To improve the recovery of tight oil, a hierarchical optimization strategy is built to maximize net present value (NPV), i.e., to initially optimize single-well fracture parameters, and then optimize well pattern parameters. On this basis, the mathematical model has been created for the hierarchical optimization of well pattern parameters in multi-stage fractured horizontal well of tight oil. First, considering the non-linear flow in matrix and the stress sensitivity of matrix and fracture, a reservoir numerical simulation method of tight oil has been established based on embedded discrete fracture model, and then through the coupling it and the covariance matrix adaptation evolution strategy (CMA-ES)algorithm to solve this mathematical model, an efficient optimization method is proposed on well pattern parameters in multi-stage fractured horizontal well of tight oil. In comparison with overall optimization, the hierarchical optimization strategy proposed in this paper can significantly improve the optimization efficiency but with no reduction in the accuracy of optimization results. In addition, if not considering the non-linear flow in matrix and the stress sensitivity of matrix and fracture permeability, the optimal results will have no guiding significance to mining. © 2019, Editorial Office of ACTA PETROLEI SINICA. All right reserved.
