Energy From Thorium Discussion Forum

It is currently Jan 18, 2018 12:57 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 48 posts ]  Go to page Previous  1, 2, 3, 4  Next
Author Message
PostPosted: Mar 16, 2012 5:39 am 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
Lars wrote:
One could say the same by putting the plutonium directly into a new LFTR and skipping the chloride machine to convert it to 233U.

You could do that, but there are solubility challenges PuF3 in FLiBe and at Hastelloy friendly temperatures implying that you would need a lot more cores to destroy the same quantity of Pu. If people are looking for a quick fix with the smallest up front investment a small number of chloride cores might do that job quite nicely. The key difference being that higher initial loading, soaking up a higher proportion of the stockpile on Day 1.

The rational part of me says fissile is a resource and Pu should be rightfully recognised as valuable fissile in the right kind of reactor. The nuclear salesman says if they want Pu destruction, then give it to them and we don't need Pu, so why not burn it off, especially if you can effectively swap it for a superior fuel.


Top
 Profile  
 
PostPosted: Mar 16, 2012 8:29 am 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3069
I'd think such a machine would need to be in a secure location. That limits the quantity of such machines. Also, I'm thinking when something unusual happens at such a plant there will be endless speculation about possibilities of moderation and super-criticality and the results thereof. I'd rather avoid such press by having just enough fast machines. So I'd put the plutonium first into thermal machines and chew up most of the plutonium them as the fuel degrades eventually put it the remainder into a few fast machines. I'm guessing we can't simultaneously have no plutonium extraction at the thermal production plants, secure sites to host the plutonium extraction technology, relatively few secure sites, and an ability to chew up all the LWR Pu in a reasonable time frame.


Top
 Profile  
 
PostPosted: Mar 16, 2012 8:39 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5057
Lindsay wrote:
Lars wrote:
One could say the same by putting the plutonium directly into a new LFTR and skipping the chloride machine to convert it to 233U.

You could do that, but there are solubility challenges PuF3 in FLiBe and at Hastelloy friendly temperatures implying that you would need a lot more cores to destroy the same quantity of Pu. If people are looking for a quick fix with the smallest up front investment a small number of chloride cores might do that job quite nicely. The key difference being that higher initial loading, soaking up a higher proportion of the stockpile on Day 1.

The rational part of me says fissile is a resource and Pu should be rightfully recognised as valuable fissile in the right kind of reactor. The nuclear salesman says if they want Pu destruction, then give it to them and we don't need Pu, so why not burn it off, especially if you can effectively swap it for a superior fuel.


If Pu is considered a resource then it costs money and you want to use it as efficiently as possible. Which means as little Pu per kWh as you can get away with. The market will favor low startup costs, and chlorides with their greater fissile startup requirement, would not do well with 10 tonnes of Pu to buy before you can get started. Using Pu to startup isobreeding LFTRs is probably the most efficient way to go. A combination of chloride reactors and fluoride reactors is potentially better but that is not how markets or investors work (nor the technological reality we have at the moment, with fluoride reactors being proven and chloride reactors being only paper reactors).

I agree with Lars. Fluoride reactors can startup on Pu good enough. The French work supplies more quantitative, detailed proof. Some of their reactor versions have really fast spectra (no graphite in the core, lots of Pu startup). This gives good neutron yield, and also allows very little processing, that makes the issue of Pu processing easier.


Top
 Profile  
 
PostPosted: Mar 16, 2012 7:29 pm 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
>If Pu is considered a resource then it costs money and you want to use it as efficiently as possible.

Thanks guys, I agree, but the reality is a more complex, for some Pu is a problematic by-product of LWR operations and nuclear weapons. Look at the extreme expense and problems associated with MOx fuel cycles for LWR's, anything has to be better than that, the worst MSR solution has to many times better than the MOx fuel cycle.

