Lead Session Chair:
Stephan Barth, Managing Director, ForWind - Center for Wind Energy Research, Germany
Shou Oh (1) F P Takeshi Ishihara (1)
(1) The University of Tokyo, Tokyo , Japan
Printer friendly version: Print
Download poster(0.75 MB)
Presenter's biographyBiographies are supplied directly by presenters at OFFSHORE 2015 and are published here unedited
Studied computational fluid dynamics and geotechnical engineering at the university of Tokyo. Currently a PhD candidate studying for accuracy improvement of dynamic analysis of wind turbines and their application to monitoring technologies.
A modified von karman model for the spectra and the spatial correlations of the
offshore wind field
The dynamic response of a wind turbine is strongly affected by the turbulence structure of wind field. As life time estimation and damage detection are expected to lead to cost reduction, the accuracy of dynamic analysis required in these techniques should be firstly achieved through the accurate modelling of the three-dimensional wind field in both spectrum and correlation form. Also the turbulence scale parameter values, which have significant effect on the accuracy of the model, should be evaluated for offshore as the values proposed in the design codes are usually based on measurements onshore.
In this study the von Karman model is chosen for modelling of the wind field because of its ability to analytically formulate both spectrum and correlation. The auto-spectrum, auto-correlation, spatial co-spectrum and spatial cross-correlation of the von Karman model are compared to measurement data from offshore ultra-sonic anemometer, and a modified von Karman model is proposed for auto-spectrum and auto-correlation to correct the difference for lateral and vertical components which have been pointed out by previous studies to be the problem of the model. Finally, the turbulence scale parameters are evaluated and compared to the values in the design codes.
Main body of abstract
Firstly a three-dimensional wind field measurement is carried out on a meteorological mast located 3km offshore Choshi, Japan. Three ultra-sonic anemometers are located at 80m, 60m, and 40m above sea level respectively. Then, auto-spectrum, auto-correlation, spatial co-spectrum and spatial cross-correlation are calculated from measured data and compared to von Karman model. For auto-spectrum and auto-correlation, the longitudinal component agreed well with measurement data and the modification for lateral and vertical components is suggested as expansion of the equation of longitudinal component. The suggested modification showed better agreement with measurement than original model. For spatial co-spectrum and spatial cross-correlation, measurements also agreed with von Karman model. In design codes, the approximate exponential form for longitudinal component is usually used with two model parameters; the length scale and the decay factor. Comparison with measurement data showed this approximation also applies to lateral and vertical component when proper decay factor is used. Finally, the turbulence scale parameters are evaluated. The integral length scale of von Karman model for auto-spectrum and auto-correlation are found to be almost same as Kaimal model in case of 80m height offshore, and to be much smaller than the values suggested in the design codes. The ratios of length scales of vertical separation and horizontal separation are found to be almost same as onshore measurement. And the decay factor for spatial co-spectrum is found to be much smaller when used to lateral and vertical components.
Offshore wind field measurement using ultra-sonic anemometers is performed to validate the original and the modified von Karman model for auto-spectrum, auto-correlation, spatial co-spectrum and spatial cross-correlation. Modified von Karman model agreed well with measurement for auto-spectrum and auto-correlation, and the turbulence scale parameters are evaluated for offshore wind field and found to be much smaller than the values suggested in the design codes. The original model agreed with measurement for spatial co-spectrum and cross-correlation and approximate equation of co-spectrum was also applicable to lateral and vertical component when smaller decay factors are used.
The audience will learn about the modelling equations of turbulence structure and the scales parameters for the offshore wind field and the difference from the onshore wind field.