Prediction of Total Organic Carbon in Organic-Rich Shale Rocks Using Thermal Neutron Parameters

Area of Science:

• Geochemistry and Petroleum Geology
• Well Logging and Formation Evaluation
• Machine Learning in Earth Sciences

Background:

• Total organic carbon (TOC) is a critical parameter for source rock evaluation and hydrocarbon potential assessment.
• Traditional TOC quantification methods include laboratory analysis and empirical correlations using well logs, each with limitations in terms of cost, time, and data coverage.
• The potential of thermal neutron logs for TOC prediction in the Horn River Formation remains largely unexplored.

Purpose of the Study:

• To estimate total organic carbon (TOC) variations in the Horn River Formation by utilizing thermal neutron logs.
• To develop and optimize an artificial neural network (ANN) model for accurate TOC prediction.
• To establish an empirical correlation derived from the ANN model for practical field applications.

Main Methods:

• Collection and analysis of over 150 datasets from the Horn River Formation.
• Development and fine-tuning of an artificial neural network (ANN) model incorporating thermal neutron log data.
• Creation of an empirical correlation based on the optimized ANN model for direct TOC prediction.

Main Results:

• The ANN model achieved a coefficient of determination (R²) of 0.73, a significant improvement from previous models (0.28).
• The developed empirical correlation provides TOC predictions with an average absolute error of 0.52 wt %.
• The model demonstrates high accuracy for TOC ranges between 0.3 and 6.44 wt %.

Conclusions:

• Thermal neutron logs are effective predictors of TOC in the Horn River Formation.
• The developed ANN model and empirical correlation offer a reliable and efficient method for real-time TOC quantification.
• This approach enhances the assessment of organic matter maturity and aids in identifying productive zones within drilled formations.