Energy From Thorium Discussion Forum

It is currently Oct 23, 2017 3:42 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 78 posts ]  Go to page 1, 2, 3, 4, 5, 6  Next
Author Message
PostPosted: Jul 18, 2008 10:52 am 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3306
Location: Alabama
I didn't see a thread on this subject, so I thought I'd start one.

The supercritical CO2 cycle is intriguing. The basic advantage of the approach is that it is easiest to compress a gas when it's dense, and CO2 near the critical point has some really nice advantages. I've been reading more and more on this cycle recently and wondering if it is advantageous over an ideal-gas Brayton cycle using helium.

Much of the recent work in this field was done by Dr. Vaclav Dostal, who was a student at MIT but is now a professor at the Czech Technical University. Here is a link to his dissertation:

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


Top
 Profile  
 
PostPosted: Jul 18, 2008 1:06 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5058
I've linked to that study a few days ago in the thread about effects of lowering peak salt temp.

Benefits are obvious:

- in the right supercritical regime, higher efficiency with lower temps
- more compact than helium cycles
- can be more simplified than helium cycles
- CO2 is much less prone to creeping and leaking than helium, it's easier to contain which allows lower cost simplified equipment (bearings, valves, connectors etc)
- CO2 is abundant and easy to obtain (not that I think helium shortage will occur - the amounts required are relatively low).


Top
 Profile  
 
PostPosted: Jul 22, 2008 4:50 pm 
Offline

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

Good idea for a separate thread as this topic has popped up numerous times. I am not sure if there is a free way to download the link you posted to Dostal's Ph.d thesis but I found another link to Dostal's work through the Wiki entry for Supercritical CO2 (first link).

It really does sound like a promising route. It needs much higher pressure and CO2 can be a little corrosive but the savings in size, cost and complexity seem huge over the Brayton cycle. It is sometimes pointed out that the helium Brayton is further along in development due to the South African work on Pebble Beds but I wonder in general if there might be even less R&D money that would need to be spent to develop a production turbine (due to simplicity of design and lower temp for blade material).

There is a quote from Dostal's work that confused me a bit.

Quote:
The supercritical CO2 cycle at 550 C achieves 46% thermal efficiency which is the same as the helium Brayton cycle at 800 C (if all losses are taken into account)


Does anyone know what he meant by taking losses into account? The typical efficiency we see quoted for helium Brayton with molten salt cooled reactors is 48% if the salt peak temperature is 705 C and the peak helium at turbine inlet is 670 C. Was Dostal perhaps referring to Brayton cycles with less intercooling or are there "losses" that are being ignored when people quote Brayton efficiencies?

I'll try to read more of Dostal's work but does anyone know off hand if the 550 C inlet temperature is a real lower limit on inlet temperature? I realize the phase transition comes into play but I wonder if the cycle works a little lower, say for a 500 C inlet temp. Even if this meant a drop to 40% (from 46% at 550 C) it might still be very attractive. I have been looking into the prospects of designs with alternate carrier salts such as NaF-BeF2 with much lower melting points than Flibe (2LiF-BeF2). The big advantage being the ability to use common stainless steels throughout the primary circuit, including in core (instead of Hastelloy N which is rather expensive and doesn't hold up well to a high neutron flux).

P.S. I just skimmed through the Wiki Dostal link and it does seem like 500 C inlet or even lower is possible with supercritical CO2.

Here is a link to the thread I started discussing the advantages of slightly lowering the operating temperature of a molten salt reactor.

http://www.energyfromthorium.com/forum/viewtopic.php?f=17&t=652


Top
 Profile  
 
PostPosted: Jul 22, 2008 11:52 pm 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3306
Location: Alabama
I got in touch with Dostal today and he says they're looking at building a 500 kW supercritical CO2 rig in the Czech Republic.


