Diagnostic analysis of acid fracturing stimulation effectiveness in fractured-vuggy carbonate reservoirs combining mechanism and data

Introduction Evaluating acid-fracturing effectiveness in fractured-vuggy carbonate reservoirs requires an understanding of the pressure responses associated with fracture-vug communication. This study integrates mechanistic modeling and data-driven analysis to investigate these responses and estimate vug characteristics. Methods A thermo-hydro-mechanical (THM) fracture-propagation model is developed using the extended finite element method (XFEM). The treatment fluid is represented through its hydraulic and thermal properties; acid-species transport and mineral-reaction kinetics are not explicitly solved. Numerical simulations examine the effects of vug size, wellbore-vug distance, injection rate, fluid viscosity, filling characteristics, and horizontal stress difference on the pressure response. Diagnostic charts are constructed using pressure-drop magnitude, pressure-drop rate, and pressure-recovery rate. A support vector regression (SVR) model is developed to estimate vug radius and filling coefficient, and the diagnostic workflow is evaluated through field net pressure matching. Results Vug size and filling characteristics are the primary factors controlling the communication-induced pressure drop. Larger vugs and lower filling coefficients produce more pronounced pressure drops. Injection rate and fluid viscosity modify the magnitude and rate of the pressure response, whereas wellbore-vug distance primarily affects its timing. Within the simulated conditions, horizontal stress difference has a comparatively limited influence. Field comparisons show agreement between the predicted and interpreted vug characteristics, supporting the applicability of the proposed workflow. Discussion Integrating THM-XFEM simulations, pressure-response charts, and SVR provides a quantitative approach to interpreting fracture-vug communication during acid fracturing. The workflow has potential to support diagnosis during treatment and inform operational decisions. Its applicability is bounded by the modeled conditions and the absence of explicitly resolved acid transport and mineral reactions.