Research On Corrosion Risk Assessment Method Of Subsea Crude Oil Pipeline Based On Degradation Modeling

Due to the complex and harsh operating environment of subsea crude oil pipelines,corrosion is one of the main reasons for their degradation and failure.This dissertation is supported by the National Natural Science Foundation of China Youth Project “Research on the Dynamic Risk Assessment Method of Leakage Disaster of Offshore Gas Pipeline Based on the Evolution of the Disaster-Causing Process” and the Open Fund of the National Engineering Laboratory for Marine Geophysical Prospecting and Exploration Equipment “Study on Diffusion Mechanism and Environmental Disaster Risk Assessment of Marine Crude Oil Pipeline Leakage”.Research on corrosion risk assessment method of subsea crude oil pipeline based on degradation modeling has been carried out.Certain research results have been obtained in the corrosion degradation prediction method of subsea crude oil pipelines,pitting corrosion growth prediction and failure probability assessment methods of subsea crude oil pipelines,and corrosion state assessment methods of subsea crude oil pipelines,which provides support for ensuring the safe and healthy operation of subsea crude oil pipelines and corrosion risk management.The main research progress is as follows:1.Research on corrosion degradation prediction method of subsea crude oil pipeline In the context of industry 4.0,the development of a data-driven corrosion rate prediction method for subsea crude oil pipelines is an urgent need to realize the digitization of pipeline systems,and is the key to improving the intelligence level of pipeline corrosion risk management.Therefore,a corrosion rate prediction method for subsea crude oil pipelines is proposed by combining three data-driven methods,namely,Principal Component Analysis(PCA),Artificial Bee Colony Algorithm(ABC)and Support Vector Regression(SVR).In the method framework,SVR is used to build the prediction model,PCA is employed to select the input variables of the model,and ABC is adopted to optimize the hyper-parameters of the model.The feasibility and effectiveness of the method are verified by simulation experiments,which can be used as a useful online tool to support the safety and digitization of subsea pipeline systems.2.Research on pitting growth prediction and failure probability assessment of subsea crude oil pipeline Randomness and uncertainty are important challenges faced by research on pitting corrosion growth prediction of subsea crude oil pipelines.The use of probabilistic methods is an effective way to study the growth process of pipeline pitting and to achieve reliable assessment of pipeline failure probability.Therefore,the failure probability assessment method of subsea crude oil pipelines based on pitting growth is proposed by integrating Bayesian network(BN)and Hierarchical Bayesian analysis(HBA).This method can effectively overcome the randomness and uncertainty of the pitting process.In the method framework,according to the linear growth model of steel pipe corrosion,the BN equation node is used to predict the growth of pitting depth of subsea crude oil pipelines,and the time distribution law of pitting pit depth of pipelines is obtained.The HBA is used to evaluate the pipeline failure probability,and the probability distribution of the time required for the subsea crude oil pipeline to reach a certain degradation level is obtained.The feasibility and applicability of the method are demonstrated through a case application analysis,which can provide support for the prediction of pitting pit depth growth and the failure probability evaluation of subsea crude oil pipelines.3.Research on corrosion state evaluation of subsea crude oil pipeline Considering three common types of corrosion,namely uniform corrosion,pitting corrosion and microbiologically influenced corrosion(MIC),a probability-based assessment method for corrosion states of subsea crude oil pipelines is proposed.In the method framework,a probabilistic network model of pipeline corrosion causality is established by combing the corrosion influencing factors of subsea crude oil pipeline and their interdependencies.The Bayesian parameter learning is applied to quantitatively evaluate the risk probability of the final nodes of the network model(i.e.,corrosion rateand pit depth)under different corrosion levels,so as to judge the corrosion risk degree of the subsea crude oil pipeline.The proposed method is illustrated by a case study analysis,which can provide support for the establishment of a corrosion risk early warning system for subsea crude oil pipelines.