Author

Hayri Sezer

Date of Graduation

2014

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Mechanical and Aerospace Engineering

Committee Chair

Ismail B Celik

Committee Co-Chair

Harry Finklea

Committee Member

Kirk Gerdes

Committee Member

Ed Sabolsky

Committee Member

Nianqiang Nick Wu

Abstract

The Solid Oxide Fuel Cell (SOFC) is a good alternative for clean and efficient power generation. These cells can be operated directly on a wide variety of fuels including biogas, hydrocarbon fuels and synthesized coal gas (syngas), which is a promising avenue for utilization of coal with much less environmental impact. One of the challenges in this technology is poisoning of SOFC anodes by trace impurities contained in coal syngas. One such impurity, phosphine is known to cause catastrophic failure of SOFC anode even at <10ppm concentrations. Fuel impurity degradation patterns can vary by different operating conditions such as humidity, applied current, temperature and anode thickness.;In the present study, more detailed models are developed to predict the typical degradation behaviors observed in SOFC anode due to phosphine by extension of an in-house one-dimensional computational code. This model is first used to predict the effect of steam concentration on phosphine induced degradation in anode supported SOFCs. The model is refined based on the experimental observation, which indicate that the phosphine degradation is less severe in the absence of steam. Simulations results showed good agreement with experimental data. Then, a sensitivity analysis, using dual numbers automatic differentiation (DNAD) is performed to investigate the influence of empirical model parameters on model outputs, electrical potential, ohmic and polarization losses. Further, the refined one-dimensional model is extended to a three-dimensional model to study the phosphine induced performance degradation in relatively large planar cells operating on hydrogen fuel. The empirical model parameters are calibrated using button cell experiments and sensitivity analysis as a guide. These parameters are then used in planar cell simulations. The results from the three dimensional model show that the contaminant coverage of nickel and fuel distribution inside the anode is highly non-uniform. These non-uniform distributions are caused by the geometrical alignment of gas channels and current collectors, as well as the variation of gas concentration along the flow direction. The non-uniform deactivation of anode gave rise to the altering of current distribution inside the planar cell such that the cell can still produce current even when some regions of the anode are partially inactive. In addition, to assess the overall cell performance at any given degradation stage, additional simulations are performed to evaluate the electrochemical behavior (polarization and impedance) of the cell. The simulation results are assessed in comparison to experimental observation whenever possible. Finally, a physics based transport model for nickel migration is formulated based on experimentally observed elemental redistribution in a SOFC anode and it is integrated into refined one-dimensional phosphine degradation code. Simulations show that the proposed mechanism of Ni diffusion driven by secondary phase formation, the electrical force, and humidity can explain the experimentally observed accumulation of Ni and secondary phases on the SOFC anode surface.;Keywords: Keywords: Solid Oxide Fuel Cells, Contaminant Degradation, Automatic differentiation, Sensitivity Analysis, Impedance, Polarization, Nickel migration, Electro-migration..

Share

COinS