UC Berkeley

Polymer-electrolyte fuel cells (PEFCs) are electrochemical devices that create electricity by consuming hydrogen and oxygen, forming water and heat as byproducts. PEFCs have been proposed for use in applications that may require start-up in environments with temperatures below 0 degrees C. Doing so requires that the cell heat up, and when its own waste heat is used to do so, the process is referred to here as ``cold start.'' However, at low temperatures the cell's product water freezes, and if the temperature does not rise fast enough, the accumulation of ice in the cathode catalyst layer (cCL) can reduce cell performance significantly, extending the time required to heat up. In addition to reducing performance during cold start, under some conditions the accumulation of ice can lead to irreversible structural degradation of the cCL.

The objective of this dissertation is to construct and verify a cold-start model for a single PEFC, use it to improve understanding of cold-start behavior, and to demonstrate how this understanding can lead to better start protocols and material properties. The macrohomogeneous model that has been developed to meet the objective is two-dimensional, transient, and nonisothermal. A key differentiating feature is the inclusion of water in all four of the possible phases: ice, liquid, gas, and membrane. In order to predict water content in the ice, liquid, and gas phases that are present in the porous media, the thermodynamics of phase equilibrium are revisited, and a method for relating phase pressures to water content in each of these phases is developed.

Verification of the model is performed by comparing model predictions for cell behavior during parametric studies to measured values taken from various sources. In most cases, good agreement is observed between the model and the experiments. Results from the simulations are used to explain the trends that are observed.

The verified cold-start model is deployed to determine a cold-start protocol and cCL properties that enable better performance. Criteria include not only minimizing start time but also exposure to high cCL ice pressures and cold-start energy while at the same time maximizing power available from the cell during the cold-start process.