Energy From Thorium Discussion Forum

It is currently Jan 19, 2018 8:08 pm

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 99 posts ]  Go to page Previous  1 ... 3, 4, 5, 6, 7  Next
Author Message
PostPosted: Feb 20, 2010 7:40 pm 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 462
Location: Teesside, UK
Further to this discussion, here's a plot of eta (neutrons out/neutrons in) for 233U and 239Pu
Attachment:
Eta-plot.png
Eta-plot.png [ 52.91 KiB | Viewed 1670 times ]

and expanded view of the high-energy end
Attachment:
Eta_detail.png
Eta_detail.png [ 43.22 KiB | Viewed 1656 times ]

The really fast neutrons can push eta well above 3, but comparing with Jaro's spectrum plot, there just aren't that many neutrons at these energies, even in a chloride reactor. Most of the work in a chloride reactor is done by the 0.1 - 1 MeV neutrons, where there is high flux and reasonable eta of 2.3-3.0. Below 0.1 MeV - where the fluoride spectrum is much higher than the chloride - eta for 239Pu drops well below 2. The inelastic behaviour of fluorine above 0.1 MeV looks 'exactly wrong' for running a U/Pu cycle.


Top
 Profile  
 
PostPosted: Feb 20, 2010 9:47 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Luke wrote:
Most of the work in a chloride reactor is done by the 0.1 - 1 MeV neutrons, where there is high flux and reasonable eta of 2.3-3.0. Below 0.1 MeV - where the fluoride spectrum is much higher than the chloride - eta for 239Pu drops well below 2.

Yes, that's exactly true: Which is why we call a chloride reactor a fast reactor.

In a fast fast reactor, if we don't get "the work done by the 0.1 - 1 MeV neutrons", we are toast, because after that the neutrons drop down into the resonance absorption region of U238, and that's the end of it.

Luke wrote:
The inelastic behaviour of fluorine above 0.1 MeV looks 'exactly wrong' for running a U/Pu cycle.

That depends on the kind of reactor.
In a thermal reactor - including U/Pu type - the vast majority of fissions happens at very low neutron energy.
To get to that low energy, we need to get lots of neutrons to slow down from the "fast" region to the thermal one, without losing too many of them in the "intermediate" energy region.
The heterogeneous lattice design is excellent fot that, since most of the slowing down collisions occur in moderator free of U238, thus greatly reducing risk of absorption during the passage from fast to thermal.

If we have a bi-modal reactor with fat fuel channels, we still need to get lots of neutrons to slow to thermal energy without getting absorbed in the increased amount of fuel.
The inelastic scattering of fluorine helps us do that, by sharply reducing neutron dwell time in the intermediate energy range.
For sure this wouldn’t work in a fast reactor, because in that case criticality depends very much on the intermediate energy group.
By contrast, all we’re looking for in the bi-modal reactor case, is a bonus neutron yield from fast fissions – we’re not dependent on them for criticality.


Top
 Profile  
 
PostPosted: Feb 21, 2010 5:23 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Luke wrote:
Further to this discussion, here's a plot of eta (neutrons out/neutrons in) for 233U and 239Pu.

By the way -- what database/website did you use for these ?

.....the reason I ask is because I'm having difficulties getting eta plots from nndc.bnl.gov

In particular, they make it difficult to combine 'nu' with (n,f) & (n,g) x-sections, to get eta.

Also, data for some nuclides, like Pu241, is sparse.

But I did piece together this graphic, where the extra neutron input from U238 is compared to the rest......


Attachments:
Pu239_U233_U235_U238_eta_fast.gif
Pu239_U233_U235_U238_eta_fast.gif [ 37.58 KiB | Viewed 1678 times ]
Top
 Profile  
 
PostPosted: Feb 21, 2010 6:14 pm 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 462
Location: Teesside, UK
jaro wrote:
Luke wrote:
Further to this discussion, here's a plot of eta (neutrons out/neutrons in) for 233U and 239Pu.

By the way -- what database/website did you use for these ?

.....the reason I ask is because I'm having difficulties getting eta plots from nndc.bnl.gov ......
I couldn't find any decent data, and ended up with the same secondary source as last time we had this discussion

jaro wrote:
In a thermal reactor - including U/Pu type - the vast majority of fissions happens at very low neutron energy.
To get to that low energy, we need to get lots of neutrons to slow down from the "fast" region to the thermal one, without losing too many of them in the "intermediate" energy region.
The heterogeneous lattice design is excellent fot that, since most of the slowing down collisions occur in moderator free of U238, thus greatly reducing risk of absorption during the passage from fast to thermal.

