- Home
- Conferences
- Conference Proceedings
- Conferences
54th EAEG Meeting
- Conference date: 01 Jun 1992 - 05 Jun 1992
- Location: Paris, France
- ISBN: 978-90-73781-04-7
- Published: 01 June 1992
81 - 100 of 405 results
-
-
Modelling mach wave propagation between boreholes in layered media
More LessCrosshole seismic experiments often yield data containing arrivals which are difficult to understand and explain, due in part to the location of the sources and receivers in boreholes. In order to use the data from these experiments to infer properties of a petroleum reservoir, an improved knowledge of the effects of source and receiver boreholes on the waves propagating in these experiments is therefore very useful. An example of this type of problem is the interpretation of the energy radiated from tube waves propagating in the source well, since in slow formations where the tube wave velocity is larger than the formation shear wave velocity, this energy can result in the recording of large amplitude Mach waves, or shock waves, in the receiver well (Meredith 1990 ; de Bruin and Huizer 1989). This phenomenon typically will occur in shallow, low velocity sedimentary layers . The conical Mach waves will not only have an amplitude much larger than that predicted for radiation directly by the source, but will also have a linear moveout velocity which cannot be explained by radiation directly Erom the source.
-
-
-
3D Seismic modelling of general material anisotropy in the presence of the free surface by a Chebyshev spectral method
Authors E. Tessmer and A. BehleA new 3-dimensional spectral method for séismic modelling in general anisotropic media is presented. An important feature of the method is its ability to handle the free surface boundary conditions very accurately . Moreover the algorithm is able to account for complete material variability, since it is a grid method, where every grid node can represent different material properties . The modelling scheme is based on the velocity-stress formulation of the equations of dynamic elasticity, but does not make use of staggered grids. For the horizontal directions spatial derivatives are carried out by the Fourier method which has periodic boundary conditions, whereas differencing with respect to the vertical direction is performed by a Chebyshev derivative operator, which is non-periodic and therefore can account for specified boundary conditions. Time integration is achieved by a fourth order Taylor expansion of the formal solution .
-
-
-
3D Surface consistent decomposition for refraction modelling: a workstation implementation
Authors W. R. Stresau, R. C. Farrell and K. P. FordRefraction analysis, modeling, and statics, are often standard components of the production processing sequence for 3-D data volumes. Many algorithms exist to define a near surface model from the refraction analysis of 2-D and 3-D production field records. A new surface consistent decomposition algorithm has been developed and implemented on the workstation that will efficiently model the near surface based upon travel times derived from 3-D production field records. The algorithm is implemented using a conjugate gradient technique. The least squares error minimization is statistically analyzed to gain confidence in the solution. This error analysis facilitates identification of picking problems, inadequately identified refractor segments, or geometry problems. The interactive implementation allows problem areas to be identified, while iterating to the final model . Capabilities exist to interactively correct the problems before proceeding further with the modelling. This method ensures an optimal near surface model is derived in an efficient manher for 3-D data volumes, crooked lines, or 2-D data. Results of the 3-D implementation of the surface consistent decomposition algorithm on a 3- D data example will be provided .
-
-
-
Angle dependent attenuation of seismic waves in a random stack of layers: examples of exponential and fractal media
Authors S. A. Shapiro and H. ZienThe effect of waves attenuation because of scattering in layered media can be observed and analyzed in data of seismic measurements like VSP and cross-hole sounding. Theoretically and numerically we consider the transmission of a pressure plane wave through an extensive fragment of a stratified disordered medium with constant density and without intrinsic attenuation. We study effects of multiple scattering.
-
-
-
Structural model uncertainty of a prospect - a geostatistical approach
More LessThe structural geometry of a Prospect is usually defined using some Time to Depth migration technique, as a function of the time map and of the corresponding seismic velocity map . For time migrated data, a simpte Time to Depth conversion is needed, with an adjustment of the depth surtace to Well data, if necessary . Depth reconstructions are always affected by quality and spatial frequency of vetocity information; if, as in the study presented, the structures are very gentle, a special precision would be necessary for a good prospect reconstruction and evaluation. For the study presented, two seismic markers were interpreted and mapped, corresponding to the top (level B ) and the bottom (level C ) of a sequence of finely stratified sand and clay layers, structured in a gentle anticline, which bears gas in correspondence of Well A1. The targets of the study were the reconstruction of their depth maps, the evaluation of the trap potentiality and the estimation of the result reliability . To reach these targets we put our main effort in improving the estimation of the velocity maps and to establish a method for providing results in term of probability .
-
-
-
Seismic wavefield fluctuations and attenuation in random media: numerical modelling and theory
Authors S. A. Shapiro and G. KneibSeismic wave propagation in random media yields complex wavefield fluctuations and apparent attenuation by scattering. Both can be observed and analyzed in VSPs and cross-hole data .
