An interpretive study of multicomponent seismic data from the Carrot Creek area of West-central Alberta

Brad Nazar

In 1987, two multicomponent seismic lines were acquired in the Carrot Creek area of west-central Alberta. Analyses of the final processed seismic sections (P-P and P-SV) showed that the interpretation of P-SV (radial-component) data can assist greatly in delineating geologic units, particularly the Cardium conglomerates of the Carrot Creek area. At the location of the conglomerate deposits, for instance, strong amplitude anomalies can be observed with the P-SV data, whereas only subtle amplitude variations can be observed with the P-P (vertical-component) data at the same locations.

The difference in Cardium responses of the P-P and P-SV stacks was found to be a result of their respective amplitude-versus-offset (AVO) responses. Forward modeling showed that, for the Cardium event, a polarity reversal occurs with offset with the P-P, