Research On Intelligent Inversion Method For Complex Hydraulic Fracture Network In Tight Reservoir

Tight oil and gas reservoirs contain abundant oil and gas resources,but the low permeability of tight reservoirs makes development difficult.Hydraulic fracturing is widely implemented.Fracture diagnostic technique is an important part of hydraulic fracturing and the fracture diagnostic map can explain the effect of hydraulic fracture and guide reservoir subsequent development.However,current fracture diagnostic technique only understand rough outline of fractures and it is difficult to describe the specific geometry of fractures.In order to improve this problem,this dissertation uses dynamic production index combined with intelligent optimization algorithm to inverse specific shape of fractures,which provides reference for evaluating hydraulic fracturing effect and subsequent reservoir development.The dissertation first propose a complex fracture network inversion method based on deterministic inversion algorithm and time-lapse seismic(4D).Oilfield workers understand the changes of saturation and formation pressure in different production periods by 4D seismic result and the fractures production performance also be reflected in the result.This provides a theoretical basis for the method proposed by this dissertation.In the deterministic inversion algorithm,the potential area of fractures is partitioned to several subspaces and the fracture network is considered to be composed of small fracture segments.The algorithm continuously optimizes the geometry of fracture segment in each subspace through the fracture network generation rule and 4D seismic result.Finally the small fracture segments in each subspace constitute the optimal fracture network.Considering the micro-seismic technology used in the hydraulic fracturing,a complex fracture network inversion method based on multi-objectivity algorithm and micro-seismic is proposed in this dissertation.This method uses the production data based on seismic source map to inverse the hydraulic fracture network morphology.A new multi-objectivity complex fracture network inversion algorithm is constructed by combining fracture network generation rule based on L-system with the decomposition method and the micro-seismic result matching effect and production data matching effect are two objective functions for fracture network inversion in dissertation study cases.The multi-objective fracture network inversion algorithm is used to obtain multiple optimal schemes that satisfy different weights of the objective functions.For determine the optimal hydraulic fracture network that meets the current production requirements,a multi-scheme decision-making method combining TOPSIS and AHP is established.Finally,two study cases of complex fracture network with different scales are established to evaluate the first inversion method and three study cases are built based on the error condition of micro-seismic source map to evaluate the second hydraulic fracture inversion method.From the inversion results of two method study cases,the shape of hydraulic fracture network inversed by the proposed method is similar to the actual fracture network,which indicates that the complex fracture network inversion methods proposed in this dissertation have certain feasibility.The shortcomings and improvement measures of two fracture inversion methods are also analyzed in the study cases.