Cyril has identified quite correctly the central question, is Pu considered a resource or something else, a dangerous waste material that needs destruction? Is it both? Depending on who you talk to you will get a range of different answers, given that diversity of view, there seems to be a market niche for a dedicated Pu/TRU burner, if it can simultaneously breed more useful and proliferation resistant denatured U233, then that sounds appealing to me.

Applying the Pu to provide startup fissile for a thorium based MSR's is fine and arguably that's a more rational and efficient response to the opportunity presented by existing Pu stockpiles making that fissile resource stretch further. For some that means not dealing with the problem of Pu quickly enough, it all comes back to your view on Pu in the first place.

One thing that I have learned as I come to understand various nuclear programmes better over time is that the choices that get made around adopting or rejecting specific design options often have nothing to do with what is technically the most rational solution. For confirmation of that we can simply look at the decisions made to shut down the ORNL MS development programme in spite of its obvious promise.

My interest is not to convince anyone that the dedicated Pu burner/U233 breeder as the best option, but I maintain that it is a potentially useful option as part of a bigger nuclear system. Beyond that it may fulfil a specific need, a market niche or perceived need.

A couple of quick responses, the moderator insertion risk is real for any fuel rich fast spectrum system, I'm not at all sure that it a safety issue of huge public interest. The safety cases have to adequately address this risk along with all the others.

Chloride vs Fluoride I don't think that we know enough to say definitively, chloride bad, fluoride good or vice versa. The ORNL paper exploring fast MS reactor options, showed that chlorides were faster and more effective than fluorides as a fast spectrum design, but fast fluorides can do a lot and seem plenty fast enough for very effective TRU burning, so maybe we don't 'need' chloride systems. Similarly chloride systems have a bad rep for potential corrosion issues, but in terms of 'actual' corrosion data that I've seen, chlorides seem quite ok, but those aren't full systems with FP's and junk. Personally I would love to see more work on chlorides and structural materials to help better understand what's possible and what's not.


Top
 Profile  
 
PostPosted: Mar 16, 2012 8:13 pm 
Offline

Joined: Apr 19, 2008 1:06 am
Posts: 2230
The good points about chlorides are:-
1. A harder spectrum. U238 will have considerable fast fission. No moderator is required making for a more compact core.
2. Chlorides are more volatile. Thorium and uranium chlorides can be distilled below 1000C. It will be a negative factor with iron/steel containing structures.
Another negative is neutronic actions of Cl35 isotope. Cl37, nearly a quarter of natural chlorine, may have to be separated and used.


Top
 Profile  
 
PostPosted: Mar 16, 2012 9:09 pm 
Offline

Joined: Jul 28, 2008 10:44 pm
Posts: 3069
Lindsay wrote:
A couple of quick responses, the moderator insertion risk is real for any fuel rich fast spectrum system, I'm not at all sure that it a safety issue of huge public interest. The safety cases have to adequately address this risk along with all the others.



My concern is not real safety. It is the damage done when there is any problem at the plant and anti's get going talking about worst case scenarios. They don't need to explain how a moderator might get there. They can just expound on how many times a Hiroshima there is inside that reactor, and that yes physically it could explode if somehow a moderator found its way in. Makes for headlines and eyeballs on the screen so the press will keep such speculation in the news cycle until things return to normal. And then revisit the topic annually as the day we almost blew up Manhatten or some such.

As such, I am coming to the conclusion we need to have the production reactors as thermal reactors and if graphite waste is a worry then we need to figure a way to recycle it. We can still have few fast reactors to clean up the even TRUs but these should be 1-2% of the total capacity and come into play only after the thermal reactors have had a go at burning down the LWR TRUs.


Top
 Profile  
 
PostPosted: Mar 16, 2012 10:09 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
jagdish wrote:
No moderator is required making for a more compact core.

Yes, very compact.

Too compact, in fact.

Fast reactors can be made VERY small - usually because of the high material density, which minimizes neutron leakage, therefore requiring small fissile startup loading.

