Geological Modeling Agent: Automated Static Model Uncertainty Assessment Using AI Agent and Geology-Aware Guidance

This abstract presents a fully automated geological modeling agent specialized in uncertainty modeling and optimization. The agent is guided by geological expertise, utilizing a set of predefined questions to ensure accurate and relevant outputs. By generating thousands of realizations, the agent comprehensively cover the full range of uncertainty parameters, including variogram ranges of porosity models, oil-water contact and seed number uncertainties by the Latin-hyper cube method. This enables robust volume assessment and sensitivity analysis workflows.

The agent’s capabilities facilitate the quantification of uncertainty in geological models, allowing for more informed decision-making. By automating the modeling process, the agent increase efficiency and reduces the risk of human error. The of realizations generated by the agent provide a comprehensive understanding of the uncertainty space, enabling the identification of key factors influencing model outcomes. This innovative approach has significant implications for the oil and gas industry, where accurate geological modeling is critical for optimizing resource extraction and minimizing uncertainty.