Energy From Thorium Discussion Forum

It is currently Oct 20, 2018 6:35 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 67 posts ]  Go to page 1, 2, 3, 4, 5  Next
Author Message
 Post subject: One-Fluid Distillation
PostPosted: Feb 05, 2008 3:28 pm 
Offline

Joined: Jan 21, 2008 9:12 pm
Posts: 317
Location: idaho falls
Some of the reservations about Weinberg’s “Chemist’s Reactor” (one-fluid, thorium-fueled, molten salt breeder reactor) seem to based upon the uncertainties about salt clean-up (“reprocessing”) expressed in ORNL’s early technical reports. From the perspective of a chemist who has had a chance to look at the totality of those reports, these uncertainties simply seem to an early example of the “conservative” waffling that often renders DOE technically incompetent today.

Basically all that the reactor’s reprocessing system would have to accomplish is to, a) recover sufficient 233U to keep it at equilibrium fuel-wise, b) retain most of the thorium and FLiBe, & c) remove sufficient FP “junk” to keep it from getting constipated. (For example, it’s necessary to remove REs because some of them have huge neutron absorption cross sections and not necessary to remove 137Cs because it doesn’t absorb neutrons and can’t accumulate to the point that it will affect bulk salt characteristics). Grime’s paper, “Molten Salt Chemistry”, (“Nuclear Applications & Technology”, Vol 8, p 137-, Feb. 1970) summarizes the state of the art in 1969. Basically, it concludes that the uranium recovery part of the system had already been pretty well worked out & that the major uncertainty remaining about junk-removal had to do with the REs. A likely reason for the latter conclusion is that a good deal of the research then being done on RE removal was devoted to trying to make the molten bismuth liquid-liquid extraction system originally developed to isolate 233Pa serve double-duty as the RE remover too – research that wasn’t going too well. (It never did pan out because the reduction potentials of the REs are almost identical to that of Th - a fact already well established by that time.) However, other ORNL studies had already generated more than enough data to safely conclude that the REs could be separated via a simple “boil off” of the bulk salts (Li/Be/Th fluorides) – a conclusion that ORNL’s spokespersons seemed reluctant to state in so many words. Here’s the rationale for my contention.

First, Hightower & McNeese’s 1969 report (ORNL 4415) established both the viability of a distillation – based separation of FLiBe from ThF4 and also that a typical RE fluoride (LaF3) would be less than one percent (0.006)as volatile as ThF4.in a molten salt mix emulating what would be left in the “pot” after most of the FLiBe has boiled away. Second, Table 6.2 of their Jan. 1968 report (ORNL 2058) pointed out that the observed vapor pressures of several RE fluorides in molten LiF followed Raoult’s law (meaning that the vapor pressure of the mixture at any temperature equals the sum of the pure salt vapor pressures times the mole fractions).

Finally, if one simply assumes that ThF4 and REF3 behave similarly upon heating except for the fact that the latter has a roughly 600C higher boiling point (which assumption is not unreasonable to a chemist), their relative volatilities work out to 1:60. Regardless of exactly which set of relative volatilities is used, computerized-modeling of what will happen during a boil-down indicates that we could recover >95% of the Th (and virtually all of FLiBe) while discarding >95% of the RE plus virtually 100% of stuff like 133Ba or 90Sr. This conclusion isn’t just an educated guess - it’s a scientific certainty.

_________________
Darryl Siemer


Top
 Profile  
 
 Post subject:
PostPosted: Feb 05, 2008 4:47 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
Very interesting points. I am no chemist but have read much of ORNLs work on reprocessing. I will try to dig up one report in particular in which they discussed using vacuum distillation for a Single Fluid system. Their conclusion was that it was not really viable but if as you say (as have others) that ORNL tended to be quite conservative then maybe there is hope for practicality.

As vacuum distillation on Thorium free salts was looked to be around 1000 C at low pressure, would you have an estimate as to what temperature things would need to be raised to to boil off ThF4 while leaving behind the rare earths?