This should be true for chloride molten salt fast reactors (LCFRs) as well, though to a slightly lesser extent, as the density of chloride salts is less than metal (possibly comparable to oxides, but I haven't checked recently)
But that's only for a "minimum reactor" which, while capable of generating a large amount of power, would be far too small to allow all that power to be transferred out of the core, while keeping temperature at a reasonable level.

So my take on this is that a practical fast reactor for power generation is far larger than that "minimum reactor", for heat transfer reasons.
Unfortunately, that large fast reactor has a similarly large fissile startup loading - many times that of the "minimum reactor".
A reactor with so much extra fissile load may then have operational stability issues that do not exist for the minimal configuration.

I believe that this problem is managed in solid fuel fast reactors by the high precision fuel rod lattice layout, and the rigidity of the assembly at all scales (from fuel pellet to rod assembly to entire core). This prevents even minute changes in density from traveling across the core by way of pressure disturbances, that may be initiated by uneven temperatures or flow turbulence.

By contrast, an LCFR would lack such constraints against density wave propagation, which could easily develop into a destructive "ringing" (oscillation) within the reactor vessel.
At least that might be the scenario for the typical LCFR concept, with a large vessel filled with nothing but molten salt.
The same would apply for a large graphite-free LFTR, except that the LCFR version would likely be more dangerous, due to the shorter neutron lifetime and resulting fast power excursions.

If the LCFR concept were changed so as to run the fuel through hundreds of rigid tubes - similar to the fuel channels in a graphite LFTR, or the rods in a solid fuel fast reactor - then the stability issue might be resolved.
The tubes would have to contact lengthwise to prevent flow between them, or else the spaces between tubes would have to be voided, with a complex manifold connecting the tube inlet/outlets.
Mechanically, either option could be a major headache, both for fabrication and for operational endurance, as the differential thermal expansion of the tubes would put a lot of shear and/or bending stress on the joints - they probably wouldn't last very long, unless some very clever arrangement were devised - likely leading to high costs, particularly when special alloys are used, for corrosion resistance in molten salt....

Has anybody seen any sort of practical LCFR concept ? ...I sure haven't, and until I do, I very much doubt that one is possible.


Top
 Profile  
 
PostPosted: Mar 17, 2012 5:02 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5057
Quote:
Cyril has identified quite correctly the central question, is Pu considered a resource or something else, a dangerous waste material that needs destruction? Is it both? Depending on who you talk to you will get a range of different answers, given that diversity of view, there seems to be a market niche for a dedicated Pu/TRU burner, if it can simultaneously breed more useful and proliferation resistant denatured U233, then that sounds appealing to me.


Right. The issue, I think, is that "the market" isn't willing to buy expensive reactors with expensive fuel cycles (eg, needing continuous feed of TRUs to operate). If the investors are not on board, the show is off.

True enough, from a system fleet perspective, having a mix of fast and slower reactors is more optimal, since the fast reactors could service the thermal reactors with fissile startup charges, and eat their thermal spectrum waste TRUs. But that is not how investment works. The investment will be made on a single plant scale. As Lars has argued, producing excess fissile isn't worth that many $$$. It would be much more interesting to optimize other cost components of the plant, such as fissile inventory, processing equipment costs, etc, rather than going for a higher breeding reactor.

So my perspective is that we ultimately need a "Daisy" reactor, one that can do it all. TRU started LFTRs seem like one option. They don't use too much of the expensive to extract TRUs to start up, but they can breakeven on breeding and have a sustainable fuel cycle. But I also think we should get started on burner MSRs such as the DMSR - even if we have to run it without thorium and TRUs at first. There's a lot of engineering work to be sorted out in the pumps, heat exchangers, passive decay heat removal systems and such, which the first build could focus on. This way we can figure out later how to process Pu online (which is vital for Pu started reactors, obviously).


