Modeling and inversion of petrophysical properties using Monte Carlo method in coquinas of the Morro do Chaves Formation (Sergipe-Alagoas Basin)

Resumo

The discovery of large accumulations of hydrocarbons in the reservoir coquinas of the Pre-Salt section of the Santos Basin has created a demand for research focused on characterizing the petrophysical properties of these rocks. However, obtaining rock samples in the Pre-Salt involves considerable complexity and costly logistics. The coquinas of the Morro do Chaves Formation in the Sergipe-Alagoas Basin (SE-AL) are considered a potential analog to the reservoir coquinas of the Pre-Salt. Therefore, they are used to understand the depositional, diagenetic, and facies processes that affect the characteristics of these rocks. The data used in this research were geophysical well logs: sonic transit time (BHC) and compressional velocity (Vp), as well as laboratory rock data such as grain density (GrainDen), porosity (PhiLab) and permeability (PermLab). These data belong to LAGESED-UFRJ and come from well 2-SMC-AL drilled in the Atol Quarry, located in the city of São Miguel dos Campos (AL), SE-AL Basin. The lack of continuous well logs prevent the application of a flowchart to characterize carbonate rocks along the well. Therefore, modeling and inversion methods were employed using the total density of the samples to estimate density and porosity logs at depth. The main objective of this study was to estimate the coefficients - a, b, c - of Gardner's (1974) polynomial equation, which would result in an adequate fit between the density data obtained in the laboratory and the observed density data, achieved by employing the Markov chains Monte Carlo method (MCMC). This statistical technique is used to simulate complex systems and make numerical estimates in order to determine the probable distribution of the coefficients and the predicted densities with the associated uncertainties. After obtaining the coefficients, it was possible to estimate a density log (MonteCarlo) over the interval of the analyzed well, and subsequently, this log was used to estimate the porosity log (PhiDENmc). Additionally, the Wyllie et al. equation (1956) was used as an alternative method for estimating the porosity log (PhiWyllie) from the sonic log. Finally, these porosity logs estimated (PhiDENmc and PhiWyllie) were correlated with PhiLab. It was possible to observe that PhiDENmc showed the best fit with the porosity data measured in the laboratory (PhiLab).