Like any other flying system, Kitepower poses risks to third-parties on the ground. Therefore, in this thesis a safety risk assessment is performed for the system of Kitepower. The research question of this thesis is: Which third-party risks are posed by Kitepower operations and how high are these? Kitepower uses a kite attached to a fixed ground station via a tether to harvest energy. This tether is reeled out by performing crosswind manoeuvres with the kite. The tether is retracted in a situation where the kite is in a gliding configuration with different aerodynamic settings. Therefore, there is a net energy gain between these two modes.

Firstly, a pure hazard brainstorm is performed to identify the hazards of the Kitepower operations. With these hazards four safety-relevant scenarios are constructed. The identified safety-relevant scenarios are: the kite, KCU, and tether fly away; the kite and KCU fly away; the kite flies away; and the complete system slides over the ground. Only the first scenario is modelled in this thesis because the first scenario is the most likely scenario.

The first scenario is modelled by using a model framework that is developed to estimate the third- party risk posed by a drone-based parcel delivery service. This model is adapted to the situation in which the kite, KCU, and tether fly away. A flight path model is used from a preceding MSc thesis. The non-nominal descend model is developed by simulating the forces on the kite. These models have been integrated in a Monte Carlo simulation that simulates the tether failure and subsequent landing of the kite multiple times at the proposed test location in Rutten, the Netherlands.

In the results it is found that the individual risk is the most critical in terms of regulation limits. The operational zone should be enlarged to a radius of 500m in the dominant wind direction at location Rutten to comply with the individual risk limits. The system operations comply with the regulations in terms of collective ground risk. However, it is also observed that the population density and wind conditions have a large impact on the individual risk and collective ground risk. Therefore, these parameters should be adapted if the risk is assessed for new locations. Furthermore, the tether failure rate should be based on operational data and not on the Kitepower expert his assumptions. Lastly, the scenarios that are not analysed yet should be analysed in further research.