Depletion calculation of UN fuel with neutron poisons

Dear experts,
I am carrying out depletion calculation for adding different neutron poisons to UN fuel of space nulcear reactor, such as Gd. The results are inconsistent with my expectations.
I made 2 calculations. case 1 is for comparison. It is fuel UN without adding neutron poison. The material settings are as follows:

un_density = 14.420    
percent_td = 0.90
un = openmc.Material(material_id=1, name='UN')
un.add_element('U', 1.0, enrichment=90.3)
un.add_element('N',1.0)
un.set_density('g/cm3',un_density*percent_td)
un.volume = 691.748

The second is case 2 with Gd added (using un_poison). The material settings are as follows, adding un_poison.depletable = True.

un_density = 14.420    
percent_td = 0.90
un = openmc.Material(material_id=1, name='UN')
un.add_element('U', 1.0, enrichment=90.3)
un.add_element('N',1.0)
un.set_density('g/cm3',un_density*percent_td)
un.volume = 691.748

poison_content = 0.02
un_poison = openmc.Material.mix_materials([un, heatpipe_gd], [1-poison_content, poison_content], 'wo', name='un_poison')
un_poison.volume = 691.748
un_poison.depletable = True

The settings for depletion are the same:

chain_path = "./chain_endfb71_sfr.xml"
model = openmc.model.Model(geometry=geom, materials=materials, settings=settings, tallies=tallies)
operator = openmc.deplete.Operator(model, chain_path, normalization_mode='fission-q')
power = 0.1e6 
time_steps = [30,60,90,185,365,2*365,2*365,2*365,2*365] 
integrator = openmc.deplete.PredictorIntegrator(operator, time_steps, power, timestep_units = 'd')
integrator.integrate()

Comparison of the results (case 1 has a large std dev):

keff results for case 1 (without poison):
[array([[1.04789228, 0.0017138 ],
       [1.04752424, 0.00167023],
       [1.04371168, 0.00185702],
       [1.04389522, 0.00162234],
       [1.04281861, 0.00175352],
       [1.0389861 , 0.00164979],
       [1.03590802, 0.00170099],
       [1.02709169, 0.00174944],
       [1.0237148 , 0.00167772],
       [1.01709204, 0.00154434]])]
burnup days:
[   0.   30.   90.  180.  365.  730. 1460. 2190. 2920. 3650.]
u235 atoms change results:
[1.96034452e+25 1.95935604e+25 1.95737872e+25 1.95441529e+25
 1.94832784e+25 1.93633080e+25 1.91241830e+25 1.88850945e+25
 1.86462332e+25 1.84071687e+25]

keff results for case 2 (with Gd added):
[array([[1.02059977e+00, 7.55974005e-05],
       [1.02039793e+00, 6.90662180e-05],
       [1.01971659e+00, 7.28710758e-05],
       [1.01917339e+00, 6.91670898e-05],
       [1.01757732e+00, 6.80452353e-05],
       [1.01427506e+00, 7.40373912e-05],
       [1.00797372e+00, 7.17044729e-05],
       [1.00160563e+00, 7.37620513e-05],
       [9.95337691e-01, 8.05775337e-05],
       [9.88860581e-01, 7.25042965e-05]])]
burnup days:
[   0.   30.   90.  180.  365.  730. 1460. 2190. 2920. 3650.]
u235 atoms change results:
[1.96034452e+25 1.96034452e+25 1.96034452e+25 1.96034452e+25
 1.96034452e+25 1.96034451e+25 1.96034451e+25 1.96034451e+25
 1.96034450e+25 1.96034450e+25]

The unreasonable places:

  1. In case 2, the amount of U-235 has almost no change.
  2. In case 2, the change of keff is a straight line, but in fact, it should be a curve, that is, with the increase of running time, neutron poisons are consumed and reactivity will be gradually released.

here are the updated results:

1 Like

Hello @chenchcc ;

It is very interesting but I’m not sure if you should necessarly get a curved keff variation. It is possible that you get a zoomed-out behavior (scale effect) since you are calculating over 10 years here.
You can try to change the power (+ or -) to see how the keff evolution will be affected.
Is that a fast neutron reactor? Gd are less sensitive in fast region as I know and you are getting shifted keff line because you are reducing the UN quantity in the core (and replacing it with ~ few % of poisons)
Probably you already did it, but how about the neutronic analysis of the fresh core (flux and reaction rate distribution(fission, capture), flux spectrum, …) what does it gave to you?

Thank you for your response and reminding, @bentridisalah
This is a typical fast neutron spectrum reactor. If the results are correct, the role of Gd in it is not as significant as I thought before.
Then I tried Er as another poison, as the above fig shows, because its cross section difference in thermal neutron energy region and fast neutron energy region is not so big as Gd. But according to the results, it still can not play the role of slowly releasing reactivity that I want. For a time, I thought that there maybe some problem with the code itself (most likely, it was the inconsistency between my two computing settings). I will think about whether there are other ways to achieve it.
Thanks again.

I almost didn’t notice this in my previous calculations, thanks for pointing it out.

1 Like

Gd,Er are spectral shift absorbers for enhancing thermal neutron absorption in the core,if your energy spectrum is hard ,burnuping results would be different from thermal speactrum.

1 Like