Energy From Thorium Discussion Forum

It is currently Feb 23, 2018 11:19 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 76 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
PostPosted: Feb 23, 2011 9:11 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3069
For what kind of reactor is this?


Top
 Profile  
 
PostPosted: Feb 23, 2011 9:49 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
Lars wrote:
But I have a hard time picturing the mechanism to cause a flow of 12 tonnes/sec of fuel salt flowing at 1-3 meters/sec to be stopped in 10mSec
Yes I'm struggling to see that too, that's a lot of mass and monemtum, that combined with some natural circulation, it is hard to imagine all of that stopping so quickly.

Our mutual friend the REBUS 3700 concept design has a mean effective neutron life time of 2.24 x 10-2 s (Ref Bokov 2005) which is a lot slower than I would have thought for a true fast spectrum design, albeit with plenty of fertile material similar to TMSR.

Now to find the speed of sound in molten salt...

Quote:
The dispersion of propagating collective excitations in molten LiF at 1287 K obtained in
the eight-variable model A(8) (spline-interpolated solid curves). Symbols show results obtained for separate sets of three dynamical variables. The dashed line shows the linear dispersion for sound excitations with c = 5320 m s-1

If the reactivity insertion is happening at the inlet and the inlet is 3.8m (REBUS core height) from the outlet with speed of sound of 5,320 m/s, the transit time for the pressure wave signalling the change in outlet flow would be 0.000714 s. So by my reckoning the pressure wave arrives at the core exit in 1/30th of the effective neutron life time. The implication being for the REBUS 3700 that the expansion of the core salt is more than fast enough to take effect and start countering any reactivity insertion before even one full neutron life time period has elapsed.

I find these figures very reassuring, especially given my recently developed interest in fast spectrum chloride reactors. But I believe that this result only holds for conditions that are short of prompt criticality. If prompt criticality is acheived then I don't think that salt expansion is fast enough and this is presumably where Jaro's post provides some additional information.
jaro wrote:
According to LA-13638,
Quote:
Analysis of data from KEWB and CRAC has led to relatively simple computer codes that follow the early transient behaviour well and rely on thermal expansion and the formation of radiolytic gas for the shutdown mechanisms.

In the KEWB systems, two quenching mechanisms seem to be dominant over a wide range of excursions.
The first of these is the rise in neutron temperature and thermal expansion as the core temperature rises, resulting in a prompt temperature coefficient equal to –2 ¢/°C at 30°C. This effect is sufficient to account for the observed yield of excursions starting near prompt criticality, but is inadequate for more violent transient experiments. The second quenching mechanism is bubble formation. The available evidence supports the contention that during the spike, void space, consisting of many very small bubbles (microbubbles) with internal pressures of from 10 to 1000 atmospheres, is created by the fission process. The bubbles later coalesce and leave the system, giving the observed gas production coefficient of about 4.4 l / MJ.
Growth of these microbubbles seems to involve the repeated interaction between fission fragments and existing microbubbles from earlier fissions. Thus, a quenching mechanism proportional to the square of the energy release can be invoked. This model is successful in describing the solution transients, notwithstanding imprecise knowledge of the manner in which the bubbles form and grow.


Top
 Profile  
 
PostPosted: Feb 24, 2011 3:34 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5060
According to Taube and Heer, a prompt criticality accident in a chloride reactor is very benign compared to solid fuel fast breeders. Figure 7.7:

http://moltensalt.org/references/static ... IR-411.pdf

Note the very high heat capacity of the molten fuel, about 100x higher than the heat capacity of the solid fuel. Combined with salt dilatation and low excess reactivity, Taube and Heer believe this makes the reactor safe. They do suggest a fused salt core catcher, just in case.


Last edited by Cyril R on Feb 24, 2011 9:27 am, edited 1 time in total.

Top
 Profile  
 
PostPosted: Feb 24, 2011 9:20 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Lars wrote:
For what kind of reactor is this?

