Machine Learning Time-to-Depth Seismic Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for converting time-domain seismic data to depth-domain data in subsurface exploration are time-consuming and require expert analysis, often resulting in inefficient and biased depth estimations due to complex geological interpretations.
Innovation Solution
The use of machine learning techniques, specifically deep neural networks, to optimize domain conversion by training models with seismic data, reducing the need for expert analysis and enabling quick, accurate, and high-resolution conversions of time-domain data to depth-domain data, even in geologically complex formations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional velocity modeling methods are used for time-to-depth conversion, then depth estimation can be achieved, but the process becomes time-consuming and requires multiple iterative steps
Solution Approach 1:
The patent replaces the conventional mechanical velocity modeling process with a machine learning-based system. A neural network is trained on seismic data to directly predict depth-domain information from time-domain seismic data, eliminating the need for iterative velocity modeling steps and well-to-seismic tie adjustments. This substitution of the mechanical modeling system with an intelligent system achieves both speed and accuracy.
Solution Approach 2:
The patent applies preliminary action by pre-training the machine learning model on extensive seismic data before actual depth conversion. The model learns velocity relationships and depth transformations during the training phase, so that during actual operation, depth conversion can be performed directly without requiring the preliminary iterative adjustments that conventional methods need.
2Measurement precision
If expert analysis is used to evaluate seismic data and velocity models, then accurate depth conversion is achieved, but the process requires highly skilled personnel and becomes prohibitively inefficient for complex geometries
Solution Approach 1:
The patent implements self-service by enabling the system to automatically perform depth conversion without requiring expert human intervention. The machine learning model independently evaluates seismic data, predicts velocity variations, and performs depth conversion autonomously. The system serves itself by learning from training data and applying that knowledge to new cases without needing expert analysts to interpret each dataset.
Solution Approach 2:
The patent changes the fundamental parameters of the conversion process by transitioning from manual expert evaluation to automated machine learning prediction. The system transforms qualitative expert judgment into quantitative automated prediction, changing how velocity models are created and how depth conversions are performed, thereby improving both accuracy and efficiency simultaneously.
3Reliability
If detailed geological interpretation by experts is performed, then accurate velocity models can be created, but the process introduces potential biases and becomes prohibitively complex for highly faulted subsurface geometries
Solution Approach 1:
The patent replaces the complex mechanical process of detailed geological interpretation with an automated machine learning system. The neural network processes seismic data directly, learning velocity variations and geological features without requiring expert interpretation. This substitution eliminates the complexity of manual modeling while maintaining or improving accuracy through data-driven predictions.
Solution Approach 2:
The patent uses copying by training the machine learning model on examples of seismic data and corresponding velocity models. The model learns to copy the relationships between seismic characteristics and velocity variations from training data, then applies this learned knowledge to predict velocity models for new datasets without requiring experts to recreate the interpretation process.
Data Source
AI summary
Optimizing seismic to depth conversion to enhance subsurface operations including measuring seismic data in a subsurface formation, dividing the subsurface formation into a training area and a study area, dividing the seismic data into training seismic data and study seismic data, wherein the training seismic data corresponds to the training area, and wherein the study seismic data corresponds to the study area, calculating target depth data corresponding to the training area, training a machine learning model using training inputs and training targets, wherein the training inputs comprise the training seismic data, and wherein the training targets comprise the target depth data, computing, by the machine learning model, output depth data corresponding to the study area based at least in part on the study seismic data; and modifying one or more subsurface operations corresponding to the study area based at least in part on the output depth data.