Loughborough University
Leicestershire, UK
LE11 3TU
+44 (0)1509 263171
Loughborough University

Loughborough University Institutional Repository

Please use this identifier to cite or link to this item: https://dspace.lboro.ac.uk/2134/20576

Title: Turbulence intensity within large offshore wind farms
Authors: Argyle, Peter
Watson, Simon J.
Montavon, Christiane
Jones, I.
Smith, Megan
Issue Date: 2015
Publisher: European Wind Energy Association
Citation: ARGYLE, P. ... et al., 2015. Turbulence intensity within large offshore wind farms. Presented at the European Wind Energy Association Conference, Paris, 17-20th Nov.
Abstract: The so-called Frandsen model forms the basis for the assessment of wind farm level turbulence intensity (TI) in the IEC standard 61400-1 edition 3. It is used in the choice of turbine suitable for a particular wind farm site. The Frandsen model was developed several years ago using field data when turbines and wind farms were of smaller scale than today. There is now an interest in the accuracy of models such as that of Frandsen when applied to the scale of the largest offshore wind farms. In this paper, we present the results of an analysis of the accuracy of the Frandsen model in predicting TI within the Greater Gabbard offshore wind farm. A comparison is made between measured data and predictions from: 1) the original Frandsen model; 2) a simplified version of the Frandsen model and 3) output from the ANSYS WindModeller CFD model. In general, the Frandsen model was found to perform well in the prediction of mean levels of TI but less well than a simplified model using either a freestream ambient TI or a turbine wake TI regardless of distance. Representative or 90% percentile TI levels are less well predicted under direct wake conditions due to the lack of consideration of turbine generated variance in turbulence and the manner in which the 90% percentile freestream TI is incorporated. ANSYS WindModeller was found to perform well in the prediction of mean TI and has the benefit of not requiring upstream TI data. The CFD model can be used to predict representative TI, when complemented with a model for the variance of turbulence. Predictions from the Frandsen model are more sensitive to the choice of freestream data than those from the CFD model.
Description: This is a conference paper.
Version: Published
URI: https://dspace.lboro.ac.uk/2134/20576
Publisher Link: http://www.ewea.org/annual2015/wp-content/uploads/files/conference/scientific-proceedings/2015-EWEA-scientific-proceedings.pdf
Appears in Collections:Conference Papers and Contributions (Mechanical, Electrical and Manufacturing Engineering)

Files associated with this item:

File Description SizeFormat
1412_ewea2015presentation.pd.pdfPublished version2.97 MBAdobe PDFView/Open

 

SFX Query

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.