271
A 2D Stationary FEM Model for Hydrocarbon Fuelled SOFC Stack Layers

Thursday, 30 July 2015: 08:40
Lomond Auditorium (Scottish Exhibition and Conference Centre)
H. Geisler, S. Dierickx, A. Weber (IAM-WET, Karlsruher Institut für Technologie (KIT)), and E. Ivers-Tiffée (IAM-WET, Karlsruhe Institute of Technology (KIT))
In previous studies a model framework has been presented, where the physically coupled processes (i) charge-transfer chemistry, (ii) electron and ion conduction and (iii) gas species transport in the porous electrodes for hydrogen (H2) operation are implemented. The model framework has been parametrized with a dataset obtained by electrochemical impedance spectroscopy (EIS), 4-point DC conductivity measurements and evaluation of 3D reconstruction data extracted via FIB-tomography of an anode-supported cell fabricated at Forschungszentrum Jülich (FZJ) [1,2].

In this study, the established model framework is extended to predict local gas composition and electrochemical performance of hydrocarbon fuelled SOFC stacks. The model geometry represents a 2D longitudinal section through a stack layer along the gas channel length. Consequently, the existing isothermal model framework is extended to meet the new requirements: (i) the pressure driven gas transport in the gas channels is modeled by the Navier-Stokes equations and in the porous electrodes by Darcy’s law, (ii) all existing and new gas transport equations are extended to six gas species and (iii) global reaction equations for steam reforming and the water-gas shift reaction are added, as proposed by Ref. [3]. In a previous publication [4] it was shown, that only H2 is electrochemically converted in the Ni/8YSZ anode structure. Hence, the electrochemical charge transfer is modeled by Butler-Volmer approach parameterized for H2 conversion from the existing model framework. Carbon monoxide (CO) as electrochemical active species also present in hydrocarbon fuels is subsequently converted via the water-gas shift reaction at the catalytic active Ni-surface within the porous anode material. The reaction kinetics for these individual reforming reactions have been determined experimentally [5], thus giving the model the capability to predict C/V characteristics dependent on local gas conversion kinetics. This will be shown by a detailed validation where gas concentration profiles recorded via gas chromatography along the gas channel length at open circuit voltage (OCV) are compared with simulated profiles for corresponding gas mixtures (see figure 1). Simulated C/V characteristics are compared with measured data recorded for reformate fuel in a temperature range of
T = 680 - 880 °C and a fuel gas utilization between 0 - 69 % (see figure 2). A high level of agreement between simulated and measured data exists throughout the investigated parameter range. The model predicts (i) where gas diffusion limitations lead to drastic performance loss and (ii) where the Ni/8YSZ anode structure most likely will suffer severe damage.

References:

[1]  H. Geisler, A. Kromp, A. Weber, and E. Ivers-Tiffée, J. Electrochem. Soc., vol. 161 (2014), pp. F778–F788.

[2]  H. Geisler, A. Kromp, A. Weber, E. Ivers-Tiffée, Proceedings of the 11th European Solid Oxide Fuel Cell Forum, A0904 (2014).

[3]  J. R. Rostrup-Nielsen, Catalysis Science and Technology, 5th ed., vol. 5. (1984).

[4]  A. Kromp, A. Leonide, A. Weber, E. Ivers-Tiffée, J. Electrochem. Soc., vol. 158 (2011), pp. B980-B986. 

[5]  A. Kromp, A. Leonide, H. Timmermann, A. Weber, and E. Ivers-Tiffée, ECS Transactions, vol. 28 (2010), pp 205-215.