Normalized-wake characterization by long-range lidars: Analysis of lidar mounting errors

Slide presentation given by Priscila M.S. Orozco at Wind Energy Science Conference (WESC) 2021, mini-symposium "Lidars and floating wind energy – Collaboration of Innovative Training Networks LIKE and FLOAWER (I)". This presentation addresses the methods and results of an analysis conducted within the scope of Mrs. Orozco's PhD project, which is inserted in the LIdar Knowledge Europe (LIKE) project H2020-MSCA-ITN-2019, funded by the European Union (Grant no. 858358). Mrs. Orozco's PhD project has been carried out at UL International GmbH (Oldenburg, DE), under the supervision of Dr. Juan José Trujillo, in collaboration with ForWind Center for Wind Energy Research at the Carl von Ossietzky University of Oldenburg (Oldenburg, DE), supervised by Prof.Dr. Martin Kühn.

Abstract submitted for WESC 2021:

Remote sensing techniques such as lidars have proven to be the most versatile tool for measuring wind turbine wakes due to their good resolution and flexibility (Wildmann et al., 2018). Indeed, they have aided in revealing that wakes exhibit large-scale dynamics, such as wake meandering, and that the measured wake field depends on which frame of reference is used to analyze it (e.g., Trujillo et al., 2011; Beck et al., 2015). These studies on far-wake tracking have shown that there is the potential of filtering out large-scale effects to obtain a wake field unaffected by external atmospheric conditions. This PhD project strives to evaluate this potential and its possible applications within the scope of the Lidar Knowledge Europe (LIKE) project, funded by the European Union Horizon 2020. The main objective is to prove to which extent external large-scale effects of the boundary layer can be separated from the far wake from full-field measurements. More specifically, we intend to reach a site-independent “normalized wake”, analogous to the wake in an environment influenced only by small-scale turbulence. The processes and methods involved in the first phase of this PhD are summarized in the conceptual flowchart from Figure 1. Analysis techniques will be tested on single-wake lidar measurements to obtain a normalized wake under different inflow conditions. Data from the RAVE PARKCAST Project in the alpha ventus wind farm (North Sea), acquired by two long-range StreamLine-XR lidars mounted on the nacelle of turbine AV04 (one looking upwind and the other downwind), will be used at this stage. After processing the data and performing a quality control, firstly we will apply different wake tracking techniques (Trujillo et al., 2011; Bastankhah and Porté-Agel, 2014b; Beck et al., 2015) to the downwind-looking lidar measurements to estimate a local quasi-steady wake, which will be represented in a meandering frame of reference. Secondly, the aim is to spot the optimal analysis method to reach a normalized wake and perform a proof of concept of its application, i.e., turning it back to an estimated local wake by considering the atmospheric conditions of the measurement site. For this objective, raw and averaged meteorological data measured by the sensors from the FINO1 mast, located at approximately 400 m from the lidar measurements, will be used. Based on the findings from Beck et al. (2015), we expect that the
normalized wake resulting from the lidar measurements will have more realistic wake deficits and wake length, representative of what a downstream wind turbine would actually perceive. We presume that the representation of the same wake but in a fixed frame of reference will underestimate the deficits as well as show a shorter length. Therefore, normalized-wake values will be more suitable for the validation of modeled wind speeds in wake.