Hello everyone !
I am performing a neutronic study in OpenMC where I replace light water (H₂O) with heavy water (D₂O) as the moderator in an identical reactor geometry.
I observe a significant decrease in keff (from ~1.00 with H₂O to ~0.73 or lower with D₂O), even after verifying material definitions, density, and thermal scattering data.
My question is:
Is it physically expected that substituting H₂O with D₂O in a fixed (non-optimized) geometry can reduce keff due to changes in moderation length and neutron spectrum?
Or should D₂O generally increase reactivity regardless of geometry?
I would appreciate any insight on whether this behavior is purely physical or if it suggests a modeling issue (materials, S(α,β), or cross-section library).
Yes, I think this is expected in your case. Deuterium has much lower absorption cross-section than protium. This leads longer mean-free-path, larger fast leakage and increases resonance absorption as well. Deuterium being a heavier isotope also means that it is slightly less effective moderator. For a reactor designed for light water this is expected.
I also replaced the light water moderator with graphite in my model, and the multiplication factor dropped significantly, reaching about 0.60. Since graphite is also a good moderator with very low absorption, I was expecting a better result.
Could you please explain what might be causing such a large decrease in k-effective? Is it mainly due to under-moderation, increased neutron leakage, changes in the neutron spectrum, or another effect related to a core originally designed for light water?
Yes, exactly. Graphite is even less effective as moderator than deuterium since its heavier, though it has about 7 times larger absorbtion cross-section (thus slightly shorter mean-free-path). The underlying physical reasons for the behaviour are pretty much the same. Leakage increases as well as fast resonance absorption, since neutrons spend more time slowing down in these moderators than in light water.
Hello !
Thank you for your explanation.
I have another question: why and how does neutron leakage increase when replacing light water with graphite or D2O? Is it mainly related to the longer neutron slowing-down path and the change in neutron diffusion properties?
In this case, should I keep the same reactor model and only change the moderator material, or would a new core design (geometry or fuel-to-moderator ratio) be required for a fair comparison with graphite moderation?
Yes, exactly. The main reason for larger leakage is the longer mean-free-path, the lower energy loss per collision and as a result the overall longer path it takes for the neutron to thermalize. The light water reactor is not designed to accomodate for that.
What exactly are you trying to achieve here? I think new core design would be necessary if you want the system to become critical.
Hi
Thank you, that makes sense.
So if I understand correctly, simply replacing the moderator is not enough. To achieve a higher k-effective or restore criticality with materials such as D₂O or graphite, the reactor geometry and core design would also need to be modified to account for the different moderation properties and neutron leakage. Is that correct?
Yes, and I’d assume that k-eff is not your quantity of interest in this case, since you want to make the system critical. Instead you may want to study the neutron spectrum or flux of a critical LWR and a HWR or a graphite moderated reactor. In that case you will need a different core design in terms of fuel and geometry. Or you can take different reactor designs and compare them, such as the Brookhaven Research Reactor in addition to your current one. Or you may want to study neutron moderation and multiplication infinite homogenous mixtures of fuel and moderator. It is difficult to say without knowing more details on what you want to achieve.
My objective is to study the effect of different moderator materials while keeping the same reactor model. I am not trying to design a new reactor or optimize the core for each moderator.
I am interested in comparing how parameters such as k-effective, neutron flux, and neutron spectrum change when the moderator is replaced. The decrease in k-effective is therefore also part of the comparison and helps illustrate the impact of the moderator on the neutronic behavior of the system.