I should mention that in discussions with Mac Toth of ORNL, he expressed serious doubts in whether any vacuum distillation system could work as well as advertised. Mac is one of the best experts still active on these salt systems. He based this pessimism on the contention that during boiling there would be a great amount of association of rare earth fluorines to the carrier fluorides leaving as vapor. I think however that the actual test runs done at ORNL show that the system can work (without ThF4 at least). Even if only 80 or 90% of the rare earths are left behind instead of >95% I think the system can work effectively.

Others on this site, Ray in particular I believe, have speculated upon going to the higher boiling points to also boil off ThF4 using carbon based stills. I have my doubts as to how practical that would be, but that is only my opinion.

Another point to keep in mind is the very low cost and huge resource base of thorium. Even if a system is wasting some thorium, that should be acceptable. The LWR competition goes through 100 to 200 tonnes of natural uranium a year per gigawatt plant. We only need burn about a tonne so if we also need to waste 5, 10 or even more tonnes of thorium a year we still have a very sustainable and economic system. For example, even 20 tonnes of thorium is only 600K to 1.2 million dollars per year. An LWR spends around 50 million per year for all the uranium, enrichment and fabrication.

Please also remember that it is the proven applicability of the vacuum distillation method if thorium is not present in the salt that have led many of us to rediscovering the advantages of the 2 Fluid designs in which thorium is kept in a separate blanket salt.


Top
 Profile  
 
 Post subject:
PostPosted: Feb 05, 2008 5:11 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
Darryl,

A little bit of discussion on this site can be found in this Reprocessing section labeled "Vapor-phase waste reprocessing". I think it was discussed elsewhere as well.

I quickly found the article I was thinking of. It was actually in terms of using vacuum distillation on single fluid systems without removing UF4 by fluoride volatility first. This was for systems with extra proliferation resistance and the article is in Kirk`s repository under ORAU-IEA-77-13. They look at various other reprocessing schemes that you might find of interest. They mention a distillation temperature of 1200 C needed to effectively boil off ThF4. That does not sound very extreme to me. They do remind the reader though, that even the simple vacuum distillation runs with the MSRE salts had many engineering difficulties and this would have been around 1000 C.


Top
 Profile  
 
PostPosted: Feb 05, 2008 7:08 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1946
Location: Montreal
darryl siemer wrote:
if one simply assumes that ThF4 and REF3 behave similarly upon heating except for the fact that the latter has a roughly 600C higher boiling point (which assumption is not unreasonable to a chemist), their relative volatilities work out to 1:60. Regardless of exactly which set of relative volatilities is used, computerized-modeling of what will happen during a boil-down indicates that we could recover >95% of the Th (and virtually all of FLiBe) while discarding >95% of the RE plus virtually 100% of stuff like 133Ba or 90Sr. This conclusion isn’t just an educated guess - it’s a scientific certainty.

Sounds like REF3 is also a good model for PuF3 -- in which case we see why there might be problems in running a denatured fluoride reactor.....

.


Top
 Profile  
 
PostPosted: Feb 05, 2008 7:23 pm 
Offline

Joined: Jan 21, 2008 9:12 pm
Posts: 317
Location: idaho falls
Thanks for the reference, I hadn't read it yet.

It seems to me that we wouldn't have to go any hotter than about 1100 C to volatilize (not "boil") off the thorium fluoride - its vapor pressure at that temp is about the same as LiF's is at 1000C & that's apparently enough to process about 1.5 ft^3 of salt per ft^2 of still pot area. Since the reactor's roughly 1600 ft^3 of salt wouldn't have to be distilled more often than about once or twice per year, this translates to a fairly small/cheap still.