-
-
-
Coupled tube waves between two boreholes: a theoretical approach
Authors J. -L. Boelle, M. Dietrich and B. PaternosterCross-hole seismic data are usually processed and interpreted by ignoring the presence of the borehole i .e., by assuming that the sources and receivers are simply buried in the subsurface . This assumption is generally acceptable for ray tomography methods using first arrivals only, but must be revised in diffraction or full waveform tomography techniques.
-
-
-
Multichannel estimation of frequency-dependent Q from VSP data
Authors P. E. Harris, C. Kerner and R. E. WhiteMeasurements of attenuation (or Q) are frequently made from VSP data. Unfortunately, whatever method of estimation is chosen, the statistics of the Q estimates are poor. To achieve low variance, a large depth interval must be used, leading to poor resolution. This in turn produces a high bias in the estimates. The optimal depth range for estimation is thus a trade-off between variance and bias .
-
-
-
Seismic attenuation measurements from crosshole data
Authors I. R. Portsmouth and M. H. WorthingtonOne compelling reason for performing crosshole seismic experiments is that relatively high frequency data are usually obtained compared to surface or VSP surveys. It may seem logical to conclude from this general observation of many published results that the seismic Q of the earth beneath the near surface weathered layer is high . However, in this experiment crosshole seismic data with a, frequency range of 400 to 2500 Hz have been analysed and estimates of the intrinsic Q of sedimentary rocks of approximately 18 have been obtained.
-
-
-
P and S well seismic attenuation studies in a fractured reservoir
Authors A. Herrenschmidt and G. OmnesThe attenuation studies in this paper are based on borehole seismic data acquired in February 1989 in a vertical producing well in the Romaskhino field in Tatary (Republic of Russia). The well produces about 2. 5m3 (21 barrels) per day from a lower carboniferous (Bashkirian) carbenate reservoir within the 788-860 m depth interval. Average total porosity is 8 .4%, . Production is controlled by vertical fractures or joints .
-
-
-
More reliable shear-wave data from VSP using CIPHER technique
Authors T. Gut, W. Sollner, E. Luschen and H. A. K. EdelmannIt has long been recognized that fracture systems in rocks can significantly influence the flow of fluids in the subsurface. Geophysical fracture characterization techniques based on seismic anisotropy and related shear-wave splitting have been under investigation for many years. It is now established as an observable effect. To keep rig time costs low, a 3-C multilevel downhole sonde has been introduced many years ago. Additional horizontal vibrators and additional survey time and processing costs for shear-wave data, however, are still important constraints. The CIPHER method (Edelmann, 1992) described below aims both at the reduction of costs and at an improvement of VSP shear-wave data.
-
-
-
Present limitations in cross-well reflection seismic
Authors B. Paternoster and V. DesouchesHigh frequency seismic propagation can improve spatial resolution in Reservoir Descripton. Cross-well seismic has great potential for this. Based on three actual cross-welt seismic surveys conducted in 1990, the objective of this paper is to discuss some of the limitations that these surveys are presentIy facing in terms of acquisition, processing and interpretation .
-
-
-
Drill bit wavefields
By J. W. RectorThe seismic body wave radiation pattern of a working rock bit can be characterized by theoretical modeling and field data examples. My model of drill-bit signal generation is a pseudo-random series of bit tooth impacts that create both axial forces and tangential torques about the borehole axis. Each drill tooth impact creates an extensional wave that travels up the drillstring and body waves that radiate into the Barth. The model predicts that P waves radiale primarily along the axis of the borehole, and shear waves radiate primarily perpendicular to the borehole axis. In a vertical hole the largest P waves will be recorded directly above and below the drill bit; whereas, the largest shear waves will be recorded in a horizontal plane containing the drill bit. In a deviated borehole, the radiation patterns should be rotated by the inclination angle of the bit. The proposed seismic body wave radiation is investigated with field data examples using multichannel arrays of vertical geophones. The modeled radiation pattern is also investigated with a three-component inverse VSP data example shown in Figure 1 . The variation in the signal-to-noise ratio witti drill bit depth and the partitioning of S-wave energy between radial and transverse components is consistent with the proposed radiation pattern. A quantitative measure of the drill bit radiation pattern, taken from calculating the rms ratio between the shear wave direct arrival on the radial component and the P-wave direct arrival on the vertical component, is also consistent with the proposed radiation pattern .
-
-
-
Acoustics on real scale models
Authors P. Gaudiani and J. L. MariA full wave form acoustics tool is a very high frequency geophysical tool, for imaging near borehole structure with an investigation of several meters. We have built a 3-dimensional real scale model, with a substratum made of concrete (the model is composed of 3 zones, figure 1). The borehole profile can be modified. For an horizontal borehole profile, the substratum'is parallel to the borehole in zone 1. It has a 1 .2° dip in zone 2, and a 4 .9° dip at the beginring of zone 3. At the end of this zone, it is parallel to the borehole. The reservoir layer is made of concrete in zone 1. Zones 2 and 3 can be filled with different types of material (i .e . sand) to simulate different types of reservoir layers .
