A High Resolution Multi-Physics Numerical Model for Planar SOFC Stack

Abstract

An SOFC stack is the basic functioning unit of SOFC as an electrical power generation device. The physical condition of an operating SOFC stack is vastly different from that of a laboratory button cell. The ability to analyze the performance characteristics of SOFC stacks is immensely helpful for advancing the SOFC technology for large scale commercial applications. Here we report a successful development of a high space resolution multi-physics mathematical model for full scale planar SOFC stacks which is, to the best of our knowledge, the first of its kind. This 3D stack model takes into account the coupled physical processes of mass, charge and heat transport, chemical and electrochemical reactions, applicable for both methane and hydrogen based fuels. The computational domain contains all components of the stack, including fuel and air channels, internal manifolds, cathode-electrolyte-anode assemblies, interconnects, seals and frames. The computational model is constructed using the commercial CFD software, FLUENT®, supplemented with in-house developed external subroutines to ensure all governing equations and physical properties are correctly coupled and solved successfully. Boundary conditions for the stack and that between material components are all properly incorporated into the computational model. To illustrate, Figure 1 shows a geometric model for a 30-cell stack with a counter- or co-flow design. The size of the stack is 136mm 143mm 133mm. The manifold consists of two/three gas inlet-outlet channels for each air or fuel flow. Every fuel cell consists of 30 channel-rib pitches, i.e., there are 900 repeating units for the electrochemically active area of the 30-cell stack. To properly resolve all the stack components involved, the computational model consists of a total of about 27 million grids. Multi-physics numerical simulations of this truly high resolution computational model of 30-cell stack have been successfully performed in a 2-CPU/16-core PC. Detailed information about the distributions of physical quantities is revealed, e.g., concentrations of chemical species, temperature profile, current density and electrical potential distributions, chemical and electrochemical reaction rates, etc. To summarize, the goal of performing detailed multi-physics simulations for realistic planar SOFC stacks has been achieved. Performance simulation with this high resolution, high computing efficiency numerical model can do much to advance the SOFC technology by improving the stack design and optimizing the operating parameters. Figure 1