_________________
Darryl Siemer


Top
 Profile  
 
PostPosted: Feb 05, 2008 8:10 pm 
Offline

Joined: Jan 21, 2008 9:12 pm
Posts: 317
Location: idaho falls
This is to address David's question about some of the anecdotal reservations about distillation that seem to hanging around out there. I went back and took another look at ORNL 4415. On p 14 it lists what are purported to be to measured, not theoretical, relative volatilities of LaF3 & ThF4 relative to LiF (0.0015:0.24). This suggests that the "chemical effects" responsible for whatever's causing those reservations were probably really due to something like droplet carry-over (that's why the stuff in the pot shouldn't "boil"). There may indeed be some black magic effect that kicks in when 100% of the LiF is gone but I can't see a mechanism for it.

Has anybody seen a reference that gives relative volatilities in LiF-free melts?

A good separation is important because most of the "high level waste" produced by the reactor is actually going to be thorium not FP.

_________________
Darryl Siemer


Top
 Profile  
 
PostPosted: Feb 05, 2008 8:23 pm 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3824
Location: Alabama
darryl siemer wrote:
A good separation is important because most of the "high level waste" produced by the reactor is actually going to be thorium not FP.


...assuming that it's a one-fluid (LiF-BeF2-ThF4-UF4) reactor. If it's a two-fluid reactor (which I really favor) the distillation residues will be entirely thorium-free.

Protactinium removal in the blanket will keep the blanket fission-product-free.


Top
 Profile  
 
PostPosted: Feb 05, 2008 8:24 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
jaro wrote:
darryl siemer wrote:
if one simply assumes that ThF4 and REF3 behave similarly upon heating except for the fact that the latter has a roughly 600C higher boiling point (which assumption is not unreasonable to a chemist), their relative volatilities work out to 1:60. Regardless of exactly which set of relative volatilities is used, computerized-modeling of what will happen during a boil-down indicates that we could recover >95% of the Th (and virtually all of FLiBe) while discarding >95% of the RE plus virtually 100% of stuff like 133Ba or 90Sr. This conclusion isn’t just an educated guess - it’s a scientific certainty.

Sounds like REF3 is also a good model for PuF3 -- in which case we see why there might be problems in running a denatured fluoride reactor.....

.


Yes PuF3 and the other higher actinides will always be the tougher species to process. They are most likely impossible to separate from the rare earths using vacuum distillation. However, fluoride volatility is not out the question, just harder. If I recall correctly, for UF4 you only need a ratio of about 3 to 1 of fluorine gas to uranium in the salt. For plutonium the main difference was needing about 70 to 1. That high concentration of fluorine was extremely corrosive to deal with. ORNL looked at a few tricks to help such as solidifying the salt, making it into a fine powder and counter flowing it in fluorine gas.

Liquid Bismuth can work but I suppose our fall back solution these days is to look to using carbon based material (graphite, carbon composites) when we run up against a high temperature and/or or corrosive situation. I.e. for use in a fluoride volatility chamber.


Top
 Profile  
 
 Post subject:
PostPosted: Feb 06, 2008 1:53 am 
Offline

Joined: Jan 21, 2008 9:12 pm
Posts: 317
Location: idaho falls
Yes David, I don't see how we could separate Pu from REF3 by distillation unless we were willing to fluorinate the heck out of it. Of course, I would recommend that we never put enough 238U (or Pu itself) into the salt for that to actually become an issue.

Kirk, it's time to explain exactly how you would go about building a practical two-fluid, thermalized neutron, thorium breeder reactor. ORNL seemed to have reached the conclusion that it would be too damned hard to do if graphite's the moderator. Has INL come up with a fabulous new breakthrough in moderator material? I can't jump onto the two-fluid bandwagon unless I can see how it's going to be done.

_________________
Darryl Siemer


Top
 Profile  
 
 Post subject:
PostPosted: Feb 06, 2008 5:18 am 
Offline
User avatar

Joined: Nov 30, 2006 3:30 pm
Posts: 3824
Location: Alabama
darryl siemer wrote:
Kirk, it's time to explain exactly how you would go about building a practical two-fluid, thermalized neutron, thorium breeder reactor. ORNL seemed to have reached the conclusion that it would be too damned hard to do if graphite's the moderator. Has INL come up with a fabulous new breakthrough in moderator material? I can't jump onto the two-fluid bandwagon unless I can see how it's going to be done.


