Identification of natural fractures in shale gas reservoirs using fracture signature function and machine learning models
Bibliographic record
Abstract
Identifying fractures is important for optimizing recovery and enhancing oil recovery techniques. Identifying natural fractures using FMI and cores is expensive and unavailable for all wells. Therefore, predictive models based on conventional wireline logs are necessary. Extreme Gradient Boosting, Decision tree, Random Forest, Support Vector Machine, Feed Forward Neural Network, and Recurrent Neural Network were applied to identify natural fractures. This study uses well logs to develop a new Fracture Signature Function equation for determining natural fractures in the shale reservoir. This unsupervised approach requires no special image log to identify natural fractures. However, the class imbalance problem between the fracture and non-fractured zone often restricts the accuracy of the machine learning models, which require a predictive model not dependent upon the special logs and class imbalance problems in the prediction of fractured zones. Synthetic Minority Oversampling (SMOTE) and Random Oversampling (ROS) were applied to solve the class imbalance problem in the data. The results show that the machine learning models did not predict the fracture and non-fracture zones with acceptable accuracy even after applying SMOTE and ROS. Relative to all machine learning models, Random Forest predicted the results with the highest accuracy of 91 % and F1-Score of 17.6 %. The Fracture Signature Function (FSFn'') predicted the natural fractures with high accuracy except in zones with very complex borehole environments. A Forward Neural Network is more efficient in identifying fracture and non-fractured zones in imbalance class problems of the dataset. The Recurrent Neural Network's predictions were biased toward the major class related to the non-fractured zone of the studied interval. The newly developed equation can be used for natural fracture identification in drilling and production strategy design by using the easily available well-log data in class imbalanced conditions.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".