AI Predicts True Sand Resistivity in Shaly Sands
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional resistivity logs fail in low resistivity laminated shaly or silty sand reservoirs due to the presence of shale and silt, leading to underestimation of hydrocarbon reserves and inaccurate water saturation calculations, especially in wells drilled before the advent of tri-axial induction resistivity logging technology.
Innovation Solution
The use of artificial intelligence (AI) models, such as machine learning algorithms, to predict true sand resistivity (RSS) logs from basic well log data, including resistivity, density, and gamma ray logs, allowing for accurate estimation of hydrocarbon properties without the need for advanced resistivity measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional resistivity logs are used in low resistivity laminated shaly or silty sand reservoirs, then the logging process is simple and fast, but the measurement precision of true sand resistivity deteriorates due to the presence of shale and silt
Solution Approach 1:
The patent introduces an artificial intelligence model as an intermediary between basic well log data and true sand resistivity prediction. The AI model processes conventional log data (resistivity, density, neutron, gamma ray) and predicts RSS values without requiring complex tri-axial induction logging tools, thereby improving measurement precision while avoiding device complexity
Solution Approach 2:
The patent creates a predictive copy of true sand resistivity data using AI models trained on existing well data. Instead of directly measuring RSS with complex tools, the system generates predicted RSS values that replicate the information obtained from advanced logging, achieving accurate measurements through data processing rather than direct physical measurement
2Measurement precision
If tri-axial induction resistivity logging tools are used to obtain accurate RSS logs, then the measurement precision improves, but the ease of operation deteriorates and operational costs increase
Solution Approach 1:
The patent replaces expensive, complex tri-axial induction logging tools with conventional, widely available logging tools combined with AI processing. The system uses basic resistivity, density, neutron, and gamma ray logs that can be obtained with standard equipment, making the process easier to operate and more cost-effective while still achieving accurate RSS predictions
Solution Approach 2:
The patent substitutes the mechanical/physical measurement system (tri-axial induction logging tools) with an information processing system (AI models). Instead of using complex physical tools to directly measure RSS, the system uses machine learning algorithms to predict RSS from conventional log data, replacing a complex mechanical measurement approach with a computational approach that is easier to implement
3Measurement precision
If more well logs and reservoir parameters are collected for AI model training, then the prediction accuracy improves, but the loss of time and resources for data collection increases
Solution Approach 1:
The patent performs preliminary action by training the AI model in advance using comprehensive data from multiple wells. The model is pre-trained on extensive datasets including resistivity, density, neutron, and gamma ray logs from numerous existing wells, so that when the model is deployed, it can quickly predict RSS for new wells without requiring additional data collection time. The heavy data processing work is done beforehand during the training phase
Solution Approach 2:
The patent creates a universal AI model that can process multiple types of well log data (resistivity, density, neutron, gamma ray) and predict RSS across different well conditions. This multi-functional model consolidates the need for various specialized processing routines into a single system that can handle diverse input data types, reducing the overall time and resources needed for data collection and processing
Data Source
AI summary
Systems, methods, and apparatus including computer-readable media for predicting true sand resistivity (RSS), for example, in laminated shaly sands, with artificial intelligence (AI) are provided. In one aspect, a computer-implemented method includes: obtaining basic log data of a target well, the basic log data including well logs of multiple types of the target well, and predicting a true sand resistivity (RSS) log of the target well using a trained artificial intelligence (AI) model with inputs including the well logs of the multiple types of the target well. The AI model was trained with well logs of the multiple types of existing wells as training inputs and known RSS logs for the existing wells as training outputs.