---

Quantifying the potential of the Raniganj Basin for shale gas exploration and CO2 sequestration using a deep learning framework.

This study proposes an advanced deep learning model for accurately predicting Total Organic Carbon (TOC) content, a key parameter in evaluating shale gas potential and CO₂ storage capacity. Unlike traditional empirical models that often oversimplify shale heterogeneity, our model learns complex, nonlinear relationships between geophysical logs and geochemical properties to predict TOC. The findings highlight the Barren Measures shale in the Raniganj Basin, India, as a promising site for unconventional hydrocarbon exploration and carbon sequestration.

Architecture of the proposed deep learning model for optimized TOC predictions. Depth-wise comparison of deep learning model predicted TOC (olive lines) with core-derived TOC data (red dots) across three wells. The predicted TOC agrees well with core derived TOC measurements, indicating the model's accuracy in capturing TOC variations at different depths.

---