Tag: carbonate rock

Estimating Rock Typing in Uncored Wells Using Machine Learning Techniques for Brazilian Pre-Salt Carbonate Reservoir

Publicado em por .
Accurate rock typing in uncored wells is essential for enhancing reservoir models, particularly in complex geological formations like the Brazilian pre-salt carbonate reservoirs. This study explores the application of machine learning (ML) techniques to estimate rock types in uncored wells. The research leveraged core data from 11 cored wells to calculate the Rock Quality Index (RQI) and Flow Zone Indicator (FZI), identifying 14 distinct rock types through the Discrete Rock Typing (DRT) method, along with well log data such as gamma ray, density, neutron, permeability, and porosity. Various machine learning models, including Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Decision Tree, Gradient Boosting, Naive Bayes, XGBoost, and Random Forest were tested, where XGBoost achieved the highest accuracy of 73.3%. Applying XGBoost to all wells resulted in accuracy ranging from 0.6 to 0.91 and the model was subsequently used to estimate rock types in over 20 uncored wells to generate reservoir simulation models. This study highlights the efficacy of machine learning (particularly XGBoost) in addressing reservoir complexities and offering significant improvements in the understanding and development of carbonate reservoirs.

Integrated Multi-Scale Pore Characterization of Carbonate Rocks in the Barra Velha Formation, Santos Basin, Brazil

Publicado em por .
Carbonate rocks feature heterogeneous porous systems that span multiple scales, from pore level to the reservoir scale. The complexity and diversity of carbonate reservoirs demand a consistent approach to their characterization. The efficient integration of multiscale imaging data and petrophysical data is increasingly important to address the challenges associated with these complex carbonate reservoirs. A crucial step in overcoming these scale gaps in reservoir modeling and simulation involves enhancing the characterization of reservoir flow units and their associations with geological and petrophysical heterogeneities at varying scales. In this study, we focus on the classification of pore types using digital rock analysis and petrophysical evaluation of pre-salt lacustrine carbonates from the Barra Velha Formation (BVF) in the Santos Basin using computerized tomography (CT), core samples description, and petrography. Eight types of pores were identified at the core scale: interparticle, stratiform-vuggy, growth framework, vuggy, vuggy-fracture, fracture, interclast, and intraclast. The distribution and characteristics of these pore types were analyzed at different scales, including thin-sections and micro-CT, and nuclear magnetic resonance (NMR), which highlights the diversity in the porous system and the impact of different pore types on porosity and permeability. NMR analyses illustrated the pore size heterogeneity to provide distinction between tight and porous samples. Hydraulic rock units (HRUs) were defined based on flow zone indicator (FZI) using the probability plot approach. Seven HRUs were defined: HRU1 and HRU2 represent samples with the highest FZI and rock quality index (RQI) values, whereas HRU3 and HRU4 denote intermediate values. HRU5, HRU6, and HRU7 represent units with the lowest values. HRU1 and HRU2 were predominantly associated with vuggy, growth framework, and interparticle porosities, which are often enhanced by dissolution processes. Conversely, HRUs with reduced reservoir qualities (5, 6, and 7), characterized by the lowest permeability values, are more prevalent in intervals with higher silicification and silica and dolomite cementation, presenting a variety of pore types at a macroscale. The integration of multiscale imaging techniques and petrophysical data underscores the complexity of pore systems, providing crucial insights into their reservoir characteristics.

Numerical Study on the Impact of Advanced Phenomena in a Fractured Carbonate Reservoir Subjected to WAG-CO2 Injection

Publicado em por .

Advanced phenomena related to water-alternating-gas (WAG) injection are usually neglected in numerical simulations. This work evaluates the impact of different physical phenomena on field indicators, considering a typical pre-salt carbonate reservoir (UNISIM-II-D-CO, a dual-por dual-perm compositional case) subjected to WAG-CO2 injection. Additionally, the computational cost incurred by each of these phenomena is evaluated, since it represents a great challenge in optimization and probabilistic studies. The following phenomena are evaluated considering a nominal base case: (i) matrix-fracture transfer calculation, (ii) relative permeability hysteresis, (iii) CO2 and CH4 solubilities in aqueous phase, (iv) diffusion, (v) numerical dispersion control models, and (vi) velocity-dependent dispersion. CO2 and CH4 solubilities in the aqueous phase, as well as molecular diffusion, did not have a significant impact on field indicators, but they increased simulation runtime more than two times. Matrix-fracture transfer modeling was the most impactful factor, followed by hysteresis and velocity-dependent dispersion. Therefore, the impact of these phenomena was also investigated in a probabilistic approach, considering an ensemble of 197 geostatistical scenarios under uncertainty. Risk curves revealed that the advanced matrix-fracture transfer models improve sweep efficiency. This effect is mainly due to gravity force which acts as a driving mechanism for the oil moving from the matrix to fractures. The capillary effect, in turn, was small compared to gravity. The impact of dispersion and hysteresis on risk curves were smaller than the effect of matrix-fracture transfer modelling. However, these phenomena are particularly interesting in UNISIM-II-D-CO due to the presence of Super-K facies. Hysteresis, when applied to low and high permeability layers, reduced gas mobility and, consequently, the gas produced, contributing to the NPV for most models under uncertainty. On the other hand, the velocity-dependent dispersion mainly affected fluid flows in the regions adjacent to Super-K layers, promoting better oil recovery. The inclusion of advanced phenomena related to WAG-CO2 injection can hold importance when modeling fractured carbonate fields, like those found in the Pre-Salt in Brazil. Nevertheless, computational costs might make their inclusion impractical in full-field simulation models employed for optimization and probabilistic studies. In such cases, it is recommended to assess low-fidelity models or alternatives to accelerate simulations, focusing mainly on the most impactful phenomena related to WAG-CO2 injection.