Showing papers in "Seg Technical Program Expanded Abstracts in 2003"

Differential Semblance Velocity Analysis By Wave-equation Migration

Abstract: Prestack wave equation migration using the double square root equation produces prestack image volumes free of artifacts, even in the presence of multipathing due to complex structure. In particular image gathers in angle or offset ray parameter are flat at correct velocity, and gathers in offset are concentrated at zero offset. Differential semblance measures the deviation from flatness or concentration, and provides a method of automatic velocity updating via optimization. The adjoint state method gives a convenient computation of the differential semblance gradient as an addendum to prestack depth extrapolation.

2‐D deconvolution imaging condition for shot‐profile migration

Abstract: To implement a better imaging condition that is feasible in p ractice it is important to understand the 3-D prestack data and wavefi elds dimensionality. 3-D prestack seismic data is defined in a 5-D continuum (t,xs , ys ,xg , yg) (Biondi, 1998), wheret is time, xs is the sourcex position,ys is the sourcey position,xg is the geophonex position, andyg is the geophoney position. After applying the first step of shot-profile migration (source and receiver wavefiel ds construction) we have for each shot position ( xs , ys ) the source and the receiver wavefieldsu(x , y,z,t) and d(x , y,z,t). After wavefields propagation a new dimension is added ( z). Then the wavefields have 6-D dimensions.

Whole Earth 4D: Reservoir Monitoring Geomechanics.

Abstract: Time-lapse (4D) seismic monitoring of pressure-induced changes in depleting gas fields reveals that detectable differences in seismic arrival times are observed above the reservoir interval. Geomechanical models of depleting reservoirs predict that as a result of reservoir compaction due to pressure depletion, changes in the long-wavelength stress and strain fields occur in the rocks bounding the reservoir. Models incorporating the geomechanical stress and strain field changes predict changes in the two-way arrival times that are compared with actual time-shift observations at a depleting gas field in the North Sea. The geomechanical-based predictions are in good agreement with the observations. Detecting geomechanical changes in the overand underburden rocks opens up new ways of using 4D data, especially in places where the signal from the reservoir rocks is small.

Determination of Rayleigh Wave Dispersion Curves For Near Surface Applications In Unconsolidated Sediments

Abstract: Multi-channel Analysis of Surface Waves (MASW) is a suitable technique for vertical shear-wave velocity profile determination whose efficiency and effectiveness depends on data acquisition parameters and processing procedures. In the present paper we compare the performances of three different methods to determine Rayleigh waves dispersion curves. Such curves are successively exploited for the inversion process, which eventually provides subsurface information of use in geological or geotechnical applications. We investigate the role of some processing procedures, with a special emphasis on the number of traces able to optimise the ratio between information content and acquisition and processing costs. We show that phase shift method is able to produce the best results in terms of accuracy and computation efficiency for the unconsolidated sediments considered in our work. Phase shift shows extremely stable results also when a reduced number of traces is considered and other methods fail due to spatial aliasing or severe noise content that prevents from unambiguous interpretation.

Source-receiver migration of multiple reflections

Abstract: Multiple reflections are usually considered to be noise and m any methods have been developed to attenuate them. However, similarly to primary reflec tions, multiple reflections are created by subsurface reflectors and contain their reflectiv ity information. We can image surface related multiples, regarding the corresponding pr imaries as the sources. Traditional source-receiver migration assumes that the source i s an impulse function. I generalize the source-receiver migration for arbitrary sources , and apply it to the migration of multiple reflections. A complex synthetic dataset is used to test the theory. Results show that my multiple migration algorithm is effective for imaging the multiple-contaminated data.

Unwrapping Instantaneous Phase to Generate a Relative Geologic Time Volume

Abstract: This presentation has three goals. The first is to introduce the Relative Geologic Time Volume (RGT Volume). This volume stores a detailed seismic interpretation using geologic time estimates for every seismic sample point. The second goal is to briefly provide some of the advantages of using a RGT Volume to store and access seismic interpretation. For example, by using such a volume, horizons can be selected using a spatial reference (i.e., inline, crossline, travel time position), such as dragging a mouse over an inline, crossline or time slice. The third goal is to describe how such a volume can be generated by unwrapping instantaneous phase. The unwrapping procedure assumes that a surface of constant instantaneous phase equals a seismic horizon, which in turn equals a constant geologic time.

