Pre-stack Seismic Inversion And Fluid Identification For Deep Reservoir

With the developing of the exploration for the hydrocarbon reservoir,the focus of exploration gradually converts the shallow reservoirs to the deep reservoirs.At present,the exploration in China of the deep reservoirs has entered to the breakthrough,deep oil and gas reservoirs have been founded in the Bohai Bay Basin as well as the northeast,southwest and northwest and other major basins.Deep hydrocarbon reservoirs will be one of the most important domain in the future.Therefore,effective seismic inversion methods for the deep reservoirs will be the hotspot in geophysics domain.Deep reservoirs are different from the shallow reservoirs in the parameters of reservoir due to the influence of deep buried,temperature,pressure and so on.The conditions above lead to a series of problems to be solved for the deep reservoir inversion,such as conventional fluid factor aiming at deep reservoir is not sensitive enough and the inversion results is worse.With the increase of depth,information of large incident angle is deficient,So a more sensitive fluid indicator aiming at small incident angle is needed to be built for deep reservoir.With the characteristics of rock physics we derived a new three term AVO approximation in terms of pore-fluid bulk modulus K_f,which is more sensitive to hydrocarbon,dry rock matrix term fm and density?.The accuracy of the new approximation is almost equivalent to exact Zoeppritz equation.And the most important is that,with this reflective approximation,the condition number of the new coefficient matrix is much smaller,which means the stability of estimating fluid factor K_f is effectively enhanced.Consequently the new approximation can meet the requirements of inversion.In addition,tests on synthetic data shows that the pore-fluid bulk modulus K_f can be estimated reliably from the pre-stack seismic data.We built a flow path of extracting pore-fluid bulk modulus K_f by elastic impedance inversion for deep reservoir with small incident angle.The energy of seismic waves is weak due to deep burial.When there is shale a strong reflection event,the existence of it makes the weak effective information under the strong reflection and makes the fluid prediction harder.To solve this problem,we adopt multichannel matching pursuit method to identify the strong reflection.So the weak effective seismic information can be prominent.And the method is proved to be accuracy,reliable and feasible after real data test.