Proton-Exchange Membrane Fuel Cell Stack Model Parameters Estimation and Model Validation

Abstract

A proton-exchange membrane fuel cell (PEMFC) is a promising, clean, and efficient power device that converts the chemical energy stored in a fuel into electricity directly. Hydrogen is supplied from an external tank and it is oxidized at the anode of the fuel cell, this activity releases electrons, which are made to flow through an external circuit. Oxygen is supplied at the cathode where it is reduced by the electrons flowing in the external circuit. The growing popularity of using PEMFC stacks in stationary, portable, and transportation applications is driving researchers to develop dynamic models to accurately capture the electrical characteristics and runtime performance. These characteristics are critical when integrating a fuel cell stack with power conditioning units (PCUs), and power electronics onboard fuel cell powered systems. Conventional PEMFC models may be complex and substantial effort is required to estimate parameters of such models. This work establishes a simple equivalent electrical circuit model of a PEMFC stack that captures the voltage-current (V-I) runtime characteristics under different load, and under different hydrogen flow rate conditions. It proposes a well-known equivalent circuit model of a battery, to be modified and used, as a model for a PEMFC stack. The existing adaptive parameters estimation (APE) technique is used to estimate the proposed PEMFC stack’s model parameters. Also, the concept of the state of charge (SoC) of a battery is thought as analogous to the amount of hydrogen available in a PEMFC stack’s supply tank. In this work, an existing battery model is modified to model the electrical performance of a 200-W PEMFC stack. All the model parameters are estimated using the APE technique that is known to require few experiments. The model is validated experimentally under different load conditions for the 200-W PEMFC stack. In addition, the model is validated with a different fuel cell stack of a smaller size, i.e. a 30-W PEMFC stack. All results show reasonably accurate terminal voltage estimation with error in the order of millivolts, and 95.84% of the all samples of estimated terminal voltage have between ± 0.1% error compared to the actual terminal voltage.