Energy From Thorium Discussion Forum

It is currently Oct 15, 2018 11:36 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 63 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next
Author Message
PostPosted: Nov 21, 2012 10:39 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Lars wrote:
The HX to the buffer salt would naturally be at a lower temp and using a lower melting temp salt - wouldn't that help?


Well, ORNL assumed the fluoroborate as coolant salt, having a melting point of 385 degrees Celsius, and found it needed to heat the feedwater to about that temperature to prevent freezing. A buffer salt could be KF-ZrF4, having a melt point of 390 degrees C, this would have similar problems. You'll be hard pressed to find a lower melt point halide. Chloride salts aren't that much lower melting, in fact the most attractive ones are higher melting than 390 C. Some of the BeF2 salts have lower melting points, but are too viscous at those lower temperatures so you can't operate that in a power cycle at too low temperatures.

Not a big deal for a supercritical power unit, but a big design issue for a subcritical unit, I've now learned from ORNL. Of course, the buffer salt isn't radioactive if designed appropriately, so could easily have another nitrate loop attached, in between the steam generator and buffer salt.

Sadly, we can't use the nitrate salt for the buffer salt, because it does not like the gamma rays.


Top
 Profile  
 
PostPosted: Nov 21, 2012 8:20 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Cyril R wrote:
ORNL 4541 talks about subcritical versus supercritical as well.

http://energyfromthorium.com/pdf/ORNL-4541.pdf

Table 5.2.

Net efficiency of supercritical: 44.5%
Net efficiency of subcritical: 41.1%

That's a big hit! Apparently, a big issue is the high power required to heat up the feedwater in the subcritical cycle, to prevent salts freezing.

Considering that, I guess that added nitrate loop looks great for a subcritical unit. Standard feedwater temperature could be used.
No that's not it Cyril, they are comparing two different steam cycle designs. The first is the SC with a mixing Tee and feedwater booster pumps, the second is a Loeffler Cycle where 100% of the steam mass flow is compressed absorbing a lot of energy in the process. The key lesson is that pumping high pressure water is far easier and less energy intensive than compressing high pressure steam. (From Table 5.2: 7.4 MW for pump vs 52 MW for steam compressor). There are other issues with the Loeffler Cycle like a lower average temperature at which thermal energy is added to the cycle which adversely affects cycle efficiency.

If the cycle designs were the same, but operating at different pressures the efficiency difference between subcritical and supercritical would be closer than most people think.

One area where I believe that ORNL could have done better is their steam system/steam cycle design, they definitely had some other ideas that they were going to investigate, but as funding started to dry up, further optimisation of the steam systems was shelved in favour of the other more important work on the nuclear island development. Which is quite understandable, but the overhang is that there is scope for improvement on what ORNL proposed, and not everyone knows that.


Top
 Profile  
 
PostPosted: Nov 22, 2012 3:34 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Lindsay, ORNL mentioned the choice of a different cycle for the subcritical design to be based on excessive SG surface area & throughput. So it seems that you can't economically do with a subcritical design what you can with a supercritical design, and that causes the lower efficiency.

You say that the cycles can be both very efficient, but MSRs have an odd requirement of much higher final feedwater temperature. This appears to have drastic consequences for subcritical designs.


Top
 Profile  
 
PostPosted: Nov 22, 2012 5:43 am 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Well there's a number of challenges, #1 water boils at temperatures less than 374C (pressure dependent) and NaBF4-NaF freezes at 384C, #2 the typical feedwater temperature is 250 - 300C, at least 84C less than the freezing point of the intermediate salt and #3 boiling heat transfer is very effective with the metal temperature close to that of the water/steam mixture, again promoting salt freeze.

All three of those challenges get easier to manage in a standard design at supercritical pressure and the last stage of feedheating can be done by using the mixing Tee and booster pumps. I'd like to think that one could actually get that same temperature or close to it without the booster pumps, but that's another story.

The Loeffler Cycle is transferring heat not into boiling water (or something like it), but superheated steam which has a much lower heat transfer coefficient, driving up the surface area requirements. Then a good portion of the superheated steam is taken away and used to boil water which further increases the required surface area in the salt driven superheater (I don't want to call it a steam generator, because steam generation is actually occurring somewhere else, in the mixing drum ref Figure 5.3 in ORNL-4541).

This is long-winded way of saying that the increased surface area and reduced efficiency are consequences of the Loeffler Cycle, not so much using subcritical steam pressure.

The three factors above, force you to do some unpalatable things, and it's a matter of choosing the least worst option as opposed to choosing the best. If one used a standard HX design with subcritical steam pressure, the salt freeze challenges are much harder to manage, so ORNL's approach is well considered and quite logical, but I want to do better.


