Energy From Thorium Discussion Forum

It is currently Sep 20, 2017 4:05 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 63 posts ]  Go to page 1, 2, 3, 4, 5  Next
Author Message
PostPosted: Jul 01, 2015 10:22 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
I have decided to spend my idle days in summer (having now passed my university exams :D ) working out a crude model of a gas cooled heavy water moderator reactor, similar in conception to the EL-4 or a cross between an AGR and a CANDU.

It would be a calandria reactor with 'Excel' (modified zircaloy) pressure tubes with insulated tube liners cooled by ~60atm carbon dioxide.

Currently aiming for a steam outlet temperature of 565C so I can use a double reheat USC turbine for ~48% thermal efficiency (for the turbine alone). However most steam generator designs will require a rather high reactor outlet temperature to allow that. I am hoping that I can get it down to 590C because it reduces the material challenges some what. It would also enable coupling to a practical Copper-Chlorine cycle plant for hydrogen production.

I have been reading up on Zirconium corrosion by carbon dioxide and it appears that I can reach temperatures approaching six hundred celsius using a Zr-0.5Cu-0.5Cr alloy plated with chromium, which would enable use of relatively neutron efficient fuel cladding and tube liners.
This would however have to be checked as I can't find copies of some studies which apparently took place on just such an alloy (I found references in documentation about the EL-4 which says that cladding choice is excellent, but their conditions were a bit lower temperature).
It might be possible to substitute Molybdenum for the Chromium which would enable use of DepMo to cut down on the neutron losses a bit, but they should not be that large considering.

Anyone got any ideas/opinions or know how low I can push the temperature drop on the steam generator stage?
Most steam generator designs for 565C supercritical steam output suggest an inlet temperature of ~640C, but then the only real samples I can find are advanced Heat Recovery Steam Generators for CCGT plants, which are once through (obviously) and are thus optimised differently to nuclear reactor units.


Top
 Profile  
 
PostPosted: Jul 03, 2015 3:36 am 
Offline

Joined: Apr 19, 2008 1:06 am
Posts: 2219
To reduce the corrosion problem, you could use Argon gas instead of carbon dioxide. You could also use it directly to run a turbine instead of a heat exchanger like the AGR. You may work out a comparison of the costs and benefits of using light water against the heavy water.
Such a reactor could use Th-Pu or Th-233U fuel as available. The Norwegians are already trying out the first.


Top
 Profile  
 
PostPosted: Jul 03, 2015 5:05 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
Problem with a direct cycle design is that most of the gas compression cycles use far higher pressures than what I am proposing.

That in turn requires significantly thicker pressure tubes which will capture more neutrons and decrease the neutron economy of the plant.


Top
 Profile  
 
PostPosted: Jul 05, 2015 7:24 am 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 256
The GCHWR is an interesting concept. The gas coolant has very little effect on neutronics, so the void coefficient is very small. We can also go to higher temperatures than water cooled reactors.

I guess that if you increase the surface of heat exchange you can drop the delta T. Sadly I don't know if the size of the steam generator will be feasible ( technically and economically ).

Concerning the comparaison with Heat Recovery Steam Generators I guess that the density of the CO2 at 60 bars is significantly higher that the exhaust gases of the gas turbine of a CCGT, that is a great advantage for thermal exchange and for the size of the steam generator.


Top
 Profile  
 
PostPosted: Jul 05, 2015 8:52 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
The insulated HEC style design means you can design the tubes to shed ~1.5% of reactor load through the tube walls into the moderator tank.
This means that after ~4 minutes from shut-down the gas circulators become unnecessary as long as the tubes remained pressurised.
Now whilst putting a pair of hold-open (normally closed) valves at the ends of every pressure tube would get very expensive very quickly, it should be possible to putting a hold open valve on each end of a group of say a dozen tubes. (380 tubes in a reactor means ~60 valves in total which is not totally ridiculous as they are relatively small).

