Evaluating fluid flow and thermal effects for fuel cell humidity sensor design

Modern fuel cells require high relative humidity (RH) of about 90% of the reactant gases hydrogen and oxygen/air in a temperature range of 70°C-90°C for optimum efficiency. Especially RH must be kept within tight tolerances, as condensed water would reduce effective area of the fuel cell electrodes, while lower humidity would dry out the fuel cell's membrane and lead to permanent damage. Humidity sensors should enable effective measurement and control of the reactants' relative humidity. However, under given harsh conditions, measuring RH poses several problems for standard, water adsorption based, polymer sensors. At high RH, measurements tend to be inaccurate and the polymer's permittivity properties tend to degrade. Furthermore, if the sensor is not well thermally isolated against the usually lower temperature of the environment, water will preferably condensate on the sensing element. Solutions to this problem require detailed modeling of heat transport by a pipe flow around a sensor socket embedded in a heat conductive and convective environment, evaluation of several constructive designs with respect to thermal sinks and inherent measurement errors. One approach investigated in this study was heating the sensor permanently, this way reducing RH and thus avoiding water condensation locally, but also introducing measurement errors which must be corrected. In this study, heat flow and humidity distribution around the sensor socket and at the sensors active area were analyzed using FE simulations coupled with additional equations for calculating RH in this harsh environment. These models allow investigating effective RH at the sensor element as function of parameters like inflow temperature, velocity and relative humidity, pipe wall temperature and sensor heating for several design variants of sensor placement in the flow as well as thermal isolation.