Parameter study of electric power production in wind farms - experiments using two model scale wind turbines

Abstract

Wind farms are widely developed even if several unsolved problems need to be faced. The rotor-wake interaction involves different physical phenomena, not yet fully understood, directly affecting the overall wind farm power production. Numerical models and engineering rules have always been used to design wind farm layout but a spread between power predictions and results is verified. In this context wind energy research assumes a "back to basic" approach, by means of wind tunnel experiments, under controlled conditions, that represent the method to calibrate and correct the theoretical simulation models. The aim of this project is to provide a useful set of wind tunnel measurements focused on the wake-rotor interaction analysis and on wind farm power output optimization. A benchmark is obtained, arranging a two-turbine wind farm, in order to calibrate numerical models and to show a wind farm case study. Two three-blade wind turbine models are used in the present study. Despite some small geometrical differences, they are both equipped with the same blade-set, based on the NREL S826 airfoil, and they have a rotor diameter of D = 0.9 m. A characterization concerning power performances and wake features of both turbines is performed, in order to obtain reference values for array efficiency assessment. The used reference velocity is set to U_ref = 11.5 m/s. Afterwards, the two models are arranged inline building a two-turbine wind farm case. Different tests are carried out varying several parameters: the separation distance between the models (3D, 5D and 9D), the inflow condition (low and high turbulence background level) and both turbines tip speed ratios. First turbine wake measurements reveal that the velocity deficit recovery and the radial expansion of the wake are dependent on the flow turbulence. Higher the turbulence, faster the velocity recovery and bigger the expansion. As a consequence, high turbulence flows allow an earlier transition from near to far wake. Turbulence generation is analysed and related to the rotor operating point. The array parametric study points out that the overall efficiency increases by moving further downstream the second turbine, rising the background turbulence level and by choosing a suitable operating point for each turbine. The analysis suggest to obtain the maximum wind farm efficiency by an accurate management of these different parameters: a strong reliance on downstream distance and on turbulence level is confirmed, smaller variations are found depending on the turbines operating point, but the relevance is still essential.