Bingjie Cheng

Multi-scale frequency and absorption attributes of seismic signals

Seismic absorption is sensitive to gas saturation, and this can lead to a number of observable features, such as low dominant frequency, absence of high frequencies and dimming of reflections. We obtain the time–frequency distribution of seismic signals at different scales based on wavelet decomposition and try to analyse the response of high-frequency sections for the characterization of hydrocarbon reservoirs. We present a case study using multi-scale (dominant) frequency and multi-scale absorption, and demonstrate that the method is effective in recognizing a deep tight-sandstone gas reservoir. By this study, we can draw the conclusion that multi-scale frequency and absorption attributes may indicate the gassy areas in reservoirs, and we have obtained many good applications in the Sichuan basin of China.

Study and application of PS-wave pre-stack migration in HTI media and an anisotropic correction method

Anisotropy correction is necessary during the processing of converted PSwave seismic data to achieve accurate structural imaging, reservoir prediction, and fracture detection. To effectively eliminate the adverse effects of S-wave splitting and to improve PSwave imaging quality, we tested methods for pre-stack migration imaging and anisotropic correction of PS-wave data. We based this on the propagation rules of seismic waves in a horizontal transverse isotropy medium, which is a fractured medium model that reflects likely subsurface conditions in the field. We used the radial (R) and transverse (T) components of PS-wave data to separate the fast and slow S-wave components, after which their propagation moveout was effectively extracted. Meanwhile, corrections for the energies and propagation moveouts of the R and T components were implemented using mathematical rotation. The PS-wave imaging quality was distinctly improved, and we demonstrated the reliability of our methods through numerical simulations. Applying our methods to three-dimensional and three-component seismic field data from the Xinchang-Hexingchang region of the Western Sichuan Depression in China, we obtained high-quality seismic imaging with continuous reflection wave groups, distinct structural features, and specific stratigraphic contact relationships. This study provides an effective and reliable approach for data processing that will improve the exploration of complex, hidden lithologic gas reservoirs.

A study and application of high-resolution methods for reef reservoir identification

Reefs represent a special type of carbonate trap that plays a key role in the migration, accumulation, and formation of a reservoir. They have commonly been the targets of exploration and development. However, reefs have complex interior structures and easily grow as thin, interbedded geological frames with reef microfacies that include the cap, core, and base of the reef. Because of the inherent drawbacks of seismic signals, including their low frequencies and narrow bandwidths, it is difficult to accurately identify reef reservoirs. Fortunately, the seismic frequency, phase, energy, waveform and other dynamic and geometrical properties can be used to compensate for the energy, expand the frequency bandwidth, and decompose and reconstruct the wavelet to obtain high-resolution seismic data. These data can highlight certain seismic responses of reefs, including boundary reflections, dome-shaped reflections from the reef outline, strong reflections from the reef cap, reflections from the reef bottom, and onlap reflections from the reef flanks. Some impedance response regularities, such as the lower impedance of the reef cap relative to the reef core and biodetritus beach and the fluctuating impedance of the reef-flat complex, are observed by combining log data with geological and high-resolution seismic data for a reef reservoir inversion. These methods were applied to the Changxing Formation in the Yuanba Gas Field. Good prediction results were obtained with a high consistency between the log and seismic data in a comparative analysis with the original seismic data and well logs.