Research On Prediction Of Pump Inspection Cycle And Downhole Equipment Working Life For Beam Pumping Wells

Pump inspection cycle(PIC)is an essential concept reflecting the working reliability of beam pump wells,and an important evaluation index for field production efficiency and level of management.Accurate prediction of pump inspection cycle can help to avoid oil well failure and to optimize the production system,showing great practical significance to improve field production efficiency and economic benefits.However,most of the existing PIC prediction methods are based on qualitative or macroscopical prediction,which are either unable to give accurate pump inspection cycle prediction for single wells,or with unacceptable calculation error,creating difficulty to provide guidance for on site operations.Regarding the issue above,this research carried out a simulation study of pump inspection cycle and downhole equipment working life based on finite element numerical simulation and data mining theories.Statistical prediction models of PIC and equipment life are established based on operation records of the tested field block,and their accuracy are discussed.Using ANSYS workbench,mechanical simulation models of sucker rod string,tubing string and pumps are established.Considering the impact of metal fatigue,corrosion,abrasion and fluid leakage,the failure processes of the above-mentioned equipments are simulated and the attenuation model of fatigue strength is established,giving the method for single equipment life prediction.Orthogonal test is designed for equipment life simulation according to the grey relevance analysis of influencing factors.Following the design of orthogonal table,a working life sample library is built using ANSYS simulation.The prediction model of life and residual life for downhole equipements is established using the tool of fuzzy evaluation,giving the mothod of quantitatively predicting the PIC of beam pumping wells.The calculation accuracy of this model is then discussed.According to the results,the average prediction error of linear regression models are27.04%,which does not meet the requirement of accuracy.The prediction model based on probabilistic distribution is accurate in describing inspection operation conditions macroscopically,yet the problem of poor pertinence exists.According to results of numerical simulation to the example oil well,fatigue life of suck rod under variable stress is 1004 days.This value declined by 24.20% and 56.72% under the impact of corrosion and eccentric abrasion.Life of oil tubing is 2625 days at its stress critical position.Under the impact of corrosion,tubing life decrease by 24.26% and 39.41% considering corrosion on inner surface or both surfaces.Under influence of abrasion,tubing life decreases by 31.8%.With the tubing surface wore through and fluid leakage start to appear,tubing failure will be triggered on day2569 due to unacceptable rate of leakage if not considering tubbing rupture failure.Dangerous position of structural failure on the pump is the movable valve cover,showing a working life of 1184 days.This value declined by 15.29% under the impact of corrosion.Without considering structural failures,pump working lives are 2017 days and 1805 days respectively considering barrel-plunger annulus leakage and valve ball leakage.Overall,structural fatigue break off is the main form of well failure,under certain circumstances,fluid leakage could also trigger well inspection operation.Prediction model of equipment life and residual life is established based on the above simulation analysis and using which 40 records of pump inspection cycles are calculated.The results show that the average relative error of PIC prediction calculation is 16.85%,which satisfies the requirement for preduction accuracy.The results of this study have realized the change of pump inspection cycle prediction from qualitative,macroscopic to quantitative and specific,providing effective guidance for the field production and well managements.