Project - park calibration

Synopsis

The project aim is to improve the already developed WTG wake models through model adjustments and recalibration of model parameters. This is done, so the models can be utilized in large offshore wind farms. The latest measurements from selected large windfarms are analysed. The data are validated and adapted so that they may be used for calibrating the existing and validated WTG wake models. The WTG wake models are mainly, developed solely for use on small land based wind farms and should be modified for use on big offshore wind farms. The focus of the project is, primarily, prediction of energy levels for single WTGs in the wind farm, and secondarily to predict mean wind speeds and turbulence intensities inside the wakes.

Purpose

The purpose of the project is to apply the latest measurements from large (offshore) wind farms in order to calibrate and improve the existing and validated wake models. The project focus is on wake effects within large wind farms. The primary part of the project is the calibration of the models for use with wind energy yield calculations and secondarily the mean wind field and the increased park turbulence.

Motivation

Most WTG wake models that are in use today are single wake models - based on the flow downstream of a single turbine. For the combination of the wake effect from several upstream WTGs, typically a simple empirical combination model is used - this is normally a 'quadratic speed-deficit model'. In connection with the PSO-project, 'The necessary distance between big wind farms at sea (STORPARK)', EMD has been establishing alternative park models as alternative to the widely used N.O. Jensen model. In total about 5 different models can now be taken into use energy-, mean wind speed and turbulence-calculations. The models are varying in complexity from simple empirical models to the more advanced CFD-models. All the models still lack - as do the N.O. Jensen model - a structured validation and calibration for use on big offshore farms, just as the wake combination model is not necessarily applicable for this purpose. It should possibly be replaced by the suggested model in the STORPARK project (which has not been validated yet) or another model.

Actions in the project

A: Analysis and identification of the available wind data for offshore wake analysis. Analysis of available WTG production data. Preparation and debugging of the data for intended for the energy yield estimates.
B: Uncertainty-evaluation of collected data. In the EU-project ENDOW it showed that there is considerable uncertainty connected to the collected data - especially when looking at modelling of actual incidents, e.g. mean wind speeds and turbulence. When data - in connection with productional estimates - is averaged over time, the uncertainties will often be canceled out.
C: Calibration of WTG wake model parameters from the available collected data, including uncertainty-evaluation of the model adjusted parameters.
D: If required, then the WTG wake models and combinational models are adjusted so that uncertainty of estimates are reduced.
E: Model predictions from the various models are evaluated and a qualitative evaluation of the models is completed on the basis of model predictions - not only on Horns Rev and Rødsand - but also at alternative sites - e.g. Vindeby, Nørrekær Enge and Klim.
F: Reporting.

Expected results

The new offshore measurements contain a powerful potential for improvement of the existing wake models. Use of known and already validated models gives a high degree of certainty for the end user / the project developer. As the project is based on new-estimation of offshore specific parameters for existing and already validated models, the results will, on a fairly short term, be able to improve the estimates and decrease the uncertainties for these models.

Duration

The project is expected to take 2.0 years. The project start was medio february 2005.

Project Documentation (in pdf-format)

• Main report
• Description of the implemented wake and turbulence models
• Overview of the wake models
• Overview of the turbulence models

Case Studies (requires WindPRO 2.6)

Below is a number of case studies with wake calculation reports. The files are all in the WindPRO 2.6 export format. A separate written pdf-report with an overview of the case study is released separately at a later stage. Note: At a later point you can download a demo-version of WindPRO 2.6 that can open these files. The version will be available here

• Horns Rev
• Klim Fjordholme
• Zafarana

EMD.DK

Back to the top