What question did this study set out to answer?

This study aims to develop a data-driven model for predicting mineral and fluid volumes from well logging data in complex gas reservoirs.

Rock composition and fluid volume prediction from well logging data in an Iranian complex gas reservoir using a data-driven modeling

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates enhanced fluid and mineral volume prediction in complex gas reservoirs, indicating improved reservoir management.

This study aims to develop a data-driven model for predicting mineral and fluid volumes from well logging data in complex gas reservoirs.
Applied machine learning and deep learning frameworks on wireline log data from ten wells.
Developed and optimized four machine learning models using Bayesian optimization: kNN, AdaBoost, RF, XGBoost.
Proposed a multi-channel neural network architecture to enhance predictive accuracy by integrating multiple model outputs.
Achieved R² = 0.94 and MAE = 0.0007 for water content prediction using the MC-NN model.
Individual models showed lower predictive performance with XGBoost at R² = 0.90 and MAE = 0.0008, RF at R² = 0.80 and MAE = 0.0012.
Quantified model uncertainties via five-fold cross-validation, with R² standard deviation of 0.012–0.028.

Mousavi et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates enhanced fluid and mineral volume prediction in complex gas reservoirs, indicating improved reservoir management.