A well-known drawback of conventional wave monitoring systems, such as wave buoys, is that they experience a loss of accuracy in extreme wave conditions. Also, most of them require important initial investment and/or high maintenance costs. Over the last few years, directional wave inference obtained from the record of vessel motions is a technique that has significantly grown as complement to traditional methods. This article presents a feasibility study on the use of the motions of a semi-submersible platform for performing wave inference. Experiments were carried out at the USP wave basin (CH-TPN) using a 1:120 scale model of a large semi-submersible platform in operational condition and five different headings. In order to provide an extensive test matrix, the experimental campaign included a set of 32 different irregular waves (sea conditions) for each heading, selected from the scatter diagram of the Norwegian sea and covering many of the sea states of interest for this research. Moreover, each sea condition was obtained using the most appropriate type of energy spectrum (JONSWAP or Torsethaugen). Bayesian inference motion-based method was adapted for the semi-submersible platform by the proper adjustment of the hyper-parameters. The estimations obtained with the Bayesian wave inference method, using the semi-submersible recorded motions, were confronted with the directional wave spectra measured during the calibration process from an array of wave probes. The results attested that the method was able to capture all of the wave conditions tested during the experimental campaign with reasonable accuracy, even the more extreme cases. They suggest that the semisubmersible platform may indeed be a promising alternative for inferring severe sea conditions.

This content is only available via PDF.
You do not currently have access to this content.