poster ewea "damping estimation of an offshore wind turbine on a monopile foundation"

1
Correlation driven analysis in the time domain By fitting an exponential function to the relative maxima of the auto-correlation functions of the measured accelerations the damping ratio can be extracted. This approach only provides good estimates for the damping when the decay consists of 1 mode. During the ambient tests this is not the case as there is strong coupling between the FA and SS mode, and therefore this approach fails. Correlation driven analysis in the frequency domain The Fast Fourier Transformation of the decaying auto- correlations can be used as input for the operational modal analysis methods in the frequency domain. This approach fits again a polynomial function with multiple modes and therefore the results are not affected by the fact that multiple modes are present. Periodogram driven analysis in the frequency domain The periodogram driven approach uses the power spectrum matrix from the measured accelerations as input data for performing operational modal analysis in the frequency domain. Note that this approach clearly identifies both the FA mode (0.358Hz) and SS (0.345Hz) mode. Time domain analysis In the time domain damping can be obtained by fitting an exponential function to the decaying time series and extracting the damping ratio from the parameters of the fitted expression. This method assumes that the decay has only the contribution of a single mode and therefore proper filtering might be required. The method was applied to the measured accelerations of the highest 3 levels in the direction of the wind. The data was pre-filtered using 3 different band-pass filters. Frequency domain analysis One can also obtain an estimate for the damping using an operational modal analysis approach to the Fast Fourier Transformations of the free decays measured during the overspeed test. This approach fits a polynomial function with multiple modes and therefore the results are not affected by the fact that multiple modes are present. . Damping Estimation of an Offshore Wind Turbine on a Monopile Foundation Christof Devriendt, Pieter Jan Jordaens, Gert De Sitter, Patrick Guillaume Vrije Universiteit Brussel – Sirris - Offshore Wind Infrastructure Lab The overall damping of the first bending mode of an offshore wind turbine consists of a combination of aerodynamic damping material damping of steel damping from wave creation damping due to inner soil friction damping due to hydrodynamic drag damping due to constructive devices The main goal of this campaign was to identify the damping ratios of the first fore-aft (FA) and side-side (SS) bending mode excluding the aerodynamic damping and damping due to constructive devices. Many large scale offshore wind farm projects use monopile foundation to realize a cost effective design. During the design of these monopile structures fatigue due to combined wind and wave loading is one of the most important problems to face. The damping significantly influences the dynamic response of the wind turbine reaction and thus also the predicted lifetime. The work presented in this poster describes a comparative study between different techniques aimed at identifying the damping values of an offshore wind turbine on a monopile foundation. Abstract Damping Effects Conclusions EWEA 2012, Copenhagen, Denmark: Europe’s Premier Wind Energy Event Offshore Measurements The measurement campaign was performed at the Belwind wind farm, which consists of 55 Vestas V90 3MW wind turbines. The wind farm is located in the North Sea on the Bligh Bank, 46 km off the coast. The tests were performed on the BBCO1-turbine. The actual water depth at the location of BBCO1 is 22.9m and the monopile has a penetration depth of 20.6m. The soil is considered stiff and mainly consists of sand. Approach Damping Estimation For the determination of the offshore damping two tests have been performed. First an emergency stop with vanishing aerodynamic damping has been examined for estimating the damping. Afterwards damping has been estimated using ambient excitation from the wind and waves, while the pitch angle was above 80 degrees in order to minimize the effect of aerodynamic damping Results Overspeed Stop Results Ambient Excitation -0.1 -0.05 0 0.05 0.1 0.15 -0.1 -0.05 0 0.05 0.1 0.15 SS-Aceleration (g) FA-Acceleration (g) Level 1 Level 2 Level 3 100 150 200 250 300 350 -1 0 1 Time (s) Acceleration (g) Data Points Curve 100 150 200 250 300 350 -0.4 -0.2 0 0.2 Time (s) Acceleration (g) Data Points Curve 100 150 200 250 300 350 -0.2 0 0.2 Time (s) Acceleration (g) Data Points Curve 0 0.5 1 1.5 -50 -40 -30 -20 -10 0 10 20 30 Frequency (Hz) Amplitude (dB) FA Level 1 FA Level 2 FA Level 3 SS Level 1 SS Level 2 SS Level 3 0.2 0.4 0.6 0.8 1 1.2 1.4 0 5 10 15 20 25 30 o o o f f o f o f o o f s s f o f s f s o s f s f s s o s f f f s o f f o s d s f f s s s f o f s s s f f f f o s f s d f s o s f f f f f s f f s f s s s f f f o f s f f f f f s f f f s o f s s o s s f s s f f f f f s o o s s f f f o s f f f f s f f s s s s f f o s f f s f f f s f f f s f s s f s s f s o s f o s s f f f s o f f f s s o s f f s f s f s f o s f s s f s Stabilization Chart frequency (Hz) model order -0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02 -0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02 SS-Aceleration (g) FA-Acceleration (g) Level 1 Level 2 Level 3 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 SS-Aceleration (g) FA-Acceleration (g) Level 1 Level 2 Level 3 Acceleration (g) Acceleration (g) 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 5 10 15 20 25 30 o s o f f f o s f s s s s f s s o f s s o o o f f f s f d s f f s f s s f o f s f f s f s s s s f f s s s s s f s s s s s f f s s f s f s o s f s s s f f s s s s s s f s f s s s s o s ss f s s s f s s f s f s s s s s s s f s f s s s s s s f s f s s s s s f s s s s s s s s s s s s s f s s s s s s s s s s s s s s s o s s s s s s s s s s s s s s s s s d s s s s s s s s d s s s s ss s s s s s s s ss s s d Stabilization Chart frequency (Hz) model order o 0 0.5 1 1.5 2 -140 -120 -100 -80 -60 -40 -20 0 20 40 Frequency (Hz) Amplitude (dB) FA Level 1 FA Level 2 FA Level 3 SS Level 1 SS Level 2 SS Level 3 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 5 10 15 20 25 30 f o f f o o f f s o o s f o f o f f f f o o f f o s s s o f f s f f f o s fs f f f s f f f s s o f f f s f f f f d s f s f ff f s s f s ff s f s f s f f s s s f s s f s f f s s s f f f s f f s f f s f f s f s s s f s f s f f s s s s s f f s f f s f f s s s s f s f s f s s s s s s s s s s s ss s s s s s s s s s s s s s s s s s s s ss s s s s s s s s Stabilization Chart frequency (Hz) model order o Measurements are taken at : 9 locations using 10 accelerometers. 4 levels at 67m, 37m, 23m and 15m above sea level. We can conclude that the ambient vibration tests together with the application of state-of-the-art output- only identification techniques can provide good estimates of the damping ratios of an offshore wind turbine. The results have been compared with the ones obtained from the commonly used over-speed tests. Ambient vibration tests have the strong advantage of being very practical and economical, as they use the freely available ambient wind wave excitation. Figure 1: location Belwind wind farm (left) park layout Belwind wind farm (right) Figure 2: measurement locations on BBCO1 Figure 3: Example measured accelerations during overspeed stop (left) and ambient excitation (right) Figure 4: Movement seen from above (left) and exponential fitting on FA acceleration on 3 levels after applying a band-pass filter of 0.3-0.5Hz (right) Figure 5: Fast Fourier transformation of the accelerations obtained on the 3 different levels in both the FA and SS direction (left) Stabilization diagram after applying operational modal analysis to the Fast Fourier Transformations. Figure 6: Movement seen from above on 3 levels in FA and SS direction during ambient excitation at 2 different moments in time Figure 7: Exponential fitting of an autocorrelation function of a sensor in the FA direction (left) and SS direction (right) Figure 8: Stabilization diagram after applying operational modal analysis to the Fast Fourier Transformations of the auto-correlations Figure 9: Auto and cross spectra with reference sensor 1 in FA direction (left) Stabilization diagram after applying operational modal analysis to the spectrum matrix (right) Both during the overspeed test and during the ambient excitation test the tuned mass damper was switched off. Acknowledgments This research has been founded by the Institute for the Promotion of Innovation by Science and Technology in Flanders (IWT) and the Fund for Scientific Research – Flanders (FWO). The authors also acknowledge Belwind NV and the OWI-lab for providing the test- facilities and support (more info: http://www.owi-lab.be)

