| Presently, oil and gas explorations have shifting from simple structure exploration tolithological exploration, and one of the most major geological exploration targets isthe strata-lithological reservoir in the future. The key problem of strata-lithologicalreservoir exploration need to solve includes identifying reservoir, analyzing lithologyand fluid properties. The quality factor estimated from seismic data can identifyreservoir, detect hydrocarbons and analyze fluid properties.From VSP data to surface seismic data, from simple model to complex strata, seismicprospecting needs more high precision and vertical resolution in estimating Q-factor.In frequency domain, this paper realizes logarithmic spectral ratio(LSR) method, peakfrequency shifting(PFS) method and analytical method. The trial computation usingtheoretical model shows estimating Q value of these three methods can bring goodeffects without noise condition, but the effect is not so reliable with noise condition.Therefore, this paper proposes a Q-factor estimation method based on BFGSalgorithm, derived the amplitude spectrum based on generalized S transform andproved the relation between amplitude spectrum of different travel time and qualityvalue. Then construct objective function based BFGS algorithm, the procedure ofinversing Q-factor is derived. Modeling computation proves that this method is betteranti-noise property compared with LSR and PFS. On this basis, this paper analysesdependency of this method upon time-frequency analysis, estimating Q value basedgeneralized S transform is more high calculation efficiency with better anti-noiseproperty. Also, the dependency of initial Q value is analyzed, the initial Q valueinfluences calculation efficiency, but do not affect the result accuracy. Finally,synthesis multilayer seismic record shows inversing Q value based BFGS algorithmcan estimate quality factor effectively with better anti-noise property.In order to analyze the validity of this Q-factor estimation method based BFGSalgorithm on reservoir prediction and application on field data, two-phase mediamodel is constructed based on real data. Firstly, analyzing the condition of real underground medium and combining the well logging data with seismic data,two-phase media parameter is gotten based BISQ theory. Then, this paper getsvelocity and density profile of real seismic data based modeling inversion, porosityand permeability through empirical formula and estimates solid substrate bulkmodulus using the modulus model of Voigt, Reuss and Hill(V-R-H). Finally, usingthis parameters to construct two-phase media model according to real strata condition.The post-stack seismic data profile is gotten using the staggered-grid high-orderfinite-difference method, which is compared with field data, seismic record is similarand reservoir situation is consistent. That shows this model approximates to real strata.The Q profile is gotten using inverse Q value method in this paper, the reservoirsituation has high attenuation on Q profile, the Q value is small, which gives that theQ value estimation method proposed in this paper can estimate Q value from complexmodel with high accuracy.Using this new method to inverse Q profile from field seismic data, the Q profile andequivalent absorption Q-inverse profile is gotten. Reservoir situation in well loggingcurve is according to high attenuation zone in Q profile and equivalent absorptionQ-inverse profile, which shows that this method can estimate quality factor from realdata effectively with better anti-noise property and high precision. |
PDF Full Text Request