Top
 Profile  
 
PostPosted: Nov 22, 2012 8:25 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Lindsay, the requirement for good heat transfer on the steam/water side and not allowing freezing of the salt in the other, appear to be contrary to each other. With low density steam in stead of water, you do need a lot of surface area but the lower heat capacity of the steam means less risk of freezing. If you have boiling water in stead of steam then the heat transfer coefficient and heat capacity on the water side are so much better that you have more risk of freezing?

You seem to have some ideas on how to avoid the worst compromises for a subcritical design. Could you elaborate or is it proprietary?


Top
 Profile  
 
PostPosted: Nov 22, 2012 7:40 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
That's a good summary Cyril, like many things MSR, initially it sounds pretty straight-forward until you start getting into some of the details, the competing and opposing design issues start to emerge.

I'm working on a steam turbine based power conversion system optimised for a 700-750C reactor outlet temperature that has high efficiency, low cost, load following capability, 5% instantaneous spinning reserve and a loading rate of at least 5%/minute. In addition I'm looking at augmenting MSR/power train capabilities to cope with grid emergencies such as grid failures and an option to re-energise and regulate the local grid following a grid emergency. Black stop and black start capability is part of the mix. This is all very ambitious and may not be allowed under regulations for NPP's, but I'm keen to establish the technical feasibility of these capabilities. Part of all that is having an extremely flexible steam generator design which can operate at subcritical or supercritical pressures with relatively normal/cool feedwater temperatures (250C or lower). Some of the important details may become proprietary information, so I can't say much.


Top
 Profile  
 
PostPosted: Jun 01, 2013 2:53 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Why is it attractive to have a low feedwater heating temperature? Modern steam cycles are sometimes over 300 degree Celsius feedwater, so it must be attractive to do this.

Less feedwater heating means less feedwater heat exchanger equipment, but it also means more heat to be transferred in the steam generators or SC water generators.

In a subcritical cycle, I can understand that it may allow a lower pressure, but it's not much. 300 degree C would require around 8.7 MPa feedwater to prevent boiling. Still only a modest pressure.

Seems to me that with the advent of printed circuit diffusion bonded heat exchangers, more feedwater heating would be the way to go. And more reheating too for that reason.


Top
 Profile  
 
PostPosted: Jun 03, 2013 4:52 am 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Cyril R wrote:
Why is it attractive to have a low feedwater heating temperature? Modern steam cycles are sometimes over 300 degree Celsius feedwater, so it must be attractive to do this.
Thermodynamically it is desirable to have a highest possible feedwater temperature if it comes from a regenerative feedwater heating system driven by bled steam. If it's not driven by bled steam it's not regenerative and you're not getting the efficiency benefit.

ORNL adopted a complex water/steam cycle to protect against salt freeze in the steam generator, it relied on supercritical pressure and high temperature preheating to protect the steam generator. The steam generator concept that I'm working on doesn't does not rely on a high feedwater temperature to avoid salt freeze, nor does it rely on supercritical steam pressure, therefore the water/steam cycle is about as simple as you can get. Simple = simple, simple = cheap, simple = operationally robust.

If you have a molten salt driven steam generator design that can operate at subcritical pressure and a feedwater temperature of 250C or lower then you've got good steam generator design. For thermodynamic reasons you may choose to run higher pressures and feedwater temperatures, but it's nice to have the choice and or lots of margin for upset conditions. And yes, a high reheat temperature is good for cycle efficiency.


Top
 Profile  
 
PostPosted: Jun 03, 2013 8:05 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Thanks Lindsay.

Using unexpanded steam or supercritical water to preheat feedwater I can see is not going to be efficient. But more bled steam going to more feedwater heaters would improve efficiency. Obviously, higher feedwater temperature means a higher max steam pressure, which may not be so good in cost-benefit, but other than that it seems just a question of economics, adding a feedwater heater costs more.

Say we use a nitrate salt loop, as third added loop, which has a melting point of 220 degrees Celsius.

Then we don't worry about freezing of steam generators anymore. But we get an upper limit of steam temperature around 550 Celsius. What efficiency can we get with this arrangement using superheat steam?

Is superheat steam simpler than supercritical? With supercritical, you have only a once through, single stage heating process. With superheated steam, there is a preheater, a steam generator, and a superheater. Also need steam seperators and steam dryers. None of that is needed with supercritical. A moisture seperator is needed later on with supercritical, superheat can avoid that as you mentioned before. So it seems to me that supercritical and superheated are similar in complexity overall, but supercritical requiring thicker walled components as an economic downside.


