Energy From Thorium Discussion Forum

It is currently Oct 17, 2018 11:18 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 52 posts ]  Go to page Previous  1, 2, 3, 4  Next
Author Message
PostPosted: Feb 23, 2017 10:27 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1553
Cyril R wrote:
They're complicated though. Tendon collection galleries, cable galleries, prestressing equipment and inspection abilities. Compared to just a metal membrane that holds pressure.

They are also easy to screw up. Look up Crystal River NPP. The utility made a mess of that (I guess that is total stupidity and that isn't much of an argument against prestressed structure).

That whole thing just stinks of corporate types ignoring what the engineers on site said and deciding to cut corners in pursuit of the bottom line.
I have never heard of anything like that occuring in any of the British or French prestressed concrete pressure vessels - which have a remarkable safety record.
I suppose the British (and to a lesser extent the French) have a much greater knowledge of such vessels due to them being at the core of the entire reactor programme from the 60s until the 80s.

Cyril R wrote:
Well, there's many ways to skin this cat. Interestingly stagnant water (heavy or light) is a good enough insulator so all you have to do is trap water so it can't flow. That's common practice in process engineering, it is typically called thermal sleeving and is widely employed in piping, especially in nozzle to vessel connections... all you have to do is weld some stainless sheets with their ends blocked as a "liner" and then heavy water gets in, and stagnates to insulate. Simply increase the number of sheets to increase the insulation. The insulation doesn't have to be very good, just decent enough that it doesnt become an annoying parasitic heat loss for the liner. Unless you have some interesting low temperature heat demand somewhere... I would say anything under 1% reactor heat loss is fine.

That is probably the most feasible option - ceramic insulation would end up enormously thick.
The question is then whether you have a light water system inside the vessel with the attendant risk of downgrading of the heavy water during a small leak (which would be incredibly difficult to fix) or if you tolerate the greater cost of using heavy water in the insulation layers.
The (relatively) low temperature of the moderator during normal operation is helpful in this regard.
Cyril R wrote:
What size vessel are you looking at? Even a quite large cast steel vessel won't need to be meters thick even to hold 200 bar and without prestress.
A 10 meter casting of 0.5 meters thick would deal with your pressures easily without prestress cables.

The CANDU 9 has a shield tank diameter of ~13.4m - although it only has 480 tubes it can apparently fit 640 without changing the shield tank.
So a 700 tube reactor would be looking at 14.5m or something conservatively.
Add a half metre on both sides for insulation and other bits and pieces. And we are looking at 15.5m inner diameter.
[This could be drastically reduced by deleting the shield as that would not be necessary, however the active core diameter of a 700 tube reactor is 8.6m or so. So at minimum you want something like 11m or so.]

Reactor core is 6m long, or something like 8m with the end shields, but since I want hoods covering the fueling machine it coudl come out at 15m or so, including the reactor end shield that is also a pressure boundary in normal operation.
And then there are the cavities for the heat transport pumps, gas heaters [the main gas heaters alone would be 200 cubic metres, not including the moderator heat recovery] and such like to consider, and the topside extension for the reactivity mechanisms and the mounting points for the emergency cooling thermosyphons.

Either way we are looking at something with an internal volume of a couple of thousand cubic metres - it is approaching the volume of some of the early gas cooled reactor vessels, albeit for a much higher power and at a much higher pressure.
Cyril R wrote:
Prestress is required for concrete high pressure vessels because the material has basically nil tensile strength. Metals are very different...

I am slightly concerned about the enormous nature of a tensile casting or the huge welds required to make such a thing work.


Last edited by E Ireland on Feb 23, 2017 11:06 pm, edited 1 time in total.

Top
 Profile  
 
PostPosted: Feb 23, 2017 11:03 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Quote:
That whole thing just stinks of corporate types ignoring what the engineers on site said and deciding to cut corners in pursuit of the bottom line.
I have never heard of anything like that occuring in any of the British or French prestressed concrete pressure vessels - which have a remarkable safety record.
I suppose the British (and to a lesser extent the French) have a much greater knowledge of such vessels due to them being at the core of the entire reactor programme from the 60s until the 80s.


Yeah it wasn't a good case in point. More a case of don't do blatantly obviously stupid things that a single apprentice engineer or interested layman can understand to be stupid. And don't put the money-grubbing techno-luddites in charge of decision making of an engineering project.

Still, if there are no cast steel/iron prestressed vessels in operation you should investigate any gremlins to great depth.

