Real time prediction of operational safety limits for dynamic positioning of an FPSO in a Deepwater Artificial Seabed system

Abstract

Aiming to overcome the limitations of conventional offshore field development concepts (dry tree or subsea tree) for petroleum production in ultra-deep water, a new alternative offshore field development solution, termed as Deepwater Artificial Seabed (DAS) system, is proposed. The DAS system works in concert with dynamic positioning (DP) floaters, such as dynamically positioned Floating Production, Storage and Offloading (FPSO) vessels. Rather than relying on the passive mooring system, the DP maintains the reliable position of the FPSO with steering and propulsion units. Nonetheless, critical DP failures, which has potential to cause the drift-off scenario for the FPSO, poses a serious threat to the structural safety of the DAS system. Therefore, it is crucial to establish operational limits for the DP FPSO to prevent such accidents. In this study, a 3-phase probabilistic modelling methodology is proposed to predict safety limits for the operation of the DP FPSO. A surrogate model is established by the Support Vector Machine (SVM) algorithm so as to decrease the computational cost due to the generation of large statistical samples. The statistical distribution of the operational safety limits of FPSO is simulated by the successive approximations through the fully-coupled drift-off analysis. The accuracy of the proposed methodology is verified by a series of mathematical tests. In order to validate the effectiveness of the methodology, the safety limit prediction of the FPSO for the DAS system is taken as a case study. The critical positions of the FPSO are predicted in real time and provides ample time and information for operators’ decision-making by the visualization of the safe moving range of the FPSO. The study contributes to the safety control of DP operations on floating production units in an efficient manner.