Abstract Acquisition footprint often poses a major problem for 3D seismic data interpretation. Ideally, footprint from acquisition is handled at the processing shop through more careful attention to trace balancing statics, noise reduction, and velocity analysis ( Hill et al. , 1999 ; Gülünay, 2000 ). Such reprocessing is not feasible on many legacy data volumes where the prestack data cannot be found or no longer exists. Seismic attributes often provide an effective means of delineating subtle geological features of interest such as channels, small faults, and fractures but can also enhance acquisition footprint. For this reason attributes can be used to both design and evaluate the effectiveness of alternative footprint suppression workflows. In this work we review, apply, and evaluate the three most popular footprint suppression workflows: structure-oriented filtering, k x - k y filtering, and limited data reconstruction using singular value decomposition. We both characterize acquisition footprint and evaluate its suppression by its attribute response through application to legacy data volumes acquired over the shelf of the Gulf of Mexico, the Central Basin Platform of Texas, the Delaware Basin of New Mexico, and the Anadarko Basin of Oklahoma. There is no silver bullet; in most cases it is useful to combine structure-oriented filtering and k x - k y filtering workflows as an attempt to remove acquisition footprint from legacy data volumes.

Abstract The demands that reservoir characterization places on seismic data far outweigh those of traditional structural interpretation. Because of this, gather conditioning is seen by many as a prerequisite to prestack inversion. This paper discusses three conditioning processes—signal/noise (S/N) improvement, stretch removal, and reflector alignment. It then seeks to document the improvements that these processes achieve in the gathers and in prestack inversion. Specifically, the gathers were measured for AVO fit using a 2-term Shuey equation and found to be improved by 20%. A comparison of wavelets extracted from each angle stack found the high frequency limit of stable phase to have increased from 30 Hz to 50 Hz. The far angle stack seismic/synthetic inversion residuals showed a 43% drop in amplitude and completely different frequency and reflector character following gather conditioning. Finally, the acoustic impedance (AI) vs. shear impedance (SI) cross-plot showed a much more compact signature that allowed more definitive lithology and pay discrimination. Conversely, the raw data cross-plot contained noisy data that erroneously entered into the area of the cross-plot where the pay signature lay. Geobodies captured from improperly conditioned data are thus (1) inflated in size by 62%, and (2) have lower impedances than is justified from well control. These errors, in turn, would lead to incorrect rock property (hydrocarbon saturation and porosity estimation) and reserve estimations.

Abstract We present a work flow for poststack seismic signal preconditioning that comprises noise attenuation and resolution enhancement steps; multiple algorithms have been adopted and their relative advantages and limitations are discussed. Specifically, we evaluate the benefit of two noise reduction approaches (LUM and SVD filters) and two resolution enhancement techniques (adaptive whitening and implicit sparse deconvolution). Such preconditioning work flows are typically used to prepare seismic data prior to detailed reservoir interpretation and characterization studies: as an example the Hewett Plattendolomite case is presented. Given the complex reflectivity of the Zechstein group (a cyclic sequence of limestones, dolostones, salt and anhydrites) and the limited thickness of the reservoir (close to seismic resolution), seismic characterization is notoriously problematic. The adoption of a resolution-enhanced seismic volume, supported by forward synthetic modeling and well data, allowed for a semi-quantitative seismic interpretation of the Plattendolomite reservoir and the delineation of its geometry and internal structure that had not been possible to recognize on the original seismic. Bandwidth-enhanced data also permitted a quantitative calibration of several seismic indicators (amplitude dimming, apparent thickness variations, and wave form), to produce porosity and net thickness maps of the new reservoir.

Abstract Seismic reflection data can be displayed to resemble illuminated apparent topography colored by amplitude or other data. In this way, seismic data can be made to look like geology. This fosters geologic intuition and aids interpretation by readily relating geophysical or stratigraphic information to the geological structure. With modern coherency filters, structural and stratigraphic seismic attributes, 3D visualization, and volume blending, it is straightforward to make such displays. Illumination is a key component of the work flow and is introduced through bump mapping or directional attributes. Two important directional attributes are relative amplitude change and seismic shaded relief. Shaded relief is especially powerful and reveals a wider variety of structural elements than other structural attributes.

