Evaluating high power density, direct-ammonia SOFC stacks for decarbonizing heavy-duty transportation applications

Ammonia is a carbon-free energy carrier and is anticipated to play a significant role in the global energy system of the future. This work is the first of a two-part paper that uses a multi-scale modeling approach to analyze a hybrid solid oxide fuel cell-gas turbine system concept that operates on NH3 fuel. The solid oxide fuel cell power module is based on a high-power density Ni-GDC/GDC/SSC cell architecture that is calibrated to experimental button cell data for the intermediate operating temperature range of 500-675 °C. Ammonia decomposition on the surface of the catalytically active Ni-particles in the anode is modeled by an elementary kinetic mechanism whereas a physically based, distributed charge transfer model accounts for the mixed-conducting behavior of the ceria-based electrolyte. Cell performance reduction when scaling from button-cells to full-scale stacks is accounted for in the modeling approach which encompasses pressurized 3-D stack simulations. The simulations disclose that the GDC-based cell design exhibits a significant performance gap between H2- and direct NH3-operation; in particular, at operating temperatures below 600 °C, which can be attributed to sluggish decomposition kinetics and mass transport resistances specific to NH3-operation. However, the simulations suggest that pressurization offers the prospective to significantly boost the cell power density by more than 55% at 0.7 V when switching from 1 to 10 atm, and the need for an external ammonia cracker can be eliminated. Operating the GDC-based cells with ammonia fuel at low temperatures also minimizes the leakage current across the electrolyte. As a prerequisite for mobile applications, the model highlights the principal feasibility of the stack design to simultaneously reach high power densities and efficiencies, yet at the same time, the simulations disclose the need for precisely controlling stack parameters on an integrated system level due to the high sensitivity of stack performance to varying operating conditions.