Study Of Fracture Property Identification Based On Shear-wave Splitting

With the rapid development of economy, the demand for oil and gas resources isincreasing year by year. As high-quality stereotype reservoir exploration anddevelopment goals declining, the direction of oil and gas exploration began to focuson the unconventional reservoirs. The fractured reservoirs widely distributed in ourcountry and have become an important exploration field. Compared with other typesof oil and gas reservoirs, fractured reservoirs with a complex fracture systemsdistribution, poor reservoir matrix properties, poor permeability and low oil recovery.Therefore, the development of fractured reservoirs have considerable difficulty, thestudy of fracture property identification is one of the important topics of the currentoil and gas exploration sector.When the shear wave propagates in the fracture medium shear-wave splittingphenomenon occurs, the shear wave will split into the fast shear wave and the slowshear wave. Propagating the multi-fractured medium, the fast shear wave and the slowshear wave will split again, so that the propagation of the wave is more complex.Using the phenomenon of shear-wave splitting to obtain fracture orientation andfracture density has important significance for the study of fractured reservoir.In this paper, we used Wang Kai’s two-dimensional three-component HTI mediaforward modeling program to simulate the phenomenon of shear-wave re-splitting,and discussed the effects of fracture density and fracture orientation between theshear-wave re-splitting phenomenon.In the fracture property identification methods, Pearson correlation coefficientmethod is a method with the strong anti-noise and high stability. Calculating thePearson correlation coefficient is an optimization problem, PSO is an effectivemethod to solve nonlinear optimization problems. Therefore we combined with particle swarm algorithm and Pearson correlation coefficient to identify fractureproperty.Firstly, we combined the particle swarm algorithm with Pearson correlationcoefficient method to identify fracture properties by the single-trace seismic data, theresults show that this method can identify fracture properties correctly. Then, in orderto validate identification results of this method in noisy environments, we chosemulti-trace seismic records added random noise as the experimental data to identifyfracture properties, and to analyze the recognition results through the statisticalmethod, the final results show that in noisy environments this method can also be aeffective method to identify fracture properties. Compared the identification results ofmulti-trace seismic data in noisy environments with the identification results of thecross-correlation method, this method has powerful noise-resistance and betterrecognition results. Finally, we identified the seismic records of the double-layermedium in which the upper medium is the isotropic medium the lower medium is HTImedium and received good recognition results.