The Research On Mechanical Analysis And Residual Life Prediction Of Casing With Corrosion Defect

In the process of oil and natural gas exploitation,associated gases such as CO₂ and H₂S experience to induce casing corrosion,resulting in local thinning and strength reduction of casing,even casing damage,and shutdown of oil and gas wells and environmental pollution.At present,empirical formulas and other methods are used to conservatively predict the casing life in engineering.How to consider the influence of corrosion on casing and determine the residual life of casing has become the difficulty and hotspot of current research.Based on experiments,finite element simulation and BP neural network,the remaining life prediction method of casing is studied in this paper.The main achievements are as follows:(1)Based on the wellbore temperature field model,the triaxial thermal stress distribution model of casing is derived.Based on this model,considering the superposition of thermal stress field and pressure field,a three-axis pressure field distribution model of casing considering corrosion defects is established.(2)The corrosion rate experiment was carried out,and the power law model was used to characterize the relationship between corrosion rate and time.The fitting R² was 0.989,indicating that the power law model can better characterize the change of corrosion rate with time.The results of microstructure analysis of corrosion products show that the corrosion pits are approximately ellipsoidal,and the length and width of corrosion pits are much larger than the depth of corrosion pits.The results of corrosion product membrane energy spectrum analysis showed that the corrosion products were mainly iron oxides and iron sulfides.Combined with previous experimental results,the corrosion rate prediction models of CO₂partial pressure,H₂S partial pressure,temperature and time were established based on BP neural network.The results showed that the prediction accuracy of the prediction model was high,and the error was within 10%.At the same time,the accuracy was further verified by comparing with the commonly used prediction models.The establishment of corrosion rate prediction model effectively improves the prediction accuracy of corrosion rate,and overcomes the limitation of single factor in the current model.(3)Combined with the microscopic morphology analysis results of corrosion test,the finite element model of corrosion defect casing-cement sheath-formation was established considering the temperature-pressure coupling effect.For single pit corrosion defects,the maximum equivalent stress increases with the increase of corrosion pit radius(0.5?2mm).With the increase of corrosion pit diameter-depth ratio(1.5:1?3.25:1),the maximum equivalent stress increases gradually,then tends to be stable and no longer changes.For composite corrosion defects,74 schemes were designed considering seven parameters including corrosion pit spacing,corrosion pit overlap spacing,corrosion pit volume ratio,temperature,internal pressure,elastic modulus and Poisson's ratio,and the sensitivity of various factors to casing strength parameters was analyzed.The results show that the larger the corrosion pit spacing(0.1?120mm)is,the maximum equivalent stress decreases sharply first and then tends to be stable.The larger the overlap distance(0.4?1.8mm),the maximum equivalent stress tends to be stable first and then increases gradually.When one aspect ratio(1.5:1)in the pit remains unchanged,the maximum equivalent stress increases with the increasing aspect ratio(1.6:1?2.3:1)in the other pit,and decreases with the increasing aspect ratio(1.5:0.2?1.5:0.9)in the other pit.In addition,the maximum equivalent stress varies linearly with temperature,internal pressure,elastic modulus and Poisson's ratio.(4)Based on BP neural network,the casing strength evaluation model is established by using 74 groups of finite element simulation results of composite corrosion defects.The calculation shows that the error between the predicted equivalent stress and the finite element simulation equivalent stress is less than 1.5%,and the error is small,which can accurately predict the equivalent stress of the casing.By combining the corrosion rate prediction model with the equivalent stress evaluation model,the residual life of casing is determined from the residual wall thickness of corrosion and the yield strength of casing,and the residual life prediction chart considering the casing temperature,casing pressure,elastic modulus and Poisson's ratio is drawn.The establishment of remaining life prediction chart has certain guiding significance for safe production of oil and gas fields.