Top
 Profile  
 
PostPosted: Jun 03, 2013 6:01 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Using nitrate salt in a third loop is certainly a solution that would work. One of the consequences is the lower peak steam temperature as you mention will be a lower conversion efficiency. Today Siemens offer 'off the shelf' steam temperature parameters of 600C on main steam and 620C on reheat, in my model for a 1 GWe machine if we go from 175 bara 600C/620C to 175 bara 550C/550C the net output drops by about 2.5% for the same thermal input as a consequence of reduced steam temperature.

Something that seems to missing from some of these conversations (not yours) is the observation that if MSR thermal power is cheap and if the more efficient power conversion systems cost more to build, then the economics will overall favour a less efficient setup which will in turn produce the lowest capital cost and lowest overall cost of electricity in spite of the lower efficiency. It's a cruel lesson, but engineering efficiency and financial efficiency can be two quite different things.

I realise that this is not an unassailable argument, one can make an argument for more expensive power conversion systems if they are more efficient, but if the nuclear island is cheap and least cost is the prime driver, it will lead to a place of adequate thermodynamic efficiency, but not ultimate thermodynamic efficiency. Something worth pondering.


Top
 Profile  
 
PostPosted: Jun 04, 2013 9:43 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
So what is the capital cost of subcritical versus supercritical steam turbines, including salt-heaters and salt-reheaters?

For the subcritical unit, is 7 MPa much cheaper than 17 MPa?


Top
 Profile  
 
PostPosted: Jun 04, 2013 7:37 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Cyril R wrote:
So what is the capital cost of subcritical versus supercritical steam turbines, including salt-heaters and salt-reheaters?

For the subcritical unit, is 7 MPa much cheaper than 17 MPa?

I can't answer those questions with any authority, but the higher the pressure and temperature, the more expensive the kit. I suspect that for any nth of a kind plant the economic analysis will favour low supercritical conditions. A NETL study that compared multiple plants of similar MW output showed that the boiler for subcritical coal plant was 10% cheaper than SC and the STG was 3.5% cheaper and the feedwater systems 6% cheaper. So based on that I would guess that the SC steam generator and reheater would cost ~10% more than the subcritical version and the STG 4% more. At those prices the SC would probably win the economic argument, but for a first of a kind the balance could swing either way, and the overall results for a subcritical configuration are likely to be quite acceptable if the nuclear island is cheap.

My comments on efficiency and cost for MSR NPP's were more related to familiar STG compared to newer higher efficiency alternative such as triple pressure helium Brayton.


Top
 Profile  
 
PostPosted: Jun 05, 2013 2:37 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Thanks again Lindsay.

If the pressure is important to capital cost, but not so much to efficiency, then how does a (say) 7 MPa subcritical unit compare to a 17 MPa subcritical one? Should be much lower in capital cost, no?


Top
 Profile  
 
PostPosted: Jun 05, 2013 5:32 am 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 823
Location: Calgary, Alberta
Cyril R wrote:
Thanks again Lindsay.

If the pressure is important to capital cost, but not so much to efficiency, then how does a (say) 7 MPa subcritical unit compare to a 17 MPa subcritical one? Should be much lower in capital cost, no?

You can see from the figures above that the change STG cost is measurable, but not massive. I tried running a 7 MPa machine and my model was not happy, but dropping from 175 bara to 120 bara dropped the output by 3.1% for the same thermal input. Temperature and temperature drop through the machine is the big thing, sometimes you need a certain inlet pressure to achieve that, so all of these things have their limits.


Top
 Profile  
 
PostPosted: Jun 05, 2013 7:19 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Lindsay wrote:
Cyril R wrote:
Thanks again Lindsay.

If the pressure is important to capital cost, but not so much to efficiency, then how does a (say) 7 MPa subcritical unit compare to a 17 MPa subcritical one? Should be much lower in capital cost, no?

You can see from the figures above that the change STG cost is measurable, but not massive. I tried running a 7 MPa machine and my model was not happy, but dropping from 175 bara to 120 bara dropped the output by 3.1% for the same thermal input. Temperature and temperature drop through the machine is the big thing, sometimes you need a certain inlet pressure to achieve that, so all of these things have their limits.


Ok, so the effect of pressure from 17 to 7 MPa is considerable. I suppose this is mostly due to the advantage of isothermal boiling at a higher temperature with higher pressures, having a big impact on pushing up the average temperature of heat addition?

175 bara is not that far from supercritical pressures (eg ORNLs 240 bara SC turbine). So we'd expect not much of an advantage there in material cost.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 63 posts ]  Go to page Previous  1, 2, 3, 4, 5  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:  
Powered by phpBB® Forum Software © phpBB Group