Computational modelling of fuel cells

Industrial applications of fuel cell technologies offer a major research opportunity. To realise the goal of practical commercial applications of this developing technology will involve the integration of the latest developments in electrical, fluid and chemical engineering systems. Although each field is well developed in its own right, the novel concept of combining these technologies within computational models offers the potential for new understanding and design procedures to be created. This multi-disciplinary approach would offer new insights into the fundamental physical phenomena, as well as creating tools that can be used for designing fuel cells and is the major focus of the research described in this thesis. Power storage fuel cells are the subject of this research. This particular fuel cell is a bipolar stack with two, single phase, liquid electrolytes. A review of the current techniques highlighted the need to address the study of the loss mechanisms present in the fuel cell. The prediction of short circuits created by the electrically conductive fluids required an electrical model. The power consumed driving the fluid system compelled a fluid dynamics model. These were combined within this work, to optimise the geometrical properties of this type of fuel cell based on general chemical kinetic and operating parameters. In order to satisfy the operating requirement of even distribution of reactants to the active areas, a finite-volume technique based on a computational fluid dynamics (CFD) code was developed. This enabled the prediction of the transport processes for fluid flow, electrical energy transfer and chemical species conversion throughout the fuel cell. The complexity of the multi-disciplinary CFD method required the use of parallel computers to reduce the lead times. The CFD approach is employed best studying small sections of the fuel cell system highlighting important phenomena. This research has already influenced design variations and has guided the direction of additional research, namely ways of electrically isolating the fluid without interrupting the flow. This work has successfully demonstrated that computational modelling can predict fuel cell electrochemical phenomena in a fraction of the time and cost of experimental procedures. This research is being used to benefit the design process and further the exploitation of fuel cell technologies. Finally, by extending the research into porous media and two phase flow these techniques could be applied to all types of fuel cell.