Search

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.

Various far-offset marine acquisition geometries have been widely deemed as promising for alleviating imaging difficulties in deepwater Gulf of Mexico (GOM). However, incorporating very far offsets into subsalt imaging remains challenging. In this case study in the GOM, we explore the benefits of far offsets using ocean bottom node (OBN) data with acquisition offsets up to 25 km and investigate the untapped potential of extremely far offsets of up to 50 km. Using far offsets above 12 km, and up to 25 km was not straightforward, but with some help, they significantly improved the subsalt images and were valuable for building the salt model in this study. Imaging benefits from extremely far offsets up to 50 km were also observed but usefulness for salt model building might be limited.

Variation of sound wave velocity in water (water velocity) and node positioning errors can cause strong 4D noise in ocean bottom node (OBN) time lapse processing if they are not addressed correctly. Here we proposed a method to jointly invert the water velocity and the node position from the recorded direct arrival time in the data. We then validated this method by synthetic and real OBN data examples from deep water Gulf of Mexico.

A workflow of azimuthal inversion is presented and successfully applied on a wide-azimuth seismic data. This workflow is based on a novel approach to quantitatively extract anisotropy information in a horizontal transverse isotropic (HTI-) medium. The method was first described by Mesdag (2016) and mimics anisotropic reflectivity behavior by isotropic forward modelling and anisotropy-transformed elastic properties. In this paper a concept of an Azimuthally varying Low Frequency Model (ALFM) is also introduced and demonstrated. The ALFM provides anisotropy information below seismic bandwidth, hereby reducing side-lobes of inverted magnitude of anisotropy, thus allowing for better anisotropy characterization.

Relatively large spectral differences exist between streamer and ocean bottom seismometer (OBS) data, mostly due to their different ghost effects – streamer data have both shot- and receiver-side ghost, while OBS data only have shot-side ghost. In a recent OBS-streamer 4D study in deepwater Gulf of Mexico, we investigated three schemes of spectral matching between OBS and streamer data: conventional 1D matching of streamer to OBS, receiver deghosting on streamer data only, and full deghosting on both surveys. We find that receiver deghosting on streamer data is better than 1D matching, providing a better match between the streamer and OBS data before migration and an increased 4D signal-to-noise ratio (S/N) after migration. However, we also found that source deghosting on both surveys does not improve 4D results; instead the spectra of streamer and OBS, especially at lower frequencies, are more different after source deghosting. Receiver deghosting alone on the streamer data gives the best 4D results among the three spectral matching schemes.

A method is described to allow quantitative usage of isotropic modeling and inversion in anisotropic media. Based on the Rüger reflectivity equations for HTI media transforms are designed for the elastic parameters used in pre-stack inversion. The transformed elastic parameters can be used in isotropic forward modeling and inversion to exactly mimic the anisotropic reflectivity behavior of the seismic data. The proposed method allows us to close the loop between well log data and seismic inversion results without having to revert to experimental anisotropic inversion algorithms. In this paper a synthetic feasibility study will be shown, indicating that we can exactly predict the outcome of a wide azimuth (WAZ) seismic inversion experiment based on isotropic elastic parameters and Thomsen’s anisotropy parameters.