◀ Back to projects overview P201101-016-TUD

 

Integrated design of far large offshore wind farm

The project investigates the design of two‐bladed wind turbines from a control engineering perspective. Two‐bladed wind turbines potentially offer an acceleration in the reduction of the cost of wind energy. Estimates predict a 10‐40% cost reduction of two‐bladed wind turbines compared to current three‐bladed wind turbines. The main reductions are achieved by saving the material of one blade, lower drivetrain torque due to a higher rotational speed yielding a lighter drivetrain and nacelle, and several installation benefits.

An important aspect of wind turbines is the control system, i.e., the controllers make sure that, among others, the wind turbine maximizes the power extraction from the wind and keeps the wind turbine loads as low as possible to guarantee its life cycle. However, whereas control for three‐bladed wind turbines has been extensively studied, only very little is done for control of two‐bladed wind turbines. This project has the aim to fill that gap. Moreover, the project considers the simultaneous design of controllers and physical parameters.

In general, the project can be categorized into three major topics:

• Linear Individual Pitch Control. The size of wind turbines has vastly increased over the past years. As a consequence,  the differences of the loads, due to for instance wind shear and tower shadow, across the rotor have increased. These loads need to be taken into account during the design of the wind turbine to ensure the turbine life cycle. Active control plays an important role in this aspect because with Individual Pitch Control (IPC) the asymmetric rotor loads can be largely accounted for. With IPC the blades are individually rotated (pitched) along their longitudinal axis according to the rotor position and the measured blade loads. Theoretical and practical studies have indicated that IPC is able to greatly reduce the blade loads and thereby other turbine loads and is therefore increasingly being implemented in state‐of‐the‐art wind turbines.
  
  The conventional approach of IPC on three‐bladed wind turbines involves a nonlinear mathematical transformation, which relates the measured blade root moments to the tilt and yaw moment of the rotor. Individual blade pitch setpoints are obtained by regulating the yaw and tilt moments and subsequently by passing the obtained signals through a reverse transformation. The advantage of using the nonlinear transformation is that the design of the controllers for the tilt and yaw moment are linear time invariant (a type of controller which is well‐known within the control engineering community). For IPC applied to two‐ bladed wind turbines, it was found that the nonlinear transformation could be replaced by a simpler and linear transformation. The advantage of applying this linear transformation is that only a single (linear time invariant) controller needs to be designed if one wants to remove the 1P blade loads (which is common). Moreover, at most two control loops are required to potentially reduce all blade load harmonics (1P, 2P, 3P, etc.). In comparison, with the conventional nonlinear transformation one requires two controllers for every blade load harmonic that one wants to mitigate. Finally, a true multivariable control design of power and load control can be obtained using the new transformation.
  
  Throughout the project an IPC strategy was developed which makes use of the linear transformation and is called Linear Individual Pitch Control (LIPC). The performance of LIPC has been evaluated in simulation studies, a wind tunnel experiment, and a measurement campaign on the NREL CART2. These studies have indicated that the load reduction performance of the LIPC strategy and the conventional IPC strategy are very similar. The measurement campaign has also indicated that LIPC can be readily applied to the next generation of two‐bladed wind turbines. 

• Integrated yaw design. A significant part of the project involved a case study with the project partner 2‐B Energy. In this part the previous described study, which focused on load reduction only, is extended with an integrated design of several controllers and a physical parameter of the turbine.  For the case study, the wind turbine developed by 2‐B Energy is used. This turbine has a number of features which are predicted to significantly reduce the cost of energy. For this project the most interesting feature is the downwind damped free‐ yaw system. For such a (damped) free‐yaw wind turbine, the rotor naturally tracks the wind direction. In order to improve the power output, the wind direction can be actively tracked. To this end, IPC can be exploited to create yaw moments around the tower and thereby track the wind.
  
  In the case study with 2‐B Energy, the choice for the amount of yaw damping provided by the yaw system is investigated as well as the yaw controller itself. The main objective of the case study was to reduce the loads on the tower. In the case study it was found that the yaw damping influences the tuning of the IPC controllers and that the objectives of the yaw misalignment control and load reduction controllers are conflicting. Hence, an integrated design approach of the yaw damping value and the yaw controllers was carried out. The case study also investigated the use of IPC to add additional damping to the yaw motion of the rotor‐nacelle assembly. It was shown that by doing so, additional load reductions were achieved in the tower torsion moment, while the other turbine loads remained rather similar. Hence, this control loop could therefore be a valuable addition to the control system of a (damped) free‐yaw wind turbine. The case study resulted on average in 5‐10% fatigue load reduction of the tower torsional loads.

• Integrated design framework. The third and final part of the project is the development of a new framework for simultaneously optimizing controller parameters and the parameters of physical systems. The proposed framework is mainly based on a result from control theory and in this context exploits the generalized Nyquist stability criterion (which is a well‐known theorem within the control community that tells whether the closed‐loop system of a controlled system is stable). With a slight modification of the latter framework, it can also be used for system identification purposes.  
  
  The framework has been successfully applied to design a LIPC controller for the NREL CART2 wind turbine. Furthermore, simulation examples have demonstrated that the method can indeed be used to simultaneously optimize controller parameters and physical system parameters. Finally, in the case study with 2‐B Energy, the framework has been used to identify the parameters of a mathematical yaw model for the state‐of‐the‐art wind turbine by 2‐B Energy.

Concluding, the project contributes on different levels to the development of (control design for) two‐ bladed wind turbines. The new linear IPC strategy reduces the number of required control loops and allows for a true multivariable design approach of the collective pitch controller, the individual pitch controller, and fore‐aft control. Moreover, the integrated design framework potentially proposes a new framework to reduce costs by combining structural and controller design. In the case study with 2‐B Energy, where several of the previous topics were applied, the design of controllers and structure was considered. On average tower torsional fatigue load reductions of 5‐10% over the baseline were achieved. With these load reductions, the final cost reduction of the project can be summarized to reduce the costs of the 2‐B Energy 2B6 support structure by 0.15%. With the 0.15% cost reduction of the support structure, the directly accountable and justifiable overall project contribution amounts to a 0.01% decrease of the cost of energy.