Created by: paulromano

To determine reaction rates to deplete materials, the method we use now is to

1. Take the sum of the fission rate of each nuclide multiplied by the Q value for fission to get an "observed power".
2. Renormalize the reaction rates based on the ratio of the user-specified power to the observed power.

Notably, the Q value that is used right now is whatever is listed in MF=3 in the ENDF file, which corresponds the recoverable energy release (total energy release from fission minus neutrinos). It does not account for indirect energy release, namely the energy released from (n,gamma) reactions, which can account for over 5 MeV of energy release per fission event.

Since we are not including indirect energy release, the normalized reaction rates tend to be too high -- with a lower Q, you need more fission reactions to reach the same power level. This results in OpenMC depleting too fast.

The simple fix is to simply increase the Q values in the depletion chain to account for the indirect energy release. This is effectively what Serpent has traditionally done; the heating value for U235 was set to 202.27 MeV by default. I understand they have a new energy deposition treatment so more recent versions might not use the constant heating values for depletion power normalization.

Longer-term, we ought to provide some options for how to do power normalization rather than just brute-force, arbitrary changes to the Q values used in the depletion chain.

Thanks to @pshriwise for helping to remind us of this through his comparisons to Serpent.