Impact of wind turbine controller strategy on deloaded operation

In the past, wind turbines were only required to provide limited ancillary services and they could inject as much power as they wanted in the grid. Nowadays the impact of wind turbines is no longer negligible. Therefore, many system operators have put in place grid codes which require renewable energy sources support the power system and provide ancillary services. These new ancillary services have for common need that wind turbines are able to decrease their power output on demand. In this paper, two controller strategies for a variable speed and variable pitch wind turbine are compared. In the first controller strategy, the rotor speed is regulated by acting on the pitch angle while the produced power is regulated by acting on the electromagnetic torque. For the second controller strategy, the rotor speed is regulated by acting on the electromagnetic torque while the produced power is regulated by acting on the pitch angle. Both strategies are assessed in an application where the power production set point of the turbine is lower than the available wind power, a situation known as deloaded operation. The core of both controller architectures is a C_p table inversion procedure allowing pitch response linearization by an adjustment of the process gain depending on the wind speed but also with respect to the power set point itself. Simulation results illustrate controller response differences for the two methods when reacting to changes of power production set point. The methods are also compared for a change of power production set point with a slope of 0.05 p.u./s. Simulations show the superiority of regulating the rotor speed by acting on the pitch angle while the produced power is regulated by acting on the electromagnetic torque. With this controller strategy, rotor speed swings and pitch angle overshoot are greatly reduced. However, when following a slow change of power production set point as the one tested, strictly with respect to following the pow- er production set point, the two controller strategy show suitable responses.