Higher-order-statistics for characterization of a wind turbine wake
Assosiate professor at HVL, Jan Bartl, will present on the use of higher order statistics when characteizing a wind turbine wake the 1st of October.
Abstract: An in-depth understanding of the wake flow behind a wind turbine is essential for effectively planning a wind farm. Therein, the first and second order statistics of the flow properties in the wake are a crucial input for models predicting the power output and fatigue loads on downstream turbines. However, the reduction of the wake flow to mean (first order) and turbulent (second order) statistics might not give the full picture or a wind turbine wake's flow properties.
The analysis of wind tunnel data recorded in the wake behind a model rotor shows that the probability density function (PDF) is not fully Gaussian at all locations in the wake. Specifically, an annual region in the periphery around the wake's mean velocity deficit shows a heavy-tailed PDF, indicating a highly intermittent flow in this area. Consequently, the inclusion of higher order statistics such as skewness ond kurtosis in the analysis of a wake flow thus results in a significantly wider wake than established wake models would suggest. As the highly intermittent flow in the area around the wake is likely to affect the loads in downstream rotors, the present results suggest also taking into account higher-order-statistics for future wake models.
The seminar will be held from 12:15 to 13:00 (we open the doors at 12:00) in "Fyrrommet" (entrance through the library) at Høyskulen på Vestlandet (HVL). It is free and open to all interested parties, so bring your lunch and join us for an interesting presentation followed by a Q&A session.