Response-Based Design Methods for Motion of Turret Moored FPSO

Abstract

In this paper, response-based methods are applied to determine response maxima for a floating structure under design environment. Two approaches are adopted to evaluate maximum heave, roll and pitch motions of turret moored FPSO: the conventional long-term simulation approach and the IFORM (Inverse First-Order Reliability Method) approach. First the distributions of environmental variables which are given in metocean data in a form of scatter diagram are described by random variables of relevant probability distribution functions. The simultaneous occurrence of wave, wind and current is taken into account using the joint probability models, and a number of sea-states are generated for the long-term simulations. Heading analysis is performed for a turret moored FPSO in order to obtain relative direction of wave. Short-term analysis is carried out for the numerous generated sea-states with the relative incidences and then long-term analysis follows to predict the maximum motion with the Weibull fitting by method of moments. The sea-state which most influences the maximum responses is chosen according to the probability contribution. In the IFORM approach, the response surface is constructed in the form of polynomials of wave height, direction and period. And the IFORM analysis is carried out using the distribution of environment variables and the response surface functions, and the maximum resultant heave, roll and pitch motions are evaluated for the given design return period. The combinations of wave height, period and direction causing response maxima are identified as well. The 100-year values of heave and pitch from the IFORM analysis is determined higher than those from the long-term simulation due to the simplified description of heading distribution whereas similar sea-states are identified by the two approaches. The 100-year roll motion can be determined using IFORM by neglecting the conditional probability of direction and period to height. Similar 100-year RP roll values are estimated from both the long-term simulation and the IFORM analysis, however, the sea-states of maxima occurrence are different because wave height governs the probability exceedence much more than period and direction in the IFORM. Finally the values from response-based methods get consistently lower than those from traditional extreme analysis which calculates responses for the 100-year extreme environment.