Data-driven analysis of well logging data for the coal mining

Tóm tắt

Coal remains one of the most widely utilized fossil fuels globally, playing a crucial role in energy production and industrial processes. As global energy demands continue to rise, the efficient and sustainable exploitation of coal resources has become increasingly important. Efficiency can be significantly enhanced through the application of geological and geophysical methods, among which well-logging holds particular significance due to its ability to provide detailed subsurface information. Well-logging data, when properly analyzed and interpreted, offer critical insights into the geological and stratigraphic characteristics of coal-bearing formations. These insights are essential for constructing accurate geological models, which, in turn, ensure that coal extraction is conducted safely, efficiently, and within planned timelines. In recent years, the integration of artificial intelligence (AI) and machine learning (ML) techniques into geoscientific workflows has opened new avenues for data-driven decision-making. These technologies are particularly valuable in handling the vast and complex datasets generated during coal assessment, exploration, and discovery. By identifying patterns and relationships within the data, ML models can enhance predictive accuracy and reduce the reliance on manual interpretation. This study applied several machine learning algorithms to predict coal seam depth and thickness using well-logging data collected from the X mine site in Quảng Ninh Province. The final model demonstrated consistently strong predictive performance when validated against actual well data, accurately identifying lithological boundaries and coal-bearing intervals. These encouraging outcomes highlight the potential of advanced computational techniques to significantly enhance coal seam characterization, offering more efficient, accurate, and cost-effective alternatives to traditional exploration methods