Deep Oil Navigation via Acoustic LWD Curve Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing deep oil and gas navigation systems face challenges in predicting missing logging data due to incomplete or unreliable acoustic LWD data, leading to inaccurate geological models and drilling trajectories, especially in situations where comprehensive logging data is not available.
Innovation Solution
A precise deep oil and gas navigation system is implemented using a drilling equipment group with advanced data acquisition and processing modules, including a data acquisition module, interference elimination module, data dimensionality reduction module, fusion feature data volume construction module, missing curve prediction module, stratigraphic structure model correction module, and trajectory adjustment module, which acquires and processes lithological, electrical, porosity, and drilling data to predict missing acoustic LWD curves and correct stratigraphic structure models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If acoustic LWD data is used to estimate mechanical properties of rocks, then measurement precision is improved, but reliability deteriorates when data is missing or reduced due to cost constraints or wellbore issues
Solution Approach 1:
The patent introduces an intermediary prediction system that uses machine learning models to bridge the gap between available logging data and the desired acoustic LWD data. The system employs multiple logging types (neutron, photoelectric factor, gamma ray, resistivity, density, etc.) as intermediate data sources to predict missing acoustic LWD curves, thereby maintaining measurement precision while overcoming data reliability issues.
Solution Approach 2:
The patent transforms the prediction problem by changing parameters from direct acoustic LWD measurements to alternative logging parameters that can be used as inputs for prediction models. The system uses parameters from different logging types (lithological, electrical, porosity, drilling, and logging groups) to predict the acoustic LWD response, effectively substituting one set of parameters for another when direct measurement is unavailable.
2Loss of information
If missing logging data is predicted through mathematical methods based on other logging types, then loss of information is reduced, but manufacturing precision deteriorates because the dependency relationship may not exist
Solution Approach 1:
The patent applies partial action by selecting only the most relevant logging data groups for prediction based on the specific well conditions and available data. The system uses an interference elimination module to remove redundant or irrelevant data groups before prediction, focusing computational resources on the most informative parameters. This selective approach improves prediction precision by avoiding the use of unrelated logging types as prediction bases.
Solution Approach 2:
The patent implements feedback mechanisms where the predicted acoustic LWD curves are validated against available actual measurements and geological knowledge. The system uses correlation analysis and geological constraint modules to verify prediction accuracy, providing feedback that can adjust prediction models and parameters. This feedback loop ensures that predictions maintain high precision by continuously verifying results against known geological relationships.
3Ease of operation
If local analysis is used to establish dependency relationships between logging types, then ease of operation is improved, but measurement precision deteriorates due to geological depth changes
Solution Approach 1:
The patent segments the geological formation into different depth intervals and lithological units, applying prediction models separately to each segment. The system divides the well data into multiple groups (lithological, electrical, porosity, drilling, logging) and processes them through separate analysis streams. This segmentation allows the system to adapt to local geological conditions at different depths while maintaining overall prediction accuracy, overcoming the limitation of uniform local analysis.
Solution Approach 2:
The patent extends the analysis from simple local depth-based relationships to multi-dimensional relationships by incorporating multiple logging types and their interrelationships. The system uses machine learning models that analyze patterns across different data dimensions (lithological properties, electrical characteristics, porosity, drilling parameters, and logging measurements) simultaneously. This multi-dimensional approach captures complex geological variations that single-dimension local analysis would miss, improving prediction accuracy while maintaining operational efficiency.
Data Source
AI summary
A precise deep oil and gas navigation system based on structural evaluation of a sand/shale formation, and a device are provided. The precise deep oil and gas navigation system is implemented by: acquiring basic data of a target well location as well as basic data and an acoustic LWD curve of an adjacent well; dividing a basic data group; retaining extremely strongly correlated data and strongly correlated data of the basic data group, and performing dimensionality reduction; constructing a three-dimensional fusion feature data volume based on a dimensionality-reduced fusion feature parameter and a time-frequency spectrum; predicting a missing curve according to the three-dimensional fusion feature data volume; and correcting a stratigraphic structure model based on the missing curve to guide the design of a drilling trajectory.