Study On Compressibility Of Shale Reservoirs In Changning Area

Due to the low porosity and permeability of shale reservoirs,effective development can only be achieved through horizontal well volume fracturing.Fractureability refers to the ability of shale reservoirs to form complex fracture networks during fracturing development.In the early stage of shale gas development,geological evaluation indicators were mainly used to analyze the impact of fracturing,and less on the influence of engineering technology controllable indicators on the fractureability.Based on the systematic analysis of the geological characteristics of shale gas reservoirs,the machine learning method is used to study the geological factors and engineering factors of shale gas reservoirs,and the impact of the reservoir on the fractureability of the reservoir,combined with BP neural network prediction model.A model for predicting the fractureability of shale gas reservoirs in the Changning area has been established,and the following understandings have been drawn:(1)Engineering geomechanical characteristics of shale gas reservoirs in Changning areaBy analyzing the on-site data of Changning area,it is worthy of fracturing mining in Changning area,which has high economic exploitation value.The mineral composition,reservoir rock mechanics and in-situ stress characteristics and natural fracture development of the Longmaxi Formation shale gas reservoir in the N3 well of a key well in the Changning area were analyzed.It is found that the brittle mineral content of the Longmaxi Formation reservoir in N3 well reaches 72.94%,and the distribution of mineral components is quite different.The average Young's modulus is 2.62×104 MPa,the Poisson's ratio is 0.201,and the horizontal stress difference coefficient is 0.01.Therefore,the shale gas reservoir in the Changning area has the conditions for volumetric fracturing.(2)Construction of the main control index system for shale gas reservoirsBased on the key geological and engineering parameters of the site,the factors affecting the shale reservoir fracturing are screened out,and the principal component analysis method and K-means clustering algorithm are applied to obtain the main index system of fracturing.The system includes natural crack development degree,total perforation number,brittle mineral,organic carbon content,gas content,Poisson's ratio,horizontal section length,brittleness index,shale content,and reservoir depth.The quantitative relationship between the post-pressurization capacity and the compressibility is: the poorly compressible ratio corresponds to the post-pressurization capacity(after normalization)is less than 1.23,and the fracturing property generally corresponds to the post-pressurization capacity(after standardization)of 1.23 to 3.44,which can be fractured.Good performance corresponds to post-pressurization capacity(after standardization)greater than 3.44.(3)Study on predictability model of shale gas reservoir fracture based on BP neural networkFor the Changning area,59 wells have been drilled,and the fracture testability method has been established through BP neural network.The geological evaluation factors 1 to 9 and the engineering factors 1 to 7 and 16 factors have been constructed.The post-pressurization capacity is predicted.Finally,the fracturing property can be judged according to the corresponding relationship between the compressibility and the post-pressurization capacity.The evaluation results show that H1 and H3 can be fractured generally(Class II,the evaluation value is between 1.23 and 3.44),not suitable for the current fracturing transformation,H2 can be good for fracturing(Class III,evaluation value ? 3.44),suitable for the present Fracturing transformation.