Study Mechanicals Properties Of Deep Shale And Prediction Of Breakdown Pressure

Deep shale oil reservoir has the characteristics such as low reservoir abundance,bad reservoir nature,low well productivity,high production cost and difficulties in well stimulation.For getting industrial productivity,these oil fields should be applied the techniques of acidizing and fracturing in the process of exploration and mining.Breakdown pressure is fundamental to optimize fracturing design parameters,such as pumping rate and the ratio of sand and fracturing fluid,and also is a prerequisite for preparing completion tubing,fracturing trucks,fracturing well head,surface manifolds.Inaccurate estimation of breakdown pressure may result in undesirable pumping rate,narrow fracturing safe pressure window,even unsuccessful hydraulic fracturing.Shale in Longmaxi formation,Southern Sichun Basin indicate nonlinear failure regime at high temperature and high confining pressure.Applying the classical elastic model for predicting the breakdown pressure can cause severe error.We have built breakdown pressure dataset through rock triaxial testing and collecting history fracturing data over one hundred well.We present a breakdown pressure prediction method based on multivariate linear regression using rigid rule to optimize the model.the obtained error mean rate is 7.23%.After that,we train a new ANN breakdown pressure prediction model,compared to multivariate linear regression model,the mean error rate is only 6.10% indicating a significant promotion.