Abstract 3D seismic attributes enable seismic interpreters to gain a more complete understanding of subsurface geology resulting in more complete and detailed interpretations. Building a thorough understanding of 3D structural variations and fault networks often requires working with multiple seismic attributes due to the fact that different attributes convey different information and that the seismic signature of faults changes through the data set. Color blending techniques have proven effective in intuitively allowing interpretation of information in multiple seismic attributes simultaneously. One of the most successful techniques uses an RGB (Red, Green, and Blue) color model to present data in a manner which is in tune with the way people perceive color. These types of blend are highly effective at visualizing data such as results of poststack frequency decomposition or offset-stack volumes. We present an alternate color-blending model based on combining attributes using the subtractive primary colors cyan, magenta, and yellow (CMY). When used with structural attributes, the subtractive model produces displays that are predominately light, and structural variations and faults are associated with darker shades of varying hues and black, and the model is aligned with the way we are accustomed to visualize fault and structural attributes, making these displays very intuitive. In this paper we provide a number of examples of how these blends can be used to show how fault character changes laterally through a fault network and relate individual faults to surrounding damage, drag zones, and areas of high fault density.

Abstract Software developed for the interpretation and analysis of geoscientific data produces no shortage of attributes designed to measure one salient factor; but as the number of attributes that can be calculated grows, the tools for visualizing these attributes must grow as well. This paper explores some early applications of multivariate visualization tools to interpreting faults, stratigraphy, and microseismic events. The specific tools we examine here are multidimensional transfer functions for volume rendering and glyph displays for multivariate visualization. We present two applications of multidimensional transfer functions: one example combines information from a fault probability volume with a fault azimuth volume to display fault trends very early in the fault interpretation work flow; the second example shows improvements to the definition of a channel boundary by combining an eigenvalue volume with raw amplitudes. Multivariate visualization is presented in the context of visualizing eigenstructure around an interpreted channel and the simultaneous display of several attributes of microseismic events. Although the tools and techniques discussed here are not entirely new, their application in the geosciences is rare, and has the potential to reveal more information to interpreters about their data.

Abstract The interpretation of depositional systems is challenging because the effects of geologic structure obscure the depositional systems in 3D seismic volumes. Even if a combination of attributes can be shown to highlight particular stratigraphy in a volume, it remains difficult to interpret due to the effects of structure (such as faulting, folding, and rotation) obscuring the stratigraphic features of interest. Domain Transformation™ removes arbitrarily complicated structure from a seismic volume, It is an interpretation-guided 3D transform that creates a volume consisting entirely of stratal-slices. The structural effects removed by the Domain Transform include differential sedimentation/compaction, folds, faults (3D displacement), unconformities (including angular unconformities), canyons, salt bodies, and carbonate buildups. The transformed volume is ideal for imaging and interpreting depositional systems because every horizontal slice more closely represents a paleodepositional surface. Depositional systems are more readily recognized from their morphology on these slices. Other stratigraphic details are restored to the approximate position and relationships that they had at the time of deposition. The use of combinations of attributes facilitates the interpretation of depositional systems and depositional stratigraphy in domain-transformed seismic volumes. “Structural attributes ( e.g. , coherence-related attributes, curvature, and horizon orientation) may be combined with other attributes derived from structure tensor analysis and with waveform related attributes ( e.g. , amplitude and frequency). Corendering these attributes provides key insights regarding the depositional systems and their internal characteristics.

Abstract Recent increases in the memory and power of graphics processing units now allow the rapid interactive computation and visualization use of 3D volumetric attributes. We show how interactive lighting and opacity using this new technology applied to traditional seismic attributes aids in the understanding of the faulting and slumping of the Barnett Shale over sink holes. The exploitation of unconventional shale reservoirs has heightened interest in the location of natural fractures and faults. The public is concerned about the leakage of reservoir and the liquids used for hydrofracturing into aquifers. In addition to aquifers, the operators are concerned about the loss of hydrofracturing fluids to anywhere outside of the reservoir. The example shows collapse features and faults may be seen with greater clarity for improved interpretation.

