A wind tunnel method for screening the interaction between wind turbines in planned wind farms Mats Sandberg 1,* , Hans Wigö 1 , Leif Claesson 1 , Mathias Cehlin 1 1 University of Gävle, Gävle, Sweden * Corresponding author. Tel: +46 26648139, Fax: +46 648181, E-mail: msg@HiG.se Abstract: The energy captured by wind farms is reduced if there is an interaction between the individual turbines. In the paper a novel method for studying the interaction between wind turbines is presented. It is based on recording the static pressure on ground in a wind tunnel provided with wind turbine models. The assumption is that the pressure distribution at ground reflects the pressure distribution at hub height. The pressure distribution at hub height is a result of the flow in the vicinity of the turbine. The pressure at ground is recorded with a pressure plate provided with 400 pressure taps. The wind turbine model is a porous disk giving a non rotating wake. At first the pressure response to one wind turbine is recorded. This is the reference case giving the characteristics of a non disturbed wind turbine. Its region of influence can therefore be determined. This provides important information on how to avoid any interaction between turbines. A nearby turbine should not be placed within the region of influence. In the paper we show how the pressure response varies with different distances between two turbines. The agreement between the static pressure on ground and at hub height has been tested by recording the static pressure at hub height with a small Prandtl tube. Keywords: Wind farms, Wind tunnel, Pressure distribution on ground, Pressure distribution at hub height, Region of influence 1. Introduction To benefit of economies of scale, wind turbines are arranged in wind farms. As wind energy capacity expands, the size of individual wind farms continues to increase requiring tens to hundreds of wind turbines typically arranged in a large array. While organizing wind turbines in wind farms help in reducing the cost of energy production they introduce another problem, aerodynamic interaction. One of the research challenges is to accurately model interactions between the individual turbines to accurately predict power output before wind farm construction. In a wind farm with N turbines the wind farm efficiency is defined as N E E T A A i = η (1) Where E A is the annual energy captured by the wind farm and E T is the annual energy captured by one isolated turbine. If there were no aerodynamic interaction the energy captured by the wind farm would be NE T and the efficiency would subsequently be 100 %. Figure 1 shows a wind turbine with rotor diameter D. Array losses can be reduced by optimization of the downwind spacing and the cross wind spacing. Rotor Sweep area D Hub height Fig. 1. Definition of some concepts 4090