Deep Fluid Identification Factor Inversion

With the increasing exploration level,there are fewer and fewer oil and gas reservoirs in the middle and shallow layers,and people gradually turn their attention to deep and ultra-deep layers,and some important discoveries have been made in these fields,such as Tahe Ordovician,northeast Sichuan Permian,Triassic and so on.However,due to the old surrounding rocks and complex structure of deep reservoirs,as well as the weak energy and low resolution of deep seismic data,deep oil and gas exploration also faces great challenges.The more challenging topic is that people are now not satisfied with structural exploration,but pay more attention to lithology and fluids.In the process of fluid identification,it is one of the common methods to search for highly sensitive fluid identification factors.Fluid factor is a parameter used to describe reservoir fluid properties based on petrophysical properties.It is an inevitable trend to use seismic data for reservoir prediction and fluid identification to establish the seismic wave reflection coefficient equation of fluid identification factor by selecting reasonable fluid identification factor and boundary conditions,and then realize the estimation process of fluid identification factor by using seismic inversion method.Generally speaking,pre-stack seismic data provide more information than post-stack seismic data,and pre-stack inversion can obtain a large number of fluid identification factors.However,most of the approximate equations of reflection coefficient are composed of three items,but the seismic data before the deep cascade lack of far-angle trace set,so the stability of parameter matrix will be reduced when the three-parameter inversion,leading to a large error of the fluid identification factor extracted in the end.In view of the fact that deep reservoir lacks large dip Angle reflection seismic data,this paper has carried out the following research and obtained corresponding results:(1)the Zoeppritz equation and its various approximate equations are elaborated and the accuracy is compared;The foundation of AVO inversion was investigated,including the generation of angular trace set and wavelet extraction.The development and advantages and disadvantages of various fluid identification factors are discussed.(2)considering the prominent instability of three-parameter inversion based on the Aki-Richard approximation,this paper aims to improve the robustness of inversion results by reducing parameter dimensions.Firstly,based on the Zeoppritz approximate formula derived by Zong et al.based on the longitudinal and transverse wave moduli,we reasonably fit the longitudinal wave moduli and density in the formula by power exponent,and deduce the approximate equations of reflection coefficients for direct inversion of longitudinal wave moduli(M)and transverse wave moduli(U),and invert them.Secondly,the relationship between the aspect wave modulus and the high sensitivity fluid identification factor is established,and the high sensitivity fluid identification factor is obtained.The model calculation and practical application show that the new method can well identify fluid properties in reservoir.(3)since the resolution of deep seismic data is low,the base tracking algorithm is extended to pre-stack seismic data to obtain the pre-stack reflection coefficient,which is restored to high-resolution seismic data by using fixed wavelet,and then pre-stack AVO inversion is carried out.The superiority of base tracking algorithm is explained by model trial calculation,and the practicability of the new method is proved by comparing the inversion profile obtained before and after data processing.(4)using the above two methods,fluid identification was carried out for the middle and deep reservoirs in the pearl river estuary basin.By improving resolution of deep seismic data and two-parameter AVO pre-stack inversion technique,a more accurate fluid identification factor profile is obtained,which provides a new method for effective fluid identification.