COST OPTIMIZATION OF PEMFCs THROUGH NUMERICAL MODELING

摘要

Aspects of the technical viability of Proton Exchange Membrane Fuel Cells (PEMFCs) have been proven. However, significant further stack technology improvements are needed before commercialization can occur. In particular, the cost and reliability of PEMFC stacks must be improved. To achieve high performance and long reliable life PEMFC stacks still require relatively high platinum loadings. Platinum loadings are further increased if the fuel used contains contaminants such as CO. Current research programs address these issues through materials development (e.g., high temperature electrolyte membranes), newer catalyst designs (e.g., smaller catalyst particles to improve catalyst utilization, use of novel catalysts or platinum alloy catalysts), novel stack operating strategies (e.g., air bleed or anode potential pulsing to enhance CO tolerance). Given the non-linear response of the cell to material, design, or operating conditions changes, as well as the cost of experimentation, research programs that rely only on empirical assessment will prove both time-consuming and costly. The use of suitable modeling tools will enhance effectiveness of the research program because the models can help: 1. Support R&D portfolio management. Models can help phrase and answer what-if questions. E.g., what improvement in CO-tolerance or power density will 120℃ operation provide, relative to 70℃ operation? Models can also help in identifying R&D goals by establishing the limits of potential performance improvements. 2. Support experimental programs directly. Models can help define a crisp null-hypothesis and thus structure the design of an effective experimental matrix. In addition, models can be used to analyze and extrapolate data, and to gain an understanding of the overall impact of complex trade-offs that result from technology changes. In order to be effective, the models must be predictive as opposed to fits to existing data. Predictive models allow one to explore a wide range of the parameter space, especially conditions that are not easily accessible by experiments. PEMFC models contain several sub-models. For the overall model to be applicable over a wide range of conditions, each of these sub-models must also be valid over the range of conditions. We have developed models for estimating the anode and cathode kinetics that are valid over a wide range of conditions. The kinetic parameters in these models have physical significance, i.e., they are related to (he underlying chemistry, and are not merely fitting parameters. A PEMFC model incorporating these kinetic models helps answer what-if questions and helps assess the effectiveness of design/material changes. We have linked the PEMFC performance model to our cost model so as to assess cost implications of performance improvements resulting from design/material changes.