Monte carlo simulation to consider uncertainty in the reliability analysis of dynamic positioning systems

Abstract

Nowadays, Dynamically Positioned (DP) units are responsible for most offshore oil exploitation operations, including drilling and maintenance campaigns. Due to the large congestion of the oil fields, keeping the vessel position, despite the environmental forces, is a critical issue. Recently, some efforts using fault trees and offshore industry-reported component failure rates were made to quantitatively model the reliability of DP systems typical configurations. Despite this approach success in bringing a numerical estimation for the fail probability of a DP system, it failed in deal with the uncertainties related to the model and to the data. The volume of fail data available in the literature differs significantly and the choice of a wrong parameter, or a combination of them, may cause the model to considerably diverge from reality. To deal with this issue, this paper introduces the use of Monte Carlo Simulation (MCS) to consider uncertainty in the Reliability Analysis of Dynamic Positioning Systems (DPS). The proposed methodology uses MCS and a fault tree approach to build a nonparametric DP system's reliability probability density functions (pdf), rather than a single reliability result. The model is then used to analyze the reliability and a path to analyze the availability of a DP system, considering the impact of data uncertainties on the system reliability, showing the effects of wrong choices regarding components fail rates on the global DP unit reliability.