New Paradigm of Fault Interpretation

Abstract: A high-level fault interpretation workflow using automatically extracted surfaces is presented with special attention on the human interaction part. The first step of the workflow is to generate a fault attribute that enhances the discontinuities in the seismic data. Fault-like surfaces are then extracted using an algorithm called “Ant Tracking”. The surfaces are then loaded into a 3D analysis tool where the interpreter, by interactively verifying, combining and deleting surfaces, decides what is to be the final interpretation. The interpreter works on two levels in the analysis tool; on the fault system level, and the individual surface level. This “top-down” approach to fault interpretation is demonstrated by a case study of two fields offshore mid Norway.

On the Inversion of Time-lapse Seismic Data

Abstract: As the number of time-lapse (4D) seismic applications increases; so do the prospects for quantitative estimates of dynamic reservoir properties, such as fluid saturation and pressure, based on time-lapse seismic data. These estimates can be obtained by the application of appropriate rock physics models that map seismic-derived estimates of timelapse Pand S-impedance changes to changes in the dynamic properties. Therefore, it is important that reliable 4D impedance models are available for this process. There are, however, a number of ways to estimate impedance change from 4D seismic data. The purpose of our study was to assess the merits of different inversion workflows by using a synthetic 4D data set and comparing models resulting from the inverse calculations with actual impedance models. This study demonstrates the importance of coupling the inversion of base and monitor surveys to mitigate the impact of inherent non-uniqueness in seismic data inversion. The results from uncoupled inversions can mislead time-lapse interpretation and compromise reservoir management decisions.

Gaussian Beam Migration For Sparse Common-shot And Common-receiver Data

Abstract: We investigate the Gaussian beam migration of commonshot and common-receiver data. The imaging of commonshot data is useful for seismic data where the receiver coordinates are well sampled, but the source coordinates are less well sampled. By reciprocity, this approach can also be applied to common-receiver data, such as from OBS experiments where the source locations are dense but the receiver locations are sparse. Since Gaussian beam migration uses smoothed, localized windowing of the data, it can provide more stable results for the inversion of sparsely sampled data.

Tomographic residual curvature analysis: The process and its components

Abstract: The image quality of 3D pre-stack depth migration in areas of complex geology depends strongly on the accuracy of the velocity model. Velocity updating by seismic tomography, in the form of either traveltime tomography or residual curvature analysis (RCA), has become an important component of the depth imaging process. In this paper, we describe the components of tomography, discussing RCA in detail. These components include: building of the tomographic updating equations; regularization in both data and model spaces; and application of the least-squares solver. Numerical examples show that the RCA algorithm works well, producing velocity models that improve the quality of depth-migrated images over models produced by vertical updating schemes.

Transformation of multiples into primary reflections

Abstract: The feedback model of multiple scattering shows that wave theory based multiple removal can be applied by a weighted convolution process of the input data with an estimate of the primary response. In this paper the feedback model is used to show that the weighted convolution process can be replaced by a weighted correlation process. This alternative formulation offers new opportunities to multiple removal as well as the transformation of multiples into primary reflections.

4D Seismic Data Processing Issues And Examples

Abstract: 4D seismic data acquisition efforts can be divided into three major categories: “legacy”, “re-shoot” and “4Ddesign ” projects. In legacy 4D seismic projects, multiple vintages of overlapping 3D seismic data sets are analyzed for time-lapse effects, but none of the 3D surveys were originally acquired with a 4D application in mind. In reshoot 4D seismic projects, the baseline 3D seismic survey was not acquired for 4D purposes, but the subsequent reshoot 3D survey was designed, at least in part, with a 4D objective in mind. Finally, in 4D-design projects, at least two of the time-lapse 3D seismic surveys were specifically designed and acquired to optimize the subsequent reservoir monitoring analysis.

Automated Geometry Extraction From 3D Seismic Data

Abstract: Classification tools are widely used in both 3D and 4D reservoir characterization, for example in mapping of 3D structures, lithological properties and production effects [1]. In this work we extent the application area of the classification methodology into automated interpretation of seismic reflectors. By classifying the seismic signal along reflectors we gain an improved automated interpretation that performs well also in structurally complex regions. Based on this classification methodology, we demonstrate automatic extraction of seismic reflectors and geo-bodies, and automatic fault displacement estimation.

