Analyzing the effectiveness of receiver arrays for multicomponent seismic exploration

Abstract

This paper uses an experimental seismic line recorded with three‐component (3C) receivers to develop a case history demonstrating very little benefit from receiver arrays as compared to point receivers. Two common array designs are tested; they are detrimental to the P‐S wavefield and provide little additional benefit for P‐P data. The seismic data are a 3C 2‐D line recorded at closely spaced (2 m) point receivers over the Blackfoot oil field, Alberta. The 3C receiver arrays are constructed by summing five (one group interval) and ten (two group intervals) point receivers. The shorter array emphasizes signal preservation while the longer array places priority on noise rejection. The effectiveness of the arrays versus the single geophones is compared in both the t−x and f−k domains of common source gathers. The quality of poststack data is also compared by analyzing the f−x spectra for signal bandwidth on both the vertical receiver component (P‐P) and radial receiver component (P‐S) structure stacks produced using these two array design philosophies. The prestack analysis shows that the two arrays effectively suppress coherent noise on both the vertical and radial geophone data and perform reasonably as spatial antialias filters. The poststack analysis reveals that, for both the P‐P and P‐S data, neither of the two arrays significantly improves the quality of the final seismic image over that obtained from point receiver data. For the P‐P data there are subtle differences between the final stacked sections, while for the P‐S data there is a significant deterioration in image quality from the application of the arrays. This P‐S image deterioration is attributed to significant variation of shear‐wave statics across the array. For this specific survey area and acquisition parameters, 3C receiver arrays are unnecessary for P‐P data and are detrimental to P‐S data.