Aqueous solution of uranyl sulphate (UO2SO4).
In an MSR the bubble quenching mechanism probably doesn't exist (no radiolysis of water).
Also, the neutron spectrum shift due to heating may not provide a significant quenching mechanism either -- as it does in a very thermal, homgeneous fuel-moderator mix.
Without these quenching mechanisms, fission pulse peak power and total energy release are likely to be higher -- though not necessarily damaging.


Top
 Profile  
 
PostPosted: Feb 24, 2011 10:08 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5060
Isn't the thermal expansion of hot molten chloride salt stronger than that of hot water? Why would you have to control a fast chloride reactor every femtosecond when you have such huge thermal capacity? Jaro, could you build a fast reactor like your HW-MSR (without the heavy water in the calandria?)


Top
 Profile  
 
PostPosted: Feb 24, 2011 1:45 pm 
Offline
User avatar

Joined: Aug 21, 2008 12:57 pm
Posts: 1058
It seems to me that a thermal expansion process will very much trail the near instantaneous prompt neutron emission and propagation of fast neutrons.


Nuclear waste, especially those that have undergone high burnup, can no longer produce an optimal range of precursor elements needed for delayed neutron production. Operational reactor control is strongly influenced by these fission fragments. Without enough delayed and/or external neutrons, the reactor must run too close to the edge of criticality to keep going on its own.


In a chloride reactor, without anything to slow the almost instantaneous propagation of fast neutrons down, a small transient event can cause and almost instantaneous power spike of say of 10e43. No finite amount of thermal mass will absorb that kind of power.


You need something that operates at nuclear speeds to slow down the propagation of these fast neutrons.


Nuclear waste burners are always subcritical so that the reactors do not approach the conditions that can cause a large power spike.


The reaction of thermal expansion no matter how fast is not on the same time level as the speed of light or a goodly faction thereof.

_________________
The old Zenith slogan: The quality goes in before the name goes on.


Top
 Profile  
 
PostPosted: Feb 24, 2011 2:39 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Cyril R wrote:
According to Taube and Heer, a prompt criticality accident in a chloride reactor is very benign compared to solid fuel fast breeders. Figure 7.7:

http://moltensalt.org/references/static ... IR-411.pdf

Note the very high heat capacity of the molten fuel, about 100x higher than the heat capacity of the solid fuel. Combined with salt dilatation and low excess reactivity, Taube and Heer believe this makes the reactor safe. They do suggest a fused salt core catcher, just in case.

Thanks Cyril -- just read the paper -- very nice !
Cyril R wrote:
Jaro, could you build a fast reactor like your HW-MSR (without the heavy water in the calandria?)

Note that one important feature of the Taube chloride fast reactor design is the spherical core: this is the most reactive configuration for a fast reactor, so re-criticality after any sort of accident is unlikely.
Not sure why you would want to build a fuel-channel-style fast reactor, but it seems to me that coming up with a practical design that would be close to the most reactive configuration possible, would be difficult (any suggestions?).


Top
 Profile  
 
PostPosted: Feb 24, 2011 2:55 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5060
Well in design I'm thinking of, the fuel tubes would be the only place in the hot cell without massive amounts of boron-10

:lol:


Top
 Profile  
 
PostPosted: Feb 24, 2011 3:38 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3069
Axil wrote:
It seems to me that a thermal expansion process will very much trail the near instantaneous prompt neutron emission and propagation of fast neutrons.

Neutron generation time is quite different for different reactor designs.
For thermal reactors there is no problem with simple salt expansion.
For fast reactors, I suspect that simple salt expansion is too slow to control a prompt neutron excursion. Assuming that the fuel contains fertile and neutron spectrum extends down into the resonsance interval (very likely) then doppler is sufficient for control. Doppler is certainly fast enough - being roughly the speed of the fission fragments hitting atoms around them - while neutrons must typically encounter many collisions before they are absorbed and being much smaller than fission fragments they travel further for each collision.

