Do I need to set the variances/residual variances in a power analysis Monte Carlo simulation?

Question

I have to do a power analysis for a longitudinal mediation model - I am writing a grant proposal for a phd milestone (so this needs to be correct enough for my professors to