Probabilistic-Based Time Function Model Of Corrosion Progress In Subsea Pipelines

Abstract

Corrosion was brought to the highest concern in marine structure safety, especially in the offshore oil and gas industry. Internal corrosion damage on subsea pipeline is unpredictable and may cause leakage and structural failure. Corrosion on crude oil and gas pipelines is needed to conduct a reliability assessment. It is necessary to have a mathematical model, such as the inverse cumulative distribution function of log-logistic distribution which can model and predict the maximum pit depth corrosion for any age of subsea pipelines. This research is aimed to propose a corrosion progress (pit depth) formulation model for subsea pipelines based on the log-logistic distribution function for any age of pipeline by changing a parameter function. The process for corrosion model development starts from collecting corrosion data for the subsea pipelines. Then the best model by carry a statistical analysis were selected. The choice was made when the log-logistic distribution gave the lowest statistical value. The log-logistic model was compared with the 2-parameter family distribution model. Parameters that were obtained by using the maximum likelihood estimation method were used to determine the best parameter for modelling.