That way no circulation loop break anywhere in the system can depressurise more than a dozen tubes, holding a large fraction of the carbon dioxide in place, limiting containment pressure rise and thus reducing the cooling problem to simply keeping the calandria water from boiling away entirely.
This means that all the tubes that are pressurised should not cause damage if they can survive the first few minutes (I will have to run calculations on that) - the problem of the others would have to be handled in another manner but it significantly reduces the scale of the problem.

And if the break is outside the core then every tube remains pressurised and the core can be cooled through the calandria/moderator cooling system. As it is likely the reactor would have a filter-vented containment it would probably be just open cycle boiling of reactor vault water.

Anyone got any opinion?


Top
 Profile  
 
PostPosted: Jul 05, 2015 5:16 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
If we assume a turbine efficiency of ~48% (typical for 31.1MPa 565/565/565C double reheat turbine) then a reactor of the same output of a CANDU 6 [2084MWt] that means we have 1000MWe gross.
If we exclude losses for circulators, losses to the moderator through gamma heating and Calandria tubes and the like that takes us to roughly 950MWe I imagine.

If we assume the same moderator levels as in the conventional CANDU core that takes us to 265t of moderator.
The Wolsong Tritium removal plant cost roughly $115m and can handle 100kg of feed per hour. The product heavy water has effectively no tritium in it (97%+ removal efficiency).
CANDU type reactors acquire tritium at a rate of about 5Ci/kg of moderator water every year, roughly ten years of allowances.
The Wolsong plant would be able to process the entire water supply on average 3.3 times per year, that means the equilibrium concentration of tritium in the core will be roughly 1.5Ci/kg.

Which means the entire moderator supply contains only 401kCi.
Which is not that big compared to the tritium release licenced by a reactor.
According to this report - releases at the Darlington reactor complex were ~3500Ci of gaseous tritiated water, which was only 0.3% of the permitted licence release for the year, which works out at roughly 1.17MCi - adjusting for the 950MWe/3500MWe capacity issue that brings the permitted release to 301kCi.
So we could blow the entire moderator inventory into steam and dump it out the plant stack and only just go over permitted limits.
If we include liquid effluent we are nowhere near the limit.

So it means if we build a TRF of that size for the reactor (which costs only about $120/kW) it means we probably don't have to keep the moderator water and the vast tonnages of water in the calandria vault separated in a serious accident.
That simplifies the design considerably - especially for a nightmarish pressure tube failure accident which would potentially dump large quantities of very high pressure carbon dioxide into the calandria.


Top
 Profile  
 
PostPosted: Jul 06, 2015 5:32 pm 
Offline

Joined: Oct 28, 2013 12:24 am
Posts: 256
You can maybe use CO2 for normal operation and an injection of borated light water in case of LOCA.

You can inject borated light water in the broken pressure tubes thanks to an accumulator. If the spectrum is very well thermalized (like a CANDU) there is no need for boron, contrary to the fast concepts of the other thread, I guess that light water will here reduce the reactivity. Look at the last accumulators ( for example for the US APWR ), they are very good, you can use several accumulators which open passively at different pressures or one accumulator at the highest pressure with CO2 and the other ones with water.

And then you use something like the passive safety systems of the ESBWR to cool the containment and the core. Water has the big advantage to be condensable.


Top
 Profile  
 
PostPosted: Jul 08, 2015 7:50 am 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
The reason I want the tritium scrubber is that if a pressure tube breaks inside the Calandria you could get a large quantity of very high pressure carbon dioxide gas venting into it - something like a dozen pressure tubes worth plus the gas in the sub headers.
This will necessitate blow-out panels in the Calandria structure to vent the gas to the vault tank, which obviously breaks containment of the heavy water and mix it in with the light water.
If we had squib valves on the top headers of each reactor we could ensure that all compromised tubes flood completely which should be enough to stop them overheating so much as to suffer fuel damage. (The squib valve would open the top sub-header to the vault tank, allowing any gas/steam trapped above the breach in the tube to vent instead of causing a vapour pocket that won't cool the fuel).

