Produção Científica

**Apresentação**

A comparison of iterative methods performance to exact and pseudo adjoint operators in least-squares migrationLeast-squares migration is often used to attenuate migration artifacts that arise in conventional migration due to, for instance, data incompleteness and limited recording aperture. It uses iterative methods to obtain a model that best fits the data, and it requires a forward/adjoint operator pair to do so. These methods use the gradient of the cost function to estimate search directions and update the current model. The gradient is obtained by migrating the residuals between estimated and observed data at each iteration. Migration operators are regarded as adjoint of forward operators but only those which pass the dot-product test are exact adjoints. These operators estimate more accurate search directions, enhancing the convergence rates of iterative methods. We test the performance of adjoint and pseudo-adjoint operators in LSM based on three different iterative methods (steepest descent, conjugate gradients and limited-memory BFGS) in order to assess their sensitivity to the adjointness of the migration operator. We then compare the performance of each method. |

**Apresentação**

Seismic trace analysis using minimum phase and singular value decomposition methods. Application to ground-roll attenuationThe spiking deconvolution aim is to correct the effect of the wavelet in seismic trace by apply an inverse filter, assumed to be minimum delay, and to get an estimate of reflectivity. In order to compensate propagation and absorption effects one way is to use a time-variant deconvolution where a different inverse filter is computed and applied for each output sample position. We modify this procedure by estimating a minimum-delay wavelet for each time-sample position of the seismic trace. This gives a decomposition of the seismic trace as a sum of minimum-delay wavelets, each multiplied by a reflectivity coefficient. We done the SVD decomposition through moving windows in the matrix which contains the minimum phase wavelets in each column. The seismic trace can be represented as a sum of the eigenimages of the wavelet matrix multiplied by the reflective function. In this way we can obtain an estimate of the reflectivity function through the inversion of this system. This seismic signal decomposition method has a range of applications in the data processing and interpretation of the seismic signal. We use this method to ground-roll attenuation present in land seismic data. Improvements in data quality are evident in prestack data panels and velocity analysis. |

**Apresentação**

Weighted AB semblance using very fast simulated annealingWe present a modification applied to a work presentend by Ebrahimi et. al, 2017, that shows the weighted AB semblance method which deals with the AVO phenomenon in velocity spectra. This method is based on the use of two weighting functions applied to the AB semblance coherence measure. The first weighting function, uses the ratio between the first and the second singular value of the time window. The second weighting function is based on the position of the seismic wavelet in the time window. Each weighting function depends on two empirical coefficients, determined by the analysis of a matrix which contains the measure called ECM (Energy Concentration af Matrix). In this work, we made a modification in the calculus of these coefficients values. We use the method Very Fast Simulated Annealing to get the best coefficient values that generates a spectrum with the best resolution. We applied the method in both synthetic and real CMP gathers. Results show that the VFSA made easier to obtain the coefficient values. These values allow to generate velocity spectra with excelent resolution. |

**Apresentação**

Verificação experimental do modelo de efetivo de Hudson-Crampin para meios anisotrópicos fissurados cujo o background apresenta anisotropia VTIA modelagem física de meios fissurados/ fraturados utilizando experimentos laboratoriais em escala reduzida reduzida tem funcionado como uma excelente alternativa para a compreensão do efeito da anisotropia na caracterização do reservatório de hidrocarbonetos fissurado/fraturado. O objetivo principal deste trabalho foi verificar experimentalmente asprevisão do modelo de meio efetivo para meios fissurados Hudson e Crampim. Para isso, foram realizadas medições ultrassônicas em amostras anisotrópicas sintéticas com diferentes densidade de fissuras distribuídas em quatro razões de aspectos. Dezesseis amostras (quatro amostra para cada razão de aspecto) fissuradas e uma amostra de referência com ansiostropia VTI e sem fissuras. As razões de aspectos variam entre Isso resulta em amostras com valores de razão de aspecto variando de 0.08 a 0.52. Todas as amostras fissuradas foram simuladas por espaços vazios na forma de discos numa matriz poro elástica com baixa anisotropia VTI homogênea feita com areia e cimento. Em relação às previsões de velocidade realizadas pelo modelos teórico, Crampin apresenta um melhor ajuste para VP e VS se propagando perpendicular ao plano de fissuras para as condições secas e saturadas. Estas previsões se tornam mais evidentes no caso em que a densidade de fissuras e razões de aspectos são baixas. A partir dos valores de velocidade, nossa comparação também foi realizada em termos dos parâmetros ε, γ e δ. |

