Energy From Thorium Discussion Forum

It is currently Dec 10, 2018 4:51 am

All times are UTC - 6 hours [ DST ]




Post new topic Reply to topic  [ 21 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Salt Analysis Tool
PostPosted: Jul 12, 2013 12:07 pm 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Well, the US Army ran out of money. To compensate, they sent me home. So, for the next eleven Fridays I need a hobby. I thought I’d disinter an old project I started here a few months ago - analyzing fission yields. I started that project in Star Office Calc, which was fraught with peril. I finally popped for MS Office at home, so now I’m switching to a database approach.

The first objective of the old project and the new is to calculate cumulative fission yields. The old data set from Lawrence Livermore, http://ie.lbl.gov/fission.html, had a column labeled Cumulative Yield, which I didn’t understand because accumulation should depend on time. Also, the yields didn’t add up to 200. In my new data source from Brookhaven, http://www.nndc.bnl.gov/sigma, the total yields are 1.99999 or 2.000000. (These sums are actually a bit suspicious in that the uncertainties in the individual yields are up to 50% for some isotopes!) This source also offers yields for different neutron energies, which we might use later to explore fast versus slow reactors.

I’m going to assume a reactor that runs at one fission per second – the height of minimalism if that’s not a contradiction in terms. I want to run the reactor for a billion seconds (31 years). Obviously, I’m not going to do the calculations a billion times. As in the original project, I’ll take an exponential shortcut. I’ll take snapshots of the salt composition at 1, 10, 100 … seconds. This has the flaw of any time-step model in that all the action takes place in the first instant. So, I’ll modify the scheme by reintroducing short-lived isotopes toward the end of the time-step. For example, in the third time-step I’ll have 90 fissions in the first second, nine more at the 90 second mark, and one at the last second. This should only take ten iterations, e0 to e9, at each time-step adding the results of the last time-step.

Before I go further, I wanted to ask the forum members’ opinions of this methodology. Has anyone heard of a similar approach?


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Feb 21, 2014 8:56 am 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Hi Steve, I calculated the fission product composition and the energies released by the different fractions of fission products for a chloride reactor.

Did you proceed?
Do you want to share your findings?

Best regards

Holger


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Feb 27, 2014 9:56 am 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Hi, Holger

I've been creeping closer. I spent a lot of time cleaning the data! I am using fission yield data from the Brookhaven Sigma pages, http://www.nndc.bnl.gov/sigma/index.jsp?as=235&lib=endfb7.1&nsub=11, and isotope data from TORI_2 at Livermore http://ie.lbl.gov/databases/databases.html, augmented by B-m data from IAEA http://www.oecd-nea.org/dbdata/nds_jefreports/jefreport-20/nea6287-jeff-20.pdf. Deconflicting these has been painful, to say the least. I really ought to publish just the cleaned up database.

I've also changed the exponential part of the methodology. I'll be doubling the time-step at every iteration. The idea is that I can reuse the data from the previous time-step in the current time-step. The fissioning nucleus doesn't care about the current content of the salt. In this regard, a week of fission is just like any other week. (I will have to decay the "old" stuff" after the doubling, to get a true snapshot of the isotopic composition.)

The problem is the first time-step - the first one second of reactor operations. The standard decay formula assumes a lump of stuff that decays, and only decays. The situation inside a working reactor is one of concurrent creation and decay. However, I only have to calculate concurrent creation and decay once. I'll be using (doubling) that data 30 times, out to a billion seconds of reactor operation or ~16.5 years.

By doubling the time-step I can recreate any duration of reactor operation. For the twenty years of the DMSR, I would add the two-year snapshot to the 16.5 year snapshot (with two more years of pure decay for the nuclei born in the first 16.5 years), then append the one year snapshot, the 145 day snapshot, etc.

Since I have data for several fuels and several neutron speeds, I was going to add a button that allows the reactor designer to vary the reactor speed. The user enters 10% fast, 20% epi, 70% slow, for instance, to describe the proportion of "effective" (non-bouncing) neutrons for the design's combination of salts, moderator and core geometry. The model makes no attempt to do neutronics, so these input numbers would need to come from some other analysis. I really have no idea of the proportion of fast reactions in a slow reactor.

I also have no concept yet for neutron absorption. Given data, which hopefully matches the neutron speed distribution from the Sigma data, I plan to evolve the fuel as well as the fission products. Currently, I could vary the fuel mix on a schedule, but I'd like to make it dynamic. Then we might try burning SNF - and maybe old weapon material - for a few years, then switch to 233-U to see how much SNF remains.

I also don't have a concept for salt cleaning yet. Otherwise, the model might be useful for solid fuel reactors, if you are willing to accept the homogenous fuel assumption.

I'm retiring and moving to Michigan tomorrow, so I'll be off-line for a few days.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Mar 17, 2014 3:48 pm 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Hi SteveMoniz....I did not understand ....do you plan to calculate the burn-up of a reactor?

