In this paper, we review a dynamical power curve concept that was introduced in earlier publications and shown to provide results for a wind turbine's power characteristic with several benefits compared with the IEC 61400-12-1 standard procedure. After summarizing the theoretical concept based on the theory of Langevin processes and their reconstruction, we enlarge on a number of specific practical issues. Special attention is paid to the convergence or robustness of the reconstructed results, and their dependence on different settings for the data analysis scheme is studied. A key issue for the procedure that is investigated in this paper is the variability of the wind speed data that may be controlled by applying a specific data filter. It is seen that the necessity for filtering depends both on the time scales present in the wind data in relation to the wind turbine power dynamics and to some degree also on the correlation between the wind and the power signal.
The observations and findings altogether suggest that the dynamical performance characteristic, as it is considered here, is definitely a promising concept, particularly since it allows much flexibility in the handling of noisy performance data, but with some technical difficulties that demand a careful consideration in order to obtain reproducible and representative results. Copyright © 2014 John Wiley & Sons, Ltd.