Cold Start of a Polymer-Electrolyte Fuel Cell II. Model Verification Using Parametric Studies

The cold-start model developed in Part I of this study is verified by comparing its predictions to a variety of experiments in the lit-erature. Using the model, the physical bases for trends observed in various parametric studies are explained. For example, it is shown that near but below 0C, liquid water flow out of the cathode catalyst layer enables the cell to operate for a significantly lon-ger time than it could if no liquid flow occurred. It is also shown that increasing the current density during start-up reduces the amount of pore space that can be utilized during cold start. On the other hand, decreasing the initial membrane water content or increasing the thickness of the cathode catalyst layer (cCL) increases the capacity of the cell to absorb product water. Finally, dur-ing a nonisothermal start the cell potential may remain constant or even decrease, even as the cell heats up. This is due to ice accu-mulation in the cCL. Once the ice in the cCL melts, allowing liquid-phase drainage, a steep increase in performance is observed