If we have a bi-modal reactor with fat fuel channels, we still need to get lots of neutrons to slow to thermal energy without getting absorbed in the increased amount of fuel.
The inelastic scattering of fluorine helps us do that, by sharply reducing neutron dwell time in the intermediate energy range.
My concern was that by the time fluorine stops being inelastic, you could be down to ~20 keV and still inside the fuel channel with all the 238U, and the resonance region 1 eV - 10 keV still to go. What are the odds of a 20 keV neutron escaping the channel, against that of a 1 MeV neutron? And the odds of a 20 keV neutron getting thermalised before it re-enters a channel, vs a 1 MeV neutron.

Thanks for the education!


Top
 Profile  
 
PostPosted: Feb 21, 2010 7:00 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Luke wrote:
I couldn't find any decent data.....

Oh.... too bad....

Its weird, the way the BNL website works:

You can get 'nu' using MF = 1 and MT = 452, and all the x-sections using MF = 3 and MT = 18 & 102.

You can even get the fraction of fissions per neutron absorbed, using the calculation feature to combine outputs, ex. y0/(y1 + y0)

But then the stupid thing doesn't let you do combinations with MF1 & MF3 data, so you have to save the output from one of them, and then enter it when you run the other, as "new data points" -- after which you can apply a calculation again, ex. eta = y0 * y1

Curses !!!!! :evil:


Top
 Profile  
 
PostPosted: Feb 25, 2010 7:34 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
A while back, Lars asked a question :
Lars wrote:
Do you know how much fissile per GWe (or 100MWe if you like) you need to make the reactor run?

....to which I replied as follows :
jaro wrote:
The above conceptual layout requires about 400kg/GWe fissile load, assuming NU (or 180kg for the ~450 MWe assumed above).

PS. If we add Th to the mix, to benefit from U233 breeding, then a big buffer tank would be needed, as discussed previously, to give Pa233 enough time to decay without absorbing neutrons.
The 5m^3 guesstimated above could easily go to 50m^3 or more.
The fissile load would grow correspondingly.
But since we're talking about NU, the added fuel cost would be negligible, compared to the equipment itself.
There would also be some colateral benefit from Np239 buffering.

The buffer tank option could get very expensive, as fuel enrichment level is increased.
At some point, it becomes more economical to go to the two-fluid core/blanket system, with no fissile in the blanket (other than what is bred in-situ).
But this option has its own added costs, such as separate cooling/pumping system, processing, etc.

The rough figures mentioned here can be compared to those provided by David for the DMSR – which makes for an interesting comparison, since both DMSR and HW-MSR are intended to be interim to an eventual (we hope) deployment of pure U233-Th two-fluid LFTRs.

David wrote:
ORNL TM7207 shows all the relative th, U238, U235, U233 concentrations at start, 15 years and 30 years. Since this has already come up, here is the table.

Image


Ignoring for the moment the HW-MSR version with Th fertile fuel and a great big decay tank, the comparison then is between 180kg total fissile and 3450kg total fissile (or possibly half that figure for the DMSR, if that numbr is for a 1GWe size).

So we’re talking from 10 to 20 times the fissile load.

More importantly, in the case of the DMSR the fuel is LEU enriched to 19.7% U235, whereas the HW-MSR fuel is plain old NU : This has implications all the way from procurement & shipping, to compliance with criticality safety regs when handling significant amounts of fuel with such high enrichment.

The table lists the total actinide mass as 127 tonnes – which implies a total HM fluoride mass of about 160 tonnes, and a total fuel salt mass (including carrier salt) of probably about 20,000 tonnes (give-or-take a few thousand).

Besides that, the fuel salt is only a fraction of the core volume, the rest being graphite – possibly around 100,000 tonnes.

That’s an awful lot of radwaste to store at EOL – compared 32 fairly thin-wall graphite tubes, and a little pile of chopped-up bits of Zr caladria vessel....

If waste volume is an issue, the choice seems clear, IMO.


