3rd EAGE St.Petersburg International Conference and Exhibition on Geosciences - Geosciences: From New Ideas to New Discoveries

To study dynamic fluidal processes taking place during the production of HC it is proposed apply seismic emission (SE) waves instead of reflected seismic waves. SE-waves are generated as a result of changes in the stress-deformed state of fluid-saturated reservoir rocks. Described is a new passive seismic technology – Seismic Location of Emission Centers (SLEC) providing a real-time identification and location of SE centers arising due to dynamic fluidal processes. Presented are examples of applying SLEC on oil fields to map the direction of hydrofrac, advance of oil displacement front, positions of washed-out zones, and residual oil zones. SLEC-technology allows determine basic directions of hydro- and oil flowing, correlation between production rate and SE-parameters. This information can be used to improve the production efficiency.

The localization of seismic events in space and time has great importance for hydro fracturing and reservoir monitoring. A new passive seismic localization technique by diffraction stacking has been introduced recently. Method has many advantages due to a performance by a modified diffraction stack of the squared amplitudes of the input seismograms resulting in the image section. The source location is obtained from the maximum of the image section. The source time is determined after source location in space. Energy is stacked during the procedure hence very weak events against the background noise could be located. Numerical 2-D tests with high-level synthetic noise give visual confirmation. Another numerical example representing the scattering overburden over homogeneous medium provides great possibilities in application to the real data which is major issue to be solved. In case of badly known velocity model additional stacking over velocity range can improve the result.

Results of geological media exploration using scattered waves extracted from multichannel seismic data are presented. The researches are based on original prestack migration method – Wave CDP method, which possesses possibility to construct two independent time sections: conventional image of reflectors and new image of space distribution of acoustic inhomogeneities - diffractors. The last are join with fracture zones – strong sources of scattered waves. Method was testes in areas with carbonate, volcanic and clay rocks.

Исследование кинематических и динамических особенностей Р- и S- волн в «больших» физических моделях, включающих пласт с варьируемыми характеристиками пустотного пространства, показало, что (а) наличие изометричных пор и сети гидравлически связанных с ними трещин снижает скорости, амплитуды и частоту проходящих Р- волн сильнее, чем наличие трещин в среде без пор, при сравнимой плотности трещин; (б) на степени расщепления S- волн различия трещинной и поро-трещинной структур порового пространства сказываются слабо; (в) в газонасыщенной поротрещинной среде снижение скорости, амплитуды и частоты проходящей Р- волны сильнее, чем в жидко-насыщенной. Все эти наблюдения качественно согласуются с накопленным опытом полевой сейсморазведки и существующими теоретическими моделями. Однако количественно эффект, зафиксированный на физической модели, оказался сильнее, чем предсказывает теория. Этот результат требует дополнительного изучения.

The objective is to demonstrate the possibility of reservoir permeability mapping using a frequency-dependent seismic attribute and analysis of log data. The coefficients of normal reflection and transmission of a planar p-wave from a permeable boundary can be expressed asymptotically as power series with respect to a small dimensionless parameter depending on the reservoir fluid mobility. The zero-order terms of the asymptotic expressions do not depend on reservoir rock permeability and are similar to the one predicted by the classical elastic model. The next, first order, term involves a factor proportional to fluid mobility, both for reflection and transmission coefficients. In case of a very thing porous permeable layer (h<<λ) transmitted-reflected slow waves can create a low frequency resonance providing an opportunity for seismic inversion. The functional structure of the first order asymptotic term suggests a frequency-dependent seismic attribute, which is proportional to reservoir fluid mobility. We have derived such attribute from real seismic data and analyzed it vs. log data. We have obtained that the possibility for seismic imaging of the reservoir transport properties, in particular mapping the lateral permeability variations, is realistic.

In this study, both poroelastic and pure elastic forward modeling have been performed on reservoir models representing Class 1 and 3 amplitude-versus-offset (AVO) responses. Thin sand layers were periodically saturated with gas and water (White’s model). The reflection amplitude for both poroelastic and elastic models were calculated from the top interface of the thick reservoir and the amplitudes were studied as a function of frequency and incident angle. We found that in the seismic frequency range, the amplitudes for the poroelastic Class 3 AVO model increase when the frequency decreases. This trend is opposite to the elastic modeling result. For the Class 1 AVO model, however, the amplitude versus frequency trend is practically the same for both the poroelastic and elastic models. The numerical modeling results are explained by using White’s theoretical patchy-saturation model. The analysis of the modeling results might explain why soft reservoir rocks that are hydrocarbon saturated have low-frequency seismic anomalies.

New technology for seismic inversion using multilayer neural network (NN) is proposed. Special set of seismic attributes based on offsets serves as an input to the neural network. Logs measured at the well serve as training data. Genetic algorithm and back-propagation are used in NN to calculate weights. This allows implicit derivation of the non-linear operator. The method naturally extends to other logs which have meaningful physical relationship to the seismic data, such as porosity or density.

Narrow-azimuth orthogonal 3D/3C surveys impose a strong acquisition footprint on the PP and PS seismic datasets. This footprint is frequently manifested as false amplitude and phase variations in time sections and slices. The distribution of bin attributes can be improved using wide-azimuth receiver templates, especially in combination with slant or zig-zag shot lines. We illustrate the correctness of these recommendations by comparing effective fold and other bin attributes for two wide-azimuth surveys with the same aspect ratio of the receiver template (0.7). One of the surveys is an orthogonal one with the “standard” grid of receiver points (50 by 300 m) and the other survey is a slant one (45°) with a reduced interval between the receiver points (40 by 240 m). We have ray-traced both surveys to compare the synthetic PP and PS cubes and to show advantages of higher density wide-azimuth surveys for illumination of steep-dip targets.

The vibroseis whitening (or application of AGC to vibroseis records before their convolution with the sweep signal) was successfully tested in Russia and abroad at the end of 1980ties by a number of processing companies including GEMOI Neftegeofizika (Naro-Fominsk), KOME Ukrgeofizika (Kiev), and GlavNIVTS of the Ministry of Natural Resources of USSR (Moscow). The efficiency of vibroseis whitening was proved for traveltime processing as well as amplitude-preserving processing using both the field and synthetic data. It was shown that relative amplitudes are preserved with very high accuracy in both directions of the time section in contrast to AGC applied after the vibroseis correlation. Nevertheless, the vibroseis whitening procedure was not routinely used, and even the opportunity to apply it is not available now, since vibroseis correlation is performed directly in the field, and the correlated records only are transferred to the processing centres. In this paper we have attempted to attract attention to the efficiency of this simple procedure and described the previously obtained results as well as new modelling studies using the modern processing and interpretation systems.

The present study from the Vienna Basin (Austria) shows the result of different processing techniques in time including (i) pre-stack time migration, (ii) post-stack time migration and (iii) Common Reflection Surface (CRS) processing. Pre-stack time migration was the preferred approach to obtain a high-resolution image of the relatively shallow Neogene section (i.e., 0 ~2.5 km). Post-stack time migration using only the low-frequency components of the seismic signal yields improved images from the deeper Cretaceous and Triassic (pre-Neogene) section at ~2.5 – ~5.0 km depth. Finally, the best image of the complex geological structures in the deep pre-Neogene target area is obtained with CRS processing. Target-oriented data re-processing should be considered as an effective tool to increase the information content of an existing data set. Focusing on the principal attributes of a dataset using robust techniques (e.g., low-frequency processing, post-stack migration approach, etc.) or targeted techniques (e.g., CRS processing) have the potential to reveal information that is not obvious in conventionally processed data.