Experimental Test Bench Showcasing an X in the Loop Strategy for Fuel Cell Testing

In the quest for reduced development times and cost of fuel cell systems for industrial applications, we identified two major issues. First, the electrochemical behaviour of fuel cell systems is inherently difficult to predict. Second, testing fuel cell systems is resource intensive. These issues compound: Setting up an accurate model of a fuel cell system incurs long testing periods and does not guarantee acceptable results outside the tested parameters or for other membrane electrode assembly compositions. Our proposal to mitigate these two major issues is the use of an X in the Loop concept. Essentially, this is the direct integration of the test sample, here a single fuel cell, into the modelling environment of the whole system. In practice, we have defined two strategies with different levels of integration. Both assume a required power profile is given. The first strategy consists of deducing the operating parameters of the sample cell from a fuel cell system model in real time and setting them to the sample cell in the test bench. The response of the test sample can be used to validate the designed system or update parameters of the fuel cell model . The second strategy is a step further towards the integration of the test sample into the model. Here, the core model of the fuel cell inside the fuel cell system model is completely replaced with the test sample, reducing the need for an accurate electrochemical model while improving the prediction of the behaviour of the fuel cell system. On the way to realising our goal, we have built an experimental test bench capable of showcasing working prototypes of the proposed operating strategies.