The Remaining Life Prediction of Buried Pipeline Based on the MaximumExtreme Value Distribution

摘要

Corrosion is recognized as one of the most important degradation mechanisms thatrnaffect integrity of metallic buried pipeline. The remaining life prediction of pipelinernis to forecast the thinning trend of wall thickness, and forecast remaining life on thernpremise that remaining strength of pipeline meets the pipeline operation and safetyrnrequirements. This paper introduced the prediction method of the most probablernmaximum corrosion depth of the entire pipeline: the maximum corrosion depths inrneach of sampling pipeline section were taken out to form a group of sequences, andrncalculated the corresponding cumulative probability; Statistical parameters ofrnGumbel distribution function were calculated by the Least-Squares method; Basedrnon the maximum extreme value distribution and return period, the calculationrnformula of the most probable maximum corrosion depth of entire pipeline that wasrnestimated by corrosion data of a small number of sampling pipeline sections wasrnobtained by formula derivation. The calculation formula of maximum depth ofrncorrosion defect was derived by the method of DNV allowable stress. For thernpipeline with one inspection, the relation between remaining wall thickness and timernwas established, and the calculation formula of remaining life prediction of pipelinernwas derived based on it. In this paper, remaining life of a domestic pipeline wasrnpredicted and analyzed by the method of extreme value distribution. By comparisonrnof linear model and exponential, the calculation result of linear model is morernconservative than that of exponential model.