Processing and Interpretation of Three-Component Borehole/Surface Seismic Data over Gabor Gas Storage Field

Analysis of a seismic dataset recorded as part of a collaborative project between Wright State University, Spectraseis, Precision Geophysical, and Dominion East Ohio over the Gabor gas storage field, Canton, Ohio, is the topic of this study. Two types of sources (vibroseis and small seismic shot-holes) as well as three types of recording systems (passive 3C broadband on the surface, experimental 3C borehole sondes, and conventional 2D surface geophone profiles) were employed with different purposes. The shot hole explosives were calibration check-shots for the 3C borehole array repeatedly deployed at multiple levels. The vibrators of the conventional 2D seismic profiles were also recorded with the borehole array, a series of 3C surface seismometers along the 2D seismic lines and a spiral array of 3C seismometers centered at the wellhead. Processed and migrated seismic data from the pair of 2D surface seismic lines crossing directly over the well location were provided by Tom McGovern of Seismic Earth Resources Technology. The particular aims of this project are to determine wave velocities from first breaks, to apply VSP processing procedures on the borehole data, and to analyze spectral attributes in the low frequency range from a beat-sweep test. This VSP study was attempted despite the fact that the data were not collected for that purpose, and unfortunately the analysis of the borehole data failed to show expected subsurface reflectors. The raw borehole records have many characteristics unique to the borehole environment, which include effects of poor geophone clamping, bad cementation and tube waves. A frequency filter combined with trace mutes was very effective in enhancing wanted events as well as improving wave shapes. Time shifts between successive traces were obtained through cross-correlation. The true wave velocities were determined based on a single layer model, and were compared to the stacking velocities obtained from surface-recorded seismic data. Several VSP processing procedures were applied attempting to track up-going reflections from the borehole records, including static time shifting, FK filtering, NMO correction, and trace stacking. The stacked VSP trace was tied to surface seismic section, however, showed poor correlation in terms of subsurface horizons. Data from a beat-sweep test was also analyzed with the hope of verifying a low frequency spectral anomaly over the gas field. During the beat sweep tests and for minutes afterward the phenomenon of distinctly increased amplitude at 3Hz was present in both surface and borehole measurements. This beat-sweep test survey provided a tantalizing first look at a way of generating low frequency seismic energy to examine the low frequency anomaly over hydrocarbon reservoirs, and in this case successfully verified the production of a 3Hz anomaly