Degradation, Efficiency, and Equilibrium of a Dead-Ended Anode Fuel Cell.

Abstract

The dead-ended anode (DEA) operation of a proton exchange membrane (PEM) fuel cell is a promising solution for applications with mild power density requirements. The hydrogen recirculation hardware can be simplified while maintaining high fuel utilization. The simplified system architecture for DEA operation reduces the cost. However, nitrogen and water accumulate in the anode, leading to decreasing voltage in galvanostatic operation due to local hydrogen depletion or starvation. Anode purging is thus necessary to release the accumulated nitrogen and water, and recover the voltage.

The thesis aims to optimize a DEA fuel cell by overcoming the disadvantages while maintaining the benefits. One of the major issues with local fuel starvation in DEA operation is the corrosion of the carbon that supports the platinum in the cathode catalyst layer, which dramatically reduces the durability. An along-channel and transient model has been developed to predict the carbon corrosion and associate irreversible voltage degradation in DEA operation. The carbon corrosion rate and voltage degradation were identified quantitatively after model tuning. Simulation results suggest that purge interval and cycle duration affect the spatiotemporal distribution of anode species and, therefore, the carbon corrosion rate in the opposite cathode; consequently, a model-based optimization of these two design variables were performed to achieve high lifetime efficiency of a DEA cell. There are three interrelated objectives in this optimization: the hydrogen loss during the purge, the average voltage output between the purges, and the voltage degradation due to the carbon corrosion. The simulation results show that the durability concern with DEA operation can be reduced when a systematic engineering optimization is performed. The highest lifetime efficiency is achieved with medium cycle duration and short purge interval. Finally, the focus was turned to DEA operation without purging, in which system equilibrium is observed under certain operating conditions. The criteria for achieving such a nitrogen-blanketing based equilibrium with reasonable power output were analyzed by solving the reduced-order model numerically and comparing with the full-order model simulation. The results suggest another way of operating a DEA cell with minimum requirements on power regulation and purge optimization.