articles

 

The article presents studies of some issues of embedding theoretical solutions into the data, observed in a seismic experiment. The investigations have been based on inverse problems solutions for thin-layer elastic models within a frequency domain. Embedding the theoretical solution into real data has been analysed taking into account the following two aspects: a transformation of the observations to improve their correspondence to the modeling assumptions and an analysis of influence of the real data characteristic properties on the theoretical solution. At the same time one has performed decomposition of the observed wave field that has resulted in its better approximation to the theoretical model. Direct problem solution in the frequency domain, taking into account the observations aperture, has been obtained with the Laplace and the Fourier-Bessel transforms. As a result, a better association of the theoretical solution and the two dimension spectra, calculated for the observed seismograms, has become available. The approach proposed opens a way for applying modern theoretical methods of direct and inverse problem solutions and extends the options of their applications for real multicomponent data processing and interpretation.

The problem of separation of surface and depth factors of a reflected field and the nonuniqueness of its solution result in nonuniqueness of the solution for the surface-consistent deconvolution problem. In the article one considers additional conditions and combinations of the model parameters that result in the deconvolution operator has been calculated in a unique way.

This paper presents an original approach of true amplitude seismic imaging by means of weighted summation of multi-component walk-away VSP data that use Gaussian beams. In order to compute these weights one should trace a Gaussian beam starting from some current point within the target area towards the acquisition system. Specifically selected Gaussian beams allow taking into account a geometrical spreading and an illumination condition for a specific acquisition system. Global Gaussian beam regularity enables one to deal with a central ray field having singularities of any kind. The algorithm also allows one to reconstruct structural disruptions (faults etc.) in the near borehole area. Using Gaussian beams prevents imaging artifacts that occur due to the limited aperture, while application of several components prevents those artifacts associated with the converted seismic waves.

The paper addresses mathematical modelling of anisotropic effective elastic properties of carbonate reservoir rocks with complex multi-scale porosity. Porosity in these rocks consists of four types of voids embedded in the matrix: pores, vugs, and vertical and horizontal cracks. Correspondingly, the model includes four structure elements for which main parameters are determined separately. The porous mineral matrix (element I) encloses channel-like vuggy zones (II) acting as fluid conduits, with their material assumed in the general case to differ from that of the matrix. Vertical and horizontal aligned cracks make elements III and IV. Elasticity is characterised using analytical relationships between elastic wave velocities and the model parameters that represent rock structure in terms of the effective medium theory. Generally, elastic properties in reservoir rocks of this kind are of orthorhombic symmetry, but the model is applicable to simpler hexagonal or isotropic cases at a certain choice of model parameters. In the reported forward modelling, the effective physical properties of reservoir rocks are estimated from porosity and geometry of voids. The forward model is used to investigate the sensitivity of elastic properties to parameters related to the void space. Inversion is performed with the model parameters of a fluid-saturated carbonate reservoir in the Yurubchen-Tokhomya field. The geometry and volume contents of matrix pores, vuggy zones, and vertical and horizontal cracks are inverted to estimate the reservoir porosity with reference to log data.

Many broadly used conventional processing algorithms for traveltime inversion and imaging in marine seismic are poorly applicable to ocean bottom seismic data. This acquisition pattern requires special schemes, including removal of multiples. As the appropriate processing procedures have been applied, the data are converted into the form typical of the conventional asquisition by redatuming in order to bring both sources and receivers to a single level.

Microseismic emission associated with hydraulic fracture can be successfully monitored on the surface using a system of autonomous six-channel stations. Data processing consists in inversion of time delays relative to the reference channel. The most reliable solutions are selected as a localized subset taking into account approximate source depths and average wave velocities. Surface seismic records of perforation shots can provide a check of known source depths and origin times to allow for the low-velocity zone at the section top and to adjust the processing scheme.

The tectonic history of the northern Russky-Chaselka arch has important implications for the right choice of development strategy given that the area has a complex local geology while oil and gas accumulations are due to vertical migration of hydrocarbons from stratigraphically lower formations. The area comprises three zones which differ in number of seismic sequences and, thus, in tectonic evolution. Two separate structural stages in the section (Triassic-Jurassic and Cretaceous) correspond to deposition in tectonic settings with different stress regimes. The large Yuribei fault of oblique-slip geometry (normal slip with a reverse component) located in the southwest underwent four extension phases and one phase of compression. According to the tectonic reconstruction based on seismic data, there are two types of isolated fault blocks in the survey area which result from the activity of the Yuribei fault during the Eocene inversion (stress reversal from extension to compression) (type 1) and from Neogene activity of small strike-slip faults (type 2).

We compare the amplitude spectra of correlated vibroseis records obtained with different sweep types: linear chirp and composite signals, and nonlinear logarithmic and dB/oct power sweeps. The calculations are performed with reference to the spectra of principal ghost components. In the case of large dynamic ranges of records, dB/oct power sweeping is more effective while logarithmic sweeping is undesirable. The reason is that logarithmic sweep shows 10–15 dB higher correlation noise in real survey conditions than the correlated dB/oct power sweep records.

Comparative statistical evaluation of signal arrival time from the pulse and vibration seismic sources have been given depending on the signal to noise ratio. The authors come to the conclusion on optional equivalence of these sources and emphasize the advantages of each of the methods.