Wave run-up on offshore wind turbine foundations

Wave run-up on appurtenances like boat landings, ladders, platforms and J-tubes of Offshore Wind Turbine foundations have in some cases caused severe and costly damage to these installations. A well-known example of large run-up was registered at the foundations of Horns Reef Offshore Wind Farm, Denmark.

As specified in DNV RP-C205 (2010), the local wave run-up should be determined based on wave run-up factors that are derived from model tests. A good presentation of model tests assessing wave run-up on cylindrical and cone-shaped foundations is given by De Vos et al. (2007) and by Lykke Andersen et al. (2010) for monopile foundations. This formulation has found widespread use in the process of secondary steel design for monopile foundations. However, both De Vos et al. (2007) and Lykke Andersen et al. (2010) applied the measured 2% and maximum wave heights for prediction of 2% and maximum run-up, respectively. Therefore, a realistic wave height distribution needs to be defined for design purposes. Assuming Rayleigh distributed waves for a shallow water site, as it has often been common practice in industry, may lead to highly conservative extreme run-up levels.

In the present paper the measured data and the methodology for determination of run-up given by De Vos et al. (2007) and Lykke Andersen et al. (2010) are reassessed with the purpose of avoiding overly conservative designs. It is shown that uncritical use of the Rayleigh distribution in combination with the conventional method for wave run-up estimation can lead to severe over-prediction of the run-up height. On the other hand use of the wave height distribution by Battjes and Groenendijk (2000) leads to very accurate predictions of the run-up levels. For conditions, in which the maximum wave height Hmax is influenced by wave breaking and the Battjes and Groendijk wave height distribution is not accepted, an alternative simple method based on energy flux is proposed. Using the significant wave height HS as an input parameter makes this alternative method extremely useful for design as the wave height distribution is typically not known. A comparison is made between wave run-up assessments according to Lykke Andersen et al. (2010) and the proposed alternative methodology is made for existing model test data as well as for extreme North Sea design conditions.