Quote:
The question is then whether you have a light water system inside the vessel with the attendant risk of downgrading of the heavy water during a small leak (which would be incredibly difficult to fix) or if you tolerate the greater cost of using heavy water in the insulation layers.
The (relatively) low temperature of the moderator during normal operation is helpful in this regard.


Obviously the liner coolant pressure will be hydrostatic, basically pool pressure, so any leak will be outward. The opposite problem - no downgrading of heavy water but having to deal with tritium in the pond. Hard to see a leak though if the channels are part of the casting. That would require a through-crack at which point you do have serious other issues to consider. I guess you would want some provisions to isolate channels so you can plug them and have redundant ones so a dead channel doesn't lead to local hot spots. The latter being a concern with concrete but probably not with a metal casting, given the conductivity and better cracking and thermal resistance.

Quote:
The CANDU 9 has a shield tank diameter of ~13.4m - although it only has 480 tubes it can apparently fit 640 without changing the shield tank.
So a 700 tube reactor would be looking at 14.5m or something conservatively.
Add a half metre on both sides for insulation and other bits and pieces. And we are looking at 15.5m inner diameter.


So, assuming a stress goal of 200 MPa for cast steel it would come to a requirement for under 4 mm thickness per atmosphere of pressure. If you don't mind me inventing new units there, it's not a lot of thickness.

Quote:
Either way we are looking at something with an internal volume of several thousand cubic metres - it is approaching the volume of some of the early gas cooled reactor vessels, albeit for a much higher power and at a much higher pressure.


This could be an issue, from a stored energy viewpoint. I'm thinking you want to bury this whole thing in a nice deep cold pool of clean light water. Any LOCA would be inherently suppressed (at the cost of having to separate the D2O from the pond water again...)

A sketch of some sort would help a lot if you have any, to explain the layout of the reactor and everything.

Quote:
I am slightly concerned about the enormous nature of a tensile casting or the huge welds required to make such a thing work.


So how do you deal with this in a pure compression situation? Presumably you would need a very strong seal face that can resist the huge compression forces without leakage. I'm thinking if you can do that, you can design a keyed casting with a seal face in a geometry and location that compresses the seal from the internal pressure in a tensile casting situation. Obviously I'm not going to suggest butt-welding a meter thick casting through-and-through. This design will have to be more clever than that...


Top
 Profile  
 
PostPosted: Feb 24, 2017 12:11 am 
Offline

Joined: Apr 19, 2008 1:06 am
Posts: 2246
Security of high cost heavy water as well safety of plant, operators and general public from radio-active matter are all very important. I feel they are best served by keeping pressures low in the reactor.
PHWR design has high pressure in tubes only. There is a need to develop a low pressure heat transfer agent. There is a need for a heavy oil or a salt eutectic. I am apprehensive about FLiBe on availability/cost issues.
Heavy construction of the type being discussed is very interesting but not very desireable in the reactor. We can have that in the generating sections. Go for supercritical systems there if you like for high efficiency.


Top
 Profile  
 
PostPosted: Feb 24, 2017 1:33 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
jagdish wrote:
Security of high cost heavy water as well safety of plant, operators and general public from radio-active matter are all very important. I feel they are best served by keeping pressures low in the reactor.
PHWR design has high pressure in tubes only. There is a need to develop a low pressure heat transfer agent. There is a need for a heavy oil or a salt eutectic.


Oil was tried in Canada. It's not stable under the intense radiation inside a reactor core. Water is nice because you can catalytically recombine it with high efficiency and effectiveness. Are there any oils that can be catalytically recombined without excessive third or nth order recombined products?

Will molten NaNO3-KNO3 eutectic be stable enough to use?

I guess none of these would be ok in an accident with heatup in terms of making lots of dissociation products?


Top
 Profile  
 
PostPosted: Feb 24, 2017 2:54 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1553
Cyril R wrote:
Oil was tried in Canada. It's not stable under the intense radiation inside a reactor core. Water is nice because you can catalytically recombine it with high efficiency and effectiveness. Are there any oils that can be catalytically recombined without excessive third or nth order recombined products?

The CANDU-OCR project dia a lot of work on it and came to the conclusion that if you use a terphenyl an on site catalytic cracking unit can recover a significant amount of material as useful coolant (even if it is not exactly the same as the reference coolant). Especially if you provide some or all of your regeneration material as toluene.

