T his second part of an article about a large 3D VSP survey in Abu Dhabi describes the interpretation effort which quantifies the value that a 3D VSP seismic image can bring when supplementing even a 640-fold, highresolution surface seismic volume.It is understood that for recovery to be optimized and bypassed resources to be minimized, especially in later stages of field production, more accurate models of a reservoir's architecture and characteristics are needed. This first 3D VSP survey in Abu Dhabi characterized details of the reservoir that could not be derived from surface data or well-log data alone. The higher-quality, higher-resolution images made it possible to map detailed stratigraphy and important but previously unknown faults. The improved structural map and updated geologic model were verified by wells drilled inside the 3D VSP image areas.
The effect of receiver array length and source effort on VSP qualityTo better understand the value in acquiring 3D VSP data with long borehole receiver arrays, processing tests using a conventional 12-level receiver configuration were conducted by using a subset of the 126-level VSP data. Using the same statics, velocity model, and other relevant processing parameters developed for the 126-level array data, a 3D VSP image was produced with data from only 12 geophone levels. Figure 1 compares VSP common depth point (VCDP) gathers at different offsets between the 12-and 126-level data. At an offset of 400 m, both the 12-and 126-level gathers show well organized energy from primary events. Good VSP images up to 400 m away from the well should be possible with both data sets. At the longer offset of 700 m, there does not appear to be any indication of primary events on the 12-level gather. However, due to higher fold, primary events are clearly visible on the 126-level gather. This result suggests the reason; in 12-level walkaway VSPs, it is difficult to image distances greater than 500 m from the wellbore, even though offsets up to 4 km