nuclear high temperature components. All you have to do is combine the code cases to get a nuclear high temperature vessel. All the data and tests are already available.

550-600C is a good range for the vessel.

One intriguing option, that we've discussed on the forum before, is to insulate the vessel on the inside. This works particularly well for drain tank concepts. It allows the vessel to operate at an already qualified nuclear vessel temperature, so regulatory speaking you don't need a new nuclear code case. Technically it eliminates a number of problems ranging from thermal creep to high temperature neutron embrittlement. From a safety viewpoint it prevents vessel overheating during transients. The only real downside is, passive cooling through the vessel (if you want that) doesn't work so well anymore. Which is why it works better for drain tank concepts.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 22 posts ]  Go to page 1, 2  Next

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 2 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