The problem with these systems is that the cost of them is reduced neutron economy due to the hydrogen in these cores, that really hurts you. Unless you went for deuterated terphenyls which are enormously expensive. It easily overwhelms the reactivity insertion you get or using higher density uranium carbide fuels in place of oxide.

Even a 1200MWe CANDU-OCR was anticipating something like 5,000-6,000 MWd/t fuel burnups.

This is why I have decided to stick with heavy water.
About the stored energy - we are better off than it might appear because most of the heavy water in the vessel is in the moderator loop and is thus at 170C or below, so it doesn't have a very large flash fraction.
And with the prestressed vessel making catastrophic failure impractical and the fueling 'hood' excluding a channel closure loss from totally emptying the reactor the largest break is either going to be a failure of a cooling jacket line or a heavy water detritiation line that is sized for a flow of ~20kg per hour anyway.


Top
 Profile  
 
PostPosted: Feb 24, 2017 3:15 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1553
Cyril R wrote:
Still, if there are no cast steel/iron prestressed vessels in operation you should investigate any gremlins to great depth.

The major work on them currently seems to be with the GT-MHR people, so I might try and reach out to someone over there who does work with the vessels and see if they have any recent work on the failure modes and such.

Cyril R wrote:
Obviously the liner coolant pressure will be hydrostatic, basically pool pressure, so any leak will be outward. The opposite problem - no downgrading of heavy water but having to deal with tritium in the pond. Hard to see a leak though if the channels are part of the casting.

Oops yes, I forgot that the pipes would be at effectively atmospheric pressure - that solves that issue, and if the pipes are as small as you suggest then the loss rate would be so small that it might be feasible to use a zinc-air battery fitted to a feed pump that could just keep pumping water into the reactor to keep it full.
Cyril R wrote:
So, assuming a stress goal of 200 MPa for cast steel it would come to a requirement for under 4 mm thickness per atmosphere of pressure. If you don't mind me inventing new units there, it's not a lot of thickness.

That is much thinner than I thought, but that does raise questions about why not everyone is simply casting vessels for ordinary light water reactors. We also have to consider that our vessel is most certainly not cylindrical with all the add ons.

Cyril R wrote:
This could be an issue, from a stored energy viewpoint. I'm thinking you want to bury this whole thing in a nice deep cold pool of clean light water. Any LOCA would be inherently suppressed (at the cost of having to separate the D2O from the pond water again...)

Even if the D2O is diluted a hundred to one it is still many times richer than ordinary water - so the reupgrade cost would be much lower than in the production of new heavy water.
Cyril R wrote:
A sketch of some sort would help a lot if you have any, to explain the layout of the reactor and everything.

Find one attached to this post - the grey areas which are the heat exchangers and circulation pumps I forsee being attached to the sides of the vessel, although it might be worth elevating the main gas heaters to try and get some thermo circulation.
The reactivity deck is to be filled with heavy water to the top, which will allow clutched control rods to fall in under gravity.
The drawing is obviously not to scale as that deck will be much taller and the vessel is much shorter. This preserves the major safety advantage over conventional reactors of having no pressure differential across the control system.
Cyril R wrote:
So how do you deal with this in a pure compression situation? Presumably you would need a very strong seal face that can resist the huge compression forces without leakage. I'm thinking if you can do that, you can design a keyed casting with a seal face in a geometry and location that compresses the seal from the internal pressure in a tensile casting situation. Obviously I'm not going to suggest butt-welding a meter thick casting through-and-through. This design will have to be more clever than that...

I assume you would just machine the edges of the castings so they fit together tightly, using keyways and keys to stop them dislocating - and then just trust that the liner will hold and prevent pressurised fluid from getting in between the castings.
But I am not an expert in such things.


Attachments:
Vessel Layout V1 - NOT TO SCALE.png
Vessel Layout V1 - NOT TO SCALE.png [ 22.14 KiB | Viewed 1672 times ]
Top
 Profile  
 
PostPosted: Feb 24, 2017 4:09 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Quote:
if the pipes are as small as you suggest then the loss rate would be so small that it might be feasible to use a zinc-air battery fitted to a feed pump that could just keep pumping water into the reactor to keep it full.


What do you mean? Isn't this liner cooling system just a passive self-drawing pond of water cooling system? I was thinking that the pool would be used to store spent fuel. So double duty, passive liner cooling and passive SNF cooling.

Quote:
That is much thinner than I thought, but that does raise questions about why not everyone is simply casting vessels for ordinary light water reactors.