Problem is a 384-tube design would need something like 32 subheader clusters - so 64 hold-open valves and 32 squib valves, although they are relatively small.
Not sure if that is relatively viable?


Also to prevent fuel overheating in the first few seconds of an accident how about the non-compromised tubes be held open using battery power and the gas circulators be kept running for as long as possible at low speed so the heat can be moved out of the tubes?
It would handle the nasty shutdown thermal-transient.
Would need a big battery bank but we are only talking about something like five minutes at 10% rated gas circulation, which will be something less than ten percent power due to the reduced pressure drop.


Top
 Profile  
 
PostPosted: Jul 08, 2015 2:56 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5048
Zirconium alloys have poor strength at 600C. Plus, you're operating a gas coolant at modestly low pressure, so that means high delta T across the cladding (or unacceptably small power density compared to heavy water coolant). Zirconium would probably have unacceptable oxidation in CO2 as well. If you want this likely you will have to go helium cooling + vented fuel elements.

I'm actually going to suggest triplex SiC-SiC-SiC cladding for this reactor, if you want to keep it CO2 cooled. The SiC cladding actually "likes" operating at 700-800C.

As for the SG, you can design for just about any pinch point. It is going to cost you in both pumping power and SG size, so I'm guessing the economic optimal with gas heated SGs will still be quite large. AGRs are at about a 100C pinch, from 640C CO2 to 540C steam. And their station pump power is 42000 kWe :shock: .


Top
 Profile  
 
PostPosted: Jul 08, 2015 5:03 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
Cyril R wrote:
Zirconium alloys have poor strength at 600C. Plus, you're operating a gas coolant at modestly low pressure, so that means high delta T across the cladding (or unacceptably small power density compared to heavy water coolant).

I hope I can win a hundred degrees back by using that tin-lead-bismuth alloy bonding, and I think I can keep cladding over-pressures down by using fuel elements that go the full height of the core and having giant fission gas plenums on the top of the fuel element.

Cyril R wrote:
Zirconium would probably have unacceptable oxidation in CO2 as well. If you want this likely you will have to go helium cooling + vented fuel elements.

The French successfully operated the EL-4 (Carbon dioxide cooled heavy water reactor) using a Zr-1.8Cu alloy and a reactor outlet temperature of ~510C.
They apparently did research that indicated that a chromium plated Zr-0.5Cu-0.5Cr alloy was drastically better on the corrosion front - which is why I am hoping I can achieve a reactor outlet of ~590-600C without massive corrosion issues.
Unfortunately the easily accessible research on this is apparently buried in archives somewhere or in French, having trouble finding much beyond passing references in some books.

Cyril R wrote:
I'm actually going to suggest triplex SiC-SiC-SiC cladding for this reactor, if you want to keep it CO2 cooled. The SiC cladding actually "likes" operating at 700-800C.

The problem with SiC cladding in my opinion is that I am trying to keep the reactor using "available today" technology and the cladding is still dogged with the 'joining' issues that have been hurting it for years. Although this reactor has potentially less trouble about that because all the fuel cans will go the full height of the core and there are not many joints to worry about.
Cyril R wrote:
As for the SG, you can design for just about any pinch point. It is going to cost you in both pumping power and SG size, so I'm guessing the economic optimal with gas heated SGs will still be quite large. AGRs are at about a 100C pinch, from 640C CO2 to 540C steam. And their station pump power is 42000 kWe :shock: .

AGRs are a bit of an interesting case.
They are pressure vessel reactors with the boilers inside the spherical pressure vessel, crammed into the gaps made by putting a cylinder inside a sphere. That puts a limitation on the volume of the generators and I imagine also causes the pipework to not exactly be optimal for low resistance gas flow.

