In a large wind farm, the downstream turbines are inevitably affected by the upwind turbine wakes, which leads to decreased velocity and increased fatigue load for the downwind turbines. So accurately predicting the wake deficits behind the turbines is critical, especially for calculating wind farm power output. The commonly used approach is to combine the single analytical wake model with the superposition model. However, the conventional superposition models are based on some simple superposition assumptions without considering the energy exchange between the ambient atmosphere and the wake region, which leads to the faster velocity recovery. In this study, the correction coefficients are introduced into the classical energy balance superposition model to consider the faster velocity recovery caused by higher turbulence intensity in the wake overlapped area. An empirical formula is proposed to calculate the correction coefficients that are the functions of the flow distances between the upwind turbines and the target turbines for a given wind direction. The measured data from the Horns Rev I offshore wind farm is used to evaluate the performance of the advanced energy balance wake superposition model. The results reveal that the predictions of the advanced energy balance superposition model fit better with the measured data in the narrow wind direction range, comparing with other conventional superposition models. What's more, in order to study the universality of the advanced energy balance model, the new model combined with a number of single analytical wake models are also tested. It is found that the approach of combining the Park wake model has the higher calculation accuracy compared with the other wake models.
In a large wind farm, the downwind turbines are inevitably located in the region of upwind turbine wakes. And it will cause the velocity loss and the increase of turbulence intensity in the wake region which is called wake effects. The power loss of downstream wind turbines due to the wake effects is up to 40% (Hanse et al., 2012; Barthelmie et al., 2003; Adaramola et al., 2011), and up to 80% increased fatigue load (Van Binh et al., 2008; Sanderse, 2009). Therefore, it is necessary to study the wind turbine wake effects in large wind farms.
