Produção Científica
Artigo em Revista
A new look at the causes of âpolarizationâ horns in electromagnetic well logging Socalled polarization horns appear in electromagnetic well logs when the sondes move through the interface between formations of different conductivities. Since the early 1990s, peaks in the logs have been attributed to the influence of surface charges due to the presence of electric field components perpendicular to the interfaces. A new analysis finds that surface charges should not be pinpointed as the immediate cause for the appearance of interface peaks in electromagnetic logs from any tool operating in any frequency range. This is accomplished by calculating the derivatives of the magnetic field components with respect to the sondeâs vertical position as it crosses the interface between two homogeneous isotropic media. The mathematical expressions reveal the components with smooth transitions and the ones with discontinuities in their rates of change across the interface. The analysis is applied to the four components needed to simulate the responses of coaxial and coplanar coil configurations in dipping logs. The results show that only the horizontal field from a horizontal dipole source suffers an influence from the current density field perpendicular to the interface between the two media, which gives rise to surface charges, but even for this component, the nonsmooth transitions are not associated with the perpendicular current. An anisotropic example gives further support to the conclusion that the polarization horns are associated with the discontinuous current density field parallel to the interface rather than with the continuous current across the interface. 

Artigo em Revista
GEMM3D: An Edge Finite Element program for 3D modeling of electromagnetic fields and sensitivities for geophysical applications This paper presents the 3D modeling program GEMM3D (Geophysical ElectroMagnetic Modeling 3D), a Fortran program for modeling frequency domain electromagnetic responses of geophysical sources in threedimensional models. The program is an implementation of the Edge Finite Element method, in an electric field formulation, to simulate any electromagnetic source of interest, at all frequencies that are used in the geophysical methods in the quasistatic regime. Its current version includes routines for plane waves and simulation of magnetotelluric data; electrical and magnetic dipoles; circular loops. Other sources and configurations may easily be included with the addition of routines for simulating primary (layered earth) fields. The program includes routines for calculating sensitivities using the Adjoint State method, which makes efficient use of direct solvers for factoring once and solving for multiple systems with the same coefficient matrix. The paper details the methods, the implementation, and the usage of the program. Validation tests are presented for different sources. The program is freely available for use in research and teaching. 

Artigo em Revista
Analysis of EshelbyChengâs model in anisotropic porous cracked medium: An ultrasonic physical modeling approach Many effective medium theories are designed to describe the macroscopic properties of a medium (the rock, or reservoir in this case) in terms of the properties of its constituents (the background matrix of the rock and the inclusions, for our scenario). A very well known effective medium theory is the EshelbyCheng model, which was studied by us in previous work, being tested for the case where the background medium was weaklyanisotropic and porous. The analysis was done testing elastic velocities and Thomsen parameters  as a function of crack density for fixed values of aspect ratio  predicted by the model with data acquired from synthetic rock samples. In this work, we aim to complete the analysis of the EshelbyCheng model capabilities when applied to rocks with porous and vertical transversely isotropic (VTI) backgrounds, testing the model for the elastic velocities as functions of aspect ratio  for fixed values of crack density  against experimental data. The data used to test the model were obtained from 17 synthetic rock samples, one uncracked and 16 cracked, the latter divided into four groups of four samples each, each group with cracks having the same aspect ratio, but with the samples having different crack densities. In these samples, ultrasonic pulse transmission measurements were performed to obtain the experimental velocities used to test the model. As was not possible to acquire data for velocity as a function of aspect ratio for fixed values of crack density, we performed interpolations of the experimental data to estimate these velocities. EshelbyCheng model effective velocities and Thomsen parameters were calculated using three formulations proposed for the crack porosity: one proposed by Thomsen, the second one proposed in our previous work (which depends only on the crack density) and the third one proposed in this work (which depends on the crack porosity and the aspect ratio, just like Thomsenâs proposal). The comparisons between elastic velocities and Thomsen parameters  as function of crack aspect ratio, for fixed values of crack density  predicted by the model and estimated from the data via interpolation showed that the third formulation produced better fittings (lower rootmeansquare errors) between model and experimental data for all ranges of aspect ratio and crack density. 

