Machine Learning Borehole Image Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Borehole images generated during well operations are often inaccurate, unclear, or suboptimal due to factors like finite resolution of imaging tools, imperfect calibration, and noise, which hinders well operators' ability to make informed decisions during drilling, completion, and production operations.
Innovation Solution
A machine-learning model, such as a deep neural network, is trained using a dataset of correlated borehole images to visually correct borehole images by enhancing their clarity, sharpness, and accuracy, allowing for real-time correction of borehole images during ongoing operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional imaging tools and processing methods are used to generate borehole images, then the imaging process is straightforward and relatively simple, but the resulting images are inaccurate, unclear, and suboptimal due to finite resolution, imperfect calibration, and noise
Solution Approach 1:
A machine learning model serves as an intermediary between the raw borehole image data and the final corrected image. The model is trained on a dataset of correlated borehole images to learn the complex relationships and transformations needed to correct inaccuracies, thereby resolving the contradiction by introducing a sophisticated intermediate processing layer that improves accuracy without requiring fundamental changes to the imaging hardware itself
Solution Approach 2:
The patent applies parameter changes by transforming the borehole image data through a machine learning model that adjusts multiple parameters simultaneously including resolution, calibration accuracy, and noise levels. The model learns optimal parameter transformations from training data, enabling it to correct images by adjusting these parameters in a coordinated manner to achieve superior image quality
2Measurement precision
If machine-learning models are used to visually correct borehole images, then image quality and accuracy are improved, but processing time and computational resources increase
Solution Approach 1:
The machine learning model is trained in advance on a large dataset of correlated borehole images before being deployed for actual image correction. This preliminary training phase allows the model to learn complex patterns and corrections, so that during actual operation, the model can quickly apply these learned transformations to new images without requiring extensive processing time, thus resolving the contradiction between high image quality and processing speed
Solution Approach 2:
The patent uses a dataset of correlated borehole images as training data, where the model learns to map between different image qualities and characteristics. By copying and learning from numerous example image pairs during training, the model acquires the ability to rapidly correct new images, transferring the computational burden from real-time operation to the training phase
3Loss of information
If traditional borehole imaging methods are used, then the equipment and process are simpler, but the ability to identify fractures and structures is hindered due to image inaccuracies
Solution Approach 1:
The machine learning model acts as an intermediary processing layer that enhances geological information extraction by correcting image inaccuracies. The model is specifically trained to preserve and enhance geological features such as fractures and structures while removing artifacts and noise, thereby resolving the contradiction by introducing a sophisticated information processing intermediary that recovers lost geological information without requiring more complex imaging hardware
Solution Approach 2:
The correction system applies parameter changes to enhance the visibility and accuracy of geological features. The machine learning model adjusts image parameters such as contrast, resolution, and feature enhancement to optimize the display of geological information, thereby resolving the contradiction by transforming the image parameters to maximize geological information content while using a software-based solution rather than more complex hardware
Data Source
AI summary
Borehole images can be corrected using machine-learning models. For example, a system can train a machine-learning model based on a training dataset. The training dataset can include a first set of borehole images correlated to a second set of borehole images, where the second set of borehole images are less precise versions of the first set of borehole images. The system can then execute the trained machine-learning model in relation to an input borehole image to receive a corrected borehole image as output from the trained machine-learning model. The corrected borehole image can be a visually corrected version of the input borehole image. The system may then perform one or more operations based on the corrected borehole image, such as generating a graphical user interface that includes the corrected borehole image for display on a display device.