While generating power by an SOFC, reactants hydrogen and oxygen are consumed; simultaneously, hydrogen is diluted with the product water-vapor. Namely, concentrations of the reactants and product vary over the electrochemical active area along the respective flow fields. The chemical potentials in the anode and cathode therefore change along the flow fields, giving rise to the reversible Nernst-loss. When the concentration variations along the anode flow field are too large, i.e., the oxygen pressure is much higher than that of hydrogen, which is likely to occur in the downstream provided that the air is supplied at sufficient rates, nickel particles, the conventional catalysts in the anode, tend to re-oxidize. As a result, the length of the three-phase boundary would shorten, limiting the electrochemical performance. Besides, the anode microstructure would expand due to the larger volume of nickel-oxide, resulting in stresses and micro-cracks [1]. To prevent the nickel re-oxidation, concentration variations are desired to be identified and mitigated. Spatial concentration variations give also rise to spatial current and temperature variations. Current variations result in performance degradation, reducing the electric efficiency of the power generation. Given that the overpotentials are released as the waste heat, temperature variations develop in relation with the involving heat transport processes, e.g., convective and radiant heat transfer processes. The temperature variations induce thermal stresses into the cell components, and they affect the current variations through the overpotentials as well. It was shown that the concentration and temperature variations couple in the counter-flow configuration, resulting in larger variations in comparison with the co-flow configuration [2]. Spatial current and temperature variations are hence of great importance from both the energy conversion efficiency and mechanical durability aspects.

Spatial characterization of concentration, current and temperature variations is rather challenging. The high operation temperature (773-1273 K) of SOFCs makes the spatial characterization more difficult. Vibrational Raman Spectroscopy [3] and IR Thermography [4] can be employed for diagnosing the spatial concentration and temperature, respectively; however, both of them are quite expensive and they require transparent materials for the gas distribution plate. Although the segmentation method is easy to implement on tubular-SOFCs [2], it is quite laborious to apply on planar-SOFCs [5]. These challenges can be circumvented by numerical tools. In principle, numerical tools are obliged to be verified by benchmark experimental data for assuring the reliability of investigations. For verifying SOFC models, we need to consider in situ measurable properties, such as voltage, current, and temperature. Among these properties, I-V (current-voltage) validation and temperature validation appear to be the most practical options, which are to ensure the computation-reliability of concentration. Even though the conventional I-V curves provide a good basis for the model-validation, they may not ensure the accurate computation of the spatial variations. It is a fact that an I-V validated model might predict a number of distinct temperature fields depending on the incorporated heat transfer processes. Thereby, the computation-accuracy of the electrochemical performance is expected to be highly affected by the inaccurate temperature fields. This study is hence devoted to investigating the role of temperature variations on the reliability of the numerical tools for computing the associated properties. Herein we present the spatial variations in the characteristic properties of a microtubular-SOFC, firstly calculated by the model validated with only the conventional I-V curve, and secondly by the model verified with temperature variations, in addition to validating with the conventional I-V curve. For these evaluations, we exploit the experimentally and numerically obtained spatial current and temperature variations in a microtubublar-SOFC. We in situ acquired the experimental data by applying the segmentation method on a microtubular-SOFC, whereas we computed the numerical data by a two-dimensional model developed for the respective experimental conditions.

References

[1] A. Faes et al. Membranes, vol.2, yy.2012, pp.585

[2] Ö.Aydın et al., J. Electrochem. Soc., vol.163, yy.2016, pp.F216

[3] G. Schiller et al., Appl. Phys. B, vol.111, yy.2013, pp.29

[4] Y. Takahashi et al., Solid State Ionics, vol.225, yy.2012, pp.113

[5] P. Metzger et al., vol.177, yy.2006, pp.2045