Beyond Common Practices in Experimental Characterisation and Model Validation of Solid Oxide Cells

Performance and reliability of state-of-the-art solid oxide cells (SOCs) need to be further improved so as to outperform, in terms of cost-effectiveness, the more traditional power and substance generating technologies, incentiving in this way their public acceptance and hence boosting market penetration. In order to achieve the aforementioned ameliorations, the use of more application-specific analysis tools, techniques and testing protocols is regarded as a crucial step. It must be noted that among these, computational modelling is especially gathering the attention of the scientific community, as it complements the more traditional analysis areas (experimental and theoretical), enabling to use hypothesis for extrapolation of their application. An exhaustive validation campaign is needed in order to ensure the soundness of the model and bounding its applicability. Yet, there is a growing feeling in the fuel cell modelling community that commonly used validation practices (basically confronting current-voltage curves) are not enough to assess the validity of the model. In order to tackle these issues, the High Temperature Fuel Cell Laboratory (HOTLAB) of ENEA has developed an in-house set-up for testing SOC single cells allowing to measure in-operando the gas composition and the temperatures in eleven spot-sampling points distributed throughout the negative electrode side of the cell. The innovative set-up has demonstrated to be an extraordinary analysis tool, mapping the distribution of temperatures and gas compositions when in stationary and transient conditions. Moreover, the geometrical characteristics of the housing enable characterising unambiguously the electrical and electrochemical performances of cell, obtaining current-voltage curves and EIS measurements barely affected by parasitic resistances. One one hand, the innovative set-up can be used to generate statistical multivariate models correlating localised conditions with global settings; and on the other hand it can be used to fully validate two or three dimensional CFD models. The synergy of both approaches can lead to more powerful monitoring methodologies and also to the generation of models capable of predicting degradation based on the operating conditions