With the rapid growing aviation industry and the high amount of produced greenhouse gases, the search in more sustainable propulsion methods continues. A possible solution is the use of hydrogen fuel cells, significantly reducing the production of NOx and CO2. Although several studies are being performed for the performance and applications of the fuel cell, for which the Airbus ZEROe program is one of them, the additional air supply system lacks research. Due to the specific operating conditions of the fuel cell and the large difference between ambient conditions, the air supply system should be analysed with care as it can require a lot of energy over an entire flight. This thesis focuses on developing a model capable of analysing the performance of the air supply system, using the Airbus ZEROe fuel cell as a reference. Furthermore, different architectures have been compared in order to analyse the effect of different components. A model has been developed capable of sizing and analysing the performance for different air supply architectures. The compressor, turbine, and heat exchangers have beenmodelled such that they provide a physical representation of what is happening inside the air supply system. Furthermore, as the model requires geometrical parameters as an input, it can be easily adapted to optimise these components for different design conditions. Therefore, the model has been used to optimise four different architectures, consisting of a variation between single or double stage compressors and the inclusion of a recuperator. The optimisation has been performed for several design conditions, analysing not only the effect for different flight conditions but also the altitude, fuel cell pressure, and air mass flow. For the optimisation, it has been found that the architecture resulting in the lowest power required has a single stage compressor and a recuperator, having a power required of 77.5 kW for cruise. However, the addition of the recuperator adds a significant amount of mass, as the total air supply system mass is estimated around 315 kg. Compared to an architecture without the recuperator (168 kg) this is a significant difference. However, the power required for this architecture is significantly higher, with a value of 100 kW. Overall it was found that although adding mass, the recuperator significantly reduces the power required. Next to the recuperator, the addition of a second stage compressor has been analysed as well. Although adding mass, it has been seen that for the higher mass flow and pressure ratio cases, such as top of climb, the double stage compressor performs better than the single stage compressor, although this difference is not as large as for the addition of the recuperator. Apart from the optimisation, the off-design performance has been analysed as well for the four different architectures, taking the cruise-, takeoffand top of climb optimised architectures as a base. It was found that the off-design performance for the different architectures is mostly limited by the choice in design point. The main component that was affected the most by this is the turbine, as a turbine bypass valve is required. The turbine bypass valve regulates the mass flow passing through the turbine, controlling the expansion ratio for a fixed shaft speed. It has been observed that the valve is required for architectures that are optimised for high pressure ratios and low mass flows, as the valve was seen opened for lower pressure ratios and higher mass flows. Although the bypass valve is required, it was observed that the performance of these architectures was always better than an architecture that did not require the turbine bypass valve, such as those optimised for takeoff. Additionally, limiting operating conditions have been identified through off-design analysis. The most critical case for sizing the heat exchangers is the hot day scenario at the top of climb. Due to the high pressure ratio and the resulting high compressor outlet temperature, adequate cooling is required to achieve the necessary fuel cell temperature. It was found that, for other optimised points, architectures with a recuperator face more challenges in this regard compared to those without a recuperator, mainly because of their smaller liquid heat exchangers. Other identified limiting conditions include low mass flow operating points for the compressor, lower-than-design expansion ratios, and higher mass flows for the turbine. These conditions necessitate the addition of bypass valves to ensure proper operation.