Simulated Prestack Local Imaging: a robust and efficient interpretation tool to control illumination, resolution, and time-lapse properties of reservoirs.

Abstract: Summary The “Simulated Prestack Local Imaging” (SimPLI 1 ) concept is a new method helping interpreters of migrated seismic data to control and constrain their interpretation of local structures such as oil/gas reservoirs. There is no need in this approach to first generate synthetic data, and then migrate to calculate the migrated seismic response of reservoirs models. Simulated prestack migrated sections are quickly obtained as functions of the survey, emitted pulse, wave-modes, and local reservoir structure. An interpreter can therefore check interactively various reservoir scenarios in terms of illumination (survey-planning, reservoir characterization), resolution, and time-lapse evolution related to rock parameter changes during the production. Input can be interpreted horizons with attributes, property grids from reservoir models and inversion results, or hypothetic models. The results are prestack migrated sections, in depth or time. The SimPLI method is a convolution technique, which goes far beyond the classic 1D convolution as being able to predict 2D/3D effects of illumination, resolution and prestack acquisition, but without requiring experts in modeling and migration.

Euler Deconvolution of Gravity Data

Abstract: Euler deconvolution of both profile and gridded magnetic data (Thompson, 1982; Reid et al, 1990) has found wide application. It has been implemented by many organisations and individuals and is commercially available from several suppliers. Applications to gravity are fewer. Marson & Klingele (1993) applied it to gravity vertical gradients. Keating applied it to gravity point data. Zhang et al (2000) applied it to gravity tensor gradient data. But Euler deconvolution requires a well-founded understanding of a critical parameter, the Structural Index (SI), which characterizes the source geometry. In the magnetic case it varies from zero (contact of infinite depth extent) to 3 (point dipole). It can be thought of as the index in the field strength fall-off with distance.

Angle Gathers For Shot-record Migration By Local Harmonic Decomposition

Abstract: It is now recognized that shot-record wave-equation migration is the best method for imaging in complex media. However, it is also recognized that producing one output image is not enough. Velocity analysis can only be done when a gather is produced, as well as AVO analysis or multiple attenuation. This paper starts by describing the local harmonic decomposition: it is an efficient algorithm which can be used to extract angle information from a wavefield. It is then shown that this algorithm can be used to produce reflection angle gathers. A synthetic test proves that these gathers are centered on zero opening angle whatever the dip and the absence of artefacts. Results from the application of this algorithm on real 3D OBC data are shown.

Multicomponent Seismic Data Registration By Least Squares

Abstract: We apply an automatic data registration (warping) algorithm to find a mapping between P-wave and converted-wave migrated images. The algorithm improves the matching of seismic volumes obtained by previous manual interpretation. There are two main products of this process. First, it improves a combined interpretation of multi-component seismic images. Second, interval velocity ratios get extracted directly from the warping function at a resolution unobtainable with other methods.

Fault detection using Hough transforms

Abstract: Faults can often be represented by straight lines on the seismic section or time slice. Image processors to enhance such alienators are the Hough transforms. We apply the Hough transform in a global and a small running window implementation. The running window implementation is efficient but more sensitive to noise. The conventional implementation is more expensive and results in long lines representing the faults. In this report, we applied the Hough transforms method on some real data. The first test was done on a time slice. The second test was done on a principle component coherence cube. Results are shown better on the figures.

Estimation of Reservoir Pressure And Saturations By Crossplot Inversion of 4D Seismic Attributes

Abstract: We present a method to simultaneously estimate pressure and saturation changes in a producing hydrocarbon reservoir using time-lapse (4D) seismic attributes. 4D seismic attributes are displayed in a crossplot domain, where pressure and saturation axes can be estimated deterministically or interpretively. These axes form the basis for a linear or nonlinear coordinate transformation to the pressure-saturation domain. A final calibration to production data is required to convert the qualitative results to quantitative estimates of pressure and saturation. Our method is applied to 4D seismic data sets acquired over producing North Sea reservoirs in the Schiehallion and Gullfaks fields.

