Subsurface Volume Modeling via Discrete Parameter Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Assisted History Match processes are inefficient in handling discrete non-sortable parameters, such as geological facies, leading to inaccuracies and inefficiencies in subsurface modeling, particularly in subsurface hydrocarbon reservoir modeling.
Innovation Solution
A method that transforms discrete parameters into indicator parameters with binary values, calculates anisotropic distance to the value transition interface, and converts these into continuous parameters for improved history matching, allowing for better correspondence with observed data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If Assisted History Match processes are used with discrete non-sortable parameters, then the modeling process can be automated, but the accuracy and reliability of the results deteriorate due to inability to handle spatial relationships
Solution Approach 1:
The patent transforms discrete non-sortable parameters into continuous parameters through a mathematical transformation process. By representing discrete facies types as continuous values based on distance to training data points, the method enables the use of automated history matching algorithms while preserving the spatial relationships and geological realism of the original discrete parameters.
Solution Approach 2:
The patent introduces an intermediary transformation layer between the discrete facies parameters and the history matching algorithm. This intermediary process converts discrete facies assignments into continuous parameter spaces, allowing automated algorithms to operate while maintaining fidelity to the underlying geological structures and spatial relationships.
2Ease of operation
If discrete parameters are treated as continuous parameters in history matching, then automated processing becomes possible, but manufacturing precision deteriorates due to rounding or truncation errors
Solution Approach 1:
Instead of简单地 converting discrete parameters to continuous and then rounding, the patent implements a sophisticated transformation that maps discrete facies types to continuous parameter spaces using distance-based interpolation. This approach allows automated processing while maintaining precision by continuously referencing the original discrete training data throughout the history matching process.
3Device complexity
If traditional history matching is applied to discrete parameters, then computational simplicity is maintained, but productivity deteriorates due to destruction of spatial relationships requiring multiple iterations
Solution Approach 1:
The patent transforms the parameter space to enable more efficient history matching convergence. By representing discrete facies as continuous parameters with preserved spatial relationships, the method allows history matching algorithms to converge faster while maintaining geological realism, avoiding the need for multiple iterative corrections.
Solution Approach 2:
The patent performs preliminary transformation of discrete parameters into a continuous representation before initiating history matching. This preliminary action preserves spatial relationships and geological structures in advance, enabling the history matching process to proceed more efficiently without needing to repeatedly correct spatial distortions during iterations.
Data Source
Figure 1
Figure 2
AI summary
Disclosed is a method of monitoring the behaviour of a subsurface volume. The method comprises transforming a single discrete parameter or an ensemble of discrete parameters describing an attribute of the subsurface volume, each discrete parameter having N possible discrete values with N ≥2, into N indicator parameters each having 2 possible discrete values; for each of the two value classes of each indicator parameter, determining the anisotropic distance to a value transition interface; transforming each of the indicator parameters into a corresponding continuous parameter using the determined anisotropic distance to the value transition interface; and using the continuous parameters in a history matching process.