-
-
-
Steep dip wave equation migration by the finite difference plus phase shift hybrid method
By C. M. HaddowTwo vital requirements of a practical migration scheme are that it should be able to accommodate velocity variations and that it should have no dip limitation. The finite differente method for seismic migration has been known for many years, and remains popular since it caters well for the first of these requirements. However, the low order approximations, usually known as the 15 degree and 45 degree equations, are dip limited and do not preserve steeply dieping events. The dip response can be improved by using higher order approximations to the wave equation; for example, Ma (1982) showed how a factorization of the continued fraction expansion could be combined with a splitting techniquè to obtain steep dip solutions. These higher order methods may be computationally inefficient, however, due to the number of iterative stages that make up a single step of wavefield extrapolation . Other techniques exist which aim for better dip handling capability. These include the explicit formulations developed by Harris (1979), Koehler (Sengbush, 1983), Beaumont et al. (1987) and Holberg (1988). They take the foren of convolutional operators and are usually table driven since the operator is velocity dependent. The accuracy of the result will depend on the length of the operator, but of course longer operators will be more expensive to use.
-
-
-
True amplitude aspects of pre-stack depth migration
More LessIn the past decade, pre-stack depth migration algorithims have been developed to image complex geologic structures in a proper way. however, in recent years it has been shown that, if seismic inversion is formulated as a sequence of processing steps, pre-stack depth migration can also be sean as one of the principal steps in seismic inversion. This means a.o. that the interest in the amplitude behaviour of seismic migration has been significantly increased. Bleistein (1987) uses the method of stationary phase to compute true amplitude reflection coefficients. In a recent paper Hubral et al (1991) discus the problem of true amplitude zero offset time migration, using raylracing arguments . In the past we have shown by wave theoretical argument show to apply true amplitude pre-stack depth migration to shot records (see e.g . Berkhout, 1984, chapter VII and Wapenaar and Berkhout, 1989, chapter XI). In this paper we will show, also by wave theoretical arguments, how to apply true-amplitude pre-stack depth migration to any subset of pre-stack data such as CMP gathers and common offset sections. Of course, the proposed process is suitable to handle irregular offsets as well.
-
-
-
Steep DIP migration by replacing vertical with horizontal propagation
By L. E. BergThe algorithm presented in this paper uses the implicit finite difference scheme (45 degree approximation), for the propagation. The advantages by choosing this alternative for a migration implementation are well known: Since the extrapolation operator is only 3 points long, it is robust in cases with considerable velocity variations. In addition, run time on computer is very short, compared to some other algorithms. The disadvantage is of course the low performance for steep dips
-
-
-
FK DMO for depth variable velocity
Authors M. A. Brzostowski, H. A. Meinardus and K. SchleicherWe have implemented a practical DMO algorithm by incorporating Hubral 's stacking velocity theory into the Jakubowicz approach. This methodology allows for arbitrary velocity variations in depth and producer DMO impulse responses with triplications. We have analyzed the operator response for a constant velocity gradient. The resulting DMO impulse response adequately matches the ray-theoretical response. The method may also be extended to 3-D.
-
-
-
Equalization of DMO for irregular spatial sampling
Authors C. J. Beasley and R. KlotzSince the introduction of dip moveout (DMO), various authors have extensively explored its amplitude and phase accuracy, the effects of variable velocity, operator aliasing, discretization, and computational efficiency. A lens familiar aspect of DMO that is often of crucial importance in everyday seismic data processing is the effect of sparse or irregular spatial sampling on DMO-processed data. While different facets of the problem have been addressed (see e.g., Black and Schleicher, 1989; Ronen, 1987; and Williams and Marcoux, 1989), a comprehensive, accurate, and efficient solution to the problem is still needed. In this paper, we propose a method based on the decomposition of DMO into its constituent dip components that accounts for the effects of irregular spatial sampling for both flat and dipping events, can be applied to any 2-D or 3-D DMO algorithm, and is computationally efficient.
-
-
-
An implementation of anisotropic migration: some issues and examples
More LessA phase-shift based migration algorithm for transverse isotropy (TI) is described where the downward continuation operators are derived from the corresponding phase-shift propagators. The operators are applied in the space-frequency domgin either explicitly or implicitly. Results demonstrate that an extension to the standard implicit stencil is required when migrating with nonelliptical velocity models. The number of coefficients required depends very much on the degree of anisotropy and the mode in question. Some conclusions are drawn as to the number of coefficients required for a given maximum phase angle . The modification of explicit operators for anisotropy is relatively straightforward and they are used for TI media with tilted symmetry axes. For this implementation (one-way downward continuation) both the ray velocity and phase velocity have to be downwand going. For media with tilted symmetry axes, where some of the upward going energy corresponds to downward going plane waves and vice versa, the range of phase angles used in the operator design has to be limited accordingly. The method has been implemented for both prestack and poststack migration. The issue of deriving the velocity model using modifications to the standard surface seismic processing sequence is also addressed .
-