Ali Bayat, Prodip K. Das, Suvash C. Saha
Green hydrogen production via proton exchange membrane water electrolyzers (PEMWEs) is a key pathway for decarbonizing industrial energy systems; however, predictive modeling accuracy is strongly influenced by membrane-level transport phenomena. In particular, hydrogen and oxygen crossover through the polymer electrolyte and electro-osmotic water drag can reduce gas purity, lower Faradaic efficiency, and introduce safety-related constraints, yet these effects are commonly neglected in single-phase numerical frameworks. In this study, a validated three-dimensional single-phase PEMWE model is extended to explicitly incorporate solution–diffusion-based gas crossover and electro-osmotic drag within the governing transport equations. Simulations performed on a representative PEMWE geometry quantify the resulting impacts on polarization behavior, membrane hydration, effective proton conductivity, and outlet gas composition under standard operating conditions. The results show that cross-membrane transport introduces a measurable voltage penalty (up to O(10 −2 ) V) and induces significant dilution of product gases, accompanied by redistribution of water content within the membrane–electrode assembly. These findings demonstrate that even under single-phase operation, membrane transport phenomena play a non-negligible role in shaping performance, efficiency, and gas purity. The presented framework provides a physically consistent basis for capturing these effects and offers a quantitative bridge between simplified single-phase models and more comprehensive multiphase PEMWE simulations, supporting improved design and safer operation of electrolyzer systems. • 3-D PEMWE model explicitly includes hydrogen and oxygen crossover. • Electro-osmotic water drag is coupled into membrane transport. • Crossover and drag reshape local hydration and conductivity fields. • Gas purity and voltage impacts are quantified vs. membrane thickness. • Framework links simplified single-phase and advanced multiphase models.