Investigation of idling instabilities in wind turbine simulations

More Info
expand_more

Abstract

Idling instabilities of large wind turbines in extreme winds is a relatively well known problem for manufacturers, at least in terms of the results in calculations. However, many are still sceptical of how realistic such instabilities are in actuality due to limitations in the current state of the art aeroelastic tools.
The overall aim of this thesis is twofold. Firstly, through the use of analytical solutions for a parked blade developed by adapting existing formulations for the aerodynamic damping forces as well as numerical means in the form of the aeroelastic tool Focus 6, Phatas, versions 9083 and 10041 (kindly provided by WMC), this thesis aims to determine how parameters relating to the local angle of attack of a wind turbine blade contribute towards inducing idling instabilities in large wind turbines and to subsequently identify the underlying mechanisms that drive such instabilities to occur in simulations. Subsequently, from the results gathered from the analyses, the second goal is to then determine if the current state of the art aeroelastic tools are able to reasonably capture such instabilities. The reference wind turbine used in this thesis was the NREL 5mW turbine.
The method used to obtain the results in the analytical solutions were first validated against existing literature and generally showed good agreement in the trends. The analytical solutions showed the aerodynamic damping values of a blade section at various nacelle yaw, rotor azimuth as well as blade pitch angles. The results from the analytical solutions showed that under certain conditions, idling instabilities were present.
The results of the numerical simulations when first run with quasi steady aerodynamics and a constant wind profile showed good agreement with the analytical results (and also existing literature) and also suggested that the instabilities were driven by the edgewise modes of the tubine. Assessment of the outputs after the inclusion of dynamic stall (where simulations were largely limited to angles of attack between -45° and 45° as this was the range where the dynamic stall model was applied in its full extent) as well as turbulence to further simulations showed that Focus 6, Phatas still calculated instabilities at settings close to the analytical solutions’ predictions.
As a final conclusion, this thesis provided a theoretical basis to argue for the case of idling instabilities being a possible issue. The numerical simulations largely support the findings of the analytical solutions where an assessment of the numerical results gave no indications to suggest otherwise.