Paulo Léo Manassi Osório

Unsupervised seismic facies analysis using wavelet transform and self-organizing maps

Unsupervised seismic facies analysis provides an effective way to estimate reservoir properties by combining different seismic attributes through pattern recognition algorithms. However, without consistent geological information, parameters such as the number of facies and even the input seismic attributes are usually chosen in an empirical way. In this context, we propose two new semiautomatic alternative methods. In the first one, we use the clustering of the Kohonen self-organizing maps (SOMs) as a new way to build seismic facies maps and to estimate the number of seismic facies. In the second method, we use wavelet transforms to identify seismic trace singularities in each geologically oriented segment, and then we build the seismic facies map using the clustering of the SOM. We tested both methods using synthetic and real seismic data from the Namorado deepwater giant oilfield in Campos Basin, offshore Brazil. The results confirm that we can estimate the appropriate number of seismic facies through the clustering of the SOM. We also showed that we can improve the seismic facies analysis by using trace singularities detected by the wavelet transform technique. This workflow presents the advantage of being less sensitive to horizon interpretation errors, thus resulting in an improved seismic facies analysis.

Wavelet transform filtering in the 1D and 2D for ground roll suppression

Among the various types of noise found in seismic land acquisition there is the one produced by surface waves. This noise is called ground roll, and it can be defined as a group of events that contaminates seismic data by forming a high dip (low velocity) and high amplitude cone, which can dominate near-surface events on the seismic records. We present in this paper a technique based on the 2D wavelet transform to remove the ground roll. The method was tested with real data and the results were superior to the ones obtained with the 1D wavelet transform.

Characterization of Thin Beds Through Joint Time-frequency Analysis Applied to a Turbidite Reservoir In Campos Basin, Brazil

A new spectral decomposition based method is presented here. We propose to use the ridges of the joint timefrequency analysis, as a new way to detect, in each trace, the maximum instantaneous frequencies and their associated amplitudes, which were applied as a tool to detect seismic geomorphologic bodies. The technique was applied to a synthesized wedge model and also to a thin bed offshore turbidite reservoir in Campos Basin, Brazil.

Unsupervised seismic reservoir characterization using wavelet transform and self organizing maps of a deep‐water field, Campos Basin, Offshore Brazil

This work presents a new alternative to extract seismic pattern attributes and a new methodology to build seismic facies maps. We propose using Wavelet Transform to identify singularities in each geological oriented segment of the temporal seismic trace and then using Self Organizing Maps (SOM) in a two-level approach. To illustrate this technique we applied it to real data from a deep-water field in the Campos Basin, Brazil.

Using Matching Pursuit and Self Organizing Maps for Seismic Reservoir Characterization of a Deep-water Field, Campos Basin, Offshore Brazil

Summary We propose to use the matching pursuit with timefrequency dictionaries algorithm applied in each geological oriented segment of the temporal seismic trace jointly with the clustering of the Self Organizing Maps (SOM) as a new alternative to build seismic facies maps. The technique was applied to a real data from a deep-water field in the Campos Basin, Brazil.

Modelagem e migração em profundidade 2D em meios com simetria polar local

This paper shows a technique based on the phase-shift method (PSM) to implement pre-stack depth migration on locally transverse isotropic media (LTI), with the symmetry axis direction varies continually along the layers. For seismic modeling, a generalization of the finite differences method for the solution of the elastic wave equation was used. With this procedure, it was possible to accommodate seismic modeling on LTI media defined by six parameters at each grid point, i.e. , density, P and S wave propagation velocities along the local symmetry axis, Thomsen parameters and the direction of the local symmetry axis itself. In order to separate from the seismograms the qP and qSV wavefields, an algorithm based on the Christoffel equation was implemented. The migration for each common shot gather is implemented solely by phase-shift based algorithms, which means that not only the depropagation of the registered wavefield, but also the generation of the time matrices involved in the imaging condition were obtained in this manner for each set of parameters at each depth level. The migration results using qP-qP and qP-qSV reflections show that the horizons were located precisely, and that the process is stable in relation to the symmetry axis variations. The proposed method is for multicomponent seismic acquisitions and might be applied to marine seismic data using streamers, or Ocean Bottom Cables or vertical cables. Since the proposed method uses phase-shift algorithms, its parallel implementation can be highly efficient.

Phase-shift anisotropic depth migration using controlled illumination applied to a model of the San Alberto field - Bolivia.

Summary We introduce a new scheme for depth migration in elastic vertical transverse isotropic media (VTI), using the concept of controlled illumination. In the proposed method the areal shots obtained from multicomponent records are extrapolated using phase-shift techniques. Through the weighted addition of delayed shots we synthesize appropriate areal shots, which increases the accuracy of the seismic imaging in the area of interest. The computational cost of the present method is much lesser, when compared to the cost of migrating all the records, since only a few areal shots are necessary to image the area selected by the interpreters. The proposed method was tested on a typical numerical 2D model from the San Alberto field in Bolivia. It was possible to correct image an exploration target located underneath a thick highly tectonic deformed anisotropic shale layer, using only 3% of the computational time necessary to migrate all the shot records.

Phase-shift anisotropic depth migration using controlled illumination: Stability in relation to addition of random noise

The present work studies the stability of a depth migration for elastic vertical transverse isotropic media (VTI), using the concept of controlled illumination. In the proposed method the areal shots obtained from multicomponent records are extrapolated using phase-shift techniques. Through the weighted addition of delayed shots we synthesize appropriate areal shots, which increases the accuracy of the seismic imaging in the area of interest. The computational cost of the present method is much lesser, when compared to the cost of migrating all the records, since only a few areal shots are necessary to image the area selected by the interpreters. The proposed method was tested on a typical numerical 2D model from the San Alberto field in Bolivia exposed to noisy conditions created by a random noise generator. Even with lower signal to noise ratio (SNR) we can correctly migrate anticline structures under a thick anisotropic shale layer. Introduction One of the great challenges in the search for new oil and gas fields has to do with seismic imaging underneath intensively tectonic deformed areas. This includes areas underneath thick salt layers, such as the coastal basins in the east of South America and areas subjected to great compressional efforts, like some coastal basins in the west of South America. In this context, the wave equation migration based methods have played a fundamental role. In fact, its simplicity and the generality of their premises render its superiority in terms of accuracy when compared to asymptotic approximation based methods. On the other hand, wave equation depth migration usually demands a computational effort that is several times superior to the capacity of most computer centers, which turns many projects unfeasible. Many propositions aim at the wave equation migration optimization. Most of them are based in the concept of wave synthesis, as initially proposed by Taner (1976) and by Schultz and Claerbout, (1978). In these works, the synthesized waves were positioned near the surface, which in many instances, were very far from the exploration targets. Berkhout (1992) proposed a method in which the synthesized waves were positioned closer to the targets, that allowed to image the interested areas more precisely. Recently, among the several works in this area, we can cite Cunha (2002), who generalizes the concept of reverse time migration for areal sources, and Wang et al. (2001) who introduced the concept of multi controlled illumination. In the present work we test the stability, in relation to random noise addition, of an algorithm for areal shot migration scheme applied to multicomponent seismic data in VTI media (Cetale Santos, et al 2005). Areal shots from multicomponent data Once we have the field seismograms for the horizontal and vertical wavefield at the surface, derived from a multicomponent 2D survey, we compute the compressional wavefield seismograms by applying the divergent operator at a datum located near the surface (Sun and Wang, 1999). From the compressional wave seismograms, represented by P(kx,z0,ω;xj), where xj is j shotpoint position at the surface, we compute the areal shots through the weighted sum of delayed shots,