Abstract We present the results of multiattribute imaging and geobody delineation applied to stratigraphic targets such as Jurassic channels and Triassic beaches and spits, imaged in data from the Norwegian sector of the North Sea. Interpretation based on the examination of seismic amplitude alone is challenging due to the complexity and subtleness of these features. To improve the definition of these Mesozoic targets, we have applied a multiattribute approach, combining frequency decomposition, seismic attribute analysis techniques, advanced visualization, and a new method of multiattribute geobody delineation. Attributes have been selected that are sensitive to the edge and magnitude response of sedimentary structures, while the use of narrow band spectral magnitude volumes allows small scale frequency variations to be analyzed. These different sources are corendered using advanced color and opacity blending, providing multiattribute composite image volumes for subsequent interpretation and as input to further geobody delineation. The use of such advanced visualization has resulted in a collection of 3D volumes that successfully distinguish the internal and overbank geometry of channels as well as the structure and extent of Triassic sandbars. A new geobody delineation system has been designed to track visible structures in color blended images. The method is semi-automatic, allowing the interpreter to interactively guide the delineation process. Application of this technique has allowed the user to isolate and extract individual Jurassic channels, the seismic response of which varies considerably, as independent 3D geobodies. We believe that the use of such advanced multiattribute visualization and delineation techniques is applicable to similar provinces globally.

Abstract Spectral decomposition is a standard tool to facilitate and accelerate seismic interpretation. Applications include highlighting changes in layer thicknesses (for example, in meandering channels and turbidite layers) as well as exploring for low-frequency gas shadows. We argue that local phase analysis serves as a complementary aid in seismic interpretation because the layer thickness, type of impedance contrast, and boundary shape determine the amplitude, peak frequency, and phase of the locally observed wavelet.

Abstract Dense reflection seismic data have become an indispensable source of spatially continuous subsurface information when supported by a correctly calibrated earth model. This has been made possible thanks to the use of multiple attributes computed from 3D seismic data combined with visualization tools enhancing both structural and stratigraphic delineation. In this study, we show the use of conventional and derived seismic attributes for a particular interval in Aderic Field, located in a deep-water offshore turbidite environment. The upper Miocene-lower Pliocene interval in Aderic Field corresponds to turbidite channel complexes (both erosive-constructive channels and depositional channels), avulsion lobes, ponded lobes structures and regional mass transport complexes (MTC). Using an a priori sedimentological concept combined with basic knowledge of rock physics, one can infer detailed architectural elements from different seismic attributes where borehole information is not available. We found that combinations of amplitude maps at different angle stacks, autocorrelation and coherency of seismic traces, and spectral decomposition of the seismic data were valuable in analyzing the interval. Such analysis provides significant information on architectural element delineation for different channel complexes and sequences in terms of morphology and width. Levee systems, the erosive character of channels, and the morphologies of lobe systems with avulsion or ponded aspects can also be characterized. In some cases, an estimation of internal heterogeneities can also be proposed. Interpretations derived from these attributes compare favorably with possible analogues in similar depositional environment, as well as with field outcrop analogues and with images from very shallow Holocene-age seismic volumes.

Abstract Seismic amplitude and instantaneous attributes as well as stratigraphic interpretation of these attributes are frequency and scale dependent. Frequency dependence offers a new dimension of seismic data that has not been fully utilized in the study of seismic picking of geologic surfaces, seismic facies, seismic geomorphology, and sequence stratigraphy. Seismic-tuning effects include thickness tuning and frequency tuning. Seismic modeling shows that, whereas thickness tuning determines seismic interference patterns and, therefore, occurrence of seismic events and seismic facies in stacked data, frequency tuning may further influence the nature of seismic-geologic time correlation and modify seismic facies. Frequency-dependent data processing and interpretation are practical ways to study the frequency-tuning effect for improved seismic interpretation and well-seismic integration. Field data examples demonstrate that by using a local geologic model and well data, we can optimize the frequency components of seismic data to a certain degree and intentionally modify seismic interference patterns and seismic facies for better seismic interpretation of geologic surfaces, facies, geomorphology, and sequence stratigraphy.

Abstract This paper presents a case study illustrating the successful utilization of the instantaneous amplitude attribute generated from the 3D seismic data in Ada Field, North Louisiana, in delineating channel-body geometry, inferring the depositional environment, and in prospecting, risk reduction, and optimization of infill well location in the Lower Cretaceous Hosston (Travis Peak in Texas) Formation, a prolific, tight-gas sandstone.