With a pressure tube reactor like this we can make the steam generators almost arbitrarily large, since they are only inside the relatively-low pressure secondary containment, and thanks to segmenting of the coolant system it seems likely that a larger secondary containment will suffer lower peak pressures despite the larger total carbon dioxide inventory inside the steam generators if they are bigger. If we accept proper isolation valves (which will be a relatively minor cost increase on a billion dollar+ reactor) then a breach in the steam generator or reactor will not depressurise the other half of the gas circuit.
EL-4 apparently managed a 20C pinch and I have seen work that suggests my 30C target has been achieved in some experimental HRSGs.

I am considering putting all the motor drive converters and even the station batteries in the "vault tank" because it will put a hard limit on the temperature they can be at during accident conditions and it provides more volume and a lower peak pressure and pressure spikes that could cause damage to equipment in a LOCA.

There is also the historical matter that the AGR outlet steam conditions were selected for a reason - they are the standard steam conditions used for all CEGB steam turbine sets of the era. They wanted commonality and used the same turboalternators and turbine rigs used in coal plants.
That is also why the reactors are relatively small compared to LWRs.

EDIT:
Rather interestingly the AGR's gas circulator cooling duty was only ~4.5MWt, which means that 44MWt of effective thermal output was added to the core (the energy has to go somewhere, and the only other place to go is into the gas).
That is about 18MWe of extra power, which means the effective gas circulator duty on the AGR was only about 30MWe, still huge but not so much.
With the higher efficiency motors available to us (canned rotors where we can keep the coils chilled easily) and our better turbines this effective amount will decrease further.
But I would have to run proper simulations.

Basic design assumption is the reactor must be safe for as long as possible assuming a catastrophic single break in the primary circuit simultaneously with the plant losing everything outside the vault containment.


Last edited by E Ireland on Jul 08, 2015 6:03 pm, edited 1 time in total.

Top
 Profile  
 
PostPosted: Jul 08, 2015 5:49 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5048
Quote:
I hope I can win a hundred degrees back by using that tin-lead-bismuth alloy bonding


That'd cut the dT between the fuel and cladding, but not do much for the cladding-coolant dT.

Quote:
I think I can keep cladding over-pressures down by using fuel elements that go the full height of the core and having giant fission gas plenums on the top of the fuel element.


Big plenums mean lower fuel power density, and the gas cooled reactor already does poorly here. Plus you still need enough strength to prevent buckling, especially early on. Pressure vessel wise, you're basically making a deepwater submarine - low pressure in the tubes especially at BOC, but 60 atm outside it. Buckling-thermal creep interaction must be pretty severe to design for I guess.

Quote:
They apparently did research that indicated that a chromium plated Zr-0.5Cu-0.5Cr alloy was drastically better on the corrosion front - which is why I am hoping I can achieve a reactor outlet of ~590-600C without massive corrosion issues.


Even if you have acceptable corrosion, you would like to have cladding surviveability after accidents and transients. Gasses are poor natural circulators especially with a monstrous SG. Loss of circulator flow means big overheating fast. Any LOCA means overheating fast. So you've got to get some short term high temp strength. If you're going to suggest something new, it'd better be accident-tolerant.

It's nice to avoid the hydrogen with H2O/D2O coolant. I'll grant you that. Too bad CO2 itself has limited condensibility, though (nothing like steam quenching in water pools, obviously).

Quote:
The problem with SiC cladding in my opinion is that I am trying to keep the reactor using "available today" technology and the cladding is still dogged with the 'joining' issues that have been hurting it for years.


High temperature brazing seems to work fine, even during temperature excursions. No need for the joint to be SiC... though it would be nice to develop a SiC joint later on, it isn't necessary now.


Top
 Profile  
 
PostPosted: Jul 08, 2015 6:06 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
I'm proposing designing the tubes so they shed 1.5% of channel power to the moderator at operating temperature.
That way if we can get over the "shutdown thermal transient" by whatever mechanism then as long as the tube pressures stay up the reactor should be safe if the moderator tank can be kept filled with water.

