Theory And Application Of Numerical Simulation Of Coupling Geomechanics And Fluid Flow Based On Embedded Discrete Fracture

The matrix in unconventional fractured reservoirs provides oil and gas storage space,and hydraulic fractures,induced fractures,and natural fractures provide fluid flow channels.At present,the actual historical data of oilfields have shown that geomechanical characteristics play an important role in the production of fractured reservoirs,which will lead to the change of some key parameters for fluid flow,such as bedrock porosity and permeability.In addition,tight reservoirs are "fracture-controlled reserves",and rock deformation will cause the change of fracture aperture,which has a great impact on the productivity evaluation of fractured horizontal wells.Therefore,to accurately simulate the dynamic fracture behavior and fluid flow mechanism in such fractured porous media,it is necessary to comprehensively consider the mechanical properties of fluid and rock.Firstly,an improved Green element method(GEM)with second-order accuracy and suitable for meshes generation such as rectangle and triangle is proposed.The unsteady mass transfer between the local matrix mesh and fracture element is obtained,and the multi-scale grid is adopted to reduce computing time.Compared with the traditional EDFM under the assumption of linear pressure distribution,the novel model coupled with improved GEM can more accurately describe the transient pressure and flux distribution of fluid flow in porous medium and can be applied to the solution of complicated fluid flow models such as the problem of heterogeneous,complex reservoir boundary,fracture networks,SRV composite zoning.Secondly,a hybrid numerical discretization method coupled flow and geomechanics is proposed,in which the extended finite element method is used to solve the elastic deformation of rock,and the improved GEM is used to accurately calculate the unsteady fluid flow between matrix and fracture.The two numerical schemes are fully coupled,and the time term of the overall calculation scheme is solved fully implicitly.The combination of EDFM and dual-porosity effective stress model realizes the dynamic fracture networks characterization of hydraulic fractures,induced fractures,and natural micro-fractures,and can accurately simulate the dynamic production characteristics in the exploitation process of tight oil reservoirs.In addition,given three limitations of traditional EDFM in dealing with complex fractures,the coupled model of multiphase flow and geomechanics is extended to a wider range of applications,including dynamic fracture propagation simulation,complex fracture geometry simulation,dynamic closure simulation of different types of fractures.Finally,the application of flow and geomechanics coupled simulation in the development of tight oil reservoirs is studied,and an integrated workflow of reconstruction and correction of real fracture networks aided by automatic history fitting is established.The geometric information of fracture networks is controlled by the data of hydraulic fracturing,core,and microseismic events.The initial morphology of complex fracture networks is generated by fracture propagation forward modeling.The uncertain parameters of reconstructed fracture networks are corrected by an intelligent algorithm named the Ensemble Kalman Filter.The whole development process is simulated by the proposed flow-geomechanics coupled model.By analyzing the evolution law of the stress field,a practical optimization method of energy replenishment before refracturing is formed,and more accurate production prediction results are obtained.This paper establishes a series of flow-geomechanics coupled mathematical models based on embedded discrete fractures,modifies and improves the corresponding solving methods of numerical simulation,realizes the high-precision and high-efficient characterization and calculation of dynamic fracture networks in unconventional fractured reservoirs,and has been successfully applied to the actual development of tight oil reservoirs.Therefore,this research results have important theoretical and application significance.