Modification of Results From Computational-Fluid-Dynamics Simulations of Single-Cell Solid-Oxide Fuel Cells to Estimate Multicell Stack Performance

Abstract

A typical single-cell fuel cell is capable of producing less than 1 V of direct current. Therefore, to produce the voltages required in most industrial applications, many individual fuel cells must typically be stacked together and connected electrically in series. Computational fluid dynamics (CFD) can be helpful to predict fuel-cell performance before a cell is actually built and tested. However, to perform a CFD simulation using a three-dimensional model of an entire fuel-cell stack can require a considerable amount of time and multiprocessor computing capability that may not be available to the designer. To eliminate the need to model an entire multicell assembly, a study was conducted to determine the incremental effect on fuel-cell performance of adding individual solid-oxide fuel cells (SOFCs) to a CFD model of a multicell stack. As part of this process, a series of simulations was conducted to establish a CFD-nodal density that would not only produce reasonably accurate results but could also be used to create and analyze the relatively large models of the multicell stacks. Full three-dimensional CFD models were then created of a single-cell SOFC and of SOFC stacks containing two, three, four, five, and six cells. Values of the voltages produced when operating with various current densities, together with temperature distributions, were generated for each of these CFD models. By comparing the results from each of the simulations, adjustment factors were developed to permit single-cell CFD results to be modified to estimate the performance of stacks containing multiple fuel cells. The use of these factors could enable fuel-cell designers to predict multicell stack performance using a CFD model of only a single cell.