Significant progress has been made in the development of wave energy converters (WECs) during recent years, with prototypes and farms of WECs being installed in different parts of the world. With increasing sizes of individual WECs and farms, it becomes necessary to consider the impacts of connecting these to the electricity network and to investigate means by which these impacts may be mitigated. The time-varying and the unpredictable nature of the power generated from wave power farms supplemented by the weak networks to which most of these farms will be connected to, makes the question of integrating a large quantity of wave power to the network more challenging. The work reported here focuses on the fluctuations in the rms-voltage introduced by the connection of wave power farms. Two means to reduce these rms-voltage fluctuations are proposed. In the first method, the physical placement of the WECs within a farm is selected prior to the development of the farm to reduce the fluctuations in the net real power generated. It is shown that spacing the WECs or the line of WECs within a farm at a distance greater than half the peak wavelength and orienting the farm at 90◦ to the dominant wave direction produces a much smoother power output. The appropriateness of the following conclusions has been tested and proven for a wave power farm developed off the Outer Hebrides, using real wave field and network data. The second method uses intelligent reactive power control algorithms, which have already been tested with wind and hydro power systems, to reduce voltage fluctuations. The application of these intelligent control methods to a 6 MW wave power farm connected to a realistic UK distribution network verified that these approaches improve the voltage profile of the distribution network and help the connection of larger farms to the network, without any need for network management or upgrades. Using these control methods ensured the connection of the wave power farm to the network for longer than when the conventional control methods are used, which is economically beneficial for the wave power farm developer. The use of such intelligent voltage - reactive power (volt/VAr) control methods with the wave power farm significantly affects the operation of other onshore voltage control devices found prior to the connection of the farm. Thus, it is essential that the control of the farm and the onshore control devices are coordinated. A voltage estimation method, which uses a one-step-ahead demand predictor, is used to sense the voltage downstream of the substation at the bus where the farm is connected. The estimator uses only measurements made at the substation and historical demand data. The estimation method is applied to identify the operating mode of a wave power farm connected to a generic 11 kV distribution network in the UK from the upstream substation. The developed method introduced an additional level of control and can be used at rural substations to optimise the operation of the network, without any new addition of measuring devices or communication means.

The dissertation presents a generalized control structure for two-stage power converters operated in a renewable energy power system for smart grid and micro grid systems. The generalized control structure is based on the two-loop average-mode-control technique, and created by reconstructing the conventional control structure and feedback configuration. It is broadly used for both dc-dc and dc-ac power conversion based on the two-stage converter architecture, while offering several functionalities required for renewable energy power systems. The generalized control structure improves the performance and reliability of renewable energy power systems with multiple functionalities required for consistent and reliable distributed power sources in the applications of the smart grid and micro grid system. The dissertation also presents a new modeling approach based on a modification of the subsystem-integration approach. The approach provides continuous-time small-signal models for all of two-stage power converters in a unified way. As a result, a modeling procedure is significantly reduced by treating a two-stage power converter as a single-stage with current sinking or sourcing. The difficulty of linearization caused by time-varying state variables is avoided with the use of the quasi-steady state concept. The generalized control structure and modeling approach are demonstrated using the two-stage dc-dc and dc-ac power conversion systems. A battery energy storage system with a thermoelectric source and a grid-connected power system with a photovoltaic source are examined. The large-signal averaged model and small-signal model are developed for the two demonstrated examples, respectively. Based on the modeling results, the control loops are designed by using frequency domain analysis. Various simulations and experimental tests are carried out to verify the compensator designs and to evaluate the generalized control structure performance. From the simulation and experimental results, it is clearly seen that the generalized control structure improves the performance of a battery energy storage system due to the unified control concept. The unified control concept eliminates transient over-voltage or over-current, extra energy losses, power quality issues, and complicated decision processes for multiple-mode control. It is also seen that the generalized control structure improves the performance of a single-phase grid-connected system through increased voltage control loop bandwidth of the active ripple current reduction scheme. As a result of the increased loop bandwidth, the transient overshoot or undershoot of the dc-link voltage are significantly reduced during dynamic load changes. / Ph. D.