The paper describes wind tunnel calibrations of cup anemometers and classification of cup anemometers based on wind tunnel measurements of their tilt angular response and rotor torque characteristics. The results are used for power-performance and wind-resource estimates, and all measurements applied follow the procedures specified in IEC 61400-12-1 and are carried out in wind tunnels being accredited and Measnet approved. 

Cup anemometer aerodynamics and wind tunnel aerodynamics are considered and used to evaluate the uncertainties involved in cup anemometer calibrations. The IEC procedure for cup anemometer calibrations specifies flow requirements of the empty working section, e.g. requirements concerning flow homogeneity and upper limit of flow turbulence, and tunnel size in form of maximum blockage ratios. The paper documents that deviation between fullscale and model-scale air turbulence and the interference effects between tunnel boundaries and rotating rotors of cup anemometers including their mounting systems may introduce a significant bias on the calibration results. This, however, is not taken into consideration by the specifications of the IEC procedure.

The importance of distance from cup anemometer rotor to tunnel boundaries is illustrated for both open and closed tunnels. It is shown that a certain minimum distance from rotor to tunnel boundary is crucial for avoiding significant contributions from these interference effects. If this distance is too small the tunnel boundary may be sucked towards the rotor in open tunnels or a friction force may be generated in closed tunnels. The suction towards the rotor in open tunnels may lead to accelerated flow in the central part of the cross section, and thereby increased rotation rate of the rotor. The friction force in closed tunnels will decrease the rotation rate of the rotor. Thus, interference effects as the ones described above will systematically bias the calibration results obtained significantly for small wind tunnels with a short distance from rotor to tunnel boundaries.

The present IEC procedure is based on a number of assumptions, which may lead to deviating calibration results. We recommend that the ongoing revision of the IEC procedure takes the new findings into consideration and that the effects shall be quantified by carrying out reference calibrations in very large wind tunnels with negligible contributions from blockage and tunnel boundary interference effects. In this way the updated procedure will reduce calibration uncertainties to a minimum and thereby minimize the uncertainties inherent in wind resources estimated for a specific site.