[MCNP] Lost too much Keff with Burnup card

[lee6853]

Hi there!
Me again.

I am doing my research about converting HEU research reactor to LEU.

I made FA and core finally and started using the burnup card to check changing of Keff and fission products.

Well, the thing was only with one-month burnup my Keff was decreased drastically from 1.118 to 1.025
I think it is too much and it is not good for using long period.
But in the output file, I found that U-235 decreased just from 2,768g to 2,396g.(U-238 was not changed.)

The reason I guess now are
1. My MCNP code has some problems.
-> I checked it but I am a beginner, I want you to check, please.(Attached Inputfile)

2. Less reflector.
-> I add additional Be reflector below the core but the result changed not much.(Common reactor has reflector below the core?)

3. Geometric Buckling. I checked 6-factor formula, but it seems that the only thing I can change is Pth which is thermal neutron non leakage possibility with Buckling. I mean original fuel assembly that I am using now is long rectangular fuel pin that has a rounded corner, but I designed it without a rounded corner just an angled corner. Do you think this difference can make such a big different Keff? But the only thing I can think now is my angled corner fuel pin has more neutron leakage than a rounded one.

4. Power. Original power pf HEU(80%) reactor was 10MWth. But with 19.9% LEU fuel, should I decrease the power?(But total U-235 mass is same...but thermal flux was decreased about 10%).


Best

 

[Alex A]

The square fuel will make no significant difference so long as the total mass is correct. I'm not familiar with BURN calculations so I may be of limited help. I started a run over night and it's not done yet, just one core on my laptop. 5 MW for 30 days burns about 186 grams of U-235 and produces about 12.5g of Pu-239. The fuel burned seems about right but that conversion ratio seems very low.

What details do you have on the reactor operating with HEU fuel? What goals are set for the operation with LEU fuel?

 

[lee6853]

The square fuel will make no significant difference so long as the total mass is correct. I'm not familiar with BURN calculations so I may be of limited help. I started a run over night and it's not done yet, just one core on my laptop. 5 MW for 30 days burns about 186 grams of U-235 and produces about 12.5g of Pu-239. The fuel burned seems about right but that conversion ratio seems very low.

What details do you have on the reactor operating with HEU fuel? What goals are set for the operation with LEU fuel?

Hi Alex A. Thanks a lot!

Well, actually I should have upload inputfile with 10MW burnup card but I did with 5MW. But it's Ok.
I am trying to reduce power level because I think 10MW is too high for this reactor.

My goal with this code is that, for the non-proliferation purpose, I want to convert HEU research reactor to LEU reactor to reduce HEU fuel in the world. Such kind of projects started long time ago but still many states are using HEU fuel for research reactor.

The reactor I designed is the reactor which was in Lybia(not now) which used 80% HEU and thermal power was 10MWth. That's why I used 10MW burnup card but it looks fuel lifetime is weird.

I searched and found that they ran reactor 20hours per day, 1day per week, 14weeks per year. That means they use this reactor only 280hours per year! Does it make sense? Then, can I assume it's lifetime is 2.57years?(Because my MCNP output says it can be used for 1month(720hours=24h*30days), and 720h/280=2.57) Is this way to calculate lifetime of fuel??

Best

 

[Alex A]

That makes a lot of sense. I see what you've done. You've taken a fuel pin in which the 'meat' thickness is 0.4mm and made it 1.6mm for use with 20% enriched fuel of the same density. This hypothetical fuel might get hotter internally, so a power derating might be appropriate - I have no idea what this might be.

I'm still at the stage of understanding the input file where I'm just making silly mistakes. I googled and found this which was very helpful. That describes a nigh identical core configuration. Some of the conclusions may be quite relevant, even though they are using a different replacement fuel. I think the 3 pin and 4 pin fuel masses are backward so I make the U-235 load of the 80% reactor 2388g, and MCNP states you're starting at 2768g. Does this match your numbers?

I'm also tinkering in ways that may go beyond what you are supposed to change. I'm trying the 4 pins in the center, though it's only improved the starting keff from 1.1186 to 1.1196. I'm disappointed with that.

I notice with a little amusement that BURN ignores keff entirely, chugging along at 0.9, because maintaining the reactor critical is someone else's problem. I wonder if there will be even fewer MWhs because right now there is no 'load'. No control rods, no experiment. Pulsed operation may help, reactivity recovers somewhat after power down but it's going to start self terminating close to the calculated end of life point.

If this rather gloomy picture is a mistake in the input file, I'm not seeing it yet.

 

[BillOnne]

I noted that your starting keff is quite high. A value of keff=1.118 is going to be in the prompt range. It's not clear that MCNP gives you sensible results in that range. Even the 1.025 is kind of high. Is it still prompt? I would need to do some careful examination. Don't have MCNP available just now.

See if you can move things around to get a value of keff much closer to 1. Depending on your "culture" you may use "dollar values" or milli-k. I am used to milli-k and I am wanting something in the "only a few" range. Calculate the reactivity value in milli-k using this formula.

$$\rho = 1000 \times (1 - 1/keff)$$

With a keff=1.118 you get more than 100 mk. You should be shooting for in the range of +/- 2 mk, as small as reasonably achievable. Because a large (positive or negative) $\rho$ value is indicating you are far from an eigenstate. It usually means that calcs involving critical systems are not very accurate, including burnup calcs.

Hmmm... Can't seem to make the LaTeX formulas work.

 

[berkeman]

Hmmm... Can't seem to make the LaTeX formulas work.

Looks okay to me, unless you wanted a different form of the equation. At PF, you have to refresh the page once when you post in LaTeX -- the forum software uses "lazy rendering" to speed things up, but it does mean that LaTeX is not rendered the first time it's posted.