Electromagnetic Inversion Using Seismic-Constrained Coarse Grids
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D electromagnetic (EM) inversion methods are computationally expensive, require significant computer resources, and result in low-resolution images due to non-uniqueness and diffusion nature of EM fields, limiting their application in hydrocarbon exploration.
Innovation Solution
A method that uses seismic data to define structural units for a coarser inversion grid, separating model discretization for inversion from forward modeling, and iteratively adjusts resistivity models based on comparisons between computed and measured EM field values, reducing the number of unknowns and computational demands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a fine mesh grid is used for forward modeling to achieve accurate EM field simulation, then modeling accuracy is improved, but the number of cells increases significantly making inversion computationally expensive
Solution Approach 1:
The patent divides the computational domain into two distinct grids: a fine mesh grid for forward modeling and a coarser inversion grid for parameter estimation. The fine mesh grid (with dimensions Nz×Nx×Ny) is used only for computing EM field responses, while the inversion grid (with dimensions Ni×Ni×Ni, where Ni < Nx, Ny, Nz) is used for storing and inverting resistivity parameters. This segmentation allows accurate forward modeling without the computational burden of inverting all fine mesh cells.
Solution Approach 2:
The patent extracts only the necessary computational details from the fine mesh grid for the inversion process. Specifically, it computes EM field responses on the fine mesh grid, then uses these responses to update resistivity parameters on the coarser inversion grid. The fine mesh structure is 'taken out' from the inversion process, leaving only the coarser grid for parameter estimation, thereby reducing computational complexity while maintaining modeling accuracy.
2Measurement precision
If the same fine discretization grid is used for both forward modeling and inversion, then modeling accuracy is maintained, but the system of linear equations becomes very large requiring supercomputers
Solution Approach 1:
The patent segments the computational task into two parts with different grid requirements: forward modeling on a fine mesh grid and inversion on a coarser grid. This allows the system to maintain high modeling accuracy using the fine mesh while reducing the inversion problem to a manageable size on the coarser grid, eliminating the need for supercomputers.
Solution Approach 2:
The patent applies different grid qualities to different computational purposes: the fine mesh grid is used locally for forward modeling where high spatial resolution is needed for accurate EM field simulation, while the coarser inversion grid is used for parameter estimation where full fine-mesh resolution is not necessary. This local differentiation of grid quality optimizes both accuracy and computational efficiency.
3Adaptability or versatility
If EM inversion is performed without using seismic structural information, then inversion flexibility is maintained, but resolution remains low due to non-uniqueness and diffusion nature of EM fields
Solution Approach 1:
The patent performs preliminary action by incorporating seismic structural information before the EM inversion process. Seismic data are used to define the geometry and location of subsurface structures, which then constrain the inversion grid and guide the resistivity parameter estimation. This preliminary constraint from seismic data reduces non-uniqueness and improves resolution without completely sacrificing inversion flexibility.
Solution Approach 2:
The patent uses seismic structural information as an intermediary to bridge the gap between EM data and subsurface resistivity distribution. The seismic data provide structural constraints that mediate the inversion process, helping to resolve the non-uniqueness inherent in EM inversion while maintaining the ability to estimate resistivity parameters from EM measurements.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces computational requirements and enhances resolution, allowing for faster and more accurate inversion of EM data, improving the ability to determine subsurface resistivity and hydrocarbon reservoir properties.
Implementation Method 1
measured data from a controlled-source electromagnetic survey
Implementation Method 2
solving Maxwell's electromagnetic field equations on the forward-modeling grid
Data Source
AI summary
Method for efficient inversion of controlled-source electromagnetic survey data to obtain a resistivity model of the subsurface of the survey area. The method extracts the dimensions and location of sub-surface structures as they may be revealed by existing seismic or other available high resolution survey data from the subsurface area (33). This structure geometry information is used to construct a discretization (grid, or mesh) for the inversion computation (34) that is different from the mesh used for the forward modeling calculations (32) in that (a) it has fewer and hence larger cells; and (b) the cells honor the assumed structural information. The inversion need only extract resistivity information (35), the geometry of the resistive structures being specified by the inversion mesh.