Long-term dynamic monitoring of offshore platform for damage assessment

This paper, through the discussion of an illustrative case study, investigates the effectiveness of a long-term dynamic monitoring system as an integrative tool for offshore platform damage assessment. The considered platform is a steel structure, named VEGA-A, operating in the Sicily channel near Siracusa. This platform, one of the largest built in Italy, has been equipped since 1988 with a monitoring system that collects both environmental (a current meter, a depth gauge and a system for detecting speed and direction of wind) and structural (by means of accelerometers installed on the top side of the platform) data. In a first part of the paper, variation of main frequencies and mode shapes over these 30 years of activity of the platform are discussed and correlated with the main storms that have affected the structure. In addition, a statistical analysis of the accelerometer signals is performed with the aim to identify potential warning thresholds and alarm. Subsequently, an identified numerical model of the platform has been built and employed to perform time-history analyses according to several (artificial) damage scenarios. The numerical results (namely, acceleration time-histories at different levels) have been employed to investigate the ability of an improved structural monitoring system to recognize the artificial damage. It is demonstrated that a modified Kernel-PCA with the inclusion of the normalized mode shapes, is able to recognize and localize damage occurrence (both on the legs and on the main horizontal and vertical bracing), provided that the monitoring system allows to characterize mode shape.