Non-Solvent Induced Phase Separation (NIPS) is a widespread technique for manufacturing hydrophobic Polyvinylidene Fluoride (PVDF) membranes. Despite extensive research, the process remains highly sensitive to minor parameter variations, necessitating experimental empirical validation for any industrial adjustment. This thesis investigates two strategic modifications to the conventional GVS S.p.A. industrial process: transitioning to a new polymer supplier and implementing a sustainable green solvent. In the first project, two Designs of Experiment (DoE) were conducted on different polymer grades from a new supplier. Predictive models were established to forecast critical performance metrics, specifically water breakthrough pressure and air flow. These models successfully identified formulations that replicate the performance of current industrial standards.