Experimental investigation of fuel cell dynamic response and control

Abstract

An experimental study of the dynamic response of a commercial fuel cell system is presented in this work. The primary goal of the research is an examination of the feasibility for using fuel cells in a load-following mode for vehicular applications, where load-following implies that the fuel cell system provides the power necessary for transient responses without the use of additional energy storage elements, such as batteries or super-capacitors. The dynamic response of fuel cell systems used in the load-following mode may have implications for safe and efficient operation of vehicles. To that end, a DC–DC converter was used to port the power output of the fuel cell to a resistive load using a pulse-width-modulating circuit. Frequency responses of the system were evaluated at a variety of DC offsets and AC amplitudes of the PWM duty cycle from 1 out to 400 Hz. Open-loop transient responses are then evaluated using transitions from 10% to 90% duty cycle levels, followed by dwells at the 90% level and then transitions back to the 10% level. A classical proportional–integral controller was then developed and used to close the loop around the system, with the result that the fuel cell system was driven to track the same transient. The controller was then used to drive the fuel cell system according to a reference power signal, which was a scaled-down copy of the simulated power output from an internal combustion engine powering a conventional automobile through the Federal Urban Driving Schedule (FUDS). The results showed that the fuel cell system is capable of tracking transient signals with sufficient fidelity such that it should be applicable for use in a load-following mode for vehicular applications. The results also highlight important issues that must be addressed in considering vehicular applications of fuel cells, such as the power conditioning circuit efficiency and the effect of stack heating on the system response.