Depletion execution

User Support
1

Running OpenMC Version 0.13.1 under Ubuntu 20.04.

I would like to be able to run a depletion, adjust the material file to “sweeten” the fuel, and then run another depletion on that revised material file. The issue I am running into can be shown as follows (Sorry for the length, but want to provide enough detail for someone to actually be able to help…):

#Generate Geometry, materials, settings (note external calls exports to xml files)

settings = DefineSetting()
Mats, materials = DefineMaterials()
geometry = DefineVesselGeometry(VesselRad,Cyl_H,VesselWall,HemiModThickness,CoolPipeRad,HexPitch,FuelBox,RodRadius,RodHeight)

import openmc.deplete

#using simple chain until I can get the system to respond properly

chain = openmc.deplete.Chain.from_xml(“/home/brian/OpenMC_data/chain_simple.xml”)

model = openmc.Model(geometry=geometry,materials=materials, settings=settings)
operator = openmc.deplete.Operator(model,“/home/brian/OpenMC_data/chain_simple.xml”)
#power is in watts - 1e6 = 1 MW - try 200 Megawatts
power = 200e6
time_steps = [60 * 60] * 5 #deplete for 5 hours
integrator = openmc.deplete.PredictorIntegrator(operator, time_steps, power)
integrator.integrate()

#This all works… I get everything I expect for the 5 hour depletion

results = openmc.deplete.ResultsList.from_hdf5(“./depletion_results.h5”)
#Save the intermediate materials files for review
for i in range(5):
mf=results.export_to_materials(i)
fn=“./materials”+str(i)+“.xml”
mf.export_to_xml(fn)
print(mf[0].get_mass(“U235”))

#Output of above is:
6587.592943820452
6578.547441670925
6569.501987610248
6560.456546983021
6551.411128389608

#showing expected depletion of U235. All good.
#Now take that last “materials4.xml” file and make it the default
#materials.xml file and run openmc.run() to get Keff:

import shutil
fn=“./materials4.xml”
shutil.copyfile(fn,‘./materials.xml’)
openmc.run()

#results from this show greatly depleted fuel:

k-effective (Collision) = 0.45918 +/- 0.00118
k-effective (Track-length) = 0.45920 +/- 0.00118
k-effective (Absorption) = 0.45804 +/- 0.00205
Combined k-effective = 0.45900 +/- 0.00114
Leakage Fraction = 0.01333 +/- 0.00047

#My expectation is that this Keff should be the same as the last iteration of keff:

time, k = results.get_eigenvalue()
time /= (60 * 60) # convert back to hours from seconds
k

array([[1.11036474, 0.00243176],
[1.11429175, 0.00236096],
[1.10776158, 0.00240418],
[1.10337545, 0.00222762],
[1.10350493, 0.00225666],
[1.10449178, 0.00228493]])

Can someone explain where I am going wrong here? Thank you.

2

OK, some additional information. It appears to me the “materials.xml” file is setup differently when the depletion models are used. Here is FuelSalt xml listing for a Keff determination (matches initiating code exactly):

cross_sections.xml</cross_sections>








Here is material file from the “zero” index of a depletion run. Note addition of the chain file isotopes, and the change in the U235/U238 values. Those were the ONLY concentrations that were changed:

So I guess my answer is, unless someone can tell me otherwise, is that the material files are not compatible between Keff determination and a depletion run.

3 
  • <density units="g/cm3" value="2.3" />

  • <nuclide ao="0.65" name="Li7" />

  • <nuclide ao="0.29" name="Be9" />

  • <nuclide ao="0.05" name="Zr91" />

  • <nuclide ao="0.002" name="U235" />

  • <nuclide ao="0.008" name="U238" />

  • <nuclide ao="1.51" name="F19" />

  • <density units="sum" />

  • <nuclide ao="0.29" name="Be9" />

  • <nuclide ao="1.51" name="F19" />

  • <nuclide ao="0.65" name="Li7" />

  • <nuclide ao="0.05" name="Zr91" />

  • <nuclide ao="1e-21" name="Cs135" />

  • <nuclide ao="1e-21" name="Gd156" />

  • <nuclide ao="1e-21" name="Gd157" />

  • <nuclide ao="1e-21" name="I135" />

  • <nuclide ao="1e-21" name="U234" />

  • <nuclide ao="6.475223363030086e-05" name="U235" />

  • <nuclide ao="0.00025900893452120345" name="U238" />

  • <nuclide ao="1e-21" name="Xe135" />

  • <nuclide ao="1e-21" name="Xe136" />

For some reason, xml didn’t paste properly in my email.