**Apresentação**

Estimativa automática de tempo de trânsiotode ondas compressionais: aplicação na determinação do parâmetro de ThomsenO trabalho tem como objetivo estimar o parâmetro elástico ε, que é conhecido como “anisotropia da onda P”. Para estimar este parâmetro, é utilizado um algoritmo que fornece o tempo de chegada da onda P a partir da teoria STA/LTA. As amostras utilizadas no trabalho foram feitas com uma mistura de cimento e areia, tem formato cilíndrico, com diâmetro de 38 milímetros alturas que variam entre 51 a 61 milímetros e inclusões cilíndricas, as amostras simulam um meio VTI de anisotropia fraca. |

**Apresentação**

Least Squares Imaging using l1 and Student residualsThis paper studies the Least Squares Migration (LSM) procedure as an optimization problem. Particularly, we study the behavior of this imaging procedure upon the use of robust loss functions. Our results demonstrate that LSM can work even under heavy noise if a suitable loss function is employed. Both the ℓ1 norm and the Student's t-norm showed themselves more stable with regard to outliers than the ℓ2 norm conventionally used in LSM. In the optimization step, we employ the hybrid deterministic-stochastic L-BFGS algorithm that exploits the structure of the objective function to reduce the computational burden of this imaging procedure. The quality of the achieved results show no deterioration over the conventional L-BFGS algorithm. |

**Apresentação**

AVO analysis using an alternative formulation for estimation of seismic properties of hydrocarbon fluids: A study on gas condensateThe equations proposed by Batzle and Wang (1992), referred as BW formulation, have been largely used by the Geoscience community since the early 90’s. These equations can be applied for the determination of fluid properties in a porous medium and the computation of the compressional and shear wave velocities within the fluid. This information is valuable for the application of Gasmmann fluid substitution and seismic reservoir characterization studies. Carvalho and Moraes (2018) compile an alternative formulation based on the equations proposed by Dranchunk and Abou-Kassem (1975) and Sutton and Hamman (2009) (referred as DASH formulation), which yields superior fit to experimental data of acoustic velocity in natural gas fluids collected from literature (Lemmon et al. 2018; Younglove et al. 1993; Plantier et al. 2005; Ahmadi et al. 2017). The present work expands this study, analyzing the differences in AVO analysis in a 2-layer model for fluid properties calculated with the use of BW and DASH formulations |

**Apresentação**

1D CMP inversion of MCSEM data to create a 3D geoelectrical modelThis paper describes the application of CMP 1D inversion of 3D MCSEM data to generate a 3D resistivity model. The method extends the CMP inversion that has been successfully applied to 2D MCSEM data. The results indicate that the method can be useful as a quick generator of approximate starting models for true 3D inversion. |

**Apresentação**

Pore topology and rock physical properties of microporosity in Lower Cretaceous carbonate sediments: insights from X-ray CT imaging Micropores (i.e., pores <30 µm diameter) are volumetrically the most important pore type in many Cretaceous sediments, comprising >50%, and often >80%, of the total pore volume. There is considerable uncertainty as to the capacity of microporosity to store and flow hydrocarbon, and little data to assess the importance of pore shape, size and connectivity on recovery efficiency. This partly reflects the limitations of technology to image and characterise micropore topology. This study focuses upon a large, multiscale dataset of X-ray CT images of 23 samples from Brazil, the Arabian Plate (both Lower Cretaceous) and the North Sea (Upper Cretaceous). The samples include skeletal peloidal pack–grainstones, four of which are pervasively cemented, skeletal wackestones and mudstones/Chalk. This has allowed a unique overview of micropore topology in three dimensions. Interparticle microporosity occurs between grains and/or within grains that have undergone boring and micritisation. Micropores have been diagenetically altered by recrystallization (neomorphism) of primary micrite, which has increased crystal size and decreased micropore volume and connectivity. Conversely, solution enhancement of micrite has enhanced micropore volumes and connectivity. |

**Apresentação**

Influence of the Pore System Properties on P-wave Attenuation in CarbonatesBiot's theory of poroelasticity provides a theoretical basis for phenomena understanding of the attenuation due to the elastic wave propagation. Carbonate rocks have complex pore system with a variety of pore types and a wide range of pore size. Many studies indicate that pore size, pore shape and geometry, and amount of micropores are important factors influencing P-wave attenuation in rock pores, and can explain the variation of attenuation at the same porosity. Therefore, the effects of the pore structure and its properties on elastic wave attenuation cannot be neglected. In this work, the petrophysical and physical properties of carbonates data set were obtained from literature and the P-wave attenuation was estimated through the Dvorkin-Mavko model to investigate the influence of petrophysical, mineralogy, geometrical parameters and porosity scale in P-wave propagation and attenuation. The results allowed to correlate the pore geometry properties of the medium such as DomSize and PoA to attenuation effects and contributed to the understanding of the attenuation in seismic studies of carbonates and complexities of the pore system. |

<< < 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 68 > >>