Supervisor:
Torsten Berning

Co-Supervisor:
Samuel Simon Araya and Jakob Hærvig

Assessment Committee:
Thomas Condra (Chair)
Dr. Steven Beale, Forschungszentrum Jülich, Germany, Queens University, Canada
Dr. Pourya Forooghi, Department of Mechanical & Production Engineering, Aarhus University, Denmark

Moderator:
Vincenzo Liso

Abstract:

The global shift toward a carbon-neutral energy system has underscored hydrogen’s pivotal role in enabling deep decarbonization across power generation, transportation, and industrial applications. Among various production pathways, alkaline water electrolysis (AWE) stands out as one of the most mature and cost-effective technologies for large-scale green hydrogen generation. Nevertheless, AWE faces persistent challenges—particularly in gas evolution, thermal regulation, and gas crossover—that are exacerbated under the dynamic load conditions typical of renewable energy integration.

This thesis presents a comprehensive computational and experimental investigation into the performance and optimization of AWE systems, centering on the development of a fully coupled, three-dimensional multiphase model. The model captures the complex interplay of electrochemical reactions, two-phase flow, ionic and electronic transport, heat transfer, and gas crossover phenomena addressing limitations in existing modeling approaches. ANSYS Fluent R1 2021 was used to develop a Eulerian-Eulerian multiphase approach. The model incorporates user-defined scalars and functions to accurately simulate gas–liquid interactions within porous electrodes and the electrolyte.

Parametric studies were conducted to evaluate the effects of key variables, including temperature, KOH concentration, flow rate, porosity, permeability, and bubble diameter. The model effectively replicates polarization behavior, elucidates the sensitivity of gas crossover to supersaturation and mass transfer kinetics, and identifies operational strategies for mitigating performance losses at low nominal power. A customized solution strategy featuring tuned under-relaxation factors and tailored initialization schemes was devised to ensure numerical stability and convergence.

Complementing the modeling efforts, experimental studies were performed using a lab-scale zero-gap AWE test cell. Electrochemical impedance spectroscopy and polarization curve measurements were used to assess the influence of electrode compression, temperature, and current collector geometry. Results highlighted the performance sensitivity to structural properties, particularly the beneficial effect of compressing fine nickel foams and the detrimental effects of compressing coarse foams.

Taken together, this work advances the state of multiphase modeling in alkaline electrolysis by introducing a predictive, physically grounded framework capable of guiding design and operational improvements. It further underscores the importance of standardized experimental methodologies to ensure reproducibility and data comparability across the alkaline electrolysis research community. The developed model lays a robust foundation for the future optimization of AWE systems supporting the scalable, efficient, and economically viable production of green hydrogen.