Seismic Wave Attenuation Characteristics And Fluid Identification In Porous Media In Mesoscopic Scale

Reservoir technology is one of the most important parts in seismic gas and oilexploration. The exploration targets turn into complex structural reservoir, lithologicreservoir and fractured reservoir. The traditional reservoir technology can not satisfythe requirement of fluid identification in these reservoirs. The underground media isconsist of not only the solid part, but also the filler in porous. Especially for gas andoil reservoirs, the seismic wave propagation characteristic is controlled directly by thefluid properties in porous. Taking full account of the poroelastic property inunderground media, it has practical implications to build new poroelastic model andfind the new fluid identification method which is more sensitive to the fluid in porous.Therefore, in this paper I study the wave field and attenuation and dispersioncharacteristics, and the related fluid identification technology.Gassmann presented the theory of wave propagation in porous media and the famousGassmann Equation was established. As a theoretical foundation for wave propagationin fluid-saturated porous media, Biot theory fully considers the two-phase property,which is consist of solid skeleton and fluid in porous, of porous medium based on theGassmann theory. As the research continues, researchers found that the squirt-flow isthe main reason of the phenomenon of strong wave attenuation and high velocitydispersion during wave propagation. A better explanation, the BISQ model wasestablished by combining the biot-flow mechanism in macroscopic scale and thesquirt-flow mechanism in microscopic scale. The BISQ model is well used to describethe attenuation phenomenon caused by porous and its saturated fluid. But in themeantime, the solid skeleton is inelastic media which is an important reason of waveattenuation phenomenon, especially after fluid infiltration, the solid shows a strongviscoelastic property.In this paper, I integrate three important mechanisms during wave propagation(poroelastic，anisotropic and viscoelastic), the viscoelastic solid-skeleton based on thegeneral Zener linear solid is introduced into anisotropic poroelastic model. Thisviscoelastic mechanism is based on a spectrum of relaxation mechanisms, which issuitable for wavefield calculations in the time domain. In this paper, the extensivedilatancy anisotropy (EDA) is only considered, which is most concerned in the oil andgas exploration. A new constitutive relation and its corresponding first-order wave equation are proposed for simulating wave propagation in time-domain. Astaggered-grid high-order difference method is applied in forward modeling, while theconvolutional relations are avoided by introducing memory variables. Based on thenew model, I simulate2-D wavefield and analyses snapshots and syntheticseismograms in both one-layered model and two-layered model. I have alsoconducted the forward modeling of degraded models(elastic BISQ-EDA model,isotropic BISQ model) as a comparison. The result shows that poro-viscoelasticmodel has an advantage in the modeling of real reservoir at seismic bank. In order tostudy how the viscoelastic skeleton work, the phase velocities and quality factors offast and slow P-waves are given through the plane-wave solution. Through thecomparison for three models with different viscoelastic material properties, pointedout that the viscoelastic skeleton introduces an additional attenuation peak into BISQmodel, whose attenuation degree and attenuation band are controlled by relaxationquality factor and relaxation center frequency, respectively.Attenuation and velocity dispersion occurs during wave propagation in porous mediaat seismic frequencies. Several studies have shown that waveinduced fluid flow at themesoscopic scale is primarily responsible for this phenomenon. The mesoscopic scaleis significantly larger than the pore size but smaller than the wavelength. Whenheterogeneities exist in the fluids in porous media, a passing compressional wavecreates pressure gradients within the fluid phase, resulting in relative movementbetween the fluid and solid phases, which causes the elastic wave attenuation andvelocity dispersion. White and White et al. first introduced the mesoscopic-lossmechanism. They developed a porous model saturated by gas and water on the basisof Biot’s theory. White’s model considers two pore structure models, the layeredpatchy model and the spherical patchy model. The layered model consists of thinporous layers alternately saturated with gas and water. The spherical model isconsidered to represent water-saturated media that contain several gas sphere pockets.However, real porous media are far more complex. Johnson generalized the Whitemodel and added two additional geometric parameters in addition to the basicgeophysical parameters. The first one is the specific area surface S/V, which isgoverned by the pore patch shape. The other one is parameter T, which is related tothe main size of the patches.The two models agree with each other very well in attenuation and dispersioncharacteristics in both patchy saturated situations. So I perform an analysis on how thegeophysical parameters affect the attenuation and dispersion characteristics.Numerical solutions to Biot’s equation are computationally intensive; thus, I build anequivalent viscoelastic model, which well approximates the wave attenuation anddispersion in porous rocks in the seismic band. Firstly, I derive the approximations of the quality factor and its minimum valueQm inusing Johnson’s model, which bettermatches the real complex patchy conditions, at the low and high frequency limits. Idevelop the approximate equations by using an exponential correction to improve theaccuracy and then compare the approximated and exact values ofQm in. TheapproximatedQm inequation consists of basic geophysical parameters in porousmedia and two geometric parameters T and S/V. I then calculate the equivalentviscoelastic attenuation factor in porous media with this equation. Finally, I approachthe phase velocity and quality factor by using the standard linear solid model (Zenermodel) in the Johnson layered and spherical patchy models. The results verify thereliability of the approximateQm inequation and the feasibility of the proposedmethod.In the actual seismic data, the high attenuation and dispersion phenomenon inreservoir is obviously observed. But so far, it is rarely used in the fluid identificationtechnology. Based on Chapman theory, the dispersion degree in gas-saturatedreservoir is much higher than it in water-saturated reservoir, which can be used as aneffective gas and water identification factor. The dispersion degree factor can becalculated directly by dispersion AVO inversion equation, which is obtained byintroducing the traditional AVO approximate equation into frequency dependentequation. In this paper, this method is used in fluid identification in2D and3Dseismic files in South China Sea, and the result is generally satisfactory.