Big Data-Driven Evaluation of Acid Fracturing Effectiveness in Deep Carbonate Reservoirs

The carbonate reservoirs in Penglai gas area have characteristics such as deep burial depth (6000-8000m), poor physical properties (porosity of 2%~6%), and strong heterogeneity. Early-stage acid fracturing results are highly disparate, and the key factors dictating treatment success remain poorly understood. This study develops a workflow to evaluate deep carbonate rock after acid fracturing by analyzing big data. Following data curation of geological, construction, and production metrics from pre-fracturing wells, feature importance was ranked using Random Forest with the SHAP (Shapley Additive Explanations) model. These rankings were amalgamated via a weighted average to produce a consensus parameter hierarchy, guiding neural network identification of key drivers. A subsequent hybrid AHP-EWM framework was developed for multivariate quantitative evaluation, elucidating the principal factors governing acid fracturing effectiveness and establishing a novel assessment protocol for Penglai gas area carbonate wells. Results demonstrate that treatment section length, acid intensity, acid volume, total hydrocarbon content, pressure drop frequency, significantly impact fracturing effectiveness. The constructed quantitative acid fracturing effectiveness index achieved over 90% linear correlation with actual well test production, verifying model validity. The acid fracturing effectiveness index ranges from 0.22 to 0.99, the geological index spans 0.15 to 0.55, and the engineering index varies between 0.18 and 0.51. Rational acid fracturing optimization and high-quality field implementation constitute dual critical factors for ensuring optimal stimulation performance in the Penglai Gas Zone. This study develops a novel, multifaceted evaluation method for assessing acid fracturing outcomes in deep carbonate reservoirs. This approach lays the theoretical groundwork for designing highly efficient deep acid fracturing treatments.