Automated Stratigraphy Interpretation From Borehole Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual interpretation of depositional facies from borehole images is user-biased, time-consuming, and challenging for highly deviated wells, as it requires extensive expertise and is non-unique, often requiring additional supporting information.
Innovation Solution
An automated method using machine learning techniques to classify sedimentary geometries and depositional environments from borehole images, involving the construction of a training set with synthetic and real images, application of noise, and utilization of decision tree-based or probabilistic graphical models to establish a depositional environment predictor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual interpretation is used to study depositional facies, then user expertise can be applied to interpret log shape, but the process becomes time-consuming and user-biased
Solution Approach 1:
The patent replaces manual mechanical interpretation with an automated computational system using machine learning algorithms. The system processes borehole images and logs to automatically identify and classify sedimentary geometries, eliminating time-consuming manual analysis while maintaining or improving interpretation accuracy through consistent algorithmic application.
Solution Approach 2:
The system enables self-service interpretation by automatically analyzing borehole data without requiring continuous human intervention. The machine learning model independently processes images, identifies patterns, and generates interpretations, allowing the system to serve itself in the interpretation task while reducing dependency on manual expertise.
2Measurement precision
If manual interpretation is used for highly deviated wells, then detailed analysis can be performed, but the process becomes very challenging and requires extensive supporting information
Solution Approach 1:
The patent replaces complex manual interpretation processes with automated machine learning algorithms that can handle highly deviated well geometries. The system automatically adjusts to different well orientations and deviations, processing borehole images and logs to identify sedimentary geometries without the complexity and challenges associated with manual analysis of such challenging cases.
3Productivity
If automated image description approaches are used, then dip picking and structural zonation can be automated, but sedimentological environment interpretation remains manual
Solution Approach 1:
The patent merges automated image description capabilities with sedimentological environment interpretation into a unified machine learning system. The system combines multiple functions including dip picking, structural zonation, and depositional facies interpretation into a single automated workflow, eliminating the need for separate manual interpretation steps and achieving comprehensive automation.
Solution Approach 2:
The machine learning system performs multiple interpretation functions simultaneously, including dip measurement, structural analysis, and sedimentological environment classification. This multi-functional approach allows a single automated system to handle various interpretation tasks that previously required separate manual processes, increasing both productivity and automation extent.
Data Source
AI summary
Embodiments of the present disclosure are directed towards systems and methods for automated stratigraphy interpretation from borehole images. Embodiments may include constructing, using at least one processor, a training set of synthetic images corresponding to a borehole, wherein the training set includes one or more of synthetic images, real images, and modified images. Embodiments may further include automatically classifying, using the at least one processor, the training set into one or more individual sedimentary geometries using one or machine learning techniques. Embodiments may also include automatically classifying, using the at least one processor, the training set into one or more priors for depositional environments.