State-of-the-art Gaidai hypersurface risk prognostics approach for FPSO and FLNG dynamics

• Safety of Liquid Natural Gas transportation systems addressed. • Novel lifetime assessment method applied to Floating Production Storage & Offloading Unit dynamics. • Precise multi-dimensional reliability assessment was carried based on simulated data, partially verified by experiments. An offshore vessel, known as the Floating Production Storage & Offloading (FPSO) unit, is designed to produce and store crude oil or Liquefied Natural Gas (LNG), prior to its transportation by accompanying Shuttle Tanker (ST). To ensure operational safety during FPSO operations, dynamic mooring/hawser stresses have to be assessed. Operational concerns might arise from increased hawser tensions during specific unloading scenarios, given in-situ environmental conditions. This case study examined FPSO vessel dynamics, subjected to hydrodynamic waves-induced loads defined by in-situ environmental conditions, utilizing AQWA software. The state-of-the-art multi-dimensional spatiotemporal structural risk evaluation method has been benchmarked and performs particularly well, even based on limited underlying multimodal datasets. The underlying system’s dynamic data can be numerically Monte Carlo (MC) simulated, either physically measured. Multimodal Gaidai hypersurface risk assessment methodology, described in this case study, provides an additional capability to effectively assess the system’s dynamic hazard, failure or damage risks, as well as structural lifetime distribution of complex multidimensional dynamic systems, accounting for multiple failure modes, incorporating nonlinear inter-correlations between system’s key parts (components). The proposed multimodal reliability method can be applied to FPSO or FLNG (Floating LNG) vessel dynamics, as well as other vessel types. The primary benefit of the presented structural reliability methodology is that currently existing structural reliability and risk assessment techniques often struggle to address a large number of the system’s primary dimensions (parts or components). Based on a representative MC numerically generated dataset, an accurate multimodal reliability evaluation was conducted, and Confidence Intervals (CI) were reported.