| With the recent commercialization of the molten carbonate fuel cell, new power systems are being proposed and designed for various applications. These applications include stationary dispersed power generation as well as marine ship service power and propulsion. Carbonate fuel cell systems are relatively new and are still being field-tested. Before these systems can be controlled, it is essential to understand their dynamic behavior and operation. This thesis work starts from fundamental principles and develops simplified, lumped-parameter dynamic models useful for analysis and design of controllers for carbonate fuel cell power systems. Two types of systems are modeled, analyzed, and controlled. The first is a utility-scale demonstration project, the Santa Clara Demonstration Project, which served as a precursor to a commercial design for dispersed power generation. The second system is based on a conceptual design proposed by FuelCell Energy for both the U.S. Navy and the U.S. Coast Guard, for ship service power and primary ship propulsion. Both systems are built around the Direct FuelCell(TM), a type of carbonate fuel cell, which is modeled in this thesis. To justify our simplified approach, transient validation of the fuel cell model is provided, comparing simulation results against experimental results from a laboratory test unit. A reduced-order fuel cell stack model is then derived. We develop balance-of-plant dynamic models for both systems, which are validated, overall, with steady-state data obtained from flowsheet analysis. A final goal of this work is to show that the Santa Clara Demonstration Project can perform adequately, under control, during cycling load operation and that the Ship Service Fuel Cell can perform adequately, under control, to instantaneous changes in load demand power. |
