Elucidating the dynamic transport phenomena of solid oxide fuel cells according to rapid electrical load change operation

Abstract

Solid oxide fuel cell (SOFC) is attracting attention as a next-generation eco-friendly power generation technology due to its advantages such as high efficiency and fuel flexibility. However, when exposed to electrical load changes, its performance deteriorates severely, which is an obstacle to commercialization. To solve this problem, many researchers study the phenomena that occur when SOFC is exposed to a dynamic load change, but the results reported in the literature seemingly conflict with each other. Some researchers report that the time it takes for SOFC to stabilize after a dynamic load change is within the 10-1 s time scale (i.e., delayed response in the short-time scale), while others report that it takes the 102 s time scale to stabilize (i.e., delayed response in the long-time scale). This study aims to provide an integrated understanding of these conflicting reports through three-dimensional dynamic simulation. Simulation results confirm that the relaxation time scale of both the short-time scale and the long-time scale are not mutually exclusive but compatible according to the observed time scale. The dynamic response of SOFC on the short-time scale is determined by gas-phase mass transfer in the anode, and the response on the long-time scale is governed by heat transfer in the entire stack. The response of fuel concentration causes the delayed response of anode concentration overpotential on the short-time scale, and the response of temperature causes the delayed response of anode activation overpotential and ohmic overpotential on the long-time scale. As a result, the cell voltage shows delayed responses simultaneously in short- and long-time scales, respectively, due to the difference in response time between these overpotentials. To confirm the effect of transfer phenomena, additional simulations are conducted by changing the geometric parameters of the stack. It is confirmed that the dynamic responses of fuel concentration and anode concentration overpotential in the short-time scale are sensitive to the thickness of anode components, and the dynamic responses of temperature, anode activation overpotential, and ohmic overpotential at the long-time scale are sensitive to the entire stack length.