Remaining Life Prediction And Maintenance Strategy Study Of Submarine Corroded Pipelines Under Imperfect Maintenance

Part of China's submarine pipeline equipment has gradually entered the aging stage,which serves as the main carrier of subsea oil and natural gas.Pipeline defects caused by corrosion factors is the main reason for the reduction of its remaining life,making it difficult and costly to repair it.To ensure the safety of its operation,preventive maintenance activities will be implemented.However,due to the constraints of economics and reliability,the number of preventive maintenance activities is usually limited,and it is difficult to repair the pipeline wall corrosion defects to the original good state.Based on the analysis of the current remaining life prediction methods for corroded pipelines,we propose that in the early stage of pipeline operation,preventive maintenance activities will be used to obtain a longer service life for the pipeline,which considering the influence of corrosion depth and corrosion rate under imperfect maintenance to predict the remaining life of the pipeline.It can provide a theoretical basis for the continued use,inspection,maintenance and decommissioning of the submarine corrosion pipeline in active service.The main research contents are as follows:(1)A staged degradation model considering the impact of imperfect maintenance activities on the corrosion depth and corrosion rate of submarine pipelines was proposed.The model parameters were calculated using maximum likelihood estimation,Markov chain Monte Carlo simulation and Bayesian methods.We comparing the imperfect maintenance model with the only considering the impact of imperfect maintenance activities on the corrosion depth and repair as new model to verify the validity of the remaining life prediction of the proposed model.(2)A phased degradation model of corroded pipelines considering the impact of different maintenance depth distribution was proposed,based on the threshold conversion idea the probability density function of the remaining life in the first hitting time is derived to realize the real-time remaining life estimation of the corroded pipeline.The maximum likelihood estimation is used to calculate the model parameters based on the state inspection data and maintenance data to realize real-time remaining life prediction of corroded pipelines.The proposed method was compared with the gamma distribution of the maintenance depth under the imperfect maintenance and the normal distribution under the condition of complete maintenance.The validity of the prediction result of the remaining life of the normal maintenance depth distribution under the imperfect maintenance was verified.(3)For the corrosion of submarine pipelines,a concept of condition detection based on the operation state of corrosion of submarine pipelines is proposed,and the limit state function of the submarine pipeline structure was established.The event tree model is used to describe for the repairs between two consecutive state inspection,the arrival time of the new corrosion defects conforms to the Poisson distribution.Judging the method of taking maintenance measures according to the severity of corrosion defects on the pipe wall,which can slow down the growth rate of the corrosion defects of the pipe wall and extend the remaining life.Considering the uncertainty of the detection tools,an optimized inspection model for the subsea pipeline was established,and the time interval for pipeline inspection was determined.By comparing the cost with repair as new state,the utility function under different reliability illustrates the economics of imperfect maintenance.It provides a certain guiding significance for enterprise pipeline managers.