A Data-Driven Framework for Fatigue Monitoring of Operating Flexible Risers

An accurate estimation of riser fatigue life plays a critical role in the efficient structural integrity management of operating flexible risers. A more reliable fatigue life estimate could, for example, enable life extension for aging flexible risers or facilitate the use of a more targeted intervention program. This paper introduces a data-centric paradigm for the fatigue monitoring of flexible risers which uses measured host vessel motion data as part of the riser global response analysis. The performance of this method is assessed by comparison with the conventional analysis method, which uses a condensed wave scatter diagram combined with vessel motion transfer functions to assess global response. In the study, Python packages Scipy and Pandas are used to process raw wave and motion data obtained from wave measuring buoys and sensors located on the host vessel, respectively, to obtain the input load data, and non-linear time-domain analysis is used to calculate the global riser response due to both irregular waves and the measured motions. Tensions and curvatures obtained from the global riser response analysis form the input to a local stress analysis, which uses a proprietary software tool to estimate the stress response at the tensile and pressure armours. The output is then combined with a rainflow cycle counting technique and an S-N curve approach to perform the local fatigue assessment at the riser critical points. This paper compares the predicted and measured key input parameters (such as vessel heading, motions etc.) that affect the estimated fatigue life of a representative flexible riser connected to an operating Floating, Process, Storage and Offloading (FPSO) host vessel. Finally, the paper provides practical recommendations for using measured host-vessel motion data as part of a rigorous fatigue monitoring of connected flexible risers.