This paper presents a sensitivity study of seismic attributes to the effects of wave propagation through a structural and stratigraphic complex model adapted from the area of Tácata, Eastern Venezuela.
A 2D full wave elastic finite different seismic modeling was performed assuming a surface source-receiver geometry configuration. Instantaneous attributes were computed over the post-stack depth migrated sections and an AVO analysis was accomplished over the CDP gathers.
Multitrace seismic attributes including coherence, dip/azimuth, amplitude gradients, and spectral decomposition provide a means of rapidly extracting structural, stratigraphic, and tuning features from high quality 3-D seismic data. Unfortunately, given cost and time constraints our 3-D seismic images often suffer from inaccurate velocity/depth models giving rise to defocused, somewhat smeared images. Likewise variability in seismic survey fold and source-receiver azimuth within cdp bins, as well as inadequate geophone array suppression of backscattered surface waves can give rise to acquisition footprint which is often exacerbated by these modern edge detection attributes.
At the 62nd meeting of the European Association of Geoscientists and Engineers, in Glasgow, a new system for marine seismic cable steering and control was described (Bittleston et al., 2000). In this paper, recent operational experience with this system is reported. Seismic data acquired using active streamer steering, to achieve repeat positioning, is demonstrated to be highly repeatable.
In recent years, the driver for towed marine acquisition systems has been cost-effective acquisition of exploration 3D data. The specifications for acquisition of these data may limit their usefulness in detailed reservoir studies. Ideally, 3D data should be acquired with uniform fold of coverage and offset, azimuth distributions.
This paper is about impulse responses of wave equation prestack depth migrations. By examining the impulse responses in detail, especially in dynamic movie form, one can see the benefits and problems associated with different wave equation migration methods, including common azimuth migration, shot domain wave equation migration, and poststack wave equation migration. Common azimuth migration has dip limitations compared to other migration methods. The prestack shot domain wave equation migration has non-symmetrical impulse responses if true amplitude preservation is used. All wave equation migration methods handle multi-pathing phenomena more or less correct, but currently only the shot domain migration treats amplitudes correctly. The shot domain wave equation migration with FFD (Fourier Finite Difference) or better propagators is the most accurate approach.
To evaluate the feasibility of using surface seismic data for fracture characterization, several geological and geophysical techniques were used to characterize fractures in a Pennsylvanian-aged sandstone reservoir at Rangely Field in Western Colorado. Crossed-dipole shear-wave logging, zero offset VSP, offset VSP and walkaround VSP data all show a preferred azimuth of anisotropy of approximately N80E, consistent with nearby core analysis and with the direction of one of three main fracture sets from regional outcrops.
Prestack time Kirchhoff migration has been known to be a flexible method to image 3D seismic data. It has traditionally used the straight ray assumption to compute the travel-times. With our new method, we can analytically estimate the travel-times for V(z) and VTI (transverse isotropy with vertical symmetry axis) media. In addition, we also calculate the appropriate amplitude weighting factors to preserve the relative amplitudes after migration. To include ray bending, the earth is assumed to consist of horizontal layers, i.e. V(z) media, so that the velocity profile used to calculate the travel-times is simply a function of depth or time. Because ray-bending and VTI are considered, the takeoff and emergence angles for shots and receivers could be calculated accurately compared to those obtained assuming straight rays. Thus, the amplitude correction terms of the operator can be determined using these takeoff and emergence angles to preserve the relative amplitudes.
A tomographic inversion scheme is introduced, that makes use of seismic kinematic wavefield attributes (NIP-wave curvatures and normal ray emergence angles) at given measurement surface positions and zero-offset traveltimes to construct a velocity macro-model. These kinematic wavefield attributes can be extracted from multi-coverage prestack data in a data-driven way by application of the Common-Reection-Surface stack. The model to be inverted for is made up of two parts: the smooth velocity model itself, given by two-dimensional B-splines on a rectangular grid, and the starting positions and starting directions at depth (i.e. local reector dips) of the normal rays corresponding to the input data.
Most traveltime approximations make an assumption about the offset: usually it is assumed that the offset is small compared to reflector depth. The accuracy of these approximations decreases strongly when this assumption is not respected. I explain here how to obtain traveltime approximations that do not make any assumption on the offset, and that are consequently accurate at all offsets.
Traveltime approximations play a key role in seismic processing: They are used by time processing tools such as moveout correction and time migration, and by velocity analysis, which provides models required by depth migration.Accurate traveltime correction is also important for AVO analysis and multiple suppression.
3D inversion of IP data provides a better result than 2D inversion. These case studies demonstrate the improvement in depth of investigation that solving the inversion as a 3D problem provides, and the reduction of artifacts caused by the 3D nature of the sources seen in the deeper parts of the 2D inversion solutions. The use of a visualizer to report results is illustrated. A significant problem for interpreters wanting 3D inversion solutions is the time required to solve the inversion problem on typical desktop computers. To address this, the UBC 3D algorithm has been re-coded to run on a parallelized Linux cluster. Practical solution times have been achieved.
A good migrated image depends strongly on the accuracy of the velocity model. For tomographic velocity analysis, pick errors, null space, velocity ambiguity, and intrinsic non-linearity all have a strong effect on the accuracy of the velocity model. Our post-migration tomography algorithm can work on either common image gathers (CIG''s) or common angle gathers (CAG''s). A large number of depth residuals and the structural dip field are automatically picked and high-graded.