Read 2703_0296_Chopra1.qxd text version

Introduction to this special section--Seismic Attributes

eismic attributes are a great aid in both the qualitative strength), as well as the Q factor could predict the porosity and quantitative mapping of subsurface geologic features. distribution more accurately. This matched the available Even before classical interpretation of picking horizons and well data. generating maps begins, animating through appropriate For spectral decomposition of seismic data, short-time seismic attribute volumes can quickly define the structural Fourier transforms and the continuous wavelet transforms fabric and delineate major stratigraphic features. Seismic are commonly used. In his paper, "Spectral decomposition attributes have proven useful in almost every geologic envi- of seismic data using CWPT," Zhang introduces the use of ronment, from clastics through carbonates to volcanic, and a continuous wavelet packet-like transform. Exploiting the from normal faulting through wrench faulting to reverse multiresolution structure of wavelet transforms, he uses faulting. Thus, seismic attributes facilitate recognition of these transforms to decompose signal into a hierarchical depositional environments and enhance the recognition of block structure. Next, different resolution blocks yield specseismic facies. Once calibrated at well locations, attributes tral components. Zhang shows that application of the CWPT can be used to identify seismic facies and provide informa- for spectral decomposition of real seismic data provides tion about both lithology and fluids, with or without the use improved resolution and should therefore lead to a more of reference horizons. For this reason, seismic attributes accurate interpretation. form an integral part of most interpretation projects comUsually the output of spectral decomposition on seispleted today. A complete book on seismic attributes, Seismic mic volumes is a series of frequency volumes, the number Attributes for Prospect Identification and Reservoir depending on the frequency bandwidth being analyzed and Characterization, has been published by SEG recently. the frequency increment at which they are generated. Broad spectrums of seismic attribute workflows have Blumentritt, in his paper, "Highlight volumes: Reducing been adopted by our industry with the primary objective the burden for interpreting spectral decomposition data," of transforming seismic data (poststack or prestack) into meaningful geological maps. Such attempts have met with varying ... seismic attributes form an integral part of most degrees of success. While the greatest hininterpretation projects completed today. drance in attribute interpretation is that seismic data are always contaminated with noise, a secondary hindrance is that we have not yet cata- introduces these volumes, which reduce the number of vollogued the full range of attribute expression of features of umes to be analyzed to just two: peak frequency volume geologic interest. The purpose of this special issue of TLE and the peak amplitude, above-average volume. The ampliis to extend this catalog through clearer explanation of how tude spectrum for each sample is computed and that freto compute, display, and interpret these attributes through quency value for which the amplitude is maximum is output, their calibration through additional case studies. thus comprising the peak frequency volume. The peak We begin this special section with "Emerging and future amplitude above average is obtained by first computing the trends in seismic attributes," a review paper by Chopra and average of the amplitude spectrum for each sample and subMarfurt. The authors explicate the advancements that have tracting it from the peak amplitude. Applications of these taken place in seismic attributes during the last three years. concepts help delineate channels that have high amplitudes Among other things, astute application of robust filtering due to thin bed tuning. techniques, computation of attributes on frequencyVisualizing more than three or four seismic attributes at enhanced seismic data, the use of volume-based curvature once in order to interpret the underlying geology is a diffiattributes in preference to surface-based curvature, the use cult task. Visualization software is available to co-render of azimuth-based attributes, suitable 3D visualization of the three attributes--using a red-green-blue, or RGB, color attributes, and multiattribute color display schemes have all model--or four attributes, using a hue-lightness-saturationhelped delineate subtle geologic features. Aided by texture opacity (HLSO) color model. However, for multiattribute analysis, seismic geomorphology models, and automatic analysis, reducing data dimensionality is essential, which fault detection, seismic attributes provide immense infor- sparks the need to present views or projections of multidimation that allows the seismic interpreter to make accurate mensional data. Wallet and Marfurt, in their paper, "A grand predictions. The authors also forecast how seismic attribute tour of multispectral components­a tutorial," first review technology will advance in the future and what exciting different projection techniques, such as scatter plots and developments are waiting to be put into practice. other attribute visualization methods (parallel coordinates In the next paper, "Stratigraphically significant attrib- plots, Andrews' curve plots, Chernoff faces); linear projecutes," Hart identifies attributes that capture changes in tion methods (including principal component analysis); waveform shape and are due to changes in stratigraphy. The nonlinear projection methods (k-means and self-organized author opines that limiting the choice of attributes to only maps); and cluster analysis. The authors also discuss the those considered "physically significant" could impede the image grand tour (IGT), an interactive method of exploring relationship between stratigraphy and the seismic wave- high-dimensional data in a dynamic fashion, which is ideal form, which could be useful for predicting the three-dimen- for multispectral image data. Applying IGT to spectral sional distribution of some properties of interest. In an decomposition data is especially valuable when the anomexample of a hydrothermally-altered dolomite play, the alous features of interest are not stronger than the surauthor mentions that in addition to using the "physically rounding, higher-amplitude signal or noise. significant" attributes (like reflection strength, rms ampliFor accurate and comprehensive interpretation of seistude, perigram and the integrated trace), use of "strati- mic data, multiattribute analysis (as mentioned above) can graphically significant" attributes (like energy half-time, be formulated as a process of volume co-rendering using cosine of instantaneous phase, and derivative of reflection RGB color model. Henderson et al., in their paper,



