Predictive Characterization of Fracture and Absorption Tests in Halite Formations: an integrated approach of numerical modeling and field data

Resumo

This study introduces a predictive model based on field data from oil wells in the pre-salt region, focusing on the analysis of fractures in saline rocks and the simulation of leak-off tests through integration with the finite element method. The analysis utilized a variety of techniques, including the construction of a correlation matrix, statistical testing, linear adjustment methods, and numerical modeling. Utilizing specific field data from halite formations, the study identified overburden stress and sediment depth as the most critical variables for minimum stress. Three linear adjustment methodsleast squares, genetic algorithms, and Bayesian methodswere employed. All methods demonstrated a strong correlation among the variables, with Bayesian methods exhibiting the lowest percentage error relative to experimental data. Residual analysis showed that the regression models provided accurate predictions, with most points closely matching the actual values. Numerical modeling further allowed for the assessment of rock behavior under different fluid conditions, indicating that the compositional fluid provides a more realistic simulation and adheres more closely to theoretical expectations and field data. The results affirm the efficacy of the proposed models, particularly highlighting the use of Bayesian methods and mathematical modeling integrated with the finite element method, in predicting the dynamics of fractures and fluid absorption in saline formations. This model significantly enhances the characterization of fractures and absorption pressures in oil wells, improving its utility in predicting behaviors in formations associated with pre-salt fields.