Ensemble Image Model for Real-Time Reservoir Boundary Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drilling systems face inefficiencies and inaccuracies in interpreting subsurface geological features, relying on manual expert interpretation and failing to provide real-time and accurate reservoir boundary determination.
Innovation Solution
A boundary identification system utilizing an ensemble image model with multiple image-to-image machine-learning models to generate accurate and real-time reservoir boundary predictions from longitudinal electromagnetic inversion result profiles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual expert interpretation is used to determine reservoir boundaries, then accuracy may be maintained through expert knowledge, but productivity is reduced due to laborious and time-consuming processes
Solution Approach 1:
The patent replaces manual expert interpretation (mechanical human analysis) with an automated machine learning system that processes electromagnetic inversion data. The system uses trained models to automatically identify reservoir boundaries, eliminating the need for manual expert analysis while maintaining or improving accuracy through consistent algorithmic application.
Solution Approach 2:
The system enables self-service by allowing the machine learning models to automatically interpret subsurface data and determine reservoir boundaries without requiring expert intervention. The models are trained once on labeled data and then autonomously perform boundary identification, making the system self-sufficient for routine interpretation tasks.
2Measurement precision
If complex manual interpretation methods are employed, then measurement precision may be maintained, but device complexity and ease of operation are worsened
Solution Approach 1:
The patent extracts the complex interpretation logic from the human expert and encapsulates it within trained machine learning models. The models contain the learned patterns and relationships needed for accurate boundary identification, separating the complex analytical function from the operational interface and making it reusable across multiple interpretations.
3Productivity
If real-time boundary determination is implemented using automated systems, then productivity is improved, but measurement precision may deteriorate without expert validation
Solution Approach 1:
The patent performs preliminary action by training the machine learning models in advance using labeled electromagnetic inversion data. The models learn accurate boundary identification patterns during the training phase, so that during real-time operation, they can immediately apply this pre-learned knowledge without requiring expert validation for each new interpretation.
Data Source
AI summary
The disclosure focuses on using a boundary identification system to actively determine borders and boundaries in subsurface geological features, such as reservoirs. In various implementations, the boundary identification system uses an ensemble image model leveraging multiple image-to-image machine-learning models to efficiently and accurately generate reservoir boundaries from inversion result profiles and images. In many instances, the boundary identification system generates reservoir boundaries from inversion results in real-time. Additionally, in some instances, the boundary identification system further improves the accuracy of the ensemble image model by diversifying the inputs and using ensembling on the individual model outputs during inference.