A Study on Leakage Pressure in Deep Marine Carbonate Formations of the Sichuan Basin

Deep marine carbonate rocks in the Sichuan Basin have been subjected to the combined effects of multiple periods of tectonic activity and paleokarst processes, resulting in highly developed fracture-cavity systems and extreme heterogeneity, which frequently lead to severe wellbore leaks during drilling. At the same time, multiple pressure systems coexist within the formation, and the phenomenon of simultaneous leakage and flow is prominent. Traditional methods for predicting leakage pressure are ill-suited to these complex reservoir conditions, severely limiting the safe and efficient drilling of deep oil and gas reservoirs. To address these engineering challenges, this paper integrates multi-source data from logging, rock mechanics, and field drilling operations. Using techniques such as simulation modeling and imaging logging interpretation, wellbore leakage is classified into pressure-differential-driven, fracture-propagation-driven, and fracturing-driven types. Corresponding leakage kinetics models are constructed for each type, resulting in a method for predicting leakage pressure and assessing risks in complex carbonate formations. The study indicates that the degree and orientation of fracture development, as well as the magnitude of formation pore pressure and in-situ stress, are the key controlling factors determining the critical leakage pressure; in sections with well-developed natural fractures, the formation’s pressure-bearing capacity decreases significantly, and variations in in-situ stress along with localized low-pressure anomalies further narrow the safe drilling fluid density window. Within the scope of the model assumptions and existing calibration data, this method can be used to analyze the control conditions for different types of fluid loss and to provide a reference for developing fluid loss prevention and control strategies for deep marine carbonate formations, as well as for evaluating drilling fluid density windows.