Integrated porosity estimation with uncertainty propagation for reservoir simulation models

Porosity is a crucial parameter in assessing reservoir performance, as its accurate estimation in simulation models impacts reserves evaluation, production forecasting, and field development. The inherent uncertainties associated with porosity estimation, namely core stress relief, logging errors, gas effects, and interpretational uncertainty, necessitate specialized approaches during model development. Traditional methods often address these uncertainties only partially, are weakly constrained by well data, may lack sufficient detail, and may not preserve geological consistency. This study presents an integrated methodology that systematically assesses these uncertainties by combining core analysis data, well logs, and automated parameter variation. To enhance reliability, the methodology ranks porosity estimates from well logs by correlation with core data. A comprehensive uncertainty assessment employs well-zone regionalization, treating each geological layer as an independent zone. All identified uncertainties are integrated into a unified database, which is then used to propagate them across the entire reservoir. In the reference interval, the proposed method exhibited an uncertainty range of 0.7–4.4% for porosity, compared to 3.1–10.4% for the four traditional approaches considered, indicating a considerable reduction in uncertainty. This approach preserves the geologically consistent porosity distribution along wellbores while ensuring sufficient granularity for the variation process. The proposed methodology was successfully applied to a real-world field case where interpretation was complicated by data scarcity and a complex geological setting. The developed approach yields a geologically sound three-dimensional model, providing a robust basis for informed decisions in hydrocarbon asset development and management, and may reduce manual effort while constraining automated history matching within data-driven porosity bounds.