Castings are used for nuclear pressure rated components, mainly in smaller components such as valve bodies. I have heard the notion that castings can have inclusions or voids that may be difficult to detect to a nuclear level of quality. Don't know enough about castings myself. I know cast stainless steel is really tough and strong and can be used for high pressure applications and also for high toughness/impact applications.

Quote:
We also have to consider that our vessel is most certainly not cylindrical with all the add ons.


I've done FEA on square pressure vessels. Trust me, you want it to be as round as you can make it.

Quote:
Even if the D2O is diluted a hundred to one it is still many times richer than ordinary water - so the reupgrade cost would be much lower than in the production of new heavy water.


That's true. Should be ok to have an upgrader onsite. If you use a liner-cooling pond for SNF storage you probably want that (otherwise you need to really carefully dry SNF bundles).

Thanks for sketch. Will look at it later when I have time.


Top
 Profile  
 
PostPosted: Feb 24, 2017 4:45 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1553
Cyril R wrote:
What do you mean? Isn't this liner cooling system just a passive self-drawing pond of water cooling system? I was thinking that the pool would be used to store spent fuel. So double duty, passive liner cooling and passive SNF cooling.

Yes, I was referring to some sort of pipe break and cracked casting leading to fluid leaking out through the liner cooling pipework.
THe pipes are small enough that even in that situation only very small amounts of water can escape.

Cyril R wrote:
That's true. Should be ok to have an upgrader onsite. If you use a liner-cooling pond for SNF storage you probably want that (otherwise you need to really carefully dry SNF bundles).

Even in normal CANDUs the fuel is transitioned from heavy water to light water a matter of a couple of minutes after the fuel bundles are out of the core.

It appears that the fuel bundles will be cool enough to go to a Wylfa/MVDS style dry store (~500W per bundle] about 100-115 days after they come out of the reactor. So the spent fuel supply in the cooling pool would be about 1500 bundles plus a buffer, I want as little fuel as possible anywhere near the reactor to reduce the total containment heat load in an accident.


Top
 Profile  
 
PostPosted: Feb 24, 2017 7:36 pm 
Offline

Joined: Apr 19, 2008 1:06 am
Posts: 2246
Cyril R wrote:
jagdish wrote:
Security of high cost heavy water as well safety of plant, operators and general public from radio-active matter are all very important. I feel they are best served by keeping pressures low in the reactor.
PHWR design has high pressure in tubes only. There is a need to develop a low pressure heat transfer agent. There is a need for a heavy oil or a salt eutectic.


Oil was tried in Canada. It's not stable under the intense radiation inside a reactor core. Water is nice because you can catalytically recombine it with high efficiency and effectiveness. Are there any oils that can be catalytically recombined without excessive third or nth order recombined products?

Will molten NaNO3-KNO3 eutectic be stable enough to use?

I guess none of these would be ok in an accident with heat up in terms of making lots of dissociation products?

There could be other molten salt coolants-like NaF-ZrF4 or NaF-BeF2, less neutron efficient than flibe but acceptable.


Top
 Profile  
 
PostPosted: Feb 24, 2017 8:05 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
jagdish wrote:
Cyril R wrote:
jagdish wrote:
Security of high cost heavy water as well safety of plant, operators and general public from radio-active matter are all very important. I feel they are best served by keeping pressures low in the reactor.
PHWR design has high pressure in tubes only. There is a need to develop a low pressure heat transfer agent. There is a need for a heavy oil or a salt eutectic.


Oil was tried in Canada. It's not stable under the intense radiation inside a reactor core. Water is nice because you can catalytically recombine it with high efficiency and effectiveness. Are there any oils that can be catalytically recombined without excessive third or nth order recombined products?

Will molten NaNO3-KNO3 eutectic be stable enough to use?

I guess none of these would be ok in an accident with heat up in terms of making lots of dissociation products?

There could be other molten salt coolants-like NaF-ZrF4 or NaF-BeF2, less neutron efficient than flibe but acceptable.


I'm sure that this means you cannot use CANDU style refueling machines. Total redesign would be required.


Top
 Profile  
 
PostPosted: Feb 24, 2017 8:11 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Quote:
It appears that the fuel bundles will be cool enough to go to a Wylfa/MVDS style dry store (~500W per bundle] about 100-115 days after they come out of the reactor. So the spent fuel supply in the cooling pool would be about 1500 bundles plus a buffer, I want as little fuel as possible anywhere near the reactor to reduce the total containment heat load in an accident.


