Study On AVO Of Dispersive Medium Using Inverse Spectral Decomposition

AVO (Amplitude versus offset) technique analysis lithological and physical characterof formation using prestack data, and then detect the hydrocarbon and describe thereservoirs. It has been widely used in the aspects of reservoir prediction. AVOanalysis gradually developed based on a variety of Zoeppritz equation approximations,but the traditional AVO analysis use elastic wave theory, and research the reflectionand transmission situation of interface which separates two elastic half-space. It doesnot take into account the impact of stratigraphic thickness, attenuation, dispersion andother factors.Firstly this paper summarize various approximations of Zoeppritz equation, pointingout their special emphasis. Then reviewed the theory of two-phase media in seismicrock physics that are closely related to the AVO technique, such as Gassmann theory,Biot theory, squit flow mechanism, BISQ theory, and then gives a detailed descriptionon the dispersive medium theory. The results show that fluid flow is an importantfactor to ensure pressure equalization, and because of the mobility and viscosity of thehydrocarbons may cause abnormal velocity dispersion and attenuation in the seismicband.With the development of time-frequency analysis technology, spectral decompositiontechnique has been widely applied on determining the stratigraphic thickness,instantaneous attributes analysis and reservoir prediction. The current time-frequencyanalysis methods are divided into two major categories：linear time-frequencyanalysis and bilinear time-frequency distribution, according to it whether meeting thesuperposition principle or linear principle. Linear time-frequency analysis methodsinclude short time Fourier transform (STFT), Gabor transform, wavelet transform,among them continuous wavelet transform (CWT) is widely used at present; bilineartime-frequency distribution is a type of time-frequency representation which takesWinger-Ville distribution as basis and core. In this paper, a spectral decompositionbased on the inversion is applied, known as the inverse spectral decomposition. Itconsider spectral decomposition as an inversion of the process, solving the inverseproblem to obtain the time-frequency spectrum. Solving the inverse problem using different constraints can get time-frequency spectrums with different resolutions,applied them to different aspects according to practical requirements.Thin reservoir AVO analysis is an important part of the development of concealmentreservoir, but if the reservoir is too thin, thin-bed tuning effect change the AVO curveshape, affecting the reservoir identification. This paper applies frequency-dependentAVO technology which is based on inverse spectral decomposition to the AVOanalysis of the thin layer. It was found that inverse spectral decomposition with highresolution removes the influence of the wavelet, thereby eliminating the influence ofthe thin-bed tuning effect on the AVO analysis.It shows that the hydrocarbons in the porous media will cause a strong attenuation andvelocity dispersion. Attenuation and dispersion make the formation velocity versusfrequency, with frequency-dependent attribute, so that the reflection coefficient versusfrequency. However, the traditional AVO approximations are based on the assumptionof elastic media, unconsidering velocity dispersion and attenuation. This paperintroduce frequency factor to Zoeppritz equation approximations, applying spectraldecomposition to the AVO inversion, that is frequency-dependent AVO inversion,obtaining the inverted dispersion and frequency-dependent properties. Theapplications of simulated data show that the inverted dispersion andfrequency-dependent properties obtaining by frequency-dependent inversion can beused as the indicator of the hydrocarbons, used for hydrocarbon detection, inparticular, applying kinds of frequency-dependent properties, because they are verydifferent between elasticity situation and dispersion situation.This thesis also study the impact of time scales and fracture density on dispersion.Theresults showed that the time scales determine the frequency range of the dispersion,meanwhile also affecting the magnitude of dispersion. As the crack density increases,the dispersion increases.Frequency-dependent AVO technique is a new method which is able to solve theproblem that the traditional AVO analysis can not deal, and using the properties offrequency-dependent to predict reservoir. But both theoretical research and practicalapplications need further research.