Mathematical modeling of solid oxide fuel cells using hydrocarbon fuels

Abstract

Solid oxide fuel cells (SOFCs) are high efficiency conversion devices that use hydrogen or light hydrocarbon (HC) fuels in stationary applications to produce quiet and clean power. While successful, HC-fueled SOFCs face several challenges, the most significant being performance degradation due to carbon deposition and the need of external reforming when using heavier HC. Modeling these devices faces these as well as other complexities such as the presence of multiple electrochemistry pathways including those of H2 and CO. The goals of this thesis are to: (1) improve the thermodynamic analysis of carbon deposition, (2) develop a multistep CO electrochemistry mechanism, and (3) apply the CO along with the H2 electrochemistry mechanisms to predict the cell performance when using syngas. Two carbon deposition mechanisms have been identified: homogeneously formed soot and catalytically grown carbon fiber. All previous thermodynamic analyses have used graphite to represent the properties of the deposited carbon regardless of the formation mechanism. However, the energetic and entropic properties of these two types of carbon are different from those of graphite. A new thermodynamic analysis is proposed that: (1) uses experimentally measured data for carbon fiber if the anode includes Ni catalyst; and (2) uses soot precursors such as CH3 and C2H2 to predict soot formation. The new approach improves the prediction of the onset of carbon deposition where previous analyses failed. A new multi-step CO electrochemistry model is proposed in which CO is directly involved in the charge-transfer steps. The model structure, with a single set of kinetic parameters at each temperature, succeeds in reproducing the characteristics of the EIS data of patterned anodes including the inductive loop at high activation overpotential. The model successfully predicts the steady-state Tafel plots, and explains the positive dependence of the exchange current density on Pco2 - Finally, a membrane-electrode-assembly (MEA) model is developed incorporating multispecies transport through the porous structure, detailed elementary heterogeneous reactions on the Ni surface, and for the first time, detailed electrochemistry models for H2 and CO. The model successfully reproduces the performance of SOFCs using pure H2 or CO. The MEA model can isolate/distinguish between the roles/contributions of the reforming chemistry and CO electrochemistry in SOFCs using syngas. Adding reforming thermochemistry improves the agreement with experiments at lower current densities, and raises the limiting current density by providing more H2 via the water-gas shift reaction. Adding CO electrochemistry improves the prediction at high current densities by the additional current generated by the CO electrochemical oxidation. The current from CO becomes comparable to that from H2 as the CO content at the TPB increases.