Well, this isn't the section to detail reactor designs, but a few things I can say about two-fluid vs. one-fluid designs.

1. I wouldn't even attempt the conventional one-fluid design based on the temperature coefficients alone. The temperature coefficients of the two-fluid reactor are strongly negative, but those of the one-fluid are weakly so, and recent French evidence points to the possibility that they might even be positive.

2. The reprocessing scheme of the two-fluid vs. the one-fluid is another strong incentive to stay this direction.

3. The graphite swelling and distortion seen that drove ORNL to shift from two-fluid to one-fluid was based on immersed graphite designs (like MSRE) where the centerline of the graphite was the hottest part of the core design. My analysis of the graphite performance in these reactors show just how strong the damage is dependent on temperature. This has led me to consider how graphite can be externally cooled in a manner that extends graphite lifetime considerably.

I've held off on detailed description of this concept until I had more detailed neutronic analysis to back it up, but my strong suspicion is that a series of graphite or SiC tubes in a lattice configuration, with the remainder of the volume filled up with a bed of graphite spheres, has a lot of promise. Some of the tubes would be core salt, some would be fertile blanket.


Top
 Profile  
 
 Post subject: Two fluid reactor
PostPosted: Feb 06, 2008 8:00 am 
Offline
User avatar

Joined: Jun 24, 2007 10:43 am
Posts: 251
Location: Dallas, TX
I believe that my father felt that protactinium extraction would be much easier in a two fluid reactor.


Top
 Profile  
 
 Post subject:
PostPosted: Feb 06, 2008 12:13 pm 
Offline

Joined: Mar 07, 2007 11:02 am
Posts: 911
Location: Ottawa
darryl siemer wrote:
Yes David, I don't see how we could separate Pu from REF3 by distillation unless we were willing to fluorinate the heck out of it. Of course, I would recommend that we never put enough 238U (or Pu itself) into the salt for that to actually become an issue.

Kirk, it's time to explain exactly how you would go about building a practical two-fluid, thermalized neutron, thorium breeder reactor. ORNL seemed to have reached the conclusion that it would be too damned hard to do if graphite's the moderator. Has INL come up with a fabulous new breakthrough in moderator material? I can't jump onto the two-fluid bandwagon unless I can see how it's going to be done.


Darryl,

If you haven`t already, I suggest you browse through the 2 Fluid concept I have proposed. The basis of the idea is that the fertile and fissile fluids do not need to be intermixed within the core as long as at least one dimension of the core is small (1 or 2 meters). ORNL was always thinking and modeling in terms of spherical or short cylinder geometry and in that case if the core only contains fissile salt (and/or graphite), then you can not obtain a useful amount of volume and have an acceptable fissile concentration at the same time. Thus they went the route of complex "plumbing" to intermix the fluids.

If however, you change geometry to an elongated cylinder or rectangular slab geometry you can have the increased volume needed and proper fissile concentration. Thus the core is about a 1 or 2 meter wide pipe of fissile salt surrounded by about 1 meter of blanket salt. The core is as long as needed, from just a few meters to 10 or more. The design works with or without graphite and I tend to favor without graphite lately. With graphite a simple thin barrier of Hastelloy N or carbon based material is needed between core and blanket. Without graphite the power density is much higher and the barrier now also acts as a simple pipe.

I have been a little secretive as of late as I am filing a patent quite soon which will have a mix of publicly disclosed content and some new variations of the concept. Once that is done I will try to post online a broader picture of the concept. For now you can check out the presentation I put online back in August (link below) or checkout the discussion under "The Modified Geometry 2 Fluid Reactor" topic within the "Reactor Design" section.

http://thoriumenergy.blogspot.com/2007/08/modified-geometry-2-fluid-molten-salt.html


Top
 Profile  
 
PostPosted: Feb 07, 2008 3:31 pm 
Offline

Joined: Jan 21, 2008 9:12 pm
Posts: 317
Location: idaho falls
For the sake of "openness" I thought that it might be useful to explain why I've just become a believer in the "two salt" concept. After reading David's POSTED description of his tube-in-shell configured reactor early yesterday morning, I did a ball park calculation that "proved" to me (a chemist) that it couldn't possibly work ; i.e., that he's claiming that it would be possible to run a 1 GWe, barely epithermal (?), power reactor with under 200 kg of total fissile stuff! To avoid "embarrassing" him, I then sent off a private note pointing this out and asking for more proof. He (and several other folks) have since provided that proof. David has given me permission to post his response to my note (see below).

