Effective Loads Quantification in Enhanced and Natural Turbulence

Knowledge of the wind conditions in a wind farm is important for a good prediction of the power performance and expected energy yield, as well as to gain insight into the response of the wind turbine, for design and understanding of potential failures. 

Wind conditions on a wind farm can be measured in several ways. Traditionally the wind is measured before construction using a met-mast, instrumented with cup or sonic anemometers at different heights. In the last decade, the use of a bottom mounted lidar has become an increasingly good substitute for, or additional installation to, the met-mast. During operation, the wind turbine can also be instrumented with lidar or anemometers on the hub or nacelle.  

Currently it is not common practice, during pre-construction site assessment or in operations, to fully investigate the shear (vertical or horizontal change in wind speed) or veer (usually vertical change in wind direction). Instead, a simple “shear profile” is fitted to the data, which may result in inaccurate estimation of both power output and turbine loading. The tools to perform such an investigation are, however, possible. This project will therefore explore the benefits of accurately measuring the wind variations (shear, veer and turbulence) across the whole rotor plane.  

The main goals of the project are:  

1. To quantify the improvement in turbine loads assessment using measurements of the wind field instead of a point measurement at hub height;  

2. To develop a method that provides the appropriate wind field characterization for loads assessment.

A new method has been developed using an existing data set – comprising met-mast anemometers, spinner anemometers, Lidar and SCADA from a large wind turbine in complex wind flow conditions.
Two main research questions were formulated:  

1. How effective are point wind measurements on the wind turbine nacelle or a met-mast for understanding wind loads?  

2. Do reconstructions of the full turbine inflow from Lidar measurements identify previouslyunexplored loading conditions; and are there therefore gaps in the current standards for preconstruction site assessment 

The expertise of ROMO Wind on spinner-based point measurements and of Leosphere on Lidar measurement is used by TNO to reconstruct wind fields based on machine learning. The focus of GE is on the consequences on the power performance and loads when using the reconstructed wind fields. Using a novel machine learning method, the 1Hz turbulent wind field will be reconstructed from the Lidar measurements. This provides the in-flow wind conditions for the wind turbine. This will be compared with turbulence models used for loading calculations, to detect whether temporal and spatial corrections are well-captured. A practical method for identifying and quantifying unusual inflow conditions when a Lidar is not present during operation has been developed. Aeroelastic modelling and the turbine SCADA measurements (e.g. blade and drive train loads and power) will be used to improve understanding of the effect of the wind field on the turbine loads. Overall this project will provide a method to assess turbine loads with higher accuracy than the current standard method, leading to a lower risk for the turbine manufacturer and the wind farm developer. 

Results

Adatabase has been created by means of measurements. The database includes measurements from met mast anemometers, a spinner anemometer, lidar and SCADA from a large wind turbine in complex wind flow conditions. 

The data from the Eloquent data base containing wind velocity and load measurements quality checked and analyzed. Detailed wind field has been has been reconstructed using a new developed method . The spinner anemometer was installed and a database was created with the measurements. The calibration of the spinner anemometer during the project indicated that a new recalibration will be required, to be aligned with the reference angle and a subsequent correction of the dataset will be applied.

The reconstructed wind fields of prior activities are used in this work package. This provides the inflow wind conditions for the wind turbine. Aeroelastic modelling and the turbine SCADA measurements were used to improve understanding of the effect of the (reconstructed) wind field on the turbine loads.