"Delineation of geological elements from RGB color blending of seismic attribute volumes," show how RGB blending can be fruitfully used to co-render spectral decomposition analyses. In addition, the authors discuss the effects of scaling (such as scaling each input component independently), noise-attenuation filtering on RGB-blended volumes (including mean, vector median, and adaptive neighborhood mean filters), and opacity. The authors also discuss segmentation and connectivity analysis process for generating a set of geobodies. Such techniques are very helpful for developing a comprehensive color-image analysis workflow for seismic interpretation. In the next two papers, geometric attributes have been used in workflows that generate enhanced images for more accurate interpretation. In their paper, "Enhanced imaging workflow of seismic data from Chicontepec Basin, Mexico," Chavez-Perez and Vargos-Meleza demonstrate a two-step workflow application composed of migration deconvolution and geometric attributes that yields enhanced imaging, leading to identification of depositional features that were indistinguishable in the original migrated seismic images. While the migration deconvolution process improved spatial resolution of seismic data and toned down acquisition footprint, as well as noise and artifacts, coherence and positive volume curvature attributes helped define the presence of the turbidite channel geometry and architecture, an objective that was set for the study. Aktepe et al. in their paper, "Imaging of basement control of shallow deformation: application to Fort Worth Basin" demonstrate the application of a workflow consisting of velocity model building, imaging, and attribute calculation to better map features at and possibly within the basement. Pre-stack depth migration eliminated the velocity pull-up and push-down effects, focused the previously unfocused reflectors, and delineated structural highs and lows and faults. The most-positive and most-negative curvature attributes suggested significant faulting in the basement, and their patterns are consistent with the results reported for a study performed in an adjacent area. In "Seismic attribute analysis and geobody visualization changes our perception about a century old highly heterogeneous field," Mahapatra and Imof demonstrate the use of instantaneous attributes and geobody analysis to identify two prominent channel systems (hitherto unreported) within the reservoir at West Coalinga oil field in California. Instantaneous amplitude, perigram, instantaneous dominant frequency, response frequency, instantaneous phase, apparent polarity, instantaneous bandwidth, and coherency attributes are utilized for marking unconformities and stratigraphic boundaries, and delineating depositional and structural features. Finally, geobody visualization analysis with opacity was utilized to identify and delineate sand sequences, which were finally tied with well data for delineation. Mukerji and Mavko, in "The flaw of averages and the pitfalls of ignoring variability in attribute interpretations" propose the use of variability of rock properties in quantitative computations of attributes. When variability and nonlinearity are present, ignoring rock property variability can lead to distorted attribute interpretations. Attribute interpretation using averages and average trends does not yield any indication of uncertainty due to variability in the properties. In contrast, Monte Carlo simulations account for attribute distributions and thereby give confidence intervals, as well as a quantitative measure of uncertainty. Determination of net-sand thickness is an important task in hydrocarbon reserve calculation. The choice of the method

adopted for this determination depends on whether the gross-sand thickness is above or below the tuning thickness. Reitsch, in his paper, "A designer attribute for net-sand thickness estimation," describes the computation and use of a "probe" function to achieve this objective, irrespective of the thickness limitation. The inner product of this probe function and the seismic reflections from the target zone, with the correct scaling, yields the net-sand thickness. Seismic wave propagation through a porous homogeneous material causes negligible relative fluid movement, as Biot's poroelastic theory predicts. However, heterogeneous permeability and saturation in rocks have substantial effects on both velocity and attenuation. Besides this fluid-flow effect, effect of scattering has an important effect in the case of finely layered porous rocks and heterogeneous fluid saturation. Goloshubin et al., in "Relative permeability from seismic attribute analysis" use both fluid flow and scattering effects to derive a frequency-dependent seismic attribute, which is proportional to fluid mobility, and then apply it for analysis of reservoir permeability. In "The use of seismic attenuation to aid simultaneous impedance inversion in geophysical reservoir characterization," Singleton discusses a workflow that shows how attenuation estimation can be used in combination with simultaneous impedance inversion for prospect evaluation. Measurement of attenuation is carried out on logs (by generating attenuation logs), synthetic seismic gathers (using full-waveform synthetic gathers), and surface-seismic data (using log spectral ratio and frequency-shift methods). These measurements of attenuation are then calibrated and visualized volumetrically. In the real data example (from the Norwegian North Sea) used for illustrating this workflow, attenuation voxel values are draped on simultaneous inversion results. Seismic attenuation anomalies appear to match the inversion geobody configuration, and the two show similar log and synthetic responses, which suggests the two are likely to measure similar phenomena. This workflow suggests attenuation is a useful complement to impedance inversion for prospect evaluation. However, care needs to be exercised when it is used in such an application, as causes of attenuation anomalies could be different from hydrocarbon saturation only. Gao, in his paper, "Application of seismic texture model regression to seismic facies characterization and interpretation" demonstrates a different approach (from conventional multiattribute classification methods) to characterize seismic facies with reference to frontier, deep-marine depositional settings. This approach uses a workflow that includes construction of a texture element as a calibration model, retrieval of data to form a texture element with the same dimension and size as the model, performs a linear leastsquares regression between the amplitude samples and those in the model, and calculates regression gradient. Applying this workflow to real data shows how lateral seismic facies distribution and vertical stacking pattern can be visualized by slicing through the texture facies volume using a series of stratigraphic horizons. These papers represent the most up-to-date reports on seismic attributes, and we hope readers find the articles in this special section both interesting and informative. --SATINDER CHOPRA Arcis Corporation, Calgary, Canada --KURT J. MARFURT The University of Oklahoma, Norman, USA




2 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

Seismic Facies Identification and Classification Using Simple Statistics
The Generation of a Rock And Fluid Properties Volume via the Integration of Multiple Seismic Attributes and Log Data
Microsoft Word - First_Break_mrd_feb06.pdf.doc