Top
 Profile  
 
PostPosted: Mar 17, 2012 5:38 am 
Offline
User avatar

Joined: May 24, 2009 4:42 am
Posts: 826
Location: Calgary, Alberta
jaro wrote:
Has anybody seen any sort of practical LCFR concept ? ...I sure haven't, and until I do, I very much doubt that one is possible.
I recommend that you read the MCFR papers in the document repository, including Eric Ottwitte's thesis and EIR-332. The most recent thing that I've read is on the REBUS-3700 which you should be able to find attached to a discussion thread post somewhere.

The REBUS concept runs at 100 kW/L, 3.8 m dia, 3.25 m high, 3686 MWt, 4.9t fissile/GWt, fuel salt is 55% NaCl,38% UCl3 and 7% PuCl3

I am intrigued by the following concern..

>By contrast, an LCFR would lack such constraints against density wave propagation, which could easily develop into a destructive "ringing" (oscillation) within the reactor vessel.

I assume that we are talking about density changes driving local power increases, driving expansion, then cooling increasing density in a repeating cycle very much like a mechanical resonance. That's a really good question I think that you are the only person who has raised it.

I think that there are a couple of things working against that effect, prompt neutrons are extremely fast, salt expansion is fast, but salt contraction (by cooling) is very slow by comparison, so I don't see how the natural frequency of one is likely to line up with the other to provide the mutual reinforcement at the same frequency that you normally see in resonant systems. Secondly fast fluoride reactors are much the same as fast chloride reactors, but not quite as quick. The French work showing the addition of 1000 pcm of reactivity (enough to achieve/surpass prompt criticality) into a small very high powered core (TMSR-NM) showed an extremely stable response to the reactivity insertion regardless of how quickly it was inserted. Note how the rate of rise is hundredths of second if the insertion is fast enough, but the rate of decline takes about 10 seconds to return to the steady-state value. The up phase and the down phase have quite different natural frequencies.

It's a great question and I look forward to knowing the answer at some point and we do know that LMFBR's can be subject to power fluctuations and in-core vibrations as experienced by the Super-Phenix, so it is definitely a bona fide issue. I'd like to think that the computational tools exist to answer that question before anyone attempts to build a MCFR.
Attachment:
1000 pcm Reactivity Insertion TMSR-NM.jpg
1000 pcm Reactivity Insertion TMSR-NM.jpg [ 400.16 KiB | Viewed 1609 times ]
Extracted from E Merle-Lucotte's Thesis, LPSC-0875


Top
 Profile  
 
PostPosted: Mar 17, 2012 10:42 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Thanks Lindsay.
Lindsay wrote:
I recommend that you read the MCFR papers in the document repository, including Eric Ottwitte's thesis and EIR-332. The most recent thing that I've read is on the REBUS-3700 which you should be able to find attached to a discussion thread post somewhere.
Yes, I have read Ottewitte's thesis and some reports about REBUS.

Lindsay wrote:
I think that there are a couple of things working against that effect, prompt neutrons are extremely fast, salt expansion is fast, but salt contraction (by cooling) is very slow by comparison, so I don't see how the natural frequency of one is likely to line up with the other to provide the mutual reinforcement at the same frequency that you normally see in resonant systems.
I think that your point about the slowness of the cooling effect would be valid in a static system, without forced (or otherwise) fuel circulation.
In an MCFR, the fuel would be circulated as fast as practically feasible, in order to minimise the size of the reactor core & fuel load.
So the density waves propagating across the core would probably always see "fresh fuel" from the HX return line inlet.

Lindsay wrote:
Secondly fast fluoride reactors are much the same as fast chloride reactors, but not quite as quick. The French work showing the addition of 1000 pcm of reactivity (enough to achieve/surpass prompt criticality) into a small very high powered core (TMSR-NM) showed an extremely stable response to the reactivity insertion regardless of how quickly it was inserted. Note how the rate of rise is hundredths of second if the insertion is fast enough, but the rate of decline takes about 10 seconds to return to the steady-state value. The up phase and the down phase have quite different natural frequencies.
Again, I believe that this applies to a static system, without fuel circulation being taken into account.
However, I also think that in the case of "a small very high powered core", the likelyhood of power oscillations should be reduced - as stated in my previous post.

