Study On Corrosion Prediction And Pre-maintenance Strategy Of Oil And Gas Pipeline In Service

Since the 1970 s,the global oil and gas pipelines in service have developed rapidly.In addition to railways,highways,waterways,and aviation,pipelines have also been listed as one of the main modes of transportation.Compared with the other four modes of transportation,pipelines are not only safe.It is stable,and it has the advantages of low operating cost and low loss,so it has been selected as the primary mode of transportation for oil and gas transportation in the global field.With the increasing demand for oil and gas on a global scale,the laying area of pipelines has become more extensive,and the laying environment has varied.The oil and gas leakage accidents caused by pipeline corrosion factors are also increasing year by year.In order to ensure social safety,fully grasp the corrosion conditions in the operation of oil and gas pipelines in operation,accurately predict and evaluate the corrosion depth and corrosion state of pipelines,and take into account the technical and economical nature of maintenance,with availability and cost factors as The foundation develops pre-maintenance strategies to maximize economic benefits.This paper mainly studies from the following aspects:(1)Combining the grey system theory and neural network,a high-precision CAGM(1,1)-BPNN in-service oil and gas pipeline corrosion depth combination prediction mode is proposed.Firstly,the original gray GM(1,1)model is constructed by using the detected target pipeline corrosion depth,and the CAGM(1,1)model is constructed by changing the background value,and the future corrosion depth of the target pipeline is calculated,and the post-test difference test is used.The method calculates the accuracy of the model;then,according to the obtained corrosion depth,the corresponding residual value is obtained,and BPNN is used to modify the training to achieve higher precision.Finally,the test section of a certain oil and gas pipeline is taken as an example to predict the corrosion of the pipeline.The results show that the combined model has higher prediction accuracy and stronger applicability when the sample data is small.(2)Using the traditional gray system to deal with the characteristics of less data and poor data and the characteristics of Markov theory to predict the future state,the structural parameter optimization GM-Markov model predicts the future corrosion development trend of the in-service oil and gas pipeline.Firstly,the corrosion degree of the target pipeline is evaluated for smoothness.Then,the parameter-optimized GM(1,1)model is constructed by changing the initial conditions,and the future corrosion depth of the target pipeline is calculated.Then,according to the predicted corrosion depth,Markov is used.The model quantitatively analyzes the future corrosion state of the in-service oil and gas pipelines and predicts its corrosion development trend.Finally,taking a submarine pipeline test section as an example,predicting the trend of pipeline corrosion development.The results show that the method is compared with the original GM(1,1).The average relative error of the model is reduced from 5.96% to 3.77%.The prediction result of the corrosion development trend is: in the second to fifth inspections,the pipeline is in a moderately corrosive state,and corresponding maintenance measures are required for the pipeline.In the next 6th to 9th inspections,the pipeline is in a state of severe corrosion and must be repaired or replaced with a new one.At the 10 th inspection,the pipeline is in a leaked state,and the new pipe must be replaced in time.The new pipe,which will pierce at any time,will bring huge losses to enterprises and society.(3)In order to improve the technical and economic effects of the periodic maintenance strategy of in-service oil and gas pipelines,an advanced maintenance strategy method for oil and gas pipelines based on availability and maintenance cost is proposed.Firstly,the CAGM-BPNN model is used to predict the pipeline corrosion depth.Then,according to the predicted corrosion depth,the Markov theory is used to divide the state interval to establish the pipeline corrosion state transition matrix.Finally,the state transition matrix is repaired according to each combined maintenance strategy.The matrix,and calculate the probability of the target pipeline in each corrosion state interval,and then calculate the availability of the target pipeline.Based on this,combined with the cost information to calculate the total cost expectation value of each combination strategy,thereby selecting the optimal pre-maintenance strategy.According to the above researches,the CAGM-BPNN-Markov hybrid corrosion prediction model can accurately predict the future corrosion status of the in-service oil and gas pipelines,and based on this,integrate the availability and cost factors to construct a set.From the corrosion prediction stage to the complete evaluation system of the maintenance strategy stage,to ensure the safe operation of the pipeline and provide a theoretical basis for the pipeline pre-protection work.