Inversion of fractures based on equivalent continuous medium model of fractured reservoirs

Abstract

Because the fractures in fractured reservoirs have a significant effect on the flow behavior of fluid, the production performance of the fractured reservoirs contains important information of the distribution of fractures. That is the basic principle of applying the production information for the inversion of fractures. It is the first step to link the distribution information of fractures with the production information to establish inversion objective function on the basis of Bayesian theory. And the geometric parameters of fractures such as midpoint coordinate (x0,y0), azimuth α and extension length L which directly determine the distribution of fractures and then affect the flow behavior are chosen as inversion parameters and some production data are chosen as inversion indexes. Secondly, in order to reconcile the demand of reflecting the flow effect of individual fractures and the use of mature simulators, equivalent continuous medium model whose equivalent permeability tensor is a crucial parameter reflecting the flow behavior of fractured porous media is applied for calculating the numerical simulation of the fractured reservoirs. Finally, SPSA algorithm is adopted to minimize the inversion objective function and then update the distribution of fractures to obtain the optimal results. Theoretical cases verify that the inversion method is effective to identify fractures while the number of identified fractures is limited because of the multi-solution of the inverse problem.