Thermal and electrical experimental characterisation of a 1 kW PEM fuel cell stack

Abstract

The present work describes the experimental characterisation of a self-humidified 1 kW PEM fuel stack with 24 cells. A test bench was prepared and used to operate a PEMFC stack, and several parameters, such as the temperature, pressure, stoichiometry, current and voltage of each cell, were monitored with a LabView platform to obtain a complete thermal and electrical characterisation. The stack was operated in the constant resistance load regime, in dead-end mode (with periodic releases of hydrogen), with 30% relative humidity air and with temperature control from a cooling water circuit. The need to operate the stack for significant periods of time to achieve repeatable performance behaviour was observed, as was the advantage of using some recuperation techniques to improve electrical energy production. At low temperatures, the individual cell voltage measurements show lower values for the cells nearer to the cooling channels. The performance of the fuel cell stack decreases at operating temperatures above 40 °C. The stack showed the best performance and stability at 30 °C, with 300 mbar of hydrogen and 500 mbar of air pressure. The optimised hydrogen purge interval was 15 s, and the most favourable air stoichiometry was 2. Between 15 A and 32 A, the maximum electrical efficiency was 40%, and the thermal energy recovery in the cooling system was 40.8%; these values are based on the HHV. Electrical efficiencies above 40% were obtained between 10 and 55 A. The variation in the electrical efficiency is explained by the variation in the following independently measured factors: the fuel utilisation coefficient and the faradic and voltage efficiencies. The deviation between the product of the factors and the measured electrical efficiency is below 0.5%. Measurements were taken to identify all the losses from the fuel cell stack; namely, the energy balance to the cooling water, which is the main portion. The other quantified losses by order of importance are the purged hydrogen and the latent and sensible heat losses from the cathode exhaust. The heat losses to the environment were also estimated based on the measured stack surface temperature. The sum of all the losses and the electrical output has a closure error below 2% except at the highest and lowest loads.