Hence the rather extreme proposed segmentation of the primary loop.


Top
 Profile  
 
PostPosted: Jul 08, 2015 6:17 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
Cyril R wrote:
Big plenums mean lower fuel power density, and the gas cooled reactor already does poorly here. Plus you still need enough strength to prevent buckling, especially early on. Pressure vessel wise, you're basically making a deepwater submarine - low pressure in the tubes especially at BOC, but 60 atm outside it. Buckling-thermal creep interaction must be pretty severe to design for I guess.


Well I'm thinking that the fission gas plenum doesn't actually have to be inside the core, it can protrude up into the gas risers (up to the sub-header). That gains you up to 50cm I would imagine.
Since the plenum's neutronics are not really a concern you could potentially insert a zircaloy or steel (if its compatible obviously) reinforcing sleeve into the top of the can to stop the plenum buckling.

Would the part of the can containing the fuel buckle however? I suppose it would depend if the alloy was sufficiently elastic to simply compress down on the fuel pellet, displacing the liquid metal bond into the plenum?
Cyril R wrote:
It's nice to avoid the hydrogen with H2O/D2O coolant. I'll grant you that. Too bad CO2 itself has limited condensibility, though (nothing like steam quenching in water pools, obviously).

I'm running with the assumption of a vented containment, at least in the first few hours of the accident, carbon dioxide has negligible activation and with a tritium scrubber of the type I'm proposing we can just vent the containment gas through a big set of passive filters.

Turns out the carbon dioxide-zirconium reaction is exothermic, however it will tend to reduce the amount (moles) of gas at high temperatures as it converts to a mix of zirconia and carbon.
So the pressure will probably drop in sealed tubes that overheat, with the temperature increase partially offsetting the precipitation out of the gas.


Top
 Profile  
 
PostPosted: Jul 08, 2015 6:27 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5048
The pressure tubes are insulated, yes I get that. It does not do much for the fuel cladding short term transient though. Its going to run hot, albeit briefly.

How do you refuel the reactor, if you have plenums outside the core?


Top
 Profile  
 
PostPosted: Jul 08, 2015 9:30 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1482
My current plan is to unbolt caps attached to the top of each pressure tube gas riser and just pull the assembly out.
This would partially obstruct the sub header as the assembly comes out but since I am not currently proposing on load refuelling this should not be a problem (I think).

Yes, the "shutdown transient" is an issue because the assemblies will heat up.
The critical time is apparently the first five minutes after shutdown - after which time the heat radiated to the moderator reaches ~1.5% and pressurised tubes should be able to shed all their decay heat without the gas circulators.

If we assume there has not been a catastrophic steam generator LOCA (and thus the main gas circuit is intact) then we could use a battery pack inside the vault to run the circulators at low power for the first five minutes after shutdown [80% of the excess thermal energy is added in the first five minutes].
If we assume a 48MWe circulator demand, five minutes at 10% power is only a 400kWh battery - which should be more than enough.
Providing a pair of 400kWh battery packs is going to be almost nothing on the plant cost, its about $170/kWh (lead-acid) which puts us at $136,000. And there should be sufficient water reserve in the steam generators to boil off to remove that heat - hell the amount of carbon dioxide in the system would probably absorb much of it just heating it uniformly to the reactor outlet temperature.

If we assume the steam generator circuit is inoperable due to a depressurisation in the steam generator "half" of the circuit or in the main gas headers the situation becomes rather more tricky.
We would have to calculate if we are going to get significant cladding damage in the first 40 minutes of the accident - at which time the amount of energy in the pressure tube should drop below that at reactor shutdown thanks to the steady 1.5% loss to the moderator.
I calculate the peak addition of thermal energy to the tube is equivalent to roughly 3.5 seconds at maximum load.
That sounds like it might be survivable.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 63 posts ]  Go to page 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