Artigo em Revista
On the seismic wavelet estimative and reflectivity recovering based on linear inversion: Welltoseismic tie on a real data set from Viking Graben, North Sea Tying seismic data to well data is critical in reservoir characterization. In general, the main factors controlling a successful seismic well tie are an accurate timedepth relationship and a coherent wavelet estimate. Wavelet estimation methods are divided into two major groups: statistical and deterministic. Deterministic methods are based on using the seismic trace and the well data to estimate the wavelet. Statistical methods use only the seismic trace and generally require assumptions about the waveletâs phase or a random process reflectivity series. We have compared the estimation of the wavelet for seismic well tie purposes through leastsquares minimization and zeroorder quadratic regularization with the results obtained from homomorphic deconvolution. Both methods make no assumption regarding the waveletâs phase or the reflectivity. The bestestimated wavelet is used as the input to sparsespike deconvolution to recover the reflectivity near the well location. The results show that the wavelets estimated from both deconvolutions are similar, which builds our confidence in their accuracy. The reflectivity of the seismic section is recovered according to known stratigraphic markers (from gammaray logs) resent in the real data set from the Viking Graben field, Norway. 

Artigo em Revista
Interval mineral and fluid densities estimation from welllogs: Application to the Norne Field dataset Formation evaluation techniques are the key to understand subsurface rocks properties from welllogs, especially those drilled in hydrocarbon exploration wells. Knowledge of the parameters related to different types of rocks is traditionally used in forward determination of lithology, porosity and water saturation, which can be refined by calibrating the input models. In this work, we perform welllog interval linear inversion with respect to formation density to investigate mineral and fluid properties in a real dataset. The method is based on an overdetermined problem, which supposes a homogeneous distribution of petrophysical parameters through stratigraphic layers and is applied in conventional reservoir rocks from the Norne Field (offshore Norway). Bulk density, gammaray and neutron porosity logs are employed in a workflow that relies on layerbylayer leastsquares regressions to estimate matrix, shale and fluid apparent densities. In this process, shale volume and the total density porosity are calculated depth by depth from empirical equations feed by the input logs. Therewith, the introduced inversion scheme stands as an alternative approach for well data interpretation that focuses on computing the densities of the rock constituents instead of fixing these parameters to invert fractional volumes. Furthermore, the application in two wellbores resulted in geological consistent individual densities in most intervals, except for a gasbearing zone observed in one of the boreholes, where porosity uncertainty caused anomalous variationin grain and fluid densities. 

Artigo em Revista
Seismic Stratigraphy of Camamu Basin, Northeastern Brazil âThe Camamu Basin is located at the northeastern Brazilian coastline and has significant hydrocarbon potential in both shallow and deep water settings. However, despite an already operating productive gas field, the basin is not well known. Herein, a regional stratigraphic interpretation of the rift, the transitional and the drift megasequences is reported, based on a data set of 152 poststacked 2D seismic lines, 1 3D seismic cube and 34 wells with eletrologs. The study revealed that the rift megasequence is much more complex than previously thought. In the southern region of the Camamu Basin, five rift sequences were mapped, while in the northern region, in the area of the regional tectonic lineament known as the Salvador Transcurrent Zone (STZ), seven rift sequences were recognized. This difference suggests a tectonic control during the rifting process, because the northern region is intensely affected by shear stress induced by transcurrent tectonics of the STZ during the crustal breakup. For the postrift or ââtransitionalââ phase associated with thermal subsidence installed after the rift, as well as for the drift succession, the tectonic control exerted by Salvadorâs Transcurrent Zone is not detectable and the tectonosedimentary evolution of the basin follows the general pattern of the Brazilian marginal basins. 
Artigo em Revista
NonHyperbolic Velocity Analysis of Seismic Data from Jequitinhonha Basin, Northeastern Brazil Normal moveout (NMO) velocity is used in seismic data processing to correct the data from the moveout effect. This velocity depends on the medium above the reflector and it is estimated from the adjustment of a hyperbolic function that approximates the reflection time. This approximation is reasonable for media formed by isotropic layers. For deeper exploration targets, which effectively behave as anisotropic media,the NMO velocity estimate from the hyperbolic approximation becomes imprecise. One possibility is the use of nonhyperbolic approximations for the reflection time and deeming the medium to be anisotropic. However, these approximations make the NMO velocity estimation a more complex problem, since the anisotropic parameters are unknown. In this study the NMO velocities for a vertical transverse isotropy medium are estimated using two nonhyperbolic reflection time approaches. For comparing the two methodologies that estimate NMO velocity, a 2D dataset from Jequitinhonha Basin is used and it presents anisotropic behavior. The results show that this approach produces more consistent results than the conventional approach, which ignores the anisotropy of the medium. 

