Michele Arena

Silicification is a major diagenetic process in carbonate reservoirs, profoundly influencing mineralogy, pore-system architecture, fluid-flow pathways, and reservoir quality. Although chert is widespread in the Aptian Barra Velha Formation of the Santos Basin pre-salt province, its origin and relationship with fluid circulation, karstification, and reservoir evolution remain poorly constrained. This study investigates cherts within a 145.65 m cored interval distributed across five cores using an integrated multiscale workflow combining core description, petrography, UV epifluorescence, SEM-EDS, portable and micro-XRF, computed tomography (CT), micro-CT, fluid inclusion petrography, and coupled silicon and sulfur stable isotope analyses. Four chert geometries were identified: tabular, nodular, domal, and irregular. Tabular and nodular cherts preserve primary carbonate fabrics and record replacement-dominated silicification followed by progressive recrystallization from opal and chalcedony to microcrystalline and macrocrystalline quartz. Domal cherts occur exclusively within microbialite-bearing facies and preserve filamentous and reticulated fabrics within microcrystalline silica, suggesting localized early silica precipitation associated with microbial substrates. In contrast, irregular cherts are concentrated within fractured, brecciated, and karstified intervals and are characterized by diffuse silicification, macrocrystalline quartz cementation, sulfate mineralization, and strong structural control. Five karstified intervals, representing approximately 37% of the studied succession, contain the most interconnected pore systems identified by CT and micro-CT, with porosity increasing from approximately 3.1% in continuous cherts to 26.7% in brecciated intervals. These zones coincide with irregular cherts, sulfate-rich mineral assemblages, and fault-controlled permeability pathways, locally producing bulk-density values up to 3.37 g/cm³. Silicon isotope compositions range from δ³⁰Si = −0.116‰ to +1.476‰, with tabular cherts consistently exhibiting the heaviest values (up to +1.476‰) and irregular cherts the lightest (down to −0.116‰), whereas sulfur isotope compositions (δ³⁴S = 8.0–33.4‰) reveal isotopically heterogeneous sulfate-bearing fluids. Associated mineral assemblages—including barite, celestine, fluorite, titanite, goyazite, dawsonite, and pyrite—record the circulation of chemically evolved mixed fluids through structurally enhanced pathways. The integrated dataset supports a multistage fluid-flow model involving early silica precipitation, replacement silicification, karstification, structurally focused fluid migration, sulfate mineralization, and late hydrocarbon charging. Chert formation is interpreted as the product of a long-lived diagenetic system in which fluid mixing, structural permeability, progressive silica recrystallization, and fault-controlled fluid flow collectively governed silicification and the development of reservoir heterogeneity in the Barra Velha Formation.

Silicification is a diagenetic process that affects the porosity and permeability of carbonate rocks, such as those in the Itapema and Barra Velha formations, Pre-salt of Santos Basin. This contribution integrates macroscopic and microscopic descriptions of silicified intervals, identifying controls and mechanisms of dissolution, replacement, and precipitation, relating them within the diagenetic history of the formations. The pre-compaction eodiagenetic silicification comprises tabular and nodular geometries, reflecting fine-grained laminated and coarser-grained massive lithologies. The post-compaction mesodiagenetic silicification comprises irregular silica disseminated throughout carbonates. Silica bodies consist of replacive (cryptocrystalline silica, mosaic microcrystalline quartz and spherulitic chalcedony) and pore-filling textures (chalcedony and macrocrystalline quartz). The highly selective nature of the replacement suggests that carbonate dissolution initiates silica precipitation as a fluid-mediated process that transfers crystallographic information. The successive zoning with decreasing crystallinity in filling/cement textures indicates a decrease in the fluid supersaturation as the precipitation advances. These features indicate that textures and silicification mechanisms reflect precipitation physicochemical conditions. The first silica event follows the dolomitization stage that replaces primary clays and carbonates, while the second event occurs after mechanical compaction and is later followed by replacive/pore-filling barite textures. Late macrocrystalline quartz veins host primary hydrocarbon fluid inclusions, linking the second silicification event with oil migration.

Considered important analogues of pre-salt reservoirs in hydrocarbon-producing basins throughout eastern Brazilian shore, the coquinas of the Morro do Chaves Formation (Barremian-Aptian of the Sergipe-Alagoas Basin) were studied in this work from a petrophysical standpoint using advanced characterization methods for cores recovered from four wells (2-SMC-01-AL, 2-SMC-02-AL, 2-SMC-03-AL e 2-SMC-01-AL) drilled in the Atoll Quarry, which is located in city of São Miguel dos Campos, Alagoas State, Brazil. To understand the relationships between lithological and diagenetic heterogeneities that control the distribution of permeability-porosity properties of this rock type, the cores underwent Dual-Energy X-ray imaging, which determines bulk density and photoelectric factor data to define seven rock types and generate effective porosity curves for the studied wells. Permeabilities in wells were modeled based on the Flow Zone Index (FZI) concept, and each reservoir interval based on the Global Hydraulic Elements (GHEs) approach. Results were calibrated with laboratory measurements in plugs, and porosity was classified by petrographic analysis by Digital Image Analysis (DIA). All wells were correlated and four internal flooding surfaces identified, with a regressive surface (RS) dividing the Morro do Chaves Formation into an upper section with better reservoir characteristics and a lower one with worse ones.

Lithofacies classification is crucial for the exploration and development of oil and gas fields, as it provides valuable information for petrophysical and seismic analysis, among others. In the case of carbonate lithofacies, however, the heterogeneity of the rocks at different scales makes their evaluation and characterization a challenge, and manual interpretation by experts is often time-consuming and subject to biases. In Brazil, the discovery of the Pre-salt interval, with its large hydrocarbon fields, has generated high interest in the origin and classification of complex carbonate rocks in the area. Recently, machine learning methods have emerged as a promising solution for automating facies classification process, and XGBoost algorithm has been demonstrated to be effective in lithofacies classification problems. In this study, XGBoost algorithm is applied to the Barra Velha Formation in the Tupi Field, Santos Basin, Brazil, to perform carbonate lithofacies prediction in a systematic and efficient manner. A two-stage methodology was developed for the classification of the Barra Velha Formation lithofacies. In the first stage, well logs were selected, quality controlled, and preprocessed to form the dataset. In the second stage, XGBoost algorithm was used to train and validate the model, and prediction and evaluation were performed. Features were ranked using ReliefF and InfoGain, and six models were tested using evaluation metrics such as accuracy, recall, F1-score, and root mean squared error (RMSE). Six different models were tested and evaluated, with model 6 showing great levels accuracy (88.57%) and recall (88.57%), as well as F1-score (88.48%) and the lowest RMSE (0.86). The results demonstrate the effectiveness of the XGBoost algorithm in the classification of the Barra Velha Formation lithofacies, and the importance of feature selection and ranking in the improvement of the models' performance.