Performance prediction of Solid Oxide Cells (SOC) by ex-situ characterisation of electrodes and physical modelling

Achieving the full potential of hydrogen energy requires the use of highly efficient devices for its production and consumption such as Solid Oxide Cells (SOCs). In-situ and ex-situ characterisation techniques can be applied to differentiate effective designs from less efficient ones. In-situ methods assess cells during
operation, while ex-situ techniques analyse individual components. Complementing these techniques, physical modelling aids in understanding cell phenomena and predicting performance. However, models in the literature often require parameter tuning. The robustness of these models improves as more parameters are independently defined. Yet, destructive tests and advanced facilities can only determine some key morphological parameters. This study enhances the characterisation of SOCs. First, a comprehensive dataset of microstructures is generated by the Plurigaussian method, and their morphological parameters are evaluated. Next, a surrogate model is developed to estimate the triple phase boundary (TPB) density and phase-specific tortuosities (τ ) using easily measurable parameters, namely phase volume fractions (ε) and mean pore/particle radius (r_p). Finally, a physical model is employed to predict cell performance. Results indicate that the ion volume fraction significantly impacts the cell performance. Additionally, reducing
particle sizes, especially electron-conductive particles, enhances cell performance by increasing TPB density. For manufacturers, optimizing electrode design with finer electron-conductive particles and composition of 60% ion and 20% electron volume fractions can notably improve SOC performance in both fuel cell and electrolyser operational modes.