Top
 Profile  
 
PostPosted: Jul 23, 2008 5:41 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 914
Location: Ottawa
I read or skimmed through much of Dostal's work. He includes very useful summary sections to each chapter which is of much help. One thing that can be confusing is the fact that it is almost always thermal efficiency is being discussed, whereas overall cycle efficiency is the value of importance. A good example on that is his discussion of the South African pebble bed work, their thermal efficiency is about 51% due to a high inlet temp of 900 C. However, when you add in all the losses the overall cycle drops down to 41 to 42% (see page 255 of the Wiki link). It is quite hard to find reference to overall cycle efficiency in Dostal's work.

He also addresses the multiple stage Brayton cycle that Per Peterson, Charles Forsberg and others are promoting for molten salt "cooled" designs. Dostal's opinion is that even with 3 stages of reheat and intercooling (and 3 extra compressors and turbines) that supercritical CO2 still is a better efficiency for the 500 to 700 C inlet range. He also contends that the incremental improvement in efficiency for each extra stage only would warrant adding 1 extra stage if the costs are properly included. He claims this is why the South African work only has a single extra stage. See Chapter 12 starting on page 254 for all the details, it is quite a compelling argument.

Again though, since Dostal rarely gives "overall cycle" efficiency it can be a bit hard to compare since he typically quotes the thermal efficiency when comparing helium Brayton to the supercritical CO2. About the only example I found was page 103-104 where a simplified version gave a thermal efficiency of 41.5% at 550 C but a "net" efficiency of ~37%. A 4.5% drop is pretty big. For comparison, in the multistage helium Brayton work the thermal efficiency at 600 C was 45.8% with a net efficiency of 44% (less than half the "drop" and why so much less a drop than the South African work?). Thus one has to be careful in going by thermal efficiency comparisons between helium and CO2.

Dostal's best case for 550 (without spending too much money on reheaters etc) was 45.3% thermal efficiency. If it has the same drop of 4.5% as the other more simplified CO2 case, then we are looking at a net efficiency of about 40.8%. I wish he would have mentioned the overall cycle efficiency much more often which perhaps indicates the real losses are a bit of an unknown.


Top
 Profile  
 
PostPosted: Jul 23, 2008 6:31 pm 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3306
Location: Alabama
Remember efficiency isn't everything in a Brayton cycle as well....I would contend that net work is nearly as important a metric.


Top
 Profile  
 
PostPosted: Jul 24, 2008 11:59 am 
Offline

Joined: Apr 24, 2008 4:54 am
Posts: 491
Location: Columbia, SC
I look at Reactor thermal power vs net MWe to the grid.

Example:

San Onofre Unit 2
Reactor = 3438 MW thermal
Main Generator output = 1180 MWe
Plant loads = ~50MWe
MW electrical to grid = 1130 MWe

Overall efficiency = 1130 / 3438 * 100 = 32.9%


Even a jump to 40% efficiency would be an enormous benefit.


Top
 Profile  
 
PostPosted: Jul 24, 2008 5:40 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1957
Location: Montreal
USPWR_RO wrote:
Even a jump to 40% efficiency would be an enormous benefit.

....and so would productive use of what is currently "waste heat."


Top
 Profile  
 
PostPosted: Jul 24, 2008 9:54 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 914
Location: Ottawa
I also found a report from MIT regarding how to integrate the supercritical CO2 cycle with a reactor in an indirect cycle (with reactor cooled by gas, lead or molten salt). There are lots of choices of how to put together a Power Conversion Unit (PCU). They were looking to a horizontal arrangement with everything except the generator within a single pressure vessel maintained at the lowest pressure of the cycle 8 Mpa (peak is 20 to 25). A single vessel for just 250 MWe is almost as big as a PWR vessel but only half the pressure or so. Much smaller and simpler than the massive "steam island" currently needed but still some big pieces of hardware needed. The CO2 turbines and compressors are tiny but it is the heat exchangers (precoolers, recuperators etc) that take up all the space.

By the way, does anyone know how CO2 would interact with our molten salts? Not that we`d probably ever want to do away with the intermediate loop but good to know what the behavior might be. I think I recall reading something about how it would interact but can`t recall what was said.


Top
 Profile  
 
PostPosted: Jan 02, 2009 8:19 pm 
Offline

Joined: Dec 31, 2008 12:09 am
Posts: 236
Location: Berkeley, CA
As Kirk notes, the largest advantage of closed gas cycles for liquid fluoride and other liquid salt systems is that it becomes much easier to prevent freezing and control over cooling transients.