Top
 Profile  
 
PostPosted: Feb 26, 2010 7:31 am 
Offline

Joined: Jul 28, 2008 5:01 am
Posts: 462
Location: Teesside, UK
jaro wrote:
......The table lists the total actinide mass as 127 tonnes – which implies a total HM fluoride mass of about 160 tonnes, and a total fuel salt mass (including carrier salt) of probably about 20,000 tonnes (give-or-take a few thousand).

Besides that, the fuel salt is only a fraction of the core volume, the rest being graphite – possibly around 100,000 tonnes......
I don't like graphite waste, it's classed as high level waste under EU regs and is a pain to dispose of, but I think you've added a zero or two here, unless you're looking at waste for the whole reactor fleet. The paper gives the DMSR core as ~8.3m diameter (300 m^3), so <1000 Te graphite.


Top
 Profile  
 
PostPosted: Feb 26, 2010 10:03 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Thanks for pointing out the discrepancy with the reactor volume.

After checking the reference document (ORNL/TM-7207), I see that the DMSR has an unusually high fuel fraction in the carrier salt -- ten times higher than one might see in other designs.
So there's a factor of ten right there.
So the total core salt volume is quite small -- only about 84 m^3, or 11% of total core volume, 785 m^3.
Their graphite mass therefore comes to 1,290 tonnes (@ 1.84 Mg/m^3 dens.)


Top
 Profile  
 
PostPosted: Mar 01, 2010 6:35 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Incidentally, in the French TMSR spectrum......

Image

......the lithium in LiF will undoubtedly play an important moderating role -- quite apart from F19 inelastic collisions.

As noted in the thread on Travelling Wave Reactors (fast spectrum, sodium-cooled), lithium injection is used to kill the fast spectrum.

LiF will of course do likewise: We don't want any in the HW-MSR fuel channels, mixed inwith the UF4/UF3 salt.


Top
 Profile  
 
PostPosted: Mar 03, 2010 1:39 am 
Online

Joined: Apr 19, 2008 1:06 am
Posts: 2231
jaro wrote:
More importantly, in the case of the DMSR the fuel is LEU enriched to 19.7% U235, whereas the HW-MSR fuel is plain old NU : This has implications all the way from procurement & shipping, to compliance with criticality safety regs when handling significant amounts of fuel with such high enrichment.

The "plain old NU" and HW moderator is one of most uranium economic designs. However, "plain old NU" and graphite moderator are what started off the first reactor.
As the NU or any fuel in a MSR is a hot liquid (UF4-UF3 eutectic in HW-MSR), The losses of HW due to evaporation could be very high. Graphite has its own life and waste problems. How would a BeO solid moderator work in a MSR in calandria configuration.
Could it reach a unity conversion ratio?


Top
 Profile  
 
PostPosted: Mar 03, 2010 8:46 am 
Offline

Joined: Jul 14, 2008 3:12 pm
Posts: 5057
jagdish wrote:
jaro wrote:
More importantly, in the case of the DMSR the fuel is LEU enriched to 19.7% U235, whereas the HW-MSR fuel is plain old NU : This has implications all the way from procurement & shipping, to compliance with criticality safety regs when handling significant amounts of fuel with such high enrichment.

The "plain old NU" and HW moderator is one of most uranium economic designs. However, "plain old NU" and graphite moderator are what started off the first reactor.
As the NU or any fuel in a MSR is a hot liquid (UF4-UF3 eutectic in HW-MSR), The losses of HW due to evaporation could be very high. Graphite has its own life and waste problems. How would a BeO solid moderator work in a MSR in calandria configuration.
Could it reach a unity conversion ratio?


That's the problem, BeO isn't good enough to allow natural uranium. It would almost certainly allow unity conversion ratio but the question is at what enrichment level.

As for cooling, we're talking 5-10 percent of total heat flux. Significant but not anywhere near a deal breaker. You'd probably want to cool the BeO moderator as well, especially for a very compact core. The aircraft reactor experiment used cooling channels in the BeO blocks. While BeO can take high temperatures, most other materials can't. Cooling a solid BeO moderator isn't quite as easy as cooling liquid water. Pumping water around for cooling is an extremely well established practice. BeO itself is a fantastic conductor of heat, but the piping for cooling will make the system crude and more complex and there will be more neutron losses to the plumbing. We'd like heat pumping capability at higher power densities; that's one of the major reasons LFTR uses a liquid fuel.


