Discussion and Analysis of Flue Gas Utilization in a Phosphoric Acid Fuel Cell Engine During Idle Operation

Fuel cell technology has in recent years won the favor of all major car manufacturers as a likely future replacement of the internal combustion engine. This has been driven by the potential for high efficiency and low emissions. Still, fuel cell engines must overcome major hurdles before being introduced into the market. One such hurdle is systems integration; in particular, with fuel cell engines that do not use hydrogen as their primary fuel. In these engines, fuels, such as methanol, are employed instead of hydrogen because of their high-energy density and ease of storage. However, these benefits are counterbalanced by the need to reform these fuels on-board the fuel cell vehicle, thus substantially increasing the complexity of the fuel cell engine. Through the course of operating a $30ft$ methanol-fueled phosphoric acid fuel cell bus, repeated overheating of the steam-reformer catalyst bed at low engine power outputs was noted. For the purpose of better understanding these overheating events, relevant data was obtained from the bus. Bus operation at low-power outputs was found to have low hydrogen consumption (⁠$∼40%$ to $∼60%$ of the stack’s incoming hydrogen) due to low electrical demands and reformate flow rate constraints. In the bus, the anode flue gas is burned in the reformer burner, which provides the heat required for the endothermic methanol steam reforming reaction (see Fig. 1 later). At low engine power, the flue gas is the reformer burner’s only fuel, but the energy carried by the flue gas is much greater than the energy required for the reforming reaction and consequently causes reformer overheating at low power. This case serves as an example of how appropriate fuel cell system integration is crucial in order to harness fuel cell benefits. To study such system integration problems, a steady-state model of the fuel cell bus engine was developed. Using this model the bus reformer overheating problem was understood and several possible improvements for utilization of the flue gas were conceited.