Deep Learning Lithological Analysis for Hydrocarbon Targeting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current drilling processes face challenges in efficiently integrating textual descriptions of rock drill cuttings with petrophysical models, leading to uncertainty in identifying hydrocarbon targets and inefficiencies in drilling operations due to the time-sensitive nature of drilling rigs and the inability to effectively combine wireline logs and cutting descriptions.
Innovation Solution
A computer-implemented method using deep learning models to preprocess textual descriptions of rock drill cuttings, generate bag-of-words vectors, and train models to output hydrocarbon potential, enabling integration with petrophysical analysis and reducing unnecessary testing by informing hydrocarbon targets during drilling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual interpretation of rock drill cuttings is used, then detailed lithological analysis can be performed, but it consumes significant time and increases operational costs
Solution Approach 1:
The patent replaces manual mechanical interpretation of rock drill cuttings with an automated deep learning system that processes images and generates lithological descriptions. The system uses convolutional neural networks to analyze cutting images and automatically produces detailed lithological reports, eliminating the need for manual examination while maintaining or improving analysis accuracy.
Solution Approach 2:
The patent creates digital copies of rock drill cuttings through imaging systems, which are then processed by deep learning algorithms. These digital representations allow for rapid automated analysis without requiring physical handling or manual inspection of the actual cuttings, significantly reducing analysis time while preserving detailed examination capabilities.
2Reliability
If textual descriptions of cuttings are integrated with petrophysical models, then hydrocarbon target identification improves, but the complexity of data integration increases
Solution Approach 1:
The patent develops a unified deep learning framework that simultaneously processes multiple data types including cutting images, textual descriptions, and petrophysical log data. This multi-functional system integrates diverse data sources through a single automated platform, improving hydrocarbon target identification while managing complexity through standardized processing pipelines and common data representations.
3Productivity
If automated deep learning models are implemented, then analysis speed increases, but the initial system complexity and training requirements increase
Solution Approach 1:
The patent implements comprehensive preprocessing pipelines that prepare cutting images and associated data before they reach the deep learning models. This includes automated image enhancement, normalization, and feature extraction that simplify the input data structure. Additionally, the system uses pre-trained models and transfer learning techniques to reduce training requirements and accelerate deployment while maintaining high analysis speeds.
Data Source
AI summary
A computer-implemented method for integrating text analysis of lithological descriptions with petrophysical models is described herein. The method includes preprocessing textual descriptions associated with cuttings to generate training data and generating bag-of-words vectors using the training data. The method also includes training a deep learning model to output hydrocarbon potential using the bag-of-words vectors as input and executing the trained deep learning model on unseen textual descriptions.