Upload: pieter-jan-jordaens

Post on 29-Nov-2014

200 views

Category:

Documents


5 download

DESCRIPTION

Best scientific poster EWEA 2012

TRANSCRIPT

  • 1. Damping Estimation of an Offshore Wind Turbine on a Monopile Foundation Christof Devriendt, Pieter Jan Jordaens, Gert De Sitter, Patrick Guillaume Vrije Universiteit Brussel Sirris - Offshore Wind Infrastructure Lab Both during the overspeed test and during the ambient Correlation driven analysis in the time domain Abstract excitation test the tuned mass damper was switched off. By fitting an exponential function to the relative maximaMany large scale offshore wind farm projects use of the auto-correlation functions of the measuredmonopile foundation to realize a cost effective design. accelerations the damping ratio can be extracted.During the design of these monopile structures fatigue 1.5 Data Points 1.5 Data Pointsdue to combined wind and wave loading is one of the 1 Curve 1 Curve Acceleration (g) Acceleration (g) 0.5 0.5most important problems to face. The damping 0 0 Figure 3: Example measured accelerations during overspeed stop (left) and ambient -0.5 -0.5significantly influences the dynamic response of the -1 -1 0 50 100 150 200 250 300 0 50 100 150 200 250 300 excitation (right) Time (s) Time (s) Figure 7: Exponential fitting of an autocorrelation function of a sensor in the FA directionwind turbine reaction and thus also the predicted (left) and SS direction (right)lifetime. The work presented in this poster describes a Results Overspeed Stop This approach only provides good estimates for thecomparative study between different techniques aimed damping when the decay consists of 1 mode. Duringat identifying the damping values of an offshore wind Time domain analysis the ambient tests this is not the case as there is strongturbine on a monopile foundation. In the time domain damping can be obtained by fitting coupling between the FA and SS mode, and therefore an exponential function to the decaying time series and this approach fails. Damping Effects extracting the damping ratio from the parameters of the Correlation driven analysis in the frequency domainThe overall damping of the first bending mode of an fitted expression. This method assumes that the decay has only the contribution of a single mode and therefore The Fast Fourier Transformation of the decaying auto-offshore wind turbine consists of a combination of proper filtering might be required. correlations can be used as input for the operational aerodynamic damping 0.15 1 modal analysis methods in the frequency domain. Acceleration (g) Level 1 Data Level 2 Points Level 3 0 Curve Stabilization Chart material damping of steel 0.1 d s s s s s s s s -1 30 s s s s s s s s s 100 150 200 250 300 350 d s s s s s s s s Time (s) d s s s s s s s s 0.05 s s s s s s s s s o s s s s s s s s FA-Acceleration (g) 0.2 damping from wave creation 25 s s s s s f s s Acceleration (g) Data s s s s s s s s 0 Points s s s s s f s s s s s f s f s s 0 Curve s s s s f s f s s -0.2 20 s s s s f s f s s model order s f s s s s f s s o s s s f s f s s damping due to inner soil friction -0.4 s s s f s f s s 100 150 200 250 300 350 s s s f s o f s -0.05 15 f s s f f s s Time (s) s s s f s s 0.2 s s f f s s s s f f s s Acceleration (g) Data f s o f f s Points 10 s fs s f f damping due to hydrodynamic drag -0.1 d fs o f f f 0 Curve os o f s o s s f ss s s 5 fs f o -0.2 f f -0.1 -0.05 0 0.05 0.1 0.15 100 150 200 250 300 350 o s o damping due to constructive devices SS-Aceleration (g) Time (s) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Figure 4: Movement seen from above (left) and exponential fitting on FA acceleration on 3 frequency (Hz) levels after applying a band-pass filter of 0.3-0.5Hz (right) Figure 8: Stabilization diagram after applying operational modal analysis to the FastThe main goal of this campaign was to identify the Fourier Transformations of the auto-correlationsdamping ratios of the first fore-aft (FA) and side-side The method was applied to the measured accelerations(SS) bending mode excluding the aerodynamic damping of the highest 3 levels in the direction of the wind. Theand damping due to constructive devices. data was pre-filtered using 3 different band-pass filters. Offshore Measurements This approach fits again a polynomial function with multiple modes and therefore the results are not The measurement campaign was performed at the affected by the fact that multiple modes are present. Belwind wind farm, which consists of 55 Vestas V90 Periodogram driven analysis in the frequency 3MW wind turbines. The wind farm is located in the domain North Sea on the Bligh Bank, 46 km off the coast. Frequency domain analysis The periodogram driven approach uses the power One can also obtain an estimate for the damping using spectrum matrix from the measured accelerations as an operational modal analysis approach to the Fast input data for performing operational modal analysis in Fourier Transformations of the free decays measured the frequency domain. during the overspeed test. This approach fits a Stabilization Chart 40 FA Level 1 FA Level 2 s ss s s s s 30 s ss s s s s polynomial function with multiple modes and therefore 20 FA Level 3 s ss s s s s SS Level 1 s ss s s s s SS Level 2 s ss s s s s 0 s ss s s s s SS Level 3 25 s fs f f s s the results are not affected by the fact that multiple s s f f s f f s f f s s -20 s s f f s s f s f s s s 20 f s f f s f Amplitude (dB) modes are present. model order f s f f s f -40 f fs f f s s s fs s f s s s ff s f s f 15 s ff s f s s -60 f ff f s f s d f f f f s f ff o s s f Figure 1: location Belwind wind farm (left) park layout Belwind wind farm (right) 30 . Stabilization Chart f fs f f f FA Level 1 -80 10 s fs o f f f FA Level 2 s ff o s s s s s f f s s o f s o ff o o f f 20 FA Level 3 30 s s f f f s f ff o o f SS Level 1 s s o f f f o s f f -100 fs o o s The tests were performed on the BBCO1-turbine. The s s o f f s o s 5 ff o SS Level 2 f o f s s f f s s 10 SS Level 3 f s f s f f f f s f f o -120 25 f s o f f f s s s s f f s f f o s f f f f 0 actual water depth at the location of BBCO1 is 22.9m s s o o f s f ff f s s f f s s o -140 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 f s o s f f s 0 0.5 1 1.5 2 frequency (Hz) Amplitude (dB) 20 f s f f f f Frequency (Hz) model order f s o f f ff s s s f f s and the monopile has a penetration depth of 20.6m. -10 15 f o s s s s f f d f s s f f f f f f f f o Figure 9: Auto and cross spectra with reference sensor 1 in FA direction (left) Stabilization -20 f s s f o s diagram after applying operational modal analysis to the spectrum matrix (right) The soil is considered stiff and mainly consists of sand. f s s f s o s d f f 10 s o f f o s f s s Note that this approach clearly identifies both the FA -30 s f f s