Quote:
Nuclear waste, especially those that have undergone high burnup, can no longer produce an optimal range of precursor elements needed for delayed neutron production. Operational reactor control is strongly influenced by these fission fragments. Without enough delayed and/or external neutrons, the reactor must run too close to the edge of criticality to keep going on its own.

It doesn't matter where the fissile comes from. U235 has around 2x the delayed neutrons of u233 or PU239. So I guess this statement refers the the lower delayed neutron production of Pu239. But it can still be sufficient - MSRE ran on u233 which is pretty similar to pu239 as far as delayed neutrons go. Absolutely, one can build reactors to keep going on their own with either Pu239 or U233- it is just a bit harder.



Quote:
In a chloride reactor, without anything to slow the almost instantaneous propagation of fast neutrons down, a small transient event can cause and almost instantaneous power spike of say of 10e43. No finite amount of thermal mass will absorb that kind of power.

Partially true. You need a control mechanism that is fast enough to contain a spike quickly. If you are too slow more heat gets dumped into the thermal mass. If the thermal mass is higher you can afford to be just a tad bit slower. But the first control knob is the speed of reducing reactivity with higher temperature.

Quote:
You need something that operates at nuclear speeds to slow down the propagation of these fast neutrons. Nuclear waste burners are always subcritical so that the reactors do not approach the conditions that can cause a large power spike.

Actually, so far as I know ALL built nuclear reactors are critical. Has anyone ever built an ADS power reactor?

Quote:
The reaction of thermal expansion no matter how fast is not on the same time level as the speed of light or a goodly faction thereof.

Actually, the doppler is. The fission products get ejected from a fission at a decent fraction of the speed of light. They don't have far to travel before colliding with an atom. But I'll agree this much - if there is no fertile (so no doppler) and the reactor is a fast reactor (so shorter neutron generation time) then I am not familiar with how they control such a beast. From the previous posts I get to do some reading and learn some more.


Top
 Profile  
 
PostPosted: Feb 24, 2011 8:29 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
This is all good stuff, but we should probably start a new thread discussing MSR response to prompt criticality events, but for now here is some more.

Quote:
In a chloride reactor, without anything to slow the almost instantaneous propagation of fast neutrons down, a small transient event can cause and almost instantaneous power spike of say of 10e43. No finite amount of thermal mass will absorb that kind of power.

Well here's a couple more things to add to the pot. In the Rebus 3700 paper they note a Doppler reactivity coefficient of -0.5pcm/C, they pretty much discount that as the thermal expansion coefficient is so much larger at -6pcm/C, but surely if you are talking about prompt criticality and a massive power spike then that -0.5pcm/C is going to be mighty handy as it acts on the same time scale as prompt neutrons (apparently).

In addition, any time that there is a massive power spike, other things must be happening at the same time gaseous FP's must be heating and expanding invoking some void coefficient, salt must be starting to expand albeit so much slower than the speed of the prompt neutrons, perhaps the salt is locally now so hot that it is starting to boil dismantling the localised promptly critical geometry of a cold fuel rich slug (or similar), the thing that created the prompt criticality event in the first place.

In the form of a question if prompt criticality is so fast and so powerful how on God's good green earth can you in a molten salt reactor assemble a promptly critical fuel slug quickly enough to maintain prompt criticality, let alone and entire core that is promptly critical?

So pulling numbers from various references how might the engineering numbers look?
In EIR-411 (thank you Cyril R) they step through a prompt criticality event one of the values used is peak power at 1500 x nominal power. I had previously calculated incorrectly that the pressure wave travel time was 0.00074 s, actually it's more like 0.00063 s (3.25m high vs 3.8m).