If you have SiC fuel cladding you can probably get the material out even faster. The stuff has basically the same strength at 1000 Celsius than at 100, oddly.

Are you going to use a passive containment cooling system of some sorts? If you have one it doesn't matter much. Old SNF is not the problem in an accident, contrary to popular opinion. Recent US analysis has shown that it can actually help by providing a heat sink for the hotter, fresh spent fuel, by surrounding fresh assemblies with old ones, helping in a pool draindown type accident.


Top
 Profile  
 
PostPosted: Feb 24, 2017 8:31 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Also I think you should be able to can up individual SNF bundles in stainless steel cans using a canning robot. Maybe double walled cans to be sure. The cans can then be placed in a rack with sufficient lattice pitch for air cooling. Then all you'd have to do is provide shielding around the rack. That could be a cast steel shield with inlet and outlet air holes. That would lead to a high surface area for cooling so you can remove the fuel to dry storage much earlier. Still I would wait say 10 half lives of I-131 just in case as a good design point.


Top
 Profile  
 
PostPosted: Feb 24, 2017 11:00 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1553
Cyril R wrote:
If you have SiC fuel cladding you can probably get the material out even faster. The stuff has basically the same strength at 1000 Celsius than at 100, oddly.

That 500W figure is based on the fuel temperatures required for a Magnox fuel element, a 37 element Zircalloy bundle would likely come out as tolerant of higher temperatures (and have greater surface area) and thus likely a shorter wet storage period. But the idea of waiting out ten 131I half lives is probably a good one. It is unlikely we could get it out much faster than that anyway.
Cyril R wrote:
Are you going to use a passive containment cooling system of some sorts? If you have one it doesn't matter much. Old SNF is not the problem in an accident, contrary to popular opinion. Recent US analysis has shown that it can actually help by providing a heat sink for the hotter, fresh spent fuel, by surrounding fresh assemblies with old ones, helping in a pool draindown type accident.

After the somewhat dubious performance of things like Isolation Condensers at Fukushima I am somewhat nervous about the performance of containment cooling systems, passive or otherwise, and think it would likely be advisable to get as much fuel out of the containment as possible. After all every watt of spent fuel not in the containment is a watt of cooling available for the reactor.


Top
 Profile  
 
PostPosted: Feb 25, 2017 1:37 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
Quote:
After the somewhat dubious performance of things like Isolation Condensers at Fukushima I am somewhat nervous about the performance of containment cooling systems, passive or otherwise, and think it would likely be advisable to get as much fuel out of the containment as possible.


The IC at Fukushima was not passive, it needed active power to operate and that was unavailable. Also it had a tiny tank of water for boiloff and that also needed a powered valve to add more water. The non-condensable management system also needed power IIRC. New reactors like ESBWR have all of this passive - big pool (>7 days worth) rather than small tank, condensate return valves that fail-on (open) on loss of power, and passive non-condensables venting to the suppression pool.

However, I will concede that none of the passive LWR systems are truely passive. That would require no valves at all.

Quote:
After all every watt of spent fuel not in the containment is a watt of cooling available for the reactor.


Not necessarily. At Fukushima they had enough water to cool the spent fuel, but no cooling for the reactor. Would have been better if some of that SNF coolant was available for boiloff in the reactor.

Not sure but I suspect SNF pool costs are not strongly a function of water volume. So you can just oversize the pool and have crazy amounts of heat sink.


Top
 Profile  
 
PostPosted: Feb 26, 2017 12:18 pm 
Offline

Joined: Jun 19, 2013 11:49 am
Posts: 1553
I wonder if I could use SERPENT to run a lattice calculation and verify my guestimates on reactor burnup.
But that is a significant amount of work that would likely be performed in an actual thesis, which is annoying.

But I think the reactor design has a great deal of potentail with a recompressed power cycle.

EDIT
Additionally - would it be possible to size the heavy water cleanup lines such that to provide 24kg/h of reactor coolant outlet for core cleanup, that they would have to drop the vast majority of the reactor pressure across the wall of the vessel?
That way even a gilluotine break of the line beyond the vessel wall would lead to only ~24-30kg/h escaping?
A suction pump could be provided immediately outside the vessel to repressurise the coolant, I assume the outlet structure would be some form of serpentine tube, it cant be simply small or it might clog easily if anything gets into the reactor somehow.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 52 posts ]  Go to page Previous  1, 2, 3, 4  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