Modelling and Simulation Analysis of a PEM Water Electrolyser for Hydrogen Production

Hydrogen energy is a renewable energy that can substitute fossil fuels in a wide application range. One of the efforts is producing hydrogen gas using a system called a Polymer Electrolyte Membrane (PEM) Electrolyser. Several efforts have been made in designing and improving a PEM electrolyser system to achieve optimum efficiency. This paper aims to develop a computational model for analysing the performance of a PEM water electrolyser, focusing on voltage losses, hydrogen production efficiency, and system optimization. This performance can be improved by analysing the effect of electrode potential on polarisation based on electrochemical and thermodynamics concepts. A MATLAB coding simulation was used in this paper to analyse the performance using combinations of Fick's Law and Darcy's Law. The first model, Model A, specifies the number of voltage losses in a PEM electrolyser with an open-circuit voltage and the three overpotentials of activation, ohmic and diffusion. Meanwhile, the second model, Model B, is known as the number of voltage and hydrogen losses due to convection and diffusion. The performance of the PEM electrolyser can be influenced by a variety of factors. This performance assessment focuses on the following relative humidity on the anode side, pressure and temperature. The results showed that Model B has a lower operating voltage than Model A, which only considered reversible voltage and voltage losses. The findings highlight the crucial role of anode relative humidity, where higher humidity lowers the operating voltage. Meanwhile, higher cathode pressure increases hydrogen crossover, raising the operating voltage but improving voltage efficiency. Model B accurately predicted output across various current densities, proving its reliability. This study underscores the importance of modelling in PEM electrolyser performance analysis. Future work should explore mass transport effects on hydrogen production using these models.