Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
Wide-azimuth Techniques For Processing High Density 3D OBC Data
Boelle, Jean Luc (TOTAL Exploration-Production) | Hugonnet, Pierre (CGGVeritas) | Navion, Sylvain (CGGVeritas) | Adler, Frank (TOTAL Exploration-Production) | Bluteau, Julien (TOTAL Exploration-Production) | Soudani, Amine (TOTAL Exploration-Production)
Summary High density wide-azimuth Ocean Bottom Cable acquisitions have proven their efficiency to better image complex structures in the North Sea. Yet conventional preprocessing sequences have not taken into account the full potential offered by such acquisitions (Boelle et al. 2008). In this paper, we present the advantages of a true wideazimuthal pre-processing regarding noise attenuation, PZ summation in the 3D ?-px-py domain, following Soudani et al. (2006), and multiple attenuation in the 3D ? qx-qy parabolic radon domain as proposed by Hugonnet et al. (2008). The wide-azimuthal information can also be used to build the velocity model following a multi-azimuth strategy. The wide azimuth processing results are compared to more conventional processing approaches in a real case study. Introduction Recent Wide Azimuth Towed Streamer (WATS) acquisitions have pointed out the importance of a wide azimuth illumination to improve the imaging of complex structures. Yet wide-azimuth designs were already used in on-shore or OBC acquisitions, and in a previous paper Boelle et al. (2008) demonstrated the advantages of following a true 3D wide-azimuth flow in the preprocessing of a high density 3D OBC campaign. In this previous work the authors have given a particular emphasis to the 3D noise attenuation in the ?-pxpy linear radon domain providing a significant improvement on final images. In the present contribution new tracks are proposed such as the 3D PZ summation in the t-pxpy domain, following Soudani et al. (2006), and the multiple attenuation using a 3D parabolic radon decomposition in the ?-qxqy domain as proposed by Hugonnet et al. (2008). The OBC acquisition shot over the Hild field (Norwegian North Sea) is used to illustrate these wide-azimuth preprocessing techniques. This work deals with the processing of only the hydrophone and vertical geophone components. The 3D OBC seismic acquisition over the Hild field A wide-azimuth 3D survey using Ocean Bottom Cables was shot in 2005. The nominal recording swath geometry consists of 4 receiver lines laid on the sea-floor 400m apart, the 4Component receiver spacing is 25m. Shot sail lines have been navigated on top of cables in a dual source flipflop sequence, in such a way that each receiver records data from a dense long offset and full azimuth 50m by 50m grid of shot points with a maximum offset of 6km. A more detailed description of the acquisition geometry is given by Vaxelaire et al. (2007). Such a design yields common receiver gathers with nearly 36000 traces, with a density which enables true 3D wide-azimuth pre-processing. As the direct wave is spatially aliased above 15Hz for a 50m spatial sampling interval, a double interpolation should be undertaken in each direction in order to process the data up to 40Hz. This leads to multiply the amount of data by 16 for each gather which was not acceptable in terms of run-time and data storage, then a two-step process was designed.
- Europe > Norway > North Sea > Northern North Sea > North Viking Graben > PL 043 > Block 30/7 > Martin Linge Field > Tarbert Formation (0.99)
- Europe > Norway > North Sea > Northern North Sea > North Viking Graben > PL 043 > Block 30/7 > Martin Linge Field > Ness Formation (0.99)
- Europe > Norway > North Sea > Northern North Sea > North Viking Graben > PL 043 > Block 30/7 > Martin Linge Field > Lunde Formation (0.99)
- (29 more...)
High Resolution 3D Parabolic Radon Filtering
Hugonnet, Pierre (CGGVeritas) | Boelle, Jean Luc (Total E&P) | Mihoub, Majda (CGGVeritas)
Summary 2D parabolic Radon filtering is a widely used method for multiple attenuation. However, for dense and wideazimuth gathers that have azimuthal anisotropy effects this approach can have problems. Because of the variation of the curvature of the events with azimuth, the bin gathers can not be processed in one go but must rather be split into sub-collections where the azimuth either has little variation or varies smoothly. We instead propose herein to take into account the azimuthal anisotropy by incorporating an elliptical model for the variations of the curvature with azimuth, to define a 3D parabolic Radon filtering. This is a more natural way of processing dense wide-azimuth gathers by honoring their actual 3D geometry. Introduction Parabolic Radon filtering has for a long time been one of the most commonly used anti-multiple processing methods. It belongs to the family of algorithms based on velocity discrimination between the primaries and multiples. The CMP gathers after NMO are modelled by a superposition of constant amplitude parabolas and the most curved parabolas, assumed to be the multiples (slower than the primaries), are retained and subtracted from the data. Thorson and Clearbout (1985) introduced the hyperbolic Radon inversion, including high resolution (HR) features that dramatically improve the capability of the method to process spatially aliased data and to preserve the primaries. Then Hampson (1986) introduced the parabolic Radon version, and took advantage of the time-invariance property of the parabolas to achieve a fast implementation, though not HR. Fast HR implementations were finally achieved by Sacchi and Porsani (1999) and Herrmann et al. (2000). All the existing implementations are 2D, time-offset. This is perfectly suited to 2D or 3D narrow azimuth (NAZ) data, or to any geometry where the azimuth varies smoothly with offset. However, problems arise with the more recent dense and wide-azimuth (WAZ) geometries, where some traces can have similar offsets but very different azimuths. If there is any azimuthal anisotropy effect (variation of the apparent velocity and hence of the curvature with the azimuth), either because of structural (dips) or intrinsic reasons, the arrival times vary randomly from trace to trace when sorted by offset and the effectiveness of the algorithm is degraded. Practical solutions typically consist in applying the 2D parabolic Radon filtering on azimuth sectors, or on suitable pseudo-2D sub-collections, but this is still often not fully satisfactory. 3D parabolic Radon In a WAZ gather, we consider the offset vector (x,y) instead of the scalar offset. The azimuthal variations of the curvature are handled by an elliptical model (fig. 1), which leads to squeezed and tilted paraboloids. Note that while with the 2D case we go from a 2D data space (t,x) to a 2D model space (?,q), with the 3D case we go from a 3D data space (t,x,y) to a 4D model space (?,qx,qy,qxy): the model space has one extra dimension. This can result in dramatically increased runtimes (to scan the whole model space) and in strongly ill-conditioned problems (many more model parameters than data samples).