PREDICTION OF IRREGULAR MOTIONS OF FREE-FALL LIFEBOATS DURING DROPS FROM DAMAGED FPSO

A Free Fall Lifeboat (FFLB) which is evacuating from a damaged host in storm conditions must to be able to safely run away from the host. For a safe evacuation, the FFLB must first of all avoid irregular motion ("log dive'), and secondly resurface at a sufficient distance away from host (headway) to be able to run away. The design standard [1] requires a controlled motion trajectory for the FFLB in waves. A drop simulator for random drops of FFLBs from a floating host in storm conditions, intact or damaged has been developed. Time histories of the host motions are generated and random drops of the FFLBs are done during three hour duration of each sea state. The FFLB point of impact in the wave is identified and the local wave height, wave length and hit point in the wave cycle is found for each of the impact points. A structured database with results from a large number of Computational Fluid Dynamics (CFD) simulations has been made for a specific FFLB. In this CFD screening process the lifeboat has been dropped with variations in wind and waves, as well as varying conditions of host during launch. This database forms the basis for a regression analysis used to estimate responses for each drop of free fall lifeboat found from the drop simulator. The present study proposes a definition and a method to identify irregular motions for FFLBs, and this definition is used as a motion indicator in the regression analysis. By using the proposed motion indicator the regression analysis provides percentage of acceptable vs. non-acceptable motion trajectories for a given intact or damaged host during a storm condition. The worst conditions are found, and can be used for further analysis of headway, i.e. the ability of the FFLB to escape from the host after resurfacing.