So we have a 1500x power spike for some time around 0.00063 s. Using Rebus salt numbers I get a temperature rise in that time of just 29.1C, which sounds too small, but there it is. Stepping back a bit further, for capacitance resistance system it takes about five time constants to see most of the result of a step change, so if we wait five time constants at 1500x power to see the change in fuel flow, where does that leave us. After five time periods (wave travelling time x 5) at 1500x power the salt temperature rise is 145.4C, a remarkably low figure. Meanwhile for a 145C temperature rise we could reasonably expect -73 pcm from Doppler effect which might be just enough to undo the prompt criticality locally. Certainly a larger increase in salt temperature of say 300 - 400C would produce -150 to -200 pcm on Doppler alone, to be very quickly followed by -1800 to -2400 pcm of negative reactivity based on thermal expansion.

To my eye that looks like a pretty amazing result, I just hope that I haven't dropped a factor of a 1000 somewhere, that would be embarrassing, but not the first time.

Aside from EIR-411 does anyone have any other references to prompt criticality events in molten salt reactors? It looks like a fascinating area.


Top
 Profile  
 
PostPosted: Feb 24, 2011 9:05 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3069
The French also did analysis of prompt criticality response times and conclusions are pretty similar. You have to postulate some pretty ridiculous reactivity insertions to get a few hundred degrees C salt temperature change. An insertion of 1000pcm extra reactivity causes a 200C rise (-5 pcm/C). Inserting this much reactivity in 10mSec could only cause an 100C overshoot. Hopefully the second Doctorate thesis uploads, pages 56-65 give the details. It is in French so google translate is in order for me (high school French isn't strong enough).

http://hal.in2p3.fr/tel-00354937/ pdf is too big to include.

I have absolutely no concern for thermal LFTRs.
A fast LFTR with fertile in the fuel salt is fine.
For fastish LFTRs, a pure two fluid machine would need to be more carefully analyzed. It is likely OK but not obvious to me.

For fast chloride reactor with no fertile in the fuel salt the concern grows but other mechanisms could help. The prompt neutron lifetime from Rebus seems too long compared to other fast reactors - so I wonder if I misunderstood something there.


Top
 Profile  
 
PostPosted: Feb 24, 2011 9:25 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
Lars wrote:
The prompt neutron lifetime from Rebus seems too long compared to other fast reactors - so I wonder if I misunderstood something there.

That puzzled me for quite some time also, I had one source saying chloride reactor prompt neutron lifetime 10-7 to 10-5 s and then the Bokov reference of 2.24 x10-2 s for the "mean (effective) neutron life time".

Given that all reactors are supposed to be subcritical WRT prompt neutrons and delayed supercritical WRT delayed neutrons, my conclusion is that the figure quoted by Bokov was to those neutrons relevant to the regulation of the reactor, ie delayed neutrons.

That's my theory as Bokov's prompt neutrons can't be different to everyone else's. And he never discusses "prompt neutron lifetime"


Top
 Profile  
 
PostPosted: Feb 24, 2011 9:30 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3069
Makes sense.

Then the insertion of a huge amount of reactivity like 1000 pcm would not be covered by his analysis and could still be a serious problem (if one came up with a mechanism to insert that much).


Top
 Profile  
 
PostPosted: Feb 24, 2011 9:38 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
I don't think so, but don't know enough to say for sure, factors in favour of it being ok are:
- The French TMSR spectrum is arguably not that far off a true fast spectrum design;
- Like the French design the Rebus has has plenty of fertile material in the core which should support the fast acting Doppler effect; and
- The physical barriers to inserting that much reactivity instantaneously, how could it be done, I can't think of anything as quick as 1/100th of a second.


Top
 Profile  
 
PostPosted: Feb 25, 2011 11:24 pm 
Offline

Joined: Jan 16, 2011 5:06 pm
Posts: 9
Ida-Russkie wrote:
How much is in spent LWR navy fuel? Don't they start out with a higher enriched fuel?


I suggested to the DOE chief scientist and the BRC.gov commission that naval spent fuel be burned in a land based or perhaps a ship based heavy water reactors :)


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 76 posts ]  Go to page Previous  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 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:  
Powered by phpBB® Forum Software © phpBB Group