The big issue for selecting the best gas for a closed gas cycle is the compressor specific speed. The threshold for getting an axial compressor to operate near its optimal performance and stability limits, for a 3000 to 3600 rpm speed occurs when the mass flow rate of gas corresponds to a thermal power of around 900 MWth, where helium becomes the most attractive gas (rather than nitrogen at lower power levels). This is why PBMR (400-500 MWth) has a gear box and GT-MHR (600 MWth) uses frequency inversion.

Even for lower power levels where nitrogen is the best fluid, one still adds some helium to increase the thermal conductivity of the gas and reduce the size of the heat exchangers. This is why Areva has picked an indirect cycle with nitrogen and helium as the working fluid.

Supercritical CO2 cycles look good in the temperature range that sodium fast reactors produce, but closed gas Brayton cycles (likely helium) look best at the higher temperatures delivered by the LFTR/AHTR/LIFE. Helium scarcity may become a big issue once we've used up most of our natural gas, but we'll have economical helium for a longer period of time if we can develop nuclear energy more rapidly to displace the use of natural gas for electricity, process heat, and hydrogen production.

For tritium control, the big issue is the fact that tritium diffuses rapidly through metals at high temperature, and thus it goes rapidly through the IHX to the intermediate salt, and through steam generator tubes into water. But as soon as one goes to closed gas cycles for power conversion, the entire gas cycle pressure boundary operates at low temperature (it is in contact with the low temperature gas coming from the compressors so that inexpensive materials can be used to contain the pressure), and the only interface with water is at the precooler and intercoolers where temperatures and tritium diffusion are very low. One must still recover tritium, but it's much easier because one can tolerate higher partial pressures in the primary and intermediate coolants without having to worry that it will be driven into water, where it is expensive to get out (e.g., CANDU's).

On the other hand, NGNP is now looking seriously at steam generation, mainly because it turns out that co-generation of electricity and process steam is what the petrochemical industry wants most in the near term. This is ok, since PBMR will demonstrate helium Brayton cycle technology earlier anyhow. In our 900 MWth PB-AHTR design work we have adopted the PBMR/Mitsubishi turbomachinery design with very small modifications (essentially ganging 3 power conversion trains together to get 3 expansion stages with reheat and 6 compression stages with intercooling, giving approximately 46% power conversion efficiency with a core inlet temperature of 600°C and outlet temperature of 704°C.

A more detailed design of this 410 MWe power conversion system design was developed by our NE 170 senior design class project,

Attachment:
PB-AHTR_Design_Isometric.jpg
PB-AHTR_Design_Isometric.jpg [ 92.53 KiB | Viewed 9159 times ]


They also developed a preliminary physical arrangement and building structural design that includes seismic base isolation, aircraft crash mitigation, HVAC zoning design, beryllium control, and radiation shielding analysis for the pebble transfer and storage systems. Most of this can of course be adapted directly for the LFTR as well.

-Per


Top
 Profile  
 
PostPosted: Jan 03, 2009 3:55 am 
Offline
User avatar

Joined: Dec 04, 2006 8:48 pm
Posts: 25
Location: Annapolis, MD
Per Peterson wrote:
As Kirk notes, the largest advantage of closed gas cycles for liquid fluoride and other liquid salt systems is that it becomes much easier to prevent freezing and control over cooling transients.

The big issue for selecting the best gas for a closed gas cycle is the compressor specific speed. The threshold for getting an axial compressor to operate near its optimal performance and stability limits, for a 3000 to 3600 rpm speed occurs when the mass flow rate of gas corresponds to a thermal power of around 900 MWth, where helium becomes the most attractive gas (rather than nitrogen at lower power levels). This is why PBMR (400-500 MWth) has a gear box and GT-MHR (600 MWth) uses frequency inversion.

Even for lower power levels where nitrogen is the best fluid, one still adds some helium to increase the thermal conductivity of the gas and reduce the size of the heat exchangers. This is why Areva has picked an indirect cycle with nitrogen and helium as the working fluid.


