Determination of optimal wind turbine alignment into the wind and detection of alignment changes with SCADA data

Abstract

Abstract. Upwind horizontal axis wind turbines need to be aligned with the main wind direction to maximize energy yield. Attempts have been made to improve the yaw alignment with advanced measurement equipment but most of these techniques introduce additional costs and rely on alignment tolerances with the rotor axis or the true north.Turbines that are well aligned after commissioning may suffer an alignment degradation during their operational lifetime. Such changes need to be detected as soon as possible to minimize power losses. The objective of this paper is to propose a three-step methodology to improve turbine alignment and detect changes during operational lifetime with standard nacelle metrology (met) mast instruments (here: two cup anemometer and one wind vane).In step one, a reference turbine and an external undisturbed reference wind signal, e.g., met mast or lidar are used to determine flow corrections for the nacelle wind direction instruments to obtain a turbine alignment with optimal power production. Secondly a nacelle wind speed correction enables the application of the previous step without additional external measurement equipment.Step three is a monitoring application and allows the detection of alignment changes on the wind direction measurement device by means of a flow equilibrium between the two anemometers behind the rotor.The three steps are demonstrated at two 2 MW turbines together with a ground-based lidar. A first-order multilinear regression model gives sufficient correction of the flow distortion behind the rotor for our purposes and two wind vane alignment changes are detected with an accuracy of ±1.4 • within 3 days of operation after the change is introduced.We show that standard turbine equipment is able to align a turbine with sufficient accuracy and changes to its alignment can be detected in a reasonably short time, which helps to minimize power losses.

Maximizing the returns of LIDAR systems in wind farms for yaw error correction applications

“…When the turbines regress more quickly, the wind farm requires more LIDAR devices to combat the yaw error problem, which increases the investment costs for LIDAR and lowers the returns. These results are a good way to understand the outcome of other studies such as previous studies discussed earlier where the focus is on optimizing the yaw error control algorithm. A 2‐year yaw error regression means the yaw error control algorithm operates better and loses its calibration more slowly.…”

“…The first approach attempts to optimize the control algorithm that uses the wind speed and direction data from cup and vane anemometer (and sometimes a meteorological mast in the wind farm) and processes the data in order to overcome the bias in the yaw system controller that causes the yaw error. This approach is discussed in detail in previous studies …”

Exploring the complexities associated with full-scale wind plant wake mitigation control experiments

“…The EET denotes the transition between the daytime convective and nocturnal stable ABLs. This evolutionary period of atmospheric stability is consistent with distinct changes in wind speed, wind direction, wind structure, atmospheric turbulence, and temperature (Mahrt, 1981;Nieuwstadt and Brost, 1986;Acevedo and Fitzjarrald, 2001;Edwards et al, 2006). To track the progression of the EET and the onset of the nocturnal stable ABL, the virtual potential temperature (i.e., θ v ) gradient between 10 and 200 m was examined using available meteorological tower data.…”

“…The values in parentheses denote the percentage of SCADA data within each experimental period (upwards of 600 observations) that were consistent with region three pitch operation. bine inflow wind direction (Mittelmeier and Kühn, 2018). Yaw error is defined as the misalignment angle between the rotor plane and the turbine inflow wind direction (i.e., quantifying the non-normal rotor orientation angle).…”

“…11) (Drobinski and Foster, 2003;Träumner et al, 2015). Although these features are elongated in the along-wind dimension, their orientation can slightly deviate (i.e., clockwise or counterclockwise) from the ABL wind direction (Morrison et al, 2005;Foster, 2005;Lorsolo et al, 2008). Although it is not fully understood what causes streak orientation to differ from the ABL wind direction, it can be hypothesized that the same boundary layer forcings also promote downstream wake deviation from the ABL wind direction.…”