Medium Voltage Generation System with Five-level NPC Converters for Kite Tidal Power

Licentiate thesis, 2019

Offshore power generation has emerged as a prominent source of energy and the installed capacity of new plants has been steadily increasing in recent years. Tidal power specifically is a promising renewable energy source which has not been highly exploited yet, despite its distinctive advantages of being predictable and independent of weather conditions. The main objective of this Licentiate thesis is to analyze and propose solutions for two common problems in offshore power production, which are the power variations due to the non-steady speed profile of the water speed flowing through the turbine and the efficient transportation of the produced power to the shore.

The tidal power application utilized in this thesis is the subsea kite, which is a recently developed tidal energy conversion technology that can increase the generated power compared to a traditional static tidal turbine. A turbine is mounted on a submerged kite and the kite moves inside the sea following a predefined trajectory and generating electric power from the tidal currents. The speed and torque of the turbine varies periodically due to the periodic movement of the kite in the sea and, therefore, the control of the generator needs to be able to handle this variable generated power. The kite studied in this thesis has rated active power of 500 kW.

In the first part of the thesis, the power generation system of the subsea kite is modelled and the profile of the generated power is extracted given a specific tidal current and turbine geometry. The control of the power converters is described and tested for the specific profile of the generated power. The speed of the generator is controlled by a properly designed Maximum Power Point Tracking algorithm, which ensures that the generator extracts the maximum power from the tidal stream. Experimental verification of the model of this innovative system is also conducted on a laboratory 35 kVA emulator of the tidal power generator.

The second part of the thesis deals with the design of a medium voltage generator drive. The use of medium voltage in the power generation system is highly advantageous for the tidal kite application, since it can reduce the current flowing through the undersea cables connecting the tidal plant to the local grid. Therefore, the size of the cables can be reduced. The drive proposed here uses two 5-level Neutral Point Clamped (NPC) converters connected back-to-back. The 5-level NPC converters can operate with high voltage, while using multiple low-voltagerated power switches. Contrarily, the typical 2-Level converters have limited voltage capability, since they would require more expensive high-voltage-rated power switches. The increased operating voltage of the power conversion system results to lower current and losses in the cables. Another advantage of the NPC converter is the low harmonics at the ac side, which reduces the requirements for passive grid filters. However, the voltage balancing of the dc-link capacitors in this converter topology is a challenge which has not been effectively solved in previous studies. Therefore, a novel voltage balancing strategy is proposed here that uses advanced Space-Vector-Modulation techniques and hardware-based voltage balancing schemes with reduced number of components and lower power losses. Finally, a laboratory prototype of the NPC-converter-based power conversion system is developed with rated power 50 kVA. SiC MOSFETs are used on the
converters to further increase the system’s efficiency and voltage capability.

This thesis presents the model, control and laboratory emulator of a kite-based tidal power generator. The experimental set-up can be utilized for conducting research on other renewable sources, such as wind power, that have similar performance. Also, the developed multilevel drive is suitable for various applications where medium voltage grid-connected drives are used and particularly in distributed renewable power generation.