A Rock Physics Modelling and Inversion Method for Shale Considering Stratigraphy

Given the mounting emphasis on shale reservoirs, stratigraphic features have attracted considerable interest from geophysicists. This was demonstrated in a number of studies that the development of bedding or lamina can significantly enhance the physical properties of the reservoir. Drilling results further corroborate that stratigraphic characteristics, such as lamina density, often play a critical role in hydrocarbon development. Notwithstanding, shale stratification has rarely been studied directly. It is imperative to reveal the microscopic rock physics mechanisms that cause the shale stratification and find out the inversion parameters that directly identify stratification properties. To quantitatively characterize the shale stratification, the concept of upper and lower boundaries in rock physics is employed to construct stratigraphic indicator factor. Based on this, a shale rock physics model that considering stratification is established. Additionally, the elastic impedance form of the approximate equation for the three-term reflection coefficient in a VTI medium is derived. A machine learning algorithm is then employed to regress the relationship between the stratigraphic indicator factor and the model parameters in the equation, thereby enabling stable inversion of the stratigraphic indicator factor.