3D Fracability Mapping with Machine Learning for Well Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying fracable areas in tight-sand, naturally fractured reservoirs is challenging due to the limitations of existing methodologies that rely on net-pay models or rock quality sand maps, which do not accurately predict hydrocarbon production potential.
Innovation Solution
A machine learning model is developed that integrates 3D petrophysical, 3D geo-mechanical, and fracture models to predict fracture closure pressure (FCP) and fracability index, using hydraulic fracture parameters, and calibrates with well logs and mechanical test data to identify favorable drilling locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If net-pay models or rock quality sand maps are used to identify fracable areas, then the methodology is simple to implement, but the accuracy of predicting hydrocarbon production potential is insufficient
Solution Approach 1:
The patent combines multiple distinct models (petrophysical model, geo-mechanical model, and fracture model) into a single integrated machine learning model. This merging allows the system to leverage the strengths of each individual model while achieving higher prediction accuracy for hydrocarbon production potential that none of the models could achieve alone.
Solution Approach 2:
The patent creates a composite modeling approach by integrating diverse data types and model components (petrophysical properties, geo-mechanical properties, fracture characteristics) into a unified fracability index model. This composite structure enables the model to capture complex subsurface relationships that single-model approaches miss, thereby improving prediction accuracy.
2Reliability
If existing methodologies are used for well placement identification, then the process is quick and simple, but the reliability of identifying high production potential areas is low
Solution Approach 1:
The patent incorporates a feedback mechanism where the machine learning model is trained using historical well data and production outcomes. The model continuously learns from past performance, adjusting its predictions to improve reliability. This feedback loop enables the system to refine its identification of fracable areas based on actual production results, thereby enhancing reliability over time.
Solution Approach 2:
The patent performs preliminary calibration of the machine learning model using available well logs and mechanical test data before deploying it for fracability prediction. This preliminary action ensures the model is properly tuned and validated against known data, increasing its reliability before it is used to identify new high-potential areas.
3Measurement precision
If a comprehensive machine learning model integrating multiple geological elements is developed, then the prediction accuracy improves, but the computational complexity and data processing requirements increase
Solution Approach 1:
The patent segments the complex prediction task into distinct computational modules: a petrophysical model processing rock properties, a geo-mechanical model handling stress and strength parameters, and a fracture model evaluating fracture characteristics. Each module processes specific aspects independently, then their results are integrated to compute the final fracability index. This segmentation reduces computational complexity by avoiding the need to process all data simultaneously in a single monolithic model.
Data Source
AI summary
Systems, methods, and software can be used for identifying fracable areas. One example of a method includes receiving at least one of 3D petrophysical property and 3D geo-mechanical property in one or more areas. The method further includes receiving a model generating at least one hydraulic fracture parameter in a wellbore in the one or more areas. The method yet further includes training a machine learning model with at least one of the 3D petrophysical property and the 3D geo-mechanical property and the at least one hydraulic fracture parameter, and generating a fracability index for the wellbore based on the trained machine learning model.