Numerical modeling and experimental validation of PEM electrolyzer

The transition from conventional fossil fuels to sustainable-clean fuels has become essential in minimizing the harmful environmental effects of greenhouse gases and in the fight against global warming. Hydrogen, abundant globally and with a high energy content, is characterized as a sustainable and environmentally friendly fuel when produced from renewable energy sources (RES). In this context, it can be said that hydrogen is promising in many sectors, such as transportation, industry, energy, aviation, and space. With the transition to renewable energy, the demand for electrolyzers that perform water electrolysis has increased significantly. Electrolysis of water is splitting water into oxygen and hydrogen gas through an electric current passing through the water. Water electrolyzers are electrochemical devices commonly used to generate hydrogen with a purity of at least 99.9% from pure water. Electrolyzers are classified according to the type of electrolyte used (solid or liquid). Liquid electrolyte is commonly employed in alkaline electrolyzers, while solid electrolyte is generally used in PEM electrolyzers and solid-oxide electrolyzers. Among the available technologies, Polymer Electrolyte Membrane (PEM) electrolyzers are known for their ability to operate at high current densities and lower specific energy consumption compared to other types of electrolysis cells. However, to benefit from the full advantage of the performance of the PEM electrolyzer, it is critical to improve its durability, reliability, and operational conditions. PEM electrolysis cell performance depends on many factors, such as electrolysis cell design and assembly, operating conditions, and bipolar plate design. The present thesis focuses on finding the critical operating conditions of the PEM electrolyzer that affect the system's performance for maximum current and hydrogen production rate. This thesis consists of 3 main parts: Modeling the complex two-phase flow in the anode bipolar plate of the electrolysis cell using Comsol Multiphysics and validating it with a reference study in the literature, performance tests of experimental electrolysis cell, determination of the optimum operational conditions of the electrolyzer by Taguchi and Response Surface Methodology (RSM), investigation of the mechanical behavior of the experimental cell using the finite element method in ANSYS Mechanical and, finally, validation of the simulation results with the experimental results. In the first part of the thesis, a three-dimensional, unstable numerical model of the anode bipolar plate of the PEM electrolyzer has been developed, validated with the experimental results of a PEM electrolyzer with parallel flow field from the literature, two-phase flow in the flow channels has been characterized, and finally, by changing the number of channels between the inlet and outlet manifolds, pressure, velocity, and the distributions of oxygen gas bubbles were simulated. Numerical results show that when the number of channels increased from N=4 to N=14, the pressure drop decreased, and the oxygen gas fraction increased. It was observed that the velocity in the inlet and outlet manifolds of the anode bipolar plate was higher than in the central region, and it was revealed that the flow velocity distribution was not uniform. The performance of PEM electrolysis cells can be significantly increased by optimizing operating and design conditions. In the experimental part of the study, firstly, the PEM electrolyzer cell with a 9 cm² active area was collected; after the conditioning process, which significantly affects the cell performance, performance tests were carried out in the temperature range of 40-80°C, and the results were recorded. The Taguchi method was applied to the experimental system to improve the PEM electrolyzer cell performance and determine the optimum operational conditions. The Taguchi method is an excellent optimization tool, significantly reducing experimentation time and costs. Temperature, water flow rate, and cell voltage are independent variables of the experimental system, while current and hydrogen flow rate are the dependent variables. Experimental results show the maximum cell performance at a temperature of 80°C, water flow rate of 16.5 ml/min, and cell voltage of 2.4 V. The efficiency values calculated based on hydrogen's lower and upper heating values under optimum operating conditions are 59.6% and 70.4%, respectively. The optimization results were visualized with three-dimensional surface and perturbation plots with a different optimization tool. Similarly, Design-Expert 13 (trial version) software was used to analyze the effects of control factors on performance and their optimum levels, and they were optimized by the RSM tool. RSM results show that the cell voltage, temperature, and interaction between the cell voltage and temperature significantly affect PEM electrolyzer cell performance. Mechanical degradation of the MEU and the gas diffusion layer (GDL) is one of the main problems affecting the lifespan of the PEM electrolysis cell. Any structural damage to the electrolyzer cell components will adversely affect the mass transfer process and cell performance. As the clamping force increases, cell performance increases, but when it exceeds the optimum value, GDL permeability decreases, and cell performance may decrease. This thesis analyzes the effect of the clamping torque used for the PEM electrolyzer assembly on the cell components. Water flow rate, clamping torque, and cell voltage, which can significantly affect cell performance, were defined as input factors of the experimental system and optimized with RSM. Statistical results show that the water flow rate is an essential factor in the hydrogen flow rate, and the cell voltage, torque, and cell voltage-torque interaction significantly affect the current. All performance tests have been performed at a temperature of 80°C. From the optimization results, the maximum current and hydrogen flow rates of the PEM electrolyzer were obtained as 23.821 A and 189.972 ml/min, respectively. The three-dimensional model geometry of the experimental PEM electrolyzer was developed in SpaceClaim software, and the material properties of each component were defined in the program. The mesh structure of the model was created in ANSYS Mechanical Mesh with the hex-dominant mesh method; then, the mathematical model was transferred to ANSYS Mechanical for numerical analysis. Three different clamping torques of 3, 6.5, and 10 Nm were applied to the PEM electrolyzer anode endplate, and the von Mises stresses and deformations generated on the MEU were numerically analyzed using the finite element method (FEM). Pressure-sensitive films were placed instead of the MEU, described as the PEM electrolyzer heart, and tested at three different values of clamping torques. Finally, numerical and experimental results were compared with each other, and it was seen that the results agreed with each other. The red intensity in the center of the pressure-sensitive films has been observed to be relatively low, and the bipolar plate channel design is less visible. This indicates that the pressure values are lower in the center of the pressure-sensitive film. When the clamping torques are 3, 6.5, and 10 Nm, respectively, the maximum stress values produced on the MEU are 34.891, 75.597, and 116.3 MPa, respectively.