Per - thank you for the detailed post. Like Kirk, I would like to be in your design class; it sounds like you and your fellow students have done some interesting and useful work.

I have approached closed cycle nuclear gas turbine design from a different angle that has led me to significantly different choices. Not claiming my design choices are better, just that they are different because the entering arguments were different. (I am looking for well developed equipment, low initial capital costs, reliability, power output sized to compete with diesel engines and combustion gas turbines, and modest initial temperatures and pressures. I also plan to minimize the number of different heat exchangers; simple cycle combustion gas turbines have achieved a great deal of commercial success without using recuperation, intercooling, and reheat.)

Unlike aeroderivative combustion gas turbines, nuclear gas turbines are not limited to direct coupling between the compressor and turbine and have no design advantage for low cross-sectional area or low weight. (People designing aircraft engines that hang under the wings, however, take those two criteria rather seriously.) Tossing out those evaluation criteria along with the much lower maximum temperature of the system - compared to modern aeroderivative gas turbines - led me to determine that centrifugal compressors with electric motors would provide a great deal of operational flexibility.

Since I made that selection, I determined that the turbine and generator arrangement best suited for my goals is an integrated turboexpander. (For more information on turboexpanders, here is an example from GE. http://www.geoilandgas.com/businesses/ge_oilandgas/en/prod_serv/prod/turboexpanders/en/index.htm

With the Adams Engine current conceptual design, the number of compressors can be varied independently of the number of turboexpanders. The compressor speeds can be varied to aid in power control, which is obtained by controlling the mass flow rate through the system. The only heat exchangers in the system are the reactor itself and a heat sink. (There is consideration for splitting the heat sink into two separate units, one that provides high temperature discharge appropriate for a steam bottoming cycle and one that is more appropriate for either direct heat discharge or low temperature heat recovery for something like desalination.)

Again, no criticism is implied or intended for any other design; I just wanted to share a bit about why I have made different choices than others in selecting our working fluid, our compressors and our turbines.

_________________
Rod Adams
Publisher, Atomic Insights
Host and producer, The Atomic Show Podcast
Founder, Adams Atomic Engines, Inc.


Top
 Profile  
 
PostPosted: Jan 04, 2009 12:31 am 
Offline

Joined: Dec 31, 2008 12:09 am
Posts: 236
Location: Berkeley, CA
Rod,

You are correct, for lower gas flow rates and power levels centrifugal compressors have the best performance. Axial compressors become optimal at higher gas flow rates and power levels, so the best selection will depend upon the desired power output of the plant. One issue for turbo-generators is over-speed control if the generator load is lost and there is no place to dump the power. This was one of the issues that caused PBMR to drop their earlier vertical turbomachinery configuration (two turbo-compressors and one turbo-expander) and go to a Mitsubishi-designed horizontal in-line configuration with the turbine, compressors and generators integrated on a single shaft.

-Per


Top
 Profile  
 
PostPosted: Jun 18, 2009 9:33 pm 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3306
Location: Alabama
I just posted on the blog about a paper I attended today at ANS09 on SCO2:

Supercritical CO2 is dense like water


Top
 Profile  
 
PostPosted: Jun 18, 2009 10:41 pm 
Offline

Joined: Jan 24, 2007 2:24 pm
Posts: 437
Location: Montreal, Quebec CANADA
Well I'm sold, CO2 is obviously the working fluid of choice, Smaller turbine diameters mean potentially much higher max RPM as well which is also an efficiency multiplier.


Top
 Profile  
 
PostPosted: Jun 18, 2009 11:16 pm 
Offline

Joined: Mar 30, 2008 8:37 am
Posts: 146
Location: Stumbling out of a van, head full of zombie
I can't get over how much smaller the turbomachinery is for SCO2. What sort of capital savings will this enable, ceteris paribus?

How much further will you be able to push overall thermal efficiency?

_________________
Circular logic - see self-referential

Self-referential - see circular logic

imbibo ergo fnord


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 78 posts ]  Go to page 1, 2, 3, 4, 5, 6  Next

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