I did calculate what happens with 1 Kg of fissioned PuCl3....fractions of fp and fp chlorides after certain time intervalls.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Mar 23, 2014 11:31 am 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Holger,

One of my goals is to calculate the isotopic composition of the salt for any time interval. That's part of the reason for the binary doubling scheme. If you want to determine the day the Platinum starts to plate out, you first look for a Pt inventory (%) greater than the saturation point of Pt in the salt. Assume you find that in the 4-year snapshot. Then go back to the 2-year snapshot and add in the Pt inventory for the 1-year snapshot. If that is too large, try adding the 6-Month snapshot, etc.

At the moment, I have data only for fission products. I will have a dynamic simulation of FP evolution, but the fuel is static. Stage Two of development will have a mix of fuels, but it will still be static. Stage Three will will include neutron absorption. (I'll need some help with that.) That should allow the fuel and heavy actinides to evolve dynamically, and fine-tune the FP evolution as well. Stage Four allows the user to vary the input fuel on a schedule. For instance, you might start with dirty SNF-based fuel, then at some point switch to 233-U, to see how much SNF you end up with compared to how much you burned.

The overall goal is to create a simple, PC-based simulation that allows a designer to test various combinations of salts, fuels, fuel schedules, and fast/slow reactor geometries. There are a whole bunch of engineering questions that can be addressed if you know what's in the salt at any given time.

Steve


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Mar 29, 2014 12:32 pm 
Offline

Joined: Sep 22, 2013 2:27 pm
Posts: 262
Steve, I did that for a MCFR (Molten Chloride Fast Reactor).

I caculated heat and fp consistance of 1 Kg fp fissioned from the fuel mixture 238 U 8%, 239 Pu 62%, 240Pu 24%, 241 Pu 4%, 242Pu 2% over 100 years.

I assumed a degassing of noble gasses and a plating out of noble metals once/hour and a further fuel reatment step daily.

Then it is to calculate the consistence of the fuel (age of fp x quantity) and at the end I get the fp energy and composition in the reactor in the gas treatment, in the plating out and in the used fuel for external reprocessing.

If you like please write a mail.

Holger


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Mar 30, 2014 12:43 pm 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Holger,

I realize that the equations can be done for a specific time frame, 100 years in your example. I am shooting for a more flexible tool. Actually, this project started in response to the Cumulative FP Yield column in the England-Rider data from Lawrence-Livermore. I thought a single column was oversimple. For example, the cumulative yield for 90-Y (from 233-U) is 6.79%, most of which comes from 90-Sr and isotopes higher on that decay chain. Well, after ~300 years this is (99.9%) true, but for shorter time frames?

The idea of building a simple tool to test reactor designs evolved from that investigation - almost by itself.

I should first ask the community on this forum (the customer) if there is a desire for such a tool. I have seen some debates here on fast vs slow reactors, so I thought I'd create a tool flexible enough to test both. I also thought, since the data are available, that the user could test different fuel mixtures. I also assumed that decay heat would be an important question. I don't have data for that yet, but I designed the database to be robust enough to incorporate that later.

The output of my model will be a series of snapshots of the isotopic composition of the salt at 1, 2, 4...one-billion seconds (16 years). There would be one table of snapshots for each design. In a "real" database (Access is something of a toy) the output would be in one table, with additional fields to describe the reactor' design and operating parameters. At the moment, those fields would be ReactorSpeed and FuelMix. (These would be named, with the design/operating parameters in other tables.)

Future design parameters might be FuelSchedule and CleansingPlan. Of course, these are just my ideas of the direction the model might go. That would depend, again, on the customer's desires.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Jul 05, 2014 5:53 pm 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Attached is a proposal for simulating isotope evolution in an MSR.

I'd appreciate your review and comments.


Attachments:
Molten Salt Reactor Simulation Concept Paper.docx [366.99 KiB]
Downloaded 173 times
Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 24, 2014 3:05 pm 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
I loaded a video presentation of my project on YouTube.

https://www.youtube.com/watch?v=D3MNsk- ... tu.be&hd=1

Let me know what you think.

Steve


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 24, 2014 6:14 pm 
Offline
User avatar

Joined: Nov 30, 2006 9:18 pm
Posts: 1938
Location: Montreal
SteveMoniz wrote:
I loaded a video presentation of my project on YouTube.

https://www.youtube.com/watch?v=D3MNsk- ... tu.be&hd=1

Let me know what you think.

Steve

Very nice !

How does your approach differ from the ORIGEN code ?


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 24, 2014 8:09 pm 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
I have never seen ORIGEN. If it uses exponential time steps, that's news to me. I've never heard of any simulation that uses exponential times steps, but my view of the entire simulation community is limited.
I get the impression that many nuclear simulations are focused on neutron economy. I am not worried about neutrons. I simply assume there are enough to keep the reactor critical. If there are 2.5 neutrons per fission, I expect I will have to deal with the 1.5 excess when I get to absorption.
I am not in the nuclear business in any way. I was an operations research analyst for the US Army. I have a lot of experience with simulations. I spent a couple of years writing code for a combat simulation. I spent more time in data analysis. I also helped design analysis plans. Those plans sometimes call for changes in the simulations, or sometimes new models.
I saw the need for a new model here. Frankly, I don't know how engineers predict what will be in a fuel rod when it is removed, or as it cools. I know even less about how they predict what's in the fuel rod when it's still in an active reactor.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 24, 2014 9:31 pm 
Offline

Joined: Apr 28, 2011 10:44 am
Posts: 250
Exponential time steps are very common if not the rule in burn up calcs.
Often the analyst will go exponential for a while
and then swtich to linear, once he gets up to say a year.
I dont think you need to start at 1 second, 1 day is fine.

