Search

NVGS, a recently acquired southward extension of the NVG survey in the northern Norwegian North Sea, is a high-quality broadband seismic data set, which reveals detailed stratigraphy from the Permian Salt to the Quaternary section for stratigraphic and lithological interpretation. In the deeper parts, old basin configurations can be mapped. Including Caledonian fold and faults and Paleozoic basins. The regional scale of the data set makes it ideal for establishing a geological model for this region. Including prediction of potential source and reservoir rocks. Although the North Sea is a mature region, the Patch Bank Ridge area is under-explored.

The characteristics of lithologic reservoirs, such as complex channels with distinctly lateral variations and variable thickness, make such reservoir features difficult to identify when using conventional geostatistical methods or applying seismic attribute methods. To decrease the uncertainty and improve the lateral distribution of the predicted map over such features, we present an improved cokriging system which combines well logs and multiple attributes directly, instead of using a data fusion method. The improved technique was applied to predict a distribution and thickness map of a channel system. The results demonstrate that improved our cokriging system can enhance the lateral resolution of the channel and reduce the uncertainty of prediction due to the use of more seismic attributes than traditionally used.

Free-surface datum variations cause complications for hydrophone only receiver deghosting algorithms that assume a horizontal free-surface. We describe a multi-shot receiver deghosting approach that derives a tau-psht-prec model of ghost free data at mean free-surface datum. The model is derived so as to simultaneously satisfy the recorded primary and ghost wavefields in the presence of a non-horizontal free-surface at the source and receiver sides. The approach is shown to produce high resolution receiver deghosting results with improved spatial consistency on a broadband variable depth streamer dataset from the North Sea.

In this paper, we develop Dynamic-warping Full Waveform Inversion (D-FWI) to address cycle skipping problem which is a well-known challenging issue in conventional FWI. The dynamic warping technique is used to detect the travel time difference between the predicted and the observed wave fields. We make use of the time shift to partially warp the observed data and thus generate a series of wave fields that connect the predicted and the observed wave fields. We then use these modified observed wave fields in the framework of FWI and solve a sequence of conventional FWIs to avoid cycle skipping issue. The applications on synthetic and real data examples show that D-FWI can converge successfully by overcoming the cycle skipping problem while conventional FWI result in a spurious model. With this new approach, we can get reasonable velocity model even starting with a smooth model at higher frequency. It will greatly save the processing duration since velocity model building can be done at the same time of pre-processing such as denoise, demultiple and interpolation.

Multiple attenuation is a key step of broadband marine processing which could be challenging for primary preservation, spatial continuities and low frequencies. Rather than a cascaded de-multiple sequence, which could lead to cumulate mistakes and uncertainties, with final results potentially questionable for further pre-stack inversion, we propose an optimized multiple attenuation methodology, as well as new Quality Control tools based on correlations and pseudo-Impedance.

We propose a 3D deblending method for towed streamer simultaneous source data based on an L1 inversion algorithm. It pursues the sparse representation of 3D coherent signals of the various sources that match the blended data in a combined transform domain of 2D TauP and 2D directional wavelet. In this method 2D sparse TauP is first applied to the data along consecutive channels for each shot, and then an L1 inversion algorithm based on 2D high angular resolution complex wavelet transform (HARCWT) is used to deblend the signals of different sources for each common P gather. Since the TauP transform along channels has good separation of the events in common shot gathers according to their slopes, and the coherent signals across shots have sparse representation in HARCWT domain, the method can achieve higher quality deblending results than 2D common channel deblending. The method was tested on numerically blended real field data, and the result was better than that of 2D deblending.

Vertical Seismic Profile (VSP) surveys rely on three-component (3C) geophones to acquire high-resolution data at target reservoirs. These 3C geophones are comprised of three independent receivers, mounted orthogonally. When inside the borehole, the orientation of each 3C geophone is unknown. To enhance image stacking power from different receivers, it is necessary to reorient all the 3C VSP receivers to a common coordinate system. We introduce a VSP coordinate reorientation workflow using elastic finite-difference modeling. The only condition required is an adequate knowledge of the overburden velocity. Since VSPs today are typically acquired to supplement existing surface seismic images, adequate velocity models already exist and this method can almost always be applied effectively. We conduct synthetic tests to demonstrate the robustness of our workflow in a variety of noise levels, velocity errors, and acquisition coverages. We also show a real data example from the deep water Gulf of Mexico (GoM).

Analyzing seismic amplitude trends requires stability of the seismic wavelet in both amplitude and phase. Unfortunately, in land acquisition, the filter applied as energy travels through the near-surface can have a dramatic impact on both. Here, we will extend previous work on proposing detailed QC methods for surface-consistent deconvolution of land dynamite data to land vibrator data. In the process, we uncover and attenuate an apparent acquisition abnormality, producing seismic data more amenable to sensitive post-migration analysis and inversion.

The presence of orthorhombic anisotropy poses serious challenges in multi azimuth (MAZ) or wide azimuth (WAZ) data imaging. Orthorhombic anisotropy makes seismic velocity vary with azimuthal direction as well as polar direction. The polar direction dependency can cause well misties and higher order moveout. On the other hand, the azimuthal dependency can cause noticeable moveout fluctuation between different acquisition directions and hence prevent constructive summation of WAZ images, especially for the fault imaging. We have presented a ray-based tomographic inversion methodology for WAZ data model building in the presence of orthorhombic anisotropy in last year SEG (Zhou et al. 2015). However, in geologically complex areas, such as in the presence of complex faults or shallow gas clouds, the effectiveness of ray-based tomographic inversion is limited in providing high resolution velocity model, which is needed to optimize the final image of the complex structures. In this paper, we developed an orthorhombic full waveform inversion approach for building high resolution velocity model. We demonstrate that our method can effectively reconstruct high resolution orthorhombic model which not only fits with geology well but also significantly improves the focusing of the fault imaging with the WAZ OBC data. The combined effect of these improvements is a step-change in the final seismic image quality.