A Floating Offshore Wind Turbine in Extreme Wave Conditions

While the design of floating offshore wind turbines (FOWT) is still at an infant stage, the general desire to realise them is strong. According to a poll conducted by GL Garrad Hassan at the HUSUM 2012 Wind Energy Trade Fair, 62% of the attendees believed that floaters will be a part of the mix and will even overtake bottom fixed foundation within the coming two decades, Bossler (2011). FOWTs are believed having a large potential of lowering the cost of energy (CoE). The CoE minimization is currently the main driver for technological development in the offshore wind industry. Therefore reliable and purposes oriented design procedures are the backbone of a cost efficient offshore wind industry. Conventional engineering procedures for the assessment of extreme event impacts, i.e. ultimate limit state (ULS) analysis of floating structures, as they have been used in the oil and gas industry, neglect two important aspects, which make them non-conservative in use for FOWT: (A) The offshore wind industry intends to install floating structures at much lower water depth (from 50m onwards), than the offshore oil & gas industry (from 300m onwards). In such cases a linear wave theory approach might not be sufficient to describe realistic wave shapes and the respective loads, especially in ULS conditions. In shallow or intermediate water depth environments, i.e. when the ratio between the water depth and the wave length becomes smaller than 0.5, waves need to be described by non-linear approaches, in order to account for their vertical asymmetry. (B) Wind turbines are dynamically sensitive structures, i.e. while the floating part can be assumed more or less rigid – the flexibility of the slender tower supporting a heavy and very dynamic rotor-nacelle assembly influences the global structure’s response significantly, especially in the pitch and roll degrees of freedom for taut moored systems. The current work evaluates the performance of engineering procedures, used in the design of bottom fixed offshore wind turbines, for the hydrodynamic ULS analysis of a FOWT. Dynamically sensitive topsides are included and water depths are considered, where wave shapes in the extreme sea states deviate from the 1st order description. A design basis is developed, which defines parametric extreme sea state caused by measured cyclonic storm conditions. The sea state parameters are defined, such that their reoccurrence probability is equal to an event occurring once in a lifetime of an offshore wind turbine structure, i.e. a 50 year return period event. It can be shown that the applied sea state is representative for harsh European offshore wind sites as well, providing hence a more general applicability. The floating substructure is represented by an industry inspired tension leg platform (TLP) equipped with an open source tower and rotor-nacelle-assembly equivalent based on the well-known NREL 5MW turbine, (Jonkman et al, 2009). An important aspect of the study is that it does not intend to verify a design. The industry inspired floater is designed for a water depth of around 100 meter and the respective tendon length was modelled to its full extent, physically as well as numerically. The waves however, are generated on a water depth that equal a prototype water depth of 56 meter. A redesign of the structure itself to the intended water depth is outside of the scope. However, the above circumstances allow a qualitative assessment of the numerical approaches to assess the behaviour of an FOWT in non–linear waves. This is also the reason, why actual measured values are not given, rather than ratios. The respectively developed numerical representation of the floating structure uses a hybrid approach of linear potential theory hydrodynamic coefficients and Morison's type loads, in order to include the wave nonlinearities. It is obvious that such an approach violates the linear potential theory assumptions, which consequently requires a validation of the hybrid model by a physical model test campaign, where an extensive set of data is acquired. The vertical asymmetry of the measured waves is numerically well covered by a 2nd order irregular sea state realisation. The physical wave realisations show maximum waves, of which one is an expected maximum wave in accordance with a Rayleigh distribution. The maximum waves are numerically represented by embedded Stream-function waves. The author compares the resulting bow tendon loading of the hybrid model to the measured responses, as a key performance indicator. 90% to 95% of the loads show a satisfying match, though the hybrid model over predicts the remaining 5% to 10% maximum loads by 32%, 34% and 29% for a linear irregular sea state, a nonlinear irregular sea state and a nonlinear irregular sea state with an embedded Stream-function wave, respectively. The limited number of sea states during the model tests as well as some model degradation during the physical testing campaign may cause some uncertainty of the measurements. However, based on the current status, it can be concluded that the non-linear wave embedded in a non-linear irregular background sea state provides the most controlled measure to assess critical ULS events for FOWT – though still towards the background of necessary further developments. The approach is similar to current state of the art ULS analysis of bottom fixed offshore wind turbines. So far it has however not been applied in floating structure designs, and the respective performance was unclear. The method is thought to be useful investigating the ultimate loading, high frequency responses from extreme transient effects and/or slack line events especially for TLP structures. The differences between the hybrid model results and the observations highlight the need for further optimization and the consequent potential to make FOWT cost competitive. Generally the study shows that the hybrid modelling approach might currently be sufficient for pre-Detailed Design stages, where higher degrees of conservatism are acceptable. However for multi-unit production the current method includes a too high degree of conservatism and requires further improvement. The presented work summarizes the current status and should be understood as a basis for subsequent developments by highlighting the challenges in FOWT ULS design. It is proposed to extend the current work by: - Increasing the number of model tests with focus on ULS behaviour in intermediate to shallow waters. - Increase the numerical model complexity by including the 2nd order inertial forces by quadratic transfer functions or by the instantaneous non-linear Froude-Krylov force. - Sensitivity study on Stream-function wave embedment parameters. - Investigate different substructures in order to understand the general applicability. Hypothesis "A hybrid model that considers linear diffractive wave excitation together with non-linear viscous wave excitation is conservative to resemble FOWT behaviour in ULS conditions".