A few things to keep in mind:
- Yes, the removal cross section is a good strategy to address (n,2n) reactions
- Use a good group structure for your deterministic method
- Beware of diffusion coefficients (if using)
Point 2 is not trivial, depending on your reactor, but there are some good group structures out there for various neutron spectrums.
Point 3 is very important to understand. Monte Carlo tallied diffusion coefficients should be treated with extreme caution.
Please see this discussion: Units of Diffusion_Coefficient? - #5 by bakingbad
and this OpenMC Pull Request to read further: Condensing diffusion instead of transport by mkreher13 · Pull Request #1916 · openmc-dev/openmc · GitHub
Furthermore, if you are using diffusion, consider using a non-linear diffusion coefficient to smooth out the differences between Monte Carlo and your diffusion solver.