--> Key Technologies for Processing of Seismic Data in Gas Cloud Area

AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Key Technologies for Processing of Seismic Data in Gas Cloud Area

Abstract

There are great difficulties in structure imaging within a gas cloud area due to serious energy attenuation and the velocity lateral variation. Gas cloud imaging can be improved by using a multi-component data, or by using a special seismic acquisition geometry to avoid propagation of seismic wave through gas cloud. By optimizing conventional P-wave processing routine and combination of various specific techniques, this paper demonstrates how to reduce the effects of gas cloud and improve structure imaging. In the proposed workflow, matching pursuit Fourier interpolation (MPFI) is applied. By using anti-leakage Fourier transform, it reconstructs near-offset data, reduces acquisition footprints and then improves the quality of pre-stack data. Besides, MPFI avoids spatial aliasing on account of the weighted high frequency data with respect to prior. The workflow groups several methods to attenuate different kinds of multiples. First, single streamer deghosting is applied to attenuate ghost waves and enhance low frequency energy. Then, deterministic water-layer demultiple and general surface multiple prediction are applied to remove seabed-related multiples and long-range multiples respectively. A key step of seismic processing in gas cloud area is energy compensation. This workflow uses Q tomography to compensate energy during depth migration, which is a specific energy compensation technique for gas cloud area. It estimates a space- and depth-variant Q field by using attenuated traveltime tomography. Then, model-driven Q compensation is realized directly during depth migration. Conventional reflection tomography can't build an accurate velocity model, where lateral variation of velocity is severe and quality of seismic data is poor. To address this problem, Diving-wave refraction tomography is applied to generate a relative accurate initial velocity model in shallow. Then, a global velocity model is generated iteratively by reflection tomographic. The low-velocity area of final velocity model matches very well with the low energy area of seismic data, and the model shows a similar trend with acoustic velocity. Processing of a real data set in gas cloud area from X oilfield, Bohai bay, proves the advantages of the proposed workflow, which improves structure imaging inside the gas cloud area and is prominently superior in faults detection.