Friday, 28 July 2017: 11:40
Atlantic Ballroom 1/2 (The Diplomat Beach Resort)
Modeling electrochemical reactions in solid oxide fuel cells (SOFCs) is a difficult and contentious topic. Recently, detailed reduced order reaction models have been presented which take into account more detailed physics than the widely used Butler-Volmer (BV) type equations. For this study, a reduced order oxygen reduction reaction (ORR) model previously developed specifically for yttria stabilized zirconia (YSZ)/lanthanum strontium manganite (LSM) cathodes is considered. The most prominent electrochemical pathways have been identified as the triple phase boundary (YSZ/LSM/pore) or surface pathway and the double phase boundary (YSZ/LSM) or bulk pathway. The ORR model considered predicts a competition between which pathway is most prominent depending on local overpotential, reactions rates and species concentrations (particularly oxygen partial pressure). However, this model and those similar to it require separate species transport equations to be solved for numerous species and surface coverages which result in a very stiff system of non-linear equations. Such a system becomes cumbersome numerically especially when trying to perform complex simulations such as impedance analyses, in comparison to BV type models. On the other hand, the traditional BV type equations are not able to distinguish between the reactions at the triple phase boundaries (TPBs) and the double phase boundaries (DPBs). In the study, the relationships between the total current, over potential, oxygen partial pressure, temperature and particularly TPB and DPB are explored based on the previously developed ORR model and experimental data from literature. Using these relationships, a new model is proposed for calculating the exchange current density as part of a BV type model which takes into account both TPB and DPB pathways as well as oxygen partial pressure. The model is then applied to situations of different TPB and DPB densities and reaction rates (exchange current densities) and compared to similar studies using the ORR model in order to verify that the model correctly handles the relationship between the two current pathways. It was found that the proposed BV model produces results comparable to the ORR model all with computational time for similar analyses being approximately 10,000 times shorter (due to decreased stiffness). The results obtained from the BV model are encouraging and indicate that a similar model can be developed for the anode side and different cathode materials, thus resulting in a robust modeling and analysis tool for SOFCs. This would amount to huge time and cost savings especially for simulations of large planar cells and cell stacks.