Large scale wind farm cluster wake effects and loss prediction

The ambitious energy transition targets demand global expansion of the installed capacity of offshore wind. This will inevitably lead to wind farms being located very close to one another in clusters and potentially affecting each other. This study focuses on the efficiency of individual farms within the cluster and the downstream wake effects of the whole cluster. A model wind farm cluster based on the Hollandse-Kust West (HKW) site was chosen for this study, which consists of three wind farms with a total of 153 IEA15MW wind turbines. Modelling of the flow is carried out using both NREL’s high fidelity LES tool SOWFA and the TNO’s in-house mid-fidelity tool FarmFlow. The power outputs of turbines and wake velocity profiles at various locations within and downstream of the wind farm cluster are compared. Further, a design power loss parameter based on Nishino’s two-scale momentum theory is defined and used to quantify the efficiency of the individual wind farms and the cluster as a whole. Results indicate that FarmFlow provides a positive outlook on the yield and capacity factor while severely limiting wake recovery. SOWFA results show larger wake deficits near wake with faster wake recovery downstream, implying higher mixing. However, due to the lack of validation data for the investigated case it is unclear which model is accurately predicting reality. An internal investigation using 6 years of measured data has revealed that a smaller wind farm cluster has effects (5% wake deficit) up to 43km downstream, which is contrary to the findings from SOWFA. © 2024 by Kisorthman Vimalakanthan.