Artigo em Revista
Inversion of satellite gravimetric data from RecĂ´ncavoTucanoJatobĂĄ Basin System Density differences among subsurface rocks cause variations in the gravitational field of Earth, which is known as gravity anomaly. Interpretation of these gravity anomalies allows assessment of the probable depth and shape of the causative body. For several decades, gravity data were acquired on the surface, but after the scientific and technological advances of the last decades, geopotential models were developed, including gravitational observations on a global scale through space satellite missions. This paper investigated the Moho structure in the region of RecĂ´ncavoTucanoJatobĂĄ riftbasin system based on the information of the terrestrial gravity field from the EIGEN6C4 geopotential model. The frequency domain inversion technique was applied, which is known as the ParkerOldenburg iterative method. Bouguer anomaly data were used in the inversion procedure to determine the thickness and geometry of the crust in the region. Data inversion considered a twolayer model with constant density contrast, in which the entire signal was related to Moho topography. In addition, data inversion was carried out to determine the basement depths. The program proved to be efficient and able to manage large data sets. The results, both of the crust thickness and the sedimentary package, validated the geodynamic evolution understanding of the basin system. 

Artigo em Revista
EVOLUTIONARY PROBLEMS OF NONLINEAR MAGNETOELASTICITY We consider mixed problems for nonlinear equations of magnetoelasticity. Our main result in the threedimensional case is the proof of an existence and uniqueness theorem; uniqueness is established under some extra restrictions on the smoothness of solutions. We also manage to prove the existence and uniqueness of a weak solution to the problem in the twodimensional case; uniqueness is established without any additional a priori assumptions on the smoothness of solutions. 

Artigo em Revista
Petrofacies classification using machine learning algorithms Carbonate reservoirs represent a large portion of the worldâs oil and gas reserves, exhibiting specific characteristics that pose complex challenges to the reservoirsâ characterization, production, and management. Therefore, the evaluation of the relationships between the key parameters, such as porosity,permeability, water saturation, and pore size distribution, is a complex task considering only welllog data, due to the geologic heterogeneity. Hence, the petrophysical parameters are the key to assess the original composition and postsedimentological aspects of the carbonate reservoirs. The concept of reservoir petrofacies was proposed as a tool for the characterization and prediction of the reservoir quality as it combines primary textural analysis with laboratory measurements of porosity, permeability, capillary pressure, photomicrograph descriptions, and other techniques, which contributes to understanding the postdiagenetic events. We have adopted a workflow to petrofacies classification of a carbonate reservoir from the Campos Basin in southeastern Brazil, using the following machine learning methods: decision tree, random forest, gradient boosting, Knearest neighbors, and naĂŻve Bayes. The data set comprised1477 wireline data from two wells (A3 and A10) that had petrofacies classes already assigned based on core descriptions. It was divided into two subsets, one for training and one for testing the capability of the trained models to assign petrofacies. The supervisedlearning models have used labeled training data to learn the relationships between the input measurements and the petrofacies to be assigned. Additionally, we have developed a comparison of the modelsâ performance using the testing set according to accuracy, precision, recall, and F1score evaluation metrics. Our approach has proved to be a valuable ally in petrofacies classification, especially for analyzing a welllogging database with no prior petrophysical information. 

<< < 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 > >>