3D Reservoir Visualization Using Machine Learning Interpolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current well logging technologies face challenges in accurately determining the characteristics of a reservoir beyond the defined range of a wellbore, limiting the effectiveness of geosteering and production estimation in oil and gas exploration.
Innovation Solution
The integration of downhole measurements from multiple wellbores using electromagnetic and other measurement tools, combined with machine learning algorithms, to interpolate and visualize three-dimensional mesh properties of the subterranean formation, enabling the creation of a comprehensive 3D reservoir model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If downhole measurements are taken only within the defined range of a single wellbore, then measurement precision is maintained, but the extent of reservoir characterization is limited
Solution Approach 1:
The patent combines measurements from multiple wellbores (first wellbore, second wellbore, etc.) to create a comprehensive 3D reservoir model. By merging data from different locations and integrating it with machine learning algorithms, the system achieves both precision (through validated measurement integration) and extensive reservoir coverage (beyond single wellbore range).
Solution Approach 2:
The patent introduces machine learning algorithms as an intermediary to process and integrate measurements from multiple wellbores. The machine learning system interpolates reservoir properties between wellbores and validates measurements, enabling extended reservoir characterization while maintaining measurement precision through algorithmic validation.
2Area of stationary object
If machine learning algorithms are used to interpolate reservoir properties between wellbores, then reservoir coverage is extended, but computational complexity increases
Solution Approach 1:
The patent performs preliminary actions by collecting and organizing measurements from multiple wellbores before applying machine learning algorithms. The system pre-processes the data, identifies measurement patterns, and prepares the computational framework in advance, which reduces the overall computational complexity when generating the 3D reservoir model.
Solution Approach 2:
The patent segments the reservoir characterization process into distinct computational stages: data collection from multiple wellbores, machine learning-based interpolation between wellbores, validation of interpolated properties, and final 3D model generation. This segmentation allows each stage to be optimized independently, managing computational complexity effectively.
3Measurement precision
If measurements are integrated from multiple wellbores, then reservoir characterization accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent implements feedback mechanisms where machine learning algorithms continuously validate measurements from multiple wellbores against interpolated reservoir properties. The system uses feedback loops to refine the integration process, improving reservoir characterization accuracy while managing data processing complexity through iterative validation and correction.
Data Source
AI summary
Methods and systems for determining 3D properties of a formation are provided. The method includes acquiring inversion results from two or more wellbores and transforming the inversion results into first 3D mesh properties, wherein the first 3D mesh properties represent one or more geological features of a formation surrounding each wellbore of the two or more wellbores within a defined range from each of the wellbores, where the one or more geological features are correlated to a 3D coordinate system. The method further includes determining, using a machine learning algorithm, one or more similar geological features among the two or more wellbores based on the first 3D mesh properties; interpolating second 3D mesh properties based on the one or more similar geological features, wherein the second 3D mesh properties are properties of the formation outside the defined range; and integrating the first 3D mesh properties and the second 3D mesh properties to acquire final 3D mesh properties.