Copper nanowire (CuNW) networks are promising candidates for flexible, transparent conductive surfaces and nanoelectronic interconnects due to their high intrinsic conductivity and compatibility with scalable fabrication methods. The electrical performance of CuNW networks is strongly influenced by network topology, nanowire geometry, and inter-nanowire junction properties. In this work, we present an efficient and scalable modeling framework for large CuNW networks with arbitrary patterns and apply it to study the sheet resistance of randomly oriented CuNW networks. Individual nanowires are described using physically accurate resistivity models that account for surface and grain-boundary effects, while inter-nanowire interactions are represented by resistive junctions. This approach is used to investigate computational limitations and the impact of leakage on sheet resistance in randomly oriented CuNW networks. The results show that up to an 8.6× speedup can be achieved by parallelization on 20 cores, but this comes at the cost of high RAM usage, up to 120 GB , which limits parallelization. Furthermore, due to leakage, sheet resistance is up to 1.7× higher than without leakage for structures with low mass density and low contact quality, while for higher mass density, its impact is lower, at the sub- 1.2× level.