This thesis presents an individual tip control (IPC) system based on blade tip deflection measurements. The controller is based on novel sensor inputs which measure flapwise tip deflection distance at a high sampling rate. IPC plays a key role in reducing fatigue loads in wind turbine components. These fatigue loads are caused by differential loads such as wind shear, yaw misalignment and turbulence. The presented controller is implemented in HAWC2 and high fidelity load measurements are produced using the DTU10MW Reference Wind Turbine. Lifetime equivalent load reductions were seen in both rotating and fixed frame components under extreme turbulence, inverse shear conditions and in normal operating conditions. A novel implementation of IPC is also presented where the blade tips are guided along a fixed trajectory to maximise blade-tower clearance. The motivation of this implementation is to reduce the chance of blade-tower interactions for large diameter turbine rotors. The theoretical background used in this study is presented first along with details of controller discretisation methods. Details of the iterative control design process is presented, and the simulated fatigue loads are compared for a number of control architectures. Finally, the implementation of the tip trajectory tracking control is presented along with an analysis of the pitch rate limits and the effect of IPC on electrical power output.

This project is a collaboration between Goldwind, Technical University of Denmark (DTU) and Delft University of Technology (TUDelft), of which focusing in developing an optimization model that could determinate the optimal operation schedule for wind turbine under the uncertainty of the electricity price and limitation in life-time equivalent fatigue loads, for the purpose of improving and maximizing the revenue or cash flow for wind farm project.

The project comprises of two main parts, including the design and validation of up- and de-rated operation modes, and the model development and optimization of the revenue. The design of the different operation modes provide varies rated power without sacrificing the life-time structural integrity, as such offer extra flexibility in wind turbine operation schedule according to the fluctuation in electricity price. Life-time equivalent loads would be calculated for the different operation modes design and a weighting factor, that indicate the rate of fatigue consumption relatively to the normal operation, would be computed and used in the optimization process. An optimization model is developed based on the turbine structural constraints as well as the financial profile. The optimization model processes the electricity price data and other wind turbine data, for example the fatigue consumption, to reproduce the optimal operation schedule and operation price thresholds that would yield utmost financial benefits.

The optimizer has a proven performance improvement in revenue generation of 1.5% and 5% reduction in the project related interest payment under one of the case study under electricity price uncertainty. The model is effective in scheduling the turbine operation upon the varying electricity price. This optimization model is providing game changing insights and operation strategies that could be used throughout the turbine life-time under the ever changing electricity price of the energy market. And further drive down the cost of energy in the highly competitive market. By improvising continuous input to the optimizer including bending moment sensors inputs, the optimization process could reproduce a more accurate revenue improvement.

Material bend-twist coupling has been widely studied in the research community as a passive control mechanism for a wind turbine. However, there is a lack of research in incorporating it in the rotor design process, with little research being conducted for its application to small wind turbines. The present study focuses on this issue by including bend-twist coupling in the design of a 500W wind turbine by using a combination of parametric studies and a multidisciplinary constrained optimisation approach. By doing so, it aims to establish the effectiveness of bend-twist coupling as a tool for passive load alleviation in small wind turbines. The effectiveness is tested through obtaining a significant decrease in the flapwise blade root bending moment accompanied by only a marginal decrease in the AEP, when compared with the baseline uncoupled turbine. The reference blade is designed under limitations imposed by the rules of the Small Wind Turbine Design Contest. Bend-twist coupling is introduced in the blades with a fixed aerodynamic design. The rotor performance is analysed inHAWCStab2. The internal structure of the blade is created with the intention of producing flexible blades with a single composite material used throughout the blade. Carbon-epoxy and glass-epoxy FRPs are considered as the material to be chosen in the unidirectional laminae. The cross-sectional stiffness analysis is conducted using BECAS for a range of fibre layup angles in both carbon-fibre and glass-fibre blades. Carbon outperformed glass for all fibre angles with regard to the amount of coupling seen in the crosssections. Apart from flapwise bend-twist coupling, other secondary torsion couplings are also present. A load response study is carried out for varying positive fibre layup angles in both carbon-fibre and glass-fibre blades, under steady wind conditions for the operational wind speed range. An increase in the flapwise blade tip displacement and a reduction in the flapwise bending loads with increasing fibrelayup angles is observed for both glass-fibre and carbon-fibre blades. The HAWTOpt2 aero-structural design tool using OpenMDAO as its core is utilised to implement the optimisation. The spanwise fibre layup angle and laminate thickness distribution are the only design variables that were allowed to be manipulated by the optimiser. The objective function is comprised of two different individual objective functions weighed according to the situation. The optimisation cases failed due to inaccurate gradients of the objective function and constraints. Due to time constraints the manufacturing of the blade and, static load and wind tunnel testing were not carried out.