Research On Corrosion Warning Of Oil And Water Gathering Pipeline In Oil Field

Pipeline transportation is one of the most widely used transportation methods of oil and gas resources at home and abroad.Most of the gathering and transportation pipelines are metal pipelines,which are prone to metal corrosion,resulting in waste of oil and gas resources and environmental pollution.In this paper,the oil and water gathering and transportation pipeline in Changqing oilfield is taken as the research object,and the corrosion factors of oil and water gathering and transportation pipeline are systematically studied by means of variance analysis and response surface experiment,and the corrosion rate prediction model based on fuzzy neural network is established.In combination with SY/T 6477-2017 Evaluation method for residual strength of oil and gas pipeline with defects,30 pipelines in service were evaluated for corrosion defects.The main achievements of this paper are as follows:（1）Variance analysis results of influencing factors of uniform corrosion:Since the F value of partial pressure of carbon dioxide,partial pressure of hydrogen sulfide,chloride ion concentration and temperature are all greater than F_0.01（3,8）=7.5910,while the F value of calcium and magnesium ion concentration and flow rate is less than F_0.05（3,8）=4.0662,Therefore,the partial pressure of carbon dioxide,the partial pressure of hydrogen sulfide,the concentration of chloride ion and the temperature have a very significant effect on the uniform corrosion rate of the pipeline,while the concentration of calcium and magnesium ion and the flow rate have no significant effect on it.（2）Variance analysis results of influencing factors of local corrosion:Since the F value of partial pressure of carbon dioxide,partial pressure of hydrogen sulfide,chloride ion concentration and temperature are all greater than F_0.01（3,8）=7.5910,while the F value of calcium and magnesium ion concentration and flow rate is less than F_0.05（3,8）=4.0662,Therefore,the partial pressure of carbon dioxide,the partial pressure of hydrogen sulfide,the concentration of chloride ion and the temperature have a very significant effect on the local corrosion rate of the pipeline,while the concentration of calcium and magnesium ion and the flow rate have no significant effect on it.（3）Under uniform corrosion rate,the interaction of influencing factors is as follows:Partial pressure of carbon dioxide and hydrogen sulfide>Partial pressure and temperature of hydrogen sulfide>Partial pressure of hydrogen sulfide and chloride ion concentration>Chloride concentration and temperature>Partial carbon dioxide pressure and the chloride ion concentration>Partial carbon dioxide pressure and the temperature.（4）Under the local corrosion rate,the interaction of influencing factors is as follows:Partial carbon dioxide pressure and Partial pressure of hydrogen sulfide>Partial pressure of hydrogen sulfide and chloride ion concentration>Partial carbon dioxide pressure and the chloride ion concentration>Chloride concentration and temperature>Partial pressure and temperature of hydrogen sulfide>Partial carbon dioxide pressure and temperature.（5）After the number of iterations reaches 1710,the mean square error（MSE）of the uniform corrosion prediction model based on fuzzy neural network reaches the accuracy requirement of less than 0.01.The actual and predicted values in the training,verification and testing stages were evenly distributed around the regression line,and the regression coefficient in the training stage was 0.978,the verification stage was 0.980,and the test stage was 0.950,which indicated that the predicted value of the model was highly correlated with the actual value,and the predicted value was more accurate.（6）After the number of iterations reached 1720,the MSE of the local corrosion prediction model based on fuzzy neural network reached the accuracy requirement of less than 0.01.The actual and predicted values in the training,verification and testing stages were evenly distributed around the regression line,with the regression coefficient of 0.990 in the training stage,0.979 in the verification stage and 0.987 in the testing stage,which indicated that the predicted value of the model was highly correlated with the actual value,and the predicted value was relatively accurate.（7）Based on the relevant corrosion data collected in the field,this paper systematically predicted the oil and water gathering and transportation pipeline by using the corrosion rate prediction model,and evaluated the corrosion defects of 30 pipelines according to the specific requirements of SY/T 6477-2017 Evaluation method for residual strength of oil and gas pipeline with defects,and the evaluation effect was good.（8）Developed according to this article research content,oil and water pipeline corrosion analysis of early warning system,the software has a monitoring data acquisition module,the multiple factor variance analysis module,the corrosion rate prediction module based on response surface method,the corrosion rate of prediction module based on fuzzy neural network and corrosion warning module that several major functional modules,such as reference for technical personnel.