Subsurface Depth Uncertainty Modeling for Drilling Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional depth uncertainty methods in subsurface characterization fail to accurately consider multiple factors affecting depth uncertainty, leading to imprecise estimations that can negatively impact resource assessments and drilling decisions.
Innovation Solution
Utilizing machine learning techniques such as gradient boost regression and neural networks to train a depth uncertainty model that accounts for parameters like depth below mud line, seismic velocity, and well type, enabling precise estimation of depth uncertainty ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional depth uncertainty methods are used, then the estimation process is simple, but the measurement precision of depth uncertainty is poor
Solution Approach 1:
The patent replaces conventional mechanical/statistical estimation methods with machine learning models (neural networks, random forests, gradient boost regression) to predict depth uncertainty. This substitution enables the system to capture complex non-linear relationships between seismic data characteristics and depth uncertainty, significantly improving measurement precision while accepting increased computational complexity.
Solution Approach 2:
The patent transforms the approach by changing from direct depth measurement to predicting depth uncertainty through multiple intermediate parameters (seismic data quality, velocity model accuracy, structural complexity). This parameter transformation allows the system to account for multiple sources of uncertainty simultaneously, improving overall estimation precision.
2Measurement precision
If multiple factors are considered in depth uncertainty estimation, then the measurement precision improves, but the device complexity increases
Solution Approach 1:
The patent develops universal machine learning models that can simultaneously evaluate multiple factors affecting depth uncertainty (seismic data quality, velocity model accuracy, structural complexity, depth range) through a single integrated system. This multi-functionality allows comprehensive consideration of multiple factors without proportionally increasing system complexity, as the model handles all parameters through unified computational frameworks.
Solution Approach 2:
The patent segments the depth uncertainty estimation problem into distinct input factors (seismic data characteristics, velocity model parameters, structural features) that can be independently evaluated and processed. This segmentation allows the system to systematically incorporate multiple factors while maintaining model manageability through modular feature engineering and independent parameter optimization.
3Reliability
If quantitative depth uncertainty ranges are provided, then the reliability of drilling decisions improves, but the loss of information increases due to model assumptions
Solution Approach 1:
The patent incorporates feedback mechanisms where predicted depth uncertainty values are validated against actual drilling outcomes and used to retrain and improve the models. This feedback loop reduces information loss by continuously refining model assumptions based on real-world data, thereby enhancing the reliability of drilling decisions while minimizing the gaps between model predictions and actual conditions.
Solution Approach 2:
The patent performs preliminary analysis of seismic data characteristics and velocity model properties before making depth uncertainty predictions. By pre-processing and characterizing the input data, the system reduces information loss from model assumptions through more accurate feature extraction and preliminary filtering of relevant parameters that better represent actual subsurface conditions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods for estimating reservoir productivity as a function of position in a subsurface volume of interest are disclosed. Exemplary implementations may: obtain an initial depth uncertainty model; obtain training depth uncertainty parameter values from the non-transient storage medium; obtain corresponding training depth uncertainty values; generate a trained depth uncertainty model by training the initial depth uncertainty model using the training depth uncertainty parameter values and the corresponding training depth uncertainty values; and store the trained depth uncertainty model.