If I were you, I'd take a look at Serpent before trying to reinvent the wheel, http:// montecarlo.vtt.fi.
The source code is availabel for free for non-commercial use, if you jump thru
a few simple hoops.

You need to start out doing the flux calculations for the geometry of the design and the fresh fuel salt
(usually via Monte carlo). The neutron flux in each cell has to be broken down into a bunch
of different energy bins. Combine this flux distribution with the energy dependent cross-sections
to get your isotope transition matrix, do the burn up for a bunch of time steps (exponential is fine)
and then repeat the Monte Carlo with the new salt compostion, and so on.
Aufiero has even modified Serpent so the code adjusts the fuel salt periodically
to keep the criticality very close to zero.

Serpent is blazingly fast,
much faster than the US govenrment codes in our experience.

Your graphics are very cool and my guess would be welcomed by the Serpent community.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 25, 2014 9:44 am 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Quote:
take a look at Serpent

Excellent reference. Thanks. The home page describes the methodology well enough to imply their priorities. They mention “depletion zones,” “detailed assembly-level burnup calculations”, and “spatial homogenization”. These imply their focus is on the depletion of individual fuel assemblies/rods/pellets. That is very much a solid fuel thing. I submit that the homogeneity of a liquid fuel makes all this moot, and the focus should be elsewhere.
They also mention “when three million neutron histories are simulated”. This implies that Serpent is an event-driven model, not a time-step model. They “create” a neutron (first event} and see what happens to it. It travels some distance (not an event). Then it might be absorbed, bounce off something, with or without shedding some momentum, or cause a fission event. All of these events are separated by a variable amount of time, so a time-step model is clumsy.
Event-driven models are well suited to Monte-Carlo (stochastic) methods. Repeat the creation event three million times, and you will get some good averages and other statistics. The alternative is to take the averages ahead of time and use a deterministic simulation. It’s faster, but you lose the ability to generate deviation statistics. You can’t report on how uncertain your answers are.
Simulation design starts with this 2x2 choice. Will this be an event-driven or time-step model? Will it be deterministic or stochastic? Your choice depends on the goal of the analysis. I’m hoping that a deterministic, time-step model has merit for some design questions.
This all depends on the more basic question of what the customer wants, which drives the analysis plan. I’ve made some assumptions in that regard. I’ve assumed an MSR designer wants to know the history of each isotope. I get the feeling that the Serpent designers only want to track certain significant isotopes, neutron poisons in particular. That’s all right for them. Their isotopes don’t travel outside the cladding/pellet. In an MSR any fission product might be a corrosion or solubility problem. I’m willing to spend CPU time on this rather than on Monte-Carlo calculations or grouping the isotopes.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 25, 2014 3:26 pm 
Offline

Joined: Apr 28, 2011 10:44 am
Posts: 250
You are looking at only a portion of Serpent's capabilities.
Serpent (as does MCNP, Scale, and others) integrates periodic neutronics calculations (event driven in yr lingo)
and burn up (time step). The cells describe the geometry of the core
which is just as important for liquid fuel reactors as solid fuel.

Serpent has been used extensively to study liquid fuel reactors
including the European MFSR project and our own DMSR effort.
The MFSR project has an on-going effort to combine Serpent and OpenFoam
into an integrated neutronics and thermo-hydraulics package
focused on liquid fuel.

I'm a customer and I need what Serpent and its competitors offer.


Top
 Profile  
 
 Post subject: Re: Salt Analysis Tool
PostPosted: Aug 26, 2014 7:39 am 
Offline

Joined: Jan 16, 2012 7:15 am
Posts: 94
Quote:
I'm a customer...

It's tough to sell to a customer that already has what they want. Of course, that never stopped a real salesman.
I see in the overview that Serpent is used to generate data for deterministic simulations. This could be helpful. I have data for fission yields from Brookhaven's Sigma database. http://www.nndc.bnl.gov/sigma/getFissio ... f=8&mt=454 for example. It has yields for (usually) three neutron energy values. I can't find similar energy dependent data for neutron absorption cross sections.
Before that, I would need aggregated neutron energy distribution estimates for each reactor design. Can Serpent fill in the blanks here, "For the DMSR x% of the neutron energies are in the thermal region, y% in the epi-thermal, and Z% in the fast region." (and "For the MFSR...For the MoSTAR...For the WAMSR"...)
Maybe I'll be a customer of theirs! I don't see anything like neutron energy distributions in the index of the Serpent users manual, unfortunately.
As for the graphics - the coolest are Pivot table charts in MS Excel. I reformatted the Sigma data for an Access database. I still have some QC to do on that, but I can attach a .csv file here in the near future.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 21 posts ]  Go to page 1, 2  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