Top
 Profile  
 
PostPosted: Mar 03, 2010 8:28 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Cyril R wrote:
As for cooling, we're talking 5-10 percent of total heat flux. Significant but not anywhere near a deal breaker.

Actually, more like 3.5% to 4% of total heat flux.

That's based on the CANDU experience, with 380 fuel channels -- an excellent heat transfer design, in a very undesirable sense.

If we switch to fat fuel channels that are ten times less numerous, we at the very least compensate for the increased radiative heat transfer that results from higher operating temperatures.....


Top
 Profile  
 
PostPosted: Mar 06, 2010 9:09 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
After a bit of study, it now appears that our assumption about inelastic scattering, especially as it relates to fluorine (see preceding posts), was completely mistaken.

This ENDF plot compares the two types of neutron scattering x-sections (elastic & non-elastic) of both F19 and U238 (as in NU) to absorbtion and fission x-sections of U238.....

Image

In both scattering cases, U238 is comparable, though generally higher than F19.
Also, a neutron scattered elastically by U238 loses almost NO energy, so its as though the scatter event never happened.
Only the inelastic scatter is of any significance.

More importantly, all scattering x-sections are higher than radiative capture of U238, from at least ~0.1MeV to 6MeV +

However, its important to remember that inelastic neutron scattering only has a significant effect on neutron energy when the target is a heavy nucleus.
Thus for an inelastic collision of a 2MeV neutron with a U238 nucleus, the neutron energy drops to 0.6MeV on average (using the formula E' = 6.4* sqrt(E/A), where A is the mass number).
For F19, the result of an inelastic neutron collision is about the same as an elastic one, for energies below about 2.5MeV ! (for the more extreme case of a 6MeV neutron, it drops to about 3.6MeV)

For U238, the scattering x-section is higher than fission (~3.5b VERSUS ~0.5b), even for energies > 1.5 MeV
But in NU (or SEU) the vast majority of HM nuclei are U238, so still a good chance for fission, after a few bounces, mainly with fluorine.

The fissile nuclei Pu239 & Pu241 have a fission x-section some 3 to 4 times higher than U238, and down to ~0.8MeV -- as do the "fissionable" nuclei (Pu238, 240 & 242).
U235 is only about 2.5 times higher for fission than U238.
But the concentration of all these is low in the fuel channel compared to U238.

In conclusion, it looks like there is a decent chance of U238 fission, even with all the different kinds of collisions.

After a few bounces inside a fuel channel, an un-absorbed neutron goes out into the surrounding D2O moderator, for a quick trip down the energy spectrum, to very low thermal energy (0.025 eV).

By contrast, the resonance absorbtion region for U238 extends only from about 5eV to 20,000eV.

The size of the fuel channels must be optimised such that the great majority of moderation in this energy range takes place in the surrounding D2O.


Top
 Profile  
 
PostPosted: Mar 08, 2010 2:45 am 
Offline
User avatar

Joined: Aug 21, 2008 12:57 pm
Posts: 1059
Quote:
The size of the fuel channels must be optimized such that the great majority of moderation in this energy range takes place in the surrounding D2O.


The devil is in the details but the details are treating you well. However…

Background, one of the ways that the CANDU achieves its high efficiency is precise sizing and positioning of is various components. The HW-MSR is a derivative of the CANDU approach.

Regarding HW-MSR optimization, is there a possibility that the HW-MSR may become a one trick pony? What happens when other possible diluents and fissile are used exclusive of U238 and U235 such as LWR wastes of various burnup levels, plutonium, U233, TH232 and various combinations thereof?

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


Top
 Profile  
 
PostPosted: Mar 08, 2010 6:37 am 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1954
Location: Montreal
Axil wrote:
What happens when other possible diluents and fissile are used exclusive of U238 and U235 such as LWR wastes of various burnup levels, plutonium, U233, TH232 and various combinations thereof?

The main problem I see with increased levels of transuranics in the core, especially the even-mass ones, is the same for all reactors: delayed neutrons cease to be effective in reactor control, because their energy at birth is less than half that of prompt neutrons from fission, so they are unable to contribute to the fission chain (even-numbered TRUs being able to fission only with fast neutrons).
This means we can't use these "lost" delayed neutrons to top-off the multiplication factor (so to speak), to avoid being critical on prompt neutrons alone -- an uncontrollable situation, as you know.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 99 posts ]  Go to page Previous  1 ... 3, 4, 5, 6, 7  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