This is because during calculation of Hirshfeld charge (or ADCH, CM5, Hirshfeld-I charges), Multiwfn automatically removes all orbitals higher than LUMO+10 to reduce cost of computing electron density. You can set "idelvirorb" in settings.ini to 0 to disable this treatment.

Best regards,

Tian

I've noticed an inconsistency when generating the orbcomp.txt: The orbital composition analysis is different when first performing a population analysis.

Steps to reproduce:
a)
> Multiwfn co.molden.input  # molden file generated with orca
> 8 # Orbital composition analysis
> 8 # Calculate atom and fragment contributions by Hirshfeld method
> 1 # Use build-in sphericalized atomic densities in free-states (recommended)
> -4 # Export composition of every atom in every orbital to orbcomp.txt

orbcomp.txt contains all orbtals

b) 
> Multiwfn co.molden.input  # molden file generated with orca
> 7 # Population analysis and atomic charges
> 1 # Hirshfeld atom population
> 1 # Use build-in sphericalized atomic densities in free-states (more convenient)
> n # If output atoms with charges to freq.molden.chg in current folder
> 0 # return to main menu
> 8 # Orbital composition analysis
> 8 # Calculate atom and fragment contributions by Hirshfeld method
> 1 # Use build-in sphericalized atomic densities in free-states (recommended)
> -4 # Export composition of every atom in every orbital to orbcomp.txt

orbcomp.txt contains only first 16 orbitals (for carbon monoxide, which has 14 occupied)

Best,
Nico