"Yes, some of the numbers that come out do seem almost too good to be true but let me give you some more data that should convince you. The standard Single Fluid design needed 1500 kg of U233, but a great deal of this was essentially being wasted in the outer under-moderated zone of the design. For comparison the standard ORNL 2 Fluid design of the mid 60s only required about 700 to 800 kg. Here is a breakdown of salt in cubic feet from table 1 of ORNL 3996 page xi .

Fuel Salt (0.22% U233F4)

166 cuft in core
26 in blanket (for cooling I assume)
147 in plena (salt collection area above and below channels)
345 in piping, heat exchangers
33 in processing

Or about 20 cubic meters all together.

So you can see how only about a quarter of the U233 is actually in the core (or 200 kg). This was for an average power density of 470 kw/liter in the salt. So my shocking value of only 187 kg in the core is better to compare to the 200 kg of a standard ORNL 2 Fluid design.

I know it is surprising that a design without graphite can run with such low fissile concentration but I am going by ORNL`s work here. Their numbers could certainly be off as U233 resonance values where not very well known until the late 70s I believe. I do not think they will be very far off the mark though as recent French work has reproduced most early ORNL work without too much variation.

What you have to remember is that with only neutronically excellent salt and U233 in the core (no fertile), it can be critical with very little fissile concentration and small dimension. The salt will perform a significant amount of neutron moderation, especially if the fissile content is as low as 0.158% UF4. Once you have criticality, after that the power output simply comes from your volume of salt in the core and how high you want to push the power density. I`ve limited things to around 400 kw/litre in my studies which is below much of ORNL work.

As you point out, my simple calculations only give about 5 cubic meters of fuel salt. Since I do not need to waste salt in collection plenums and with the advent of compact heat exchangers I can indeed get a design down to very low fissile requirement even without graphite. French studies assume 20 cubic meters total which might be something I should look at to give better overall heat capacity. At 20 cubic meters of salt we are still only talking about 500 kg of U233 if the concentration is 0.158%

Again, if ORNL was using poor cross sectional data, we might need to raise the U233 concentration somewhat but I do not think by a large amount.

Anyhow, I hope that answers some of your initial questions. Feel free to contact me with any other questions and stay tuned for more details once I get a patent filed as I have taken the work in some new directions."

_________________
Darryl Siemer


Top
 Profile  
 
PostPosted: Nov 30, 2012 7:50 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5045
One thing that I wonder about in a single fluid vacuum distillation unit is the condenser temperature. ORNL wanted to keep the condenser at a temperature of 850 degrees Celsius, which is slightly above the freezing point of the still's highest freezing major component (LiF). Thus, the still product line and condenser face will never see meaningful amounts of solids (and associated problems).

If one wants to distill the thorium out - and this goes for UF4 as well - then at what temperature must the condenser be operated at? Specifically, is it ok to have some ThF4 and UF4 freeze onto the condenser face, if the carrier salt comes over that to transport it to the product line via liquid phase dissolution?


Top
 Profile  
 
PostPosted: Nov 30, 2012 11:26 am 
Offline

Joined: Jun 05, 2011 6:59 pm
Posts: 1335
Location: NoOPWA
If we use a reasonable fractional distillation unit, the condensates should never solidify, except maybe some precip in the still bottoms.

_________________
DRJ : Engineer - NAVSEA : (Retired)


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 67 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:  
cron
Powered by phpBB® Forum Software © phpBB Group