In this computational work, a new simulation tool on the graphene/polymer nanocomposites electrical response is developed based on the finite element method (FEM). This approach is built on the multi-scale multi-physics format, consisting of a unit cell and a representative volume element (RVE). The FE methodology is proven to be a reliable and flexible tool on the simulation of the electrical response without inducing the complexity of raw programming codes, while it is able to model any geometry, thus the response of any component. This characteristic is supported by its ability in preliminary stage to predict accurately the percolation threshold of experimental material structures and its sensitivity on the effect of different manufacturing methodologies. Especially, the percolation threshold of two material structures of the same constituents (PVDF/Graphene) prepared with different methods was predicted highlighting the effect of the material preparation on the filler distribution, percolation probability and percolation threshold. The assumption of the random filler distribution was proven to be efficient on modelling material structures obtained by solution methods, while the through-the -thickness normal particle distribution was more appropriate for nanocomposites constructed by film hot-pressing. Moreover, the parametrical analysis examine the effect of each parameter on the variables of the percolation law. These graphs could be used as a preliminary design tool for more effective material system manufacturing.