Increasing Reliability in the Design of Offshore Wind Turbines by Improved Foundation Models
,2, Kristoffer Skjolden Skau1
,2, Hendrik Sturm1, Hans Petter Jostad1
1Norwegian Geotechnical Institute, Oslo, Norway, 2Norwegian University of Science and Technology, Trondheim, Norway
Designing offshore wind turbines is a complex task as these dynamic systems are subjected to variable cyclic loads. Improving the accuracy of analysis tools used in the design and optimisation process can increase the reliability and thus reduce uncertainties and risks. Although considerable development has been made on representing the interaction between wind, waves and structural aspects in integrated analysis tools, modelling of the foundation remains unrefined. The aim of this contribution is to illustrate the inherent uncertainties of current industry design practice and to demonstrate the impact of improved foundation models on increasing the reliability.
In current industry practice, the foundation is modelled by means of linear or non-linear springs. Shortcomings of the simplified models representing foundation behaviour in integrated analyses have been identified and discussed in previous literature. As a result noticeable differences in predicted and actual measured natural frequencies are systematically observed. These discrepancies introduce uncertainties in the fatigue lifetime prediction, which leads to increased risk and potentially increased costs for O&M. Thus, improved foundation models that can accurately represent the foundation behaviour are needed.
Two new models, for shallow and deep foundations, have been developed. The models are formulated as kinematic hardening macro-elements, where the soil and foundation behaviour is reduced to a local load-deflection relation. The models include foundation hysteretic damping and corresponding foundation stiffness both continuously updating based in the loading history. This leads to more accurate predicted loads, reduced uncertainties in the estimated fatigue lifetime and therefore reduced risk in the design.
The performance of the improved foundation models is illustrated at foundation level and the capabilities and limitations of the models are listed. In particular, the possibilities of modelling different foundation stiffness and corresponding foundation damping as a function of the current load levels and recent load history are explained in detail.
The models have been implemented in a time-domain aero-hydro-servo-elastic tool, called 3DFloat. Calibration of the models by finite element analyses is described. The effectiveness of the improved foundation models, over to the simplified models used in current industry design practice, is demonstrated by comparing displacement and rotation time histories of an OWT subjected to different typical mean wind speeds.
In contrast with the current industry practice, the improved models can reproduce different foundation stiffness for unloading and reloading and foundation damping depending on the loading history. This behaviour has been reported in cyclic field- and model tests as well as in centrifuge tests. Both stiffness and damping properties of the foundation have a noticeable effect on the fatigue damage, especially during idling where the contribution from aerodynamic damping is relatively low. At mudline, accumulated fatigue damage reduces up to 16% depending on the foundation model used.
The current industry practice in foundation modelling fails to reproduce full-scale measurements, leading to uncertainties in the design that increase the risk and cost of offshore wind turbine design. More accurate foundation models have been developed and implemented in a software to perform integrated analyses of offshore wind turbines. The improved foundation models can reproduce the overall observed foundation behaviour. Comparison against the current industry practice illustrates differences with the current foundation modelling and highlights the benefits of modelling the foundation behaviour more accurately. A more accurate modelling of the foundation response in integrated analyses of offshore wind turbines will reduce the uncertainty in the predicted loads and in the estimated fatigue lifetime and therefore reduced risk in the design.