Envelopes of 2D And 3D Magnetic Data And Their Relationship to the Analytic Signal: Preliminary Results

Abstract: Unlike the 2D case, the total gradient (magnitude of the gradient vector) in 3D is not the envelope of either the vertical or horizontal derivatives of the total-field anomaly over all possible directions of the earth’s field and source magnetization. Instead, the total gradient of the reduced-to-pole data defines the envelope for the vertical derivative while the total gradient of the reduced-to-equator data defines the envelope for the horizontal derivative.

Velocity And Attenuation Anisotropy In Reservoir Rocks

Abstract: Sedimentary rocks are generally characterized by textural anisotropy that gives rise to a directional dependence of acoustic velocities. This textural anisotropy in sedimentary rocks can lead to attenuation anisotropy. Our analyses of the directional dependence of velocity and attenuation in different reservoir rocks show that attenuation anisotropy can be quite large. Directional dependences of velocity and attenuation can be very different and are not influenced in the same manner by textural variations. Depending on the texture, the attenuation anisotropy can be opposite to the velocity anisotropy.

Three-dimensional Illumination Analysis Using Wave Equation Based Propagator

Abstract: A wave-equation based method is developed for seismic illumination analysis. The approach uses the one-way screen propagator to provide fast and accurate wavefield extrapolation. A local plane wave analysis is used to determine energy propagation directions. The method can avoid the singularity problem usually linked to the high frequency asymptotic method. It provides a practical tool for three-dimensional full-volume illumination analysis in complicated structures. To demonstrate the potential application of this method, numerical examples for the 3D SEG/EAGE salt model are calculated.

Fast And Rigorous Inversion of Triaxial Induction Logging Data to Determine Formation Resistivity Anisotropy, Bed Boundary Position, Relative Dip And Azimuth Angles

Abstract: In this paper, we present a fast and rigorous inversion method to determine formation resistivity anisotropy, formation bed boundary location and relative dip and azimuth angles from a recently developed triaxial induction tool. This tool is fully triaxial and has sets of three orthogonal coils which are located at the same point. The sensitivity matrix with respect to horizontal conductivity, vertical conductivity, formation bed boundary locations is computed rapidly with the help of its analytical expression in a 1D TI anisotropic medium. The relative dip azimuth is solved first by rotating the full tensor measurement. The rest of the parameters are solved simultaneously by using Gauss-Newton minimization. The inversion results on synthetic data and field data show the accuracy and efficiency of the inversion algorithm.

Gabor Deconvolution Revisited

Abstract: Gabor deconvolution has been updated and experience has been obtained on real data. The updates are (1) a new method of spectral smoothing called hyperbolic smoothing, (2) a Gabor transform using compactly supported windows that improves run times by one to two orders of magnitude (3) a post-deconvolution time-variant bandpass filter whose maximum frequency tracks along a hyperbola in the timefrequency plane. We discuss the technical details of these improvements and present a data example.

Flattening 3-D Seismic Cubes Without Picking

Abstract: As a seismic interpretation aid, I present different methods to convert an entire 3D data cube into a cube of horizon slices without picking. These methods essentially sum local dip estimates into total time-shifts. Thus far, an efficient Fourier-based method that sums the dips along time slices to flatten unfaulted data has been developed and tested on a simple field dataset. Another method in the time-space (T-X) domain that flattens vertically faulted data, again by summing along slices, has been developed and tested on a simple sythetic data. This T-X domain method requires a fault model as input. A computationally efficient and practical way of globally flattening 3-D data volumes in the presense of noise, nonvertical faults, and unconformities can be developed.

Seismic pore pressure prediction with uncertainty using a probabilistic mechanical earth model

Abstract: Summary We propose a methodology to propagate uncertainties in seismic pore pressure prediction using a 3-D Probabilistic Mechanical Earth Model (P-MEM). An extended form of Bowers formula is used to link pore pressure to seismic velocity, overburden stress, porosity and clay volume. Probability Distribution Functions (PDFs) for all input variables are stored as attributes in the 3-D MEM. An output PDF for pore pressure is then calculated point by point in the 3-D model, using either a linearized Gaussian approximation or a sequential stochastic simulation approach that fully accounts for nonlinearities in the velocity to pore pressure transform and spatial correlation between the different input variables. The linearized and stochastic approaches are compared in the context of a seismic pore pressure prediction study involving overpressured reservoir sands.