Estimating on-site wave spectra from the motions of a semi-submersible platform: An assessment based on model scale results

This article presents the results of a preliminary investigation on the use of the motions of a large-volume semi-submersible platform as the basis for estimating waves on-site. The main focus of the application is on the harsher wave conditions, for which other measurement devices are known to present limitations. As a first step in this investigation, a series of tests were performed with a scaled model in a wave basin. A wide range of waves was tested in order to evaluate how the estimation errors behave with respect to wave period and amplitude. Moreover, since the statistical inference relies on a linear dynamic model, the influence of uncertainties on the unit's RAOs must be assessed. Since the non-linear viscous effect on resonant responses is known to be one of the main sources of error in this regard, different approaches for predefining the viscous damping levels are tested and compared. The results attest that, at least in the controlled conditions of the tests, the inference method is able to provide accurate estimations of the extreme waves conditions that were tested. A simple method for predicting the damping based on the recorded motions is shown to improve the estimations.