is this true that when in a system there are both of nanofluids(cu) and water
we cannot use NEMD OR OTHER METHODS to calculate thermal conductivity and we should modify compute heat flux?
It’s up to you to decide whether or not you can use a particular method.
If by NEMD you mean boundary driven MD, in which you heat one part of the box and cool another with thermostats, then you absolutely can do that with a Cu nanofluid in LAMMPS. Again, though, it’s up to you to determine whether or not that gives you sensible results.