A three-dimensional numerical model of an anion exchange membrane fuel cell

Abstract

The global demand for alternative energy conversion procedures has increased considerably in the past few decades. As a result, increasing attention has been given to proton exchange membrane fuel cells as they offers several advantages over fossil fuel engines such as high efficiency, high power density and the absence of greenhouse gas emission to name a few. Polymer exchange membranes can be classified as either proton exchange membrane or anion exchange membrane (AEM). The latter has several distinct advantages including the possibility of precious metal free catalysts, enhanced oxygen reduction catalysis, and a wider option of fuels. Even though significant progress has been made on the experiment analysis of the AEM fuel cells, further improvement of the current numerical models is necessary in order to better describe the transportation phenomena within the catalyst layer and to enhance the accuracy of the numerical results. The numerical simulation presented in this thesis is performed applying the methodologies of computational fluid dynamics. A finite volume method-based methodology was used and conservation equations of mass, momentum, species, liquid water, membrane water content, electronic and ionic potential and energy were solved in a coupled manner. The SIMPLE algorithm was used to link velocity and pressure. Grid sensitivity and convergence criteria analysis were performed in order to ensure that the grid size does not significantly affect the solution and that the numerical solution is converged. In this thesis, firstly, a three-dimensional multiphase macro-homogeneous models is proposed which was subsequently used to evaluate the effects of operating temperature, inlet relative humidity, and flow direction (anode and cathode flowing in the same and opposite direction) on the overall performance of the fuel cell. Secondly, a three-dimensional multiphase agglomerate model for an AEM fuel cell is proposed and, in addition to the length scale present in the macro- homogeneous model (i.e. catalyst layer thickness), an additional length scale is introduced (i.e. ionomer thickness) to the numerical model. Subsequently, a direct comparison is made between the proposed agglomerate model and the previously developed macro-homogeneous model and a detailed discussion between both models is presented. Finally, further improvement on the agglomerate model is proposed to mimic the transport phenomena within the catalyst layer in a more realistic manner and an investigation of the effects of the catalyst layer composition (i.e. platinum and carbon loading, ionomer loading), structural parameters (i.e. catalyst layer thickness, porosity) and operating parameters is performed.