Machine Learning Based Integrated Approach to Estimate Total Organic Carbon in Shale Reservoirs – A Case Study from Duvernay Formation, Alberta Canada
Bibliographic record
Abstract
Abstract Shale gas reservoirs have become prominent contributors to the world's hydrocarbon resources and production. They exhibit multiple storage mechanisms, two of which are linked to the free and adsorbed gas phase. Since the adsorbed gas may be stored as a denser phase than the free gas, the contribution of the adsorbed phase can be significant. The adsorbed volume is related to the total organic carbon (TOC) and thus, higher TOC can indicate higher hydrocarbon inplace. Furthermore, productivity can be linked to TOC through the potential for overpressure and conversion of kerogen to pore space. However, estimation of the TOC is not a trivial problem, as it depends on geological factors such as depositional environment. In this study, we propose an integrated workflow using concepts of machine learning to estimate TOC. The workflow is divided into 3 sections which are area selection, sub-region categorization, and prediction modeling. Firstly, 3 active exploration and development areas (Kaybob, Pembina, and East shale basin) of the Duvernay Formation are highlighted and the geology of each specific area is analyzed. Thereafter, using the available core data and average properties of the attributes (Gamma Ray, resistivity, density, and distance from mean vitrinite reflectance line), each area is clustered into sub-regions using SVM, logistic regression, and k-means clustering. Finally, using Random Forest prediction, models for each sub-region are developed and ranked with average mean square errors and standard deviations. It is observed that the Kaybob area can be clustered into 2 regions. This observation is supported by the principal component plot (PC1 vs PC2), which shows a dual cloud structure. This is further supported through clustering analysis, which also revealed the same observation. Results of the prediction modeling found random forest as the best predictor, achieving a match wiht the core data with a error less than 10% and in some cases only a 1% deviation. Shale reservoir characterization requires estimation of the key parameters such as TOC. However, it is difficult to estimate TOC with purely physics-based or purely statistical models, as it requires limited specialized data and is impacted by subtle variations in the reservoir. This study suggests that TOC can be accurately estimated by combining geological interpretation and machine learning based algorithms without bearing cost of the specialized data.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| 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 source (direct Gemma or distilled Codex), 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".