Dear Expert @sourena @paulromano @AlexandreTrottier , when I compare the Xe135 production for Candu reactor with Serpent and Openmc, I get different results. How can I solve this problem?
candu_openmc.py (5.7 KB)
candu_serp.py (3.2 KB)
candu_openmc.py (5.7 KB)
candu_serp.py (3.2 KB)
1 Like
I think you have defined your boundary types differently. But I have no idea how different the results are. In your input files, while the “periodic” boundary is defined in Serpent, you have defined it as “reflective” in Openmc. But I still recommend you to get advice from an expert.This Expert @sourena can help you in more detail on this matter.
1 Like
Hello. I have not looked over the input files, however SERPENT has an option to solve the burnup problem while also accounting for equilibrium concentrations of fission products. Was this used? To my knowledge, there is no corresponding feature in OpenMC but @paulromano may correct me here.
Are other production rates in better agreement? For example, Pu-239 production is critical in CANDU physics. If Pu-239 generation rates are consistent, I would look to the treatment of equilibrium fission products. If the generation rates differ significantly, I would check the power densities used in the normalization of the flux,
1 Like
Yes, I used the same file in the code, but the results were still the same. I think this file is '…/data/depletion/serpent_fissq.json’Even though I changed the Boundary conditions type, my results do not change.
If the power set is nominal then both curves look off. Both should clearly saturate after 3 days.
In serpent you set powerdensity (kW/g) equal to power in openMC. Without really knowing openMC syntax that looks suspicious to me. Are you sure 271*linear power is correct for both?
1 Like
Thank you for your answer. I may have missed this mistake. I will try to compare the results by doing the calculations again.
Apologies, I missed this message earlier. Yes, as @awHOR pointed out, Serpent is depleting with a power density of 38.6 W/g, but in OpenMC you are setting the power, not the power density. You can fix it by doing:
integrator = openmc.deplete.PredictorIntegrator(
op, timesteps, power_density=38.6, timestep_units='d')
but note that the units for power_density in OpenMC are W/g, so you need to use 38.6 not 38.6e-3.
On the 38.6 value itself: it is a little bit high, but it’s fairly close to what it should be for a typical CANDU lattice (I’ve seen values ranging from ~24 to ~35 depending on where in the core the bundle is supposed to be located). Note that the maximum power limit for a 37-element CANDU bundle is 935 kW.
1 Like