L. Estimated ultimate recovery prediction of fractured horizontal wells in tight oil reservoirs based on deep neural networks.

No abstract

No abstract

As an unconventional resource, tight oil reservoirs hold abundant geological reserves and great development potential. However, tight oil reservoirs suffer from high heterogeneity and low permeability. To obtain commercial oil and gas flow, multistage hydraulically fractured horizontal well technology is often adopted. In this paper, an embedded discrete fracture model (EDFM) is adopted to study the fractured reservoir. First, the relationship between the EDFM connection list is studied and its coefficient matrix of the equation system is solved. EDFM connection list is a method to record the connections between the grids. Thereafter, the reasons for the sparsity of the coefficient matrix and the symmetry of the locations of nonzero elements in the matrix are explained, and a simple method to establish and assemble the coefficient matrix is proposed. Using this method, a simulator for tight oil reservoir with EDFM is developed. By comparing the simulation results of the self‐programming simulator with the commercial software SAPHIR, the usefulness of the simulator is verified. Finally, several simulations and sensitivity analysis are performed to determine the key factors affecting production. Modeling studies show the influence of fractures, the horizontal well, and stress sensitivity, providing very useful information for the design of hydraulic fracturing and production development programs.