Lindsay wrote:
I'd like to think that the computational tools exist to answer that question before anyone attempts to build a MCFR.
I'm pretty sure that the computational tools do NOT exist.
That's why I believe that theoretical studies such as those of Ottwitte's thesis need to be taken with a great big pinch of salt, so to speak.....


Top
 Profile  
 
PostPosted: Mar 17, 2012 12:09 pm 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5057
Jaro, there is a question I've been meaning to ask you. You've voiced a concern over power oscillations. However, from my limited understanding of the neutronics stuff, heterogeneous reactors like your HW-MSR would likely suffer from oscillations as well. Essentially you have two different neutron "worlds" which have different time constants and reactivity values for their own neutron "populations". By contrast, a perfectly homogeneous reactor has very tight neutron coupling, resulting in a predictable reactivity performance.

Is my understanding incorrect?


Top
 Profile  
 
PostPosted: Mar 17, 2012 12:23 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Cyril R wrote:
....heterogeneous reactors like your HW-MSR would likely suffer from oscillations as well. Essentially you have two different neutron "worlds" which have different time constants and reactivity values for their own neutron "populations". By contrast, a perfectly homogeneous reactor has very tight neutron coupling, resulting in a predictable reactivity performance.

Is my understanding incorrect?
Yes, I think your understanding is correct in the general sense.

However, you have to keep in mind that the fast neutron population is too small to have gross effects on reactivity.
As I keep reiterating, the heterogeneous aspect of an HW-MSR is to boost the conversion ratio (fissile breeding). The reactivity is still dominated by the much larger thermal neutron population.


Top
 Profile  
 
PostPosted: Mar 17, 2012 3:31 pm 
Offline

Joined: Dec 05, 2008 8:50 am
Posts: 337
jaro wrote:

I'm pretty sure that the computational tools do NOT exist.
That's why I believe that theoretical studies such as those of Ottwitte's thesis need to be taken with a great big pinch of salt, so to speak.....


Have you ever tried to contact in some way Ottewitte in order that he joined the forum ? I think his point of view about fast chlorides may be very usefull and interesting


Top
 Profile  
 
PostPosted: Mar 17, 2012 8:39 pm 
Offline

Joined: Apr 19, 2008 1:06 am
Posts: 2230
If the LCFR becomes too compact for heat transfer, you could dilute the fuel with NaCl-MgCl2 to the required extent. you could also put it in CANDU-like tubes and transfer heat with a clean salt keeping the fuel and coolant separate , or to put it more simply, put the fuel in the fire tubes of a boiler configuration. Each tube can be sub-critical and the assembly could be critical.
To retain the MSR advantages of flushing out Xe and the overflow on over-heating, keep the number of tubes small, in 3-7 range.
Comparing it with HW-MSR, coolant substitutes for moderator.


Top
 Profile  
 
PostPosted: Mar 17, 2012 9:28 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
jagdish wrote:
Comparing it with HW-MSR, coolant substitutes for moderator.

Yes, that MIGHT work, with some very special coolant.

The problem arises when we ask what happens if the coolant is lost through a leak : is the reactivity going to increase ?

This is actually a very pertinent question, since the suggested configuration is not unlike that of standard fast reactors : The concern has always been precisely this event where the coolant is lost, and the reactivity goes way up, because a neutron absorbing mass is lost.
This is quite different from the HW-MSR case, where moderator loss absolutely kills reactivity (in fact, moderator dump has been used as an emergency shutdown mechanism, in some of the early Candu reactors... it was abandoned in later designs, because the regulatory authority considered it too slow-acting, to be of much use....)


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 48 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:  
cron
Powered by phpBB® Forum Software © phpBB Group