Machine Learning Drilling Parameter Analysis for Elastic Properties
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional drilling methods are inefficient due to suboptimal drilling program designs and crew mistakes, leading to increased rig time, costs, and drilling incidents, as they fail to effectively determine the elastic properties of geological formations in real-time.
Innovation Solution
A machine learning-based system that extracts feature vectors from drilling parameters, logging while drilling logs, and bit vibrations to classify the drilling environment and select appropriate regression algorithms for predicting elastic properties of geological formations, enabling real-time graphical representation and optimization of the drilling process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional drilling methods are used, then drilling operations can be performed with simple equipment, but drilling efficiency is reduced and costs increase due to inability to determine elastic properties in real-time
Solution Approach 1:
The patent replaces traditional mechanical drilling optimization methods with a machine learning-based computational system. The system uses neural networks to process drilling parameters (weight on bit, rotary speed, pump rate, vibrations) and automatically determine elastic properties of geological formations, substituting complex mechanical trial-and-error approaches with intelligent algorithms that improve drilling efficiency while managing system complexity through software-based solutions.
2Measurement precision
If direct measurements of elastic properties are taken, then accurate formation characterization is achieved, but drilling costs and time increase
Solution Approach 1:
The patent uses drilling parameters (weight on bit, rotary speed, pump rate, bit vibrations) as intermediary measurements that can be obtained during normal drilling operations without stopping. These intermediate measurements are fed into machine learning models to indirectly determine elastic properties, serving as a mediator between easily measurable drilling conditions and the desired formation characteristics, thereby achieving accurate characterization without direct formation testing that would consume time and resources.
Solution Approach 2:
The system creates a virtual model or copy of the formation's elastic properties by training neural networks on drilling parameter data. Instead of directly measuring formation properties, the system generates a computational replica that predicts elastic moduli based on patterns learned from drilling vibrations and operational parameters, enabling accurate formation characterization as a digital copy without physical formation testing.
3Adaptability or versatility
If multiple machine learning algorithms are evaluated, then the most suitable algorithm is selected for specific drilling conditions, but computational complexity increases
Solution Approach 1:
The patent segments the drilling environment into distinct categories (e.g., hard rock, soft rock, variable conditions) and assigns specific machine learning algorithms to each segment. By dividing the complex problem of algorithm selection into manageable segments based on drilling conditions, the system achieves adaptability across different geological formations while keeping the overall system complexity controlled through modular algorithm deployment.
Data Source
AI summary
Methods for determination of elastic properties of geological formations using machine learning include extracting a first feature vector from data acquired during drilling. The data includes at least drilling parameters. The first feature vector is indicative of a drilling environment classification. A machine learning classification algorithm determines the drilling environment classification based on the first feature vector. A machine learning regression algorithm is selected from multiple machine learning regression algorithms based on the drilling classification. A second feature vector is extracted from the data acquired during drilling based on the drilling classification and the selected machine learning regression algorithm. The second feature vector is indicative of elastic properties of a geological formation. The selected machine learning regression algorithm determines the elastic properties of the geological formation based on the second feature vector.