"This research focuses on studying the feasibility of tethered undersea kite (TUSK) systems for power generation. Underwater tethered kite systems consist of a rigid wing that moves in a circular or figure-8 path below the surface. The tether can connect to a platform mounted either on the surface or anchored to the seafloor. On the kite is a turbine that extracts energy from the kite’s forward motion, which has the potential to be several times the current velocity. This speed multiplication combined with the density increase of water as opposed to air is one of the main benefits of this class of systems over wind turbines. A scale-model TUSK kite was designed. Testing was conducted in a water flume at Alden Research Labs (ARL). Model scale factors were determined from a real world prototype TUSK system currently in commercial development. The scale-model kite was primarily constructed out of ABS plastic using 3D printing rapid prototyping methods. Other components of the system were either repurposed from prior projects or constructed with traditional methods. Testing was conducted at current speeds of 0.15 m/s, 0.31 m/s, and 0.46 m/s; kite pitch angles of 80?, 85?, and 90?; and over circular and figure 8 trajectory shapes. Data collected included the azimuth and declination angles of the rigid tether as well as the power output of the generator on board the kite. Filtering techniques were employed on the data to generate graphs of kite position, velocity, and output for analysis. Relationships between current velocity, kite velocity, kite pitch angle, and power output have been measured. Inaccuracies in the model and areas for improvement in future work have been identified."

This thesis has its starting point in an existing computer model of an electricity generating kite, from Heidelberg University. The modelled kite has an area of 500 m2 and is tethered to a generator at sea. A control unit steers the kite in an optimised trajectory. The design and trajectory that maximise mean power output per loop had been found using the optimisation software MUSCOD-II. Firstly, the model is investigated in order to find possible adjustments to make it closer to reality. Then a method to take the economic aspect into account in the optimisation has been developed. The most important findings in the model survey concerned wind speed. The original model overrated the wind speed at high altitudes and it used a mean wind speed instead of including yearly variations. Adjustments are made and a new objective function aiming at maximising the yearly average power output per invested Euro is used. Furthermore, the revised model has a preset wind speed range within which the kite can operate, and a maximum power output of the generator (the nominal power) which is found through optimisation with respect of cost. Cable strength and other production limitations are included as well. Using cost estimations for relevant parts, the revised optimisation model results in a system with a tethering cable about half the original length, and a steadier power output over the loop. The yearly production sums up to 16.8 GWh, as compared to the original model which would have given 42.9 GWh yearly.