In using TCMs based on metallic nanowires it is important to determine the effect of nanowire geometry and spatial arrangement on the resulting network. To this end we have extensively simulated the effect of wire length and device size on the percolation properties of the network produced. We have performed Monte Carlo simulations of 2D conductive stick networks including for the first time stick lengths approximating nanowires which are produced experimentally. Each simulation is performed based on an average stick length but the actual lengths of the nanowires in the simulation are randomly generated with a normal distribution around the defined average length. The effects of density and length distribution on the percolation threshold are also explored. The results of such simulations are also employed to determine an elementary representative volume, which can be directly applied to a device design by allowing the determination of the nanowire density required to produce a conductive network associated with a characteristic length, such as diffusion length or pixel size. We also extend this work to the specific application of metallic nanowire networks as front electrodes in dye sensitized solar cells (DSSC), allowing a calculation of the collection efficiency as a function of network density. These calculations were based on the diffusion length of electrons generated within a DSSC and a spatial mapping of the collection efficiency function on the solar cell surface.