3D Rock Imaging Segmentation for Reservoir Property Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods in hydrocarbon exploration struggle to accurately determine relationships among seismic, electrical, and reservoir properties, particularly in complex rocks where current methods cannot isolate the effects of each compositional class, and often require multiple rock samples or rely on inaccurate process simulations.
Innovation Solution
A method that uses 3D imaging and segmentation of rock samples to create local properties volumes, calibrating models to predict effective material properties based on volume fractions of compositional classes, allowing for the assessment of subsurface reservoir potential with a single rock sample.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current methods are used to determine relationships among seismic, electrical, and reservoir properties, then multiple rock samples are required, but measurement precision and accuracy are insufficient
Solution Approach 1:
The patent segments the rock sample into compositional classes using 3D imaging and segmentation techniques, allowing analysis of individual mineral phases (quartz, feldspar, calcite, clay) and their contributions to effective properties. This segmentation enables accurate determination of relationships among properties from a single sample by isolating the effects of each compositional class.
Solution Approach 2:
The patent transitions from traditional 2D or bulk property measurements to 3D imaging and segmentation, adding a spatial dimension to the analysis. This allows visualization and quantification of the spatial distribution and volume fractions of compositional classes throughout the sample, enabling more accurate property relationships to be determined from a single sample.
2Measurement precision
If process simulations are used to predict reservoir properties, then computational models are required, but accuracy is insufficient for complex rocks
Solution Approach 1:
The patent replaces complex computational process simulations with direct 3D imaging and numerical analysis of actual rock samples. Instead of using computational models to simulate rock formation processes, the method directly images the actual 3D structure and composition of the sample, then uses numerical methods to calculate effective properties, providing more accurate results for complex rocks without requiring complex process simulations.
3Loss of information
If traditional methods are used to analyze rock composition, then bulk properties are measured, but the effects of individual compositional classes cannot be isolated
Solution Approach 1:
The patent applies segmentation to divide the 3D image of the rock sample into distinct compositional classes based on their physical and chemical properties. This allows the effects of each class (quartz, feldspar, calcite, clay) to be isolated and analyzed separately, determining their individual contributions to effective properties such as elastic moduli, permeability, and electrical conductivity.
Solution Approach 2:
The patent assigns different properties to different compositional classes within the 3D image, allowing each class to have its own local properties (density, elastic moduli, permeability, electrical conductivity). This local quality approach enables accurate calculation of effective properties by considering the specific characteristics of each compositional class and its spatial distribution.
Data Source
AI summary
A hydrocarbon exploration method is disclosed for developing a model of at least one effective material property of a subsurface reservoir as a function of the composition and structure of the reservoir rock. In one embodiment, the method comprises: obtaining a 3D image (102) of a rock sample characteristic of a reservoir of interest (101); segmenting the 3D image into compositional classes (103) based on similarities in mineralogy, structure and spatial distribution; selecting a model (105) that relates an effective material property of interest to the volume fractions of each compositional class; and determining the parameters of the model (106). The model may be used to assess the commercial potential of the subsurface reservoir (107).