Predicted Trend Logs for Stratigraphic Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for sequence stratigraphic correlation in subsurface formations are limited in identifying and correlating geologic patterns across multiple wells, often resulting in inconsistent and inaccurate interpretations due to noise sensitivity and lack of global pattern recognition.
Innovation Solution
The method involves generating predicted trend logs from well log data using a sliding window to select maximum and minimum values, computing mean values, estimating a mean drift trend, configuring a recursive filter, and calculating a relative trend log, which is then integrated and normalized to produce a predicted trend log for cross-sectional rendering, facilitating consistent pattern identification across the subsurface formation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used for sequence stratigraphic correlation, then the process is simpler, but the accuracy and consistency of pattern identification across multiple wells deteriorates
Solution Approach 1:
The method segments the well log data processing into distinct computational stages: calculating maximum and minimum envelopes, computing mean trends, estimating drift components, and generating predicted trend logs. This segmentation allows each stage to address specific aspects of pattern identification, improving overall accuracy while maintaining systematic complexity management.
Solution Approach 2:
The patent introduces predicted trend logs as an intermediary representation that mediates between raw well log data and final stratigraphic correlation. These predicted trend logs serve as a standardized intermediate product that enhances consistency across multiple wells while providing a foundation for accurate pattern identification without requiring direct complex comparison of all raw data.
2Reliability
If noise-sensitive methods are used, then the processing is faster, but the reliability of geologic pattern identification deteriorates
Solution Approach 1:
The method extracts and removes noise components from well log data through envelope calculation and mean trend computation. By separating the signal (geologic patterns) from noise (measurement variations) through mathematical operations on maximum and minimum envelopes, the method improves reliability of pattern identification while maintaining processing efficiency through algorithmic optimization.
Solution Approach 2:
The patent employs iterative refinement where predicted trend logs are generated and then used to improve subsequent correlation analyses. The feedback loop allows the system to learn from previous analyses and continuously improve pattern identification reliability, with the predicted trends serving as both output and input for refined interpretations.
3Measurement precision
If global pattern recognition is not implemented, then the method is easier to apply, but the consistency of cross-well correlation deteriorates
Solution Approach 1:
The predicted trend log methodology serves multiple functions simultaneously: it identifies local geologic patterns, establishes global correlation frameworks, and provides standardized representations for cross-well comparison. This multi-functionality achieves consistent cross-well correlation while avoiding the need for separate specialized methods for different correlation scales.
Solution Approach 2:
The method transforms well log data into predicted trend logs through parameter transformations that emphasize geologically significant variations. By changing the representation parameters from raw log values to trend-based predictions, the method enhances consistency of cross-well correlation while managing analytical complexity through mathematical transformation.
Data Source
AI summary
Methods and devices for correlating stratigraphic sequences in a subsurface formation include: measuring a property of a subsurface formation at a plurality of wells using a logging tool; and generating a predicted trend log from the well log for the well. Generating the predicted trend log includes: selecting a set of maximum values and a set of minimum values using a sliding window; computing a set of mean values of the property; estimating a mean drift trend based on the set of mean values and the values of the property; configuring a recursive filter; and convolving the recursive filter to obtain a relative trend log.