OpenMC-MCNP heating validation

I am calculating neutron heating of components in a 3D tomamak geometry. I used both MCNP and OpenMC to obtain the values but they don’t match. I tried used simple reduced order model ( cylinders ) to see if the problem was in the geometry and still numbers are off. Materials are exactly same using FENDL3.2b. I am still not sure what could be the problem. Any thoughts?