Posters

Abstract

Wind farms reduce the speed of the wind upstream of them, which is known as wind farm blockage. Wind farm developers and owners have recently become aware of the significant impact this effect can have on estimates of wind farm production - with wind farm yields being over-estimated. Traditional wind energy prediction methods have not accounted for wind farm blockage, and there is currently no industry consensus on the best approach to represent this physical effect within engineering models. Branlard and Meyer Forsting have developed a novel engineering model, based on a vortex cylinder, which accounts for wind farm blockage effects and have provided source code which is publicly available. This engineering model displays remarkable accuracy at a low computational cost and has been used in the analysis. This engineering model is more complex than parametric approaches and takes into account a variety of factors, such as wake diffusion and expansion, turbulence, and the interaction between wakes from multiple turbines. The coupling of the vortex cylinder model with engineering wake models allows for a more accurate representation of the wind farm flow field than either model could do on its own. The vortex cylinder model captures the upstream induction effects, which are required for determining blockage effects while the engineering wake models capture the downstream wake effects. The upstream turbine's loading is affected by the induction from the other turbines, while the downstream turbine's loading is affected by the wake from the upstream turbines. Therefore, several iterations are run to ensure that the wind farm flow field and the turbines' loading are consistent. Wind farm blockage is a complex phenomenon that is influenced by a number of factors. Nevertheless, the publicly available information on the sensitivity of the blockage effect to these factors is very limited or inexistent. This poster aims to investigate how wind farm blockage varies with turbine spacing, and to share these findings with the industry to stimulate discussion. As the base case for the analysis, we have considered the NREL generic 10MW turbine, a hub height of 150m, air density of 1.225 kg/m3, power law shear exponent of 0.1 and turbulence of 8%. With these conditions, we have estimated the blockage effect for a regular gridded wind farm layout of 9x9 turbines with a spacing ranging from 3 to 12 rotor diameters. In order to use realistic wind conditions, we have used the wind speed distribution of the met mast Greater Gabbard, located offshore 23km to the coast of Suffolk England. Nevertheless, we have sifted the wind rose from the met mast to ensure that the predominant direction is perpendicular to the first row of the layout. The estimated blockage effect ranged from 2.0% to 0.2% for 4 and 12 rotor diameters spacing, respectively. This range agrees with the values that are overall accepted by the industry. Nevertheless, a more detailed validation against real measurements or actuator disc CFD would improve the confidence on these figures.