Dynamic Time Warping for Geological Formation Tops Picking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for identifying geological formation tops in wells are manual, time-consuming, and require high-level geological expertise, limiting their efficiency and accuracy in hydrocarbon exploration and reservoir characterization.
Innovation Solution
The use of dynamic time warping (DTW) techniques to automatically predict geological formation tops from well logs, enabling the alignment and correlation of well logs between key master and training wells, thereby reducing cycle times and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual methods are used for identifying geological formation tops, then high-level geological expertise can be applied, but the process becomes time-consuming and less efficient
Solution Approach 1:
The patent replaces manual mechanical analysis by geoscientists with an automated computer-based system using dynamic time warping algorithms. The system automatically correlates well logs between master and training wells to predict formation tops, eliminating the need for manual interpretation while maintaining or improving accuracy through consistent algorithmic application.
Solution Approach 2:
The patent creates a digital model by copying the geological formation data from the master well and applying it to training wells through DTW correlation. The system generates predicted formation tops for training wells by matching patterns from the master well, effectively copying and adapting geological knowledge across multiple wells automatically.
2Reliability
If manual processes are used for picking well tops, then geological expertise can be utilized, but manpower resources are not optimized and quality at corporate database level is reduced
Solution Approach 1:
The system enables self-service by allowing the automated DTW algorithm to independently perform formation tops picking without requiring continuous human intervention. Once the master well is established, the system automatically correlates and predicts formation tops across multiple training wells, freeing manpower for higher-value activities while maintaining data quality through systematic processing.
Solution Approach 2:
The patent creates a universal system that can be applied across multiple wells and different geological scenarios. The DTW algorithm serves multiple functions: correlating well logs, predicting formation tops, updating 3D reservoir models, and generating indices for various training wells, thereby optimizing both data quality and manpower utilization across the entire exploration portfolio.
3Measurement precision
If conventional manual methods are used, then detailed geological interpretation can be performed, but the process cannot be scaled to multiple wells simultaneously
Solution Approach 1:
The patent implements a dynamic system using DTW algorithms that can adaptively correlate well logs across varying geological conditions. The system dynamically adjusts the correlation by generating warping indices that optimize the matching between master and training wells, allowing it to handle multiple wells with different characteristics while maintaining consistent quality standards across the entire dataset.
Data Source
AI summary
Systems and methods include a method for predicting geological formation tops. First well log data associated with a key master well is received. Formation data identifying tops of formations confirmed in the key master well is received. Merged key master well and formation data is generated in a dynamic time warping (DTW)-readable format by merging the first well log data with the formation data. Second well log data associated with a training well located in geographic proximity to the key master well is received. The second well log data is formatted into the DTW-readable format. A DTW function is executed to generate indices associated with the formation tops. The DTW function uses the merged key master well and formation data and the formatted second well log data as DTW function inputs. Predicted geological formation tops for the training well are predicted using the generated indexes.