Session: Model Testing-1

Paper Number: 165248

165248 - Low-Frequency Extreme Response Waves: An Experimental Verification for Catenary-Moored Floating Offshore Wind Turbines 

Abstract: 

Ultimate limit state design loads traditionally involves time-domain assessment of multiple 3-hour short-term random seed sea-states, following an environmental contour with a 50 year return period. For catenary moored floating offshore wind substructures in parked conditions, mooring loads are driven by slowly-varying wave drift forces. The use of full quadratic transfer functions (QTFs) to capture these slowly-varying forces remains computationally costly, such that time-domain analysis of 3-hour sea-states is not guaranteed to capture the statistical maximum occurring low-frequency wave load. For fixed-bed structures, deterministic wave approaches, such as NewWave theory, and Most-Likely Extreme Response (MLER) waves, have been introduced, to expose the structure to all possible wave frequencies within a short time-window, to induce a maximum, focused wave amplitude (NewWave) and linear response (MLER). These approaches have been tested on floating offshore wind substructures, e.g. Brown et al. (2023), including in irregular background sea-states in an attempt to incorporate the historical effects of the wave train; Conditional Random Response Waves. However, all of these approaches assume that maximum wave loading is driven by wave-frequency loads, and cannot account for the low-frequency response. In Lande-Sudall et al. (2024), a theoretical approach was outlined to define a Low-Frequency Extreme Response (LFER) wave, whereby the structure’s QTF is used to design a deterministic wave train with wave components phase-shifted onto multiple focus times, to create a regular difference-frequency wave forcing. It was shown that an extreme low-frequency response in surge, equivalent to the analytical maximum from several hundred hours sea-state duration, could be generated within fewer than ten oscillations at the surge natural frequency.

In this work, we present the initial results of experiments conducted on a 1:100 scale model of the INO WINDMOOR semi-submersible, exposed to LFER waves designed to oscillate maximum surge response, with varying number of oscillations at the surge natural frequency; 3, 5, 10 and 20 oscillations. The 50 m-long MarinLab towing tank at Western Norway University of Applied Sciences, with full-scale water depth of 220 m is used for testing, with a lightweight linear spring mooring to the tank walls providing a surge natural period of 104 s, full-scale. Motion response is recorded by a Qualisys motion-capture camera system, and mooring line loads are measured using three load cells mounted at the tank walls. LFER waves are designed based on multiple JONSWAP sea-states of varying significant wave height, Hs=3.74-11 m, peak period, Tp=7-12 s, and peakedness parameter, gamma=4.9, allowing direct comparison to 3-hour stochastic JONSWAP waves based on the same definitions. Preliminary results indicate that mooring loads exceeding those from the 3-hour random sea-states can be obtained. For example, with Hs=6.2 m and Tp=9 s, the maximum surge excursion from the 3 h random sea-state was approximately 7.8 m, with corresponding maximum mooring force of 1242 kN, whilst the corresponding LFER wave based on five surge oscillations gave a maximum excursion of approximately 10 m, and corresponding peak mooring load of 1564 kN. Nevertheless, experimental implementation of the LFER wave trains is not straight-forward: Careful linear calibration of wave components is required and the time-series elevation for longer duration runs, e.g. with 20 surge periods, can become contaminated by bound waves. Work is on-going to assess these effects further and to evaluate optimal number of surge oscillations.

Brown et al. (2023) On the selection of design waves for predicting extreme motions of a floating offshore wind turbine. Ocean Eng. 290.

Lande-Sudall, D. and Stansby, P. K. (2024) Short-duration design waves for modelling of extreme second-order surge response with spar substructure test case. Appl. Ocean. Res. 153.

Presenting Author: David Lande-Sudall Western Norway University of Applied Sciences - Department of Mechanical Engineering and Maritime Studies

Presenting Author Biography: Dr David Lande-Sudall is an Associate Professor in Ocean Engineering at the Western Norway University of Applied Sciences (HVL), Bergen in Norway. David started his current position after completing his PhD on co-located offshore wind and tidal stream turbines from the University of Manchester, UK. His research is focussed on ocean renewable energy and particularly experimental hydrodynamics within HVL’s MarinLab towing tank facility. David leads the research group, Wind, Water and Waves (W3) and is currently Principal Investigator on the Research Council of Norway funded project, HYDROMORE (grant no. 324388).

Authors:

David Lande-Sudall Western Norway University of Applied Sciences - Department of Mechanical Engineering and Maritime Studies
Gloria Stenfelt Western Norway University of Applied Sciences - Department of Mechanical Engineering and Maritime Studies
Peter Stansby The University of Manchester