Experimental study on the dynamic characteristics of kW-scale molten carbonate fuel cell systems

The aim of this work is to develop dynamic models for two types of kW-scale molten carbonate fuel cell (MCFC) systems on the basis of experimental data. The dynamic models are represented as a 3×3 transfer function matrix for a multi-input and multi-output (MIMO) system with three inputs and three outputs. The three controlled variables which severely affect the stack performance and lifetime are the temperature difference in the stack and the pressure drop at the anode and the cathode. Three manipulated variables, namely, current load, fuel and oxidant utilization, are selected to keep the three controlled variables within their safety limits for the reliable operation and protection of the system in case of emergency. Each element in the transfer function matrix is in the form of a first-order model using a simple, unit step, response test during operation. The non-zero off-diagonal elements in the transfer function matrix show that some interactions exist among the operating variables, and two zeros show no interaction between fuel and oxidant flow without gas cross-over. The stability of both dynamic models is analyzed using the relative gain array (RGA) method. Large diagonal elements in the RGA matrix show that the pairing between the manipulated and controlled variables is appropriate. Proper pairing is also proven by the singular value analysis (SVA) method with a smaller singular value in each system.