Methodologies Of Seismic Stochastic Inversion For Tight Sandstone Reservoirs

With the development of oil and gas exploration in China,the exploration of the medium-deep-depth tight reservoirs have become the focus of oil and gas exploration.Tight sandstone has the characteristics of deep burial,low porosity and permeability,complex pore structure,heterogeneous saturation of pore fluid,and strong heterogeneity.How to describe the tight sandstone reservoir through the seismic inversion has become current research hotspots.The "Sweet spot" in tight sandstone reservoirs is relatively high porosity and permeability area under the background of low porosity and permeability,which is the most attention geological target in tight sandstone reservoir prediction.Aiming to an offshore gas field in east China,based on the rock physics theory of tight sandstone,this paper studies the seismic response characteristics of the "Sweet spot" in tight reserviors,and develops prestack seismic stochastic inversion methods to achieve the effective prediction of tight sandstone reservoirs and "sweet spot".The porosity of permeability of tight sandstone reservoirs are low,and the coexistence of the reservoirs with different physical characteristics reflects strong reservoir heterogeneity.Therefore,with reference to classification criterion of "Sweet spot",a new seismic rock physical model suitable for highly heterogeneous tight sandstone reservoirs is established by combining the rock physical models for different physical characteristics.The "sweet spot" in tight sandstone is relatively thin,and thin interbeds are widely developed.For conventional linear inversion,it is often difficult to meet the identification requirements of the "sweet spots".Compared with deterministic inversion,stochastic inversion has a higher vertical resolution.On the basis of Bayesian theory,this paper proposes a global iterative geostatistics inversion method based on optimizing local correlation coefficients and a Bayesian Markov chain Monte Carlo(MCMC)probabilistic inversion method based on joint distribution constraints,which have higher convergence and stability.The elastic parameters characteristics of tight sandstone reservoir and different types of "sweet spot" in tight sandstone reservoir is different.Therefore,this paper proposes a prediction method based on the synchronous inversion method for elastic sensitive parameters of tight sandstone reservoir and its internal "sweet spot" was established by deriving the AVO modeling formula about the sensitive parameters of tight sandstone reservoir and "sweet spot".Finally,according to the identification of "Sweet spot",we realize the prediction of petrophysics parameters based on an optimal geostatistics inversion method..Studies on a tight sandstone gas field in eastern China show that the prediction of tight sandstone reservoir and its internal "sweet spot" based on the stochastic inversion can well identify the different type of "sweet spot",which provides an important geophysical support for the tight sandstone gas field exploration and development.