Residual Life Prediction For Long-distance Transportation Oil&Gas Corroded Pipeline

Oil and gas pipelines embedded in soil throughout the year,. they can unavoidably be corroded due to the soil environments around them. As corrosion phenomena of the pipelines is more and more serious, pipeline operation has some danger.So, residual strength evaluation and life prediction of the oil &gas pipelines are very important for pipeline maintenance and replacement of decision-making and reliability management.Based on widely collected site investigation data, The paper analyses the corrosion mechanisms and factors of long-distance transportation pipelines. The factors in soil include water content, salt content (including Cl^-,SO₄^2-,HCO₃^-), electronic resistivity, pH value, soil oxidation-reduction potential. Using artificial neural network and support vector machine (SVM) method to analyse the corrosion rate data, and forecaset pipeline corrosion rate . The results can be seen, support vector machine (SVM) method for the forecasting corrosion rate results is better than artificial neural network.According to the characteristics of the form of corrosion ,corrosion are classified. and use the finite element method to solve the minimum remaining wall thickness. The cumulative method is used to determine the relationship between Mises stress of corrosion pipeline and corrosion defects size. The author develops software for calculating limits thickness of pipeline corrosion by Visual Basic, providing convenient calculations and tools for engineers and technicians as in the practical projects.