Well Log Quality Improvement via Bad Hole Data Replacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Well logs obtained during oil and gas exploration often contain erroneous or missing data, which hinders accurate interpretation and can lead to suboptimal exploration results.
Innovation Solution
A well log quality improvement method and apparatus that uses a determination model to identify and replace bad hole sections with alternative data, and normalizes the data distribution based on a reference well log, employing techniques such as machine learning models and auto trend matching to enhance data accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If well log data is used directly for interpretation, then analysis time is short, but data accuracy is reduced due to erroneous or missing data
Solution Approach 1:
The patent applies preliminary action by performing quality control operations before the actual well log interpretation. The system automatically identifies bad hole sections and generates alternative data through conditioning operations, so that when the well log is analyzed, the data is already cleaned and prepared. This preliminary data preparation reduces the time needed during interpretation while maintaining high data accuracy.
Solution Approach 2:
The patent introduces an intermediary system that acts as a mediator between the raw well log data and the final interpretation. This intermediary includes a determination model that identifies erroneous data, a conditioning operation that generates alternative data, and a quality control module that validates the data. This intermediary layer filters out bad data automatically, providing clean data for interpretation without requiring manual intervention, thus reducing analysis time while improving accuracy.
2Measurement precision
If manual quality control is performed on well log data, then data accuracy is improved, but operation complexity increases
Solution Approach 1:
The patent implements self-service by enabling the system to automatically perform quality control operations on well log data without requiring manual intervention. The determination model autonomously identifies bad hole sections by analyzing multiple parameters, and the conditioning operation automatically generates alternative data. The quality control module self-validates the data quality. This automation maintains high data accuracy while significantly reducing operational complexity compared to manual quality control processes.
Solution Approach 2:
The patent applies parameter changes by transforming the well log data through systematic parameter modifications. The system changes data parameters by generating alternative data with different statistical properties, applying transformations to normalize distributions, and adjusting parameters to match reference well logs. These automated parameter changes improve data accuracy without requiring complex manual operations, as the system handles the complexity of parameter transformation internally.
3Measurement precision
If multiple determination models are used to identify bad hole sections, then detection accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies merging by combining multiple determination models into a unified quality control system. Instead of running models sequentially, the system integrates multiple models to work together in a coordinated manner, where each model contributes specific detection capabilities. The results from multiple models are merged and validated together, allowing the system to achieve high detection accuracy through ensemble approaches while optimizing processing time through parallel execution and efficient integration of model outputs.
Data Source
AI summary
Proposed are a well log quality improvement apparatus and a well log quality improvement method. In an embodiment, a well log quality improvement method includes performing a quality controlling operation on a well log by inputting the well log to a well logging data processing model to train the model and by determining a bad hole section corresponding to a log section associated with a bad hole, performing a conditioning operation on the well log by replacing the bad hole section included in the well log with alternative data, and normalizing a distribution of data of the well log according to a distribution of data of a reference well log obtained from a reference well. By improving the quality of the well log, the overall accuracy of oil and gas exploration may be improved, and time required for data analysis may be reduced.