PO066

Comparison of 10MW Turbine Dynamics Between Bottom Fixed And Floating Foundations

Andreas Manjock, Kimon Argyriadis
DNV GL RC, Hamburg, Germany

Abstract

The load simulations in this study are intended to demonstrate the dynamic impact to a generic 10 MW wind turbine when transferring it from an offshore bottom fixed foundation to a deep water floating station keeping system. As foundation variants a bottom fixed jacket for 50 m water depth and a tension leg platform (TLP) at 180 m water depth are selected. Special focus is given herein on the different dynamics of the two foundation types and their typical effects to structural loading to the rotor-nacelle-assembly (RNA) and its tower.

The wind turbine model applied in this study is the INNWIND 10MW reference turbine with a three bladed upwind rotor of diameter of 178.3 m, a hub height of 119.0 m above LAT and a total rotor-nacelle mass of 674 tons. For the bottom mounted turbine model the INNWIND reference jacket design with a total mass of 2,120 tons is modelled. The TLP design is based on the OC5 TLP design characterised by a cylindrical floater and transition piece with four horizontal lever arms to connect the tendons. The total mass of the TLP foundations accounts to 2,350 tons.

The loading for specific components of the 10MW RNA and the tower are addressed for the different foundation concepts and are analysed in detail.

Method

Based on a TLP tank test performed within the EU Framework 7 research project INNWIND.EU the simulations with Bladed 4.7 are calculated in test scale 1:60. The measurements have been applied for a comprehensive calculation model validation. Structural properties and hydrodynamic coefficients are adjusted that the dynamic behaviour of the TLP in terms of eigenfrequencies, modes, damping and coupled motions are matched.

For the bottom fixed jacket model the identical environmental test conditions (sea states adopted to water depth) are used and subsequently compared to the TLP simulations.

Because of the 1:60 scale the results show rather qualitative motions and loading trends than real size load values. However, Since the model is consequently dimensioned according to Froude scaling laws the results are suitable for upscaling.

Results

The Bladed 4.7 jacket model is validated against the FE calculations of the designer while the TLP model is calibrated by the tank test data. Both foundation types are calculated applying decay tests, regular waves, irregular sea states, embedded extreme waves (focussed waves) including wind and waves combinations/misalignments.

The simulations for the TLP show in general an excellent representation of the dynamic behaviour of the TLP model compared measured motions and loading. However, the coupling of certain deflections and in particular the phase angles, for instance surge and sway coupling, cannot be simulated in sufficient accuracy.

Finally, loads and deflection comparisons for both foundation types are shown in graphs as time domain plots, as power spectral density (PSD) comparisons, and as response amplitude operators (RAO's).

Conclusions

The load and deflection comparison of the two foundation types revealed that the RNA does not receive significant higher loads on the TLP than on the bottom fixed jacket. Some increased loading for the 10MW RNA exists for thee rotor thrust, yaw bearing bending moments as well for the nacelle yaw moments. The tower on the other hand suffers higher torsional loading on the TLP foundation due to weak stiffness properties of tendons for the yaw mode. Large gyroscopic loading during normal operation in extreme sea state results in dimensioning loads for the tower.

In the presentation a full picture of the load differences at blade sections, blade root, hub centre, yaw bearing/tower top, tower sections and tower bottom/transition piece is given.

Objectives

The simulation tool used in this study, Bladed 4.7, shows in this comparison that it is capable to model bottom fixed/jacket and floating/TLP offshore wind turbines in high model resolution and accuracy. However, in the 1:60 model experiments some couplings of floating degree of freedoms (DOF's) are observed which are not reproduced completely by the simulations. The coupling effects could be analysed in more detail regarding their relevance for full size floating structures.

Model tests including realistic RNA power control algorithms and pitch actions are still rare. The improvement of laboratory wind conditions in line with more detailed pitch actuator models would support the better prediction of load effects under lab conditions.