Enhancing Operational Efficiency in Hydraulic Fracturing: A Predictive Modeling Approach for Fracturing Design in the Sirikit Oil Field

Abstract This paper presents a predictive modeling approach aimed at enhancing operational efficiency in hydraulic fracturing design, specifically tailored for tight sand reservoirs in the Sirikit Oil Field, Thailand. The methodology integrates deterministic data from pre-fracture diagnostic tests, including Mini-Falloff Tests (MFTs) and Step-Rate Tests (SRTs), to develop empirical correlations between fluid efficiency and optimal pad percentage. A comprehensive dataset encompassing over 200 fracturing stages across 100 wells was analyzed, revealing a strong linear relationship between fluid efficiency from SRTs and pad volume requirements. To improve predictive accuracy and applicability, the dataset was stratified into geological compartments based on common depositional and structural characteristics. This compartmentalized modeling enabled the use of offset well data for design purposes, allowing the omission of calibration injection steps in appropriate scenarios. Field validation in Well-A demonstrated the model's accuracy, with only a 2% deviation between predicted and actual pad percentages. Following this, the model was deployed across 15 wells and 26 stages in the N Area during 2023–2024, resulting in zero screen-out events, an average of 5–10 hours saved per stage, and a 7.6% reduction in fracturing costs. The results confirm that combining empirical modeling with geological segmentation can significantly improve treatment design precision, reduce operational risk, and enhance efficiency. This approach establishes a foundation for future advancements, including integration with multivariate regression, theoretical models, and machine learning techniques to support broader field-wide and cross-basin applications.