Geophysical Inversion for Unknown Lithology and Geological Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for inverting geophysical data to infer subsurface geological properties and lithology are limited by the need for a priori assumptions about lithology classes, which may not be known, and often require explicit rock physics equations, restricting their applicability when multiple lithologies are present in a single volume.
Innovation Solution
A method that subdivides the subsurface into discrete cells with a discrete lithology parameter, allowing the lithology class to be determined during inversion using a set of possible lithology classes and associated rock physics equations, enabling the use of appropriate equations at each cell to calculate geological parameters and minimize a misfit function.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a priori assumption of lithology class is made, then the inversion process can proceed with a specific rock physics relationship, but the method becomes inapplicable when the actual lithology is unknown or when multiple lithologies are present
Solution Approach 1:
The patent applies dynamics by making the lithology parameter variable rather than fixed. The inversion process allows the lithology class to change during iterations, transitioning from a static a priori assumption to a dynamic parameter that adapts based on the data fit, enabling the system to handle unknown or mixed lithologies effectively
Solution Approach 2:
The patent changes the lithology from a fixed input parameter to a variable parameter that can take different discrete values during the inversion process. By allowing the lithology parameter to vary and be updated based on data misfit, the method can adapt to different lithological conditions without requiring prior knowledge
2Measurement precision
If discrete lithology classes are used with explicit rock physics equations, then the inversion can estimate geological parameters, but the method requires prior knowledge of lithology classes which is often unavailable
Solution Approach 1:
The patent applies preliminary action by establishing a framework with multiple possible lithology classes and their associated rock physics equations before the inversion begins. This preparatory setup allows the system to have all necessary relationships ready, and then during inversion, the appropriate lithology class is selected based on which provides the best data fit, eliminating the need for prior lithology knowledge
3Adaptability or versatility
If continuous lithology parameter is used instead of discrete classes, then the inversion can find lithology variations, but the method loses the ability to identify distinct lithologic classes
Solution Approach 1:
The patent applies segmentation by dividing the continuous lithology space into discrete classes or facies types. The inversion process estimates a discrete lithology parameter that assigns each volume element to a specific lithology class, thereby maintaining the ability to identify distinct lithologic classes while still capturing spatial variations through the distribution of these discrete classes
Data Source
AI summary
Method for estimating porosity and water saturation or other geological parameters of a subsurface region when the lithology of the region is unknown, requiring only geophysical data from remote surveys. The geophysical data are inverted (14-18) treating lithology as a third, and discrete, unknown model parameter (12) to be solved for in the inversion. A technique for solving mixed integer non-linear programs may be used. Suitable rock physics relationships (13) are used to relate the desired geological parameters to the geophysical model parameters required for simulating the geophysical data in the inversion process.