Probabilistic Analysis and Risk Assessment of Deep Water Composite Production Riser Against Fatigue Limit State

Abstract

Deep water composite risers are subjected to randomly fluctuating loads, induced by wind and waves in the presence of fluctuating axial tension which may be critical in deep sea conditions. Therefore, risers experience the extreme bending and randomly fluctuating stresses throughout their service life. Cumulative fatigue damage is a critical assessment of riser life in the presence of large dynamic stresses. Probabilistic analysis and risk assessment of composite risers for cumulative fatigue is a vital design requirement for its satisfactory service and survival for stipulated period. Without addressing the reliability assessment, composite risers may not be recommended for deep water oil and gas exploration and production. Hence, the reliability assessment is a critical issue that is to be addressed for the safety of the deep water composite riser. It is studied for the entire system for all possible sea states occurring in the exploration region. Unlike conventional risers, the wall structure of a composite riser is more complicated. Therefore, multiple failure mechanisms are used jointly to assess the safety of the composite riser. Fatigue reliability is a challenging task due to complex nature of dynamic response and associated uncertainties caused by the material and external loads. The present study is focused on reliability assessment using stochastic finite element analysis. Response time histories for random sea plus current have been obtained. Requisite numbers of sea states are considered for the simulation of a wide range of off-shore environment and estimation of accumulated damage. By using the S-N data, damage fractions are calculated then summed linearly using Miner-Palmgren rule. The total damage has been obtained by summing the accumulated damages over all the sea states under consideration. Non-linear limit state function is derived based upon the above given approach to calculate the fatigue life. Important uncertainties associated with random variables are considered while deriving the limit state function. Numerical methods, such as Monte Carlo simulation and Advanced First Order Reliability Method, are used for the calculation of the reliability. The sensitivities of various random variables on overall probability of failure have been studied and design points have been located on failure surface. Probabilities of failure for important parameters are investigated.