Produção Científica

**Apresentação**

Deconvolution and correlation-based interferometric redatuming by wavefield inversionSeismic interferometry is a method to retrieve Green's functions for sources (or receivers) where there are only receivers (or sources, respectively). This can be done by correlation-or deconvolution-based methods. In this work we present a new approach to reposition the seismic array from the earth's surface to an arbitrary datum at depth using the one-way reciprocity theorems of convolution and correlation type. The redatuming process is done in three steps: (a) retrieving the downward Green's function for sources at the earth's surface and receivers at the datum, (b) retrieving the corresponding upward Green's function, and (c) retrieving the reflected upward wavefield for sources and receivers at the datum. Input for steps (a) and (b) are the surface data and wavefields simulated in a velocity model of the datum overburden. Step (c) uses the responses of steps (a) and (b) as input data in the convolution-based interferometric equation. The method accounts for inhomogeneities in the overburden medium, thus reducing anticausal events and artefacts as compared to a purely correlation-based procedure. |

**Apresentação**

ComparaÃ§Ã£o entre o algoritmo ε de Wynn e o ∆2 de Aitken no cÃ¡lculo de componentes eletromagnÃ©ticas de fontes geofÃsicas dipolares.Este trabalho apresenta os procedimentos utilizados na determinaÃ§Ã£o de componentes de campo de algumas fontes geofÃsicas em modelos geoelÃ©tricos unidimensionais por meio de quadratura com extrapolaÃ§Ã£o (QWE - Quadrature With Extrapolation). Os resultados com QWE foram determinados a partir de duas tÃ©cnicas de aceleraÃ§Ã£o de convergÃªncia: o algoritmo \epsilon de Wynn e o processo \Delta^2 de Aitken. Com anÃ¡lises no emprego ao Dipolo ElÃ©trico Horizontal na direÃ§Ã£o x (DEHx), o Dipolo MagnÃ©tico Vertical (DMV) e o Dipolo MagnÃ©tico Horizontal na direÃ§Ã£o x (DMHx) quando possuem soluÃ§Ã£o analÃtica, demonstra-se que o algortimo Delta^2 de Aitken tambÃ©m pode ser utilizado na modelagem de dipolos, tendo erro relativo na mesma ordem do algoritmo \epsilon, mas em geral maior nÃºmero de avaliaÃ§Ãµes. |

**Apresentação**

Full-waveform inversion using an efficient preconditioning method for the gradient vectorFull-waveform inversion (FWI) is a efficient method and it has been used successfully to invert subsurface parameters. It consists basically on the minimization of the difference between the predicted and observed data. However, its application using finite-difference schemes is limited to low frequency content and the increase of the range of higher frequency will demand a high computational cost of the wavefield propagation procedure and also the whole inversion scheme. To overcome this problem, we apply the rapid expansionmethod (REM) for numerical wavefield extrapolation inside the FWI workflow thus increasing the frequency content of the inversion process. Besides that, an efficient preconditioningmethod using source-receiver illumination (PSRI) for computing the gradient vector in order to increase resolution of the models and better convergence of the objective function has also been proposed. Beyond that, we compared the performance of the preconditioning method with relation to conventional gradient method for two types of source signature. To test the efficiency of our proposed FWI approach, we apply it using a frequency multiscale scheme for a synthetic data set with a complex velocity model. The inversion results show satisfactory inverted velocity models which can be used to produce depth imaging of high quality. Thus we demonstrate the effectiveness and applicability of our FWI scheme using REM combined with a multiscale approach. |

**Apresentação**

Multichannel Predictive Deconvolution and Parabolic Radon Transform for Multiple Reflection's Filtering on Land Seismic Data From SolimÃµes BasinThe SolimÃ´es basin became the target of oil exploration campaigns after the discovery of the JuruÃ¡ field in the 1970s. Its geological evolution is marked by magmatic spills represented by the diabolic sills, which, in addition to preventing seismic energy, generate Multiple reflections, creating problems in processing. This work proposes the application of multichannel predictive deconvolution for the attenuation of multiple reflections after ground-roll filtration. The results showed that although the periodicity of these events was not totally perfect, the method was effective in terrestrial data, attenuating multiple reflections and preserving the primary reflections. |

**Apresentação**

Iterative SSA method to filtering of the land seismic dataNormally the singular value decomposition (SVD) filtering is applied in the time Ã— distance domain, exploring the spatial correlation between a set of seismic traces. In the present work we explore the Singular Spectrum Analysis (SSA) method in the time direction using a iterative algorithm. The SSA method extract the correlation between reflected events along the time variable and decompose seismic traces in the low (first eigenimage only) and high frequency parts (sum the rest of the eigenimages). We illustrate the effectiveness of this new approach to the prediction and subtraction of the ground roll and improve velocity analisys and stacking of the seismic data. |

**Apresentação**

Stabilized unidimensional deconvolution-based imaging conditions in Marchenko imagingMultiple reflections are typically not accounted for in seismic migration processes, which can lead to the emergence of artifacts. In Marchenko imaging, we retrieve the complete up- and downgoing wavefields in the subsurface to construct an image without such artifacts. The quality of this image depends on the type of imaging condition that is applied. In this work, we introduce an imaging condition that is based on stabilized unidimensional deconvolution (SUD). Two specific approaches are considered. In the first approach, we use the full up- and downgoing wavefields for deconvolution. Although this leads to balanced and relatively accurate amplitudes, the crosstalk is not completely removed. The second one is to incorporate the initial focusing function in the deconvolution process, in such a way that the retrieval of crosstalk is avoided. We compare images with the results of the classical cross-correlation imaging condition, which we apply to reverse-time migrated wavefields and to the up- and downgoing wavefields that are retrieved by the Marchenko method. |

**Apresentação**

Conventional seismic processing flow analysis: 2D line of the Rio Grande Chain region Southeastern BrazilSeismic reflection is one of the most important tools for the oil and gas industry, both in the exploratory and in the production phase. It is divided into three components: acquisition, processing and interpretation. This work is inserted in the context of seismic data processing. For this, the processing of a real 2-D line acquired in 1979, located in the region of the Rio Grande Chain, southeast of Brazil, was made using the software ProMAX/SeisSpace. In order to generate a representative image of the geological framework of the area, a conventional processing flow that includes two different procedures was applied: procedure 1 and procedure 2, which are respectively related to the application of post and pre-stack migrations. The steps of the processing flow were submitted to different parameterizations, and the inputs and outputs were compared aiming to evaluate how they work to improve the seismic data. Based on the processing flow used, the best result obtained for the final seismic section was the one corresponding to the Kirchhoff pre-stack time migration, due to its better seismic aspect. |

**Apresentação**

Estimation of P-wave dispersion and attenuation under pore-texture complexities in carbonatesBiot's theory of poroelasticity provides a theoretical basis for phenomena understanding of low-frequency seismic P-wave dispersion and attenuation due to the wave-induced fluid flow. An important mechanism to explain the P-wave attenuation in saturated porous media is the fluid flow induced by the wave on the mesoscopic scale, i.e., a scale larger than the pore size and smaller than the wavelength. Carbonates exhibit complex pore structure and texture that hinder the understanding about P-wave dispersion and attenuation at dry and fluid-saturated conditions. Therefore, this work treats the influence of petrophysical, depositional texture and geometrical parameters in P-wave dispersion and attenuation in clean limestones. Data set is composed by 6 samples of oolitic and cemented grainstones â€“ Albian age from Campos basin, Brazil, with textural characterization and laboratory measurements that allowed us applied Gassmannâ€™s and Geertsma and Smith (1961) approaches to estimate P-wave dispersion, Dvorkin and Mavko (2006) model to predict the P-wave attenuation based on characteristic frequency determination at mesoscopic scale by interlayer flow White et al. (1975) model. |

**Apresentação**

Mercury intrusion porosimetry and 2D DIA applied to the estimation of pore-structure parameters and permeability Reservoir characterization requires an understanding of rockâ€™s textural and mineralogical characteristics as well as geometric properties of the porous space. Pore geometry also control key petrophysical properties that determine reservoir quality of reservoir with respect to the productibility. In this study were quantified pore structure parameters and their influence on absolute permeability in a set of carbonate samples, using the integration of Mercury Intrusion Porosimetry (MIP) and 2D Digital Image Analysis (DIA) Methods. The man aim was to verify the interdependence between geometric properties and porosity to estimate the absolute permeability, using a multiple linear regression methodology (MLR). |

**Apresentação**

Effects of the dip angle in the induction anisotropy logsThis paper performs an analysis of the effects of the wellâ€™s deviated angle on the tensor triaxial induction tool signals within a thinly sand-shale laminated reservoirs and their equivalent intrinsic anisotropic models. The responses from coaxial and coplanar coil arrays in inclined wells are studied in detail, including the analysis of their apparent anisotropy logs, as well as their estimation of sand conductivity in the environments with a structural anisotropy.The dip angle effects are modeled in simple geometries as one-dimensional (1D) models, neglecting the presence of the borehole and the invasion zones, since they provide basic insight for understanding tool responses in more complex models. The results show a strong sensitivity of both the coaxial and coplanar signals to the deviated angle. It is verified that the anisotropy values are significantly reduced when the well is inclined as compared to what is found for the true vertical case, even for inclinations small enough for the wells to be classified as technically vertical (30 degrees or less). Therefore, the angle effects must be carefully considered, even for technically vertical wells. Otherwise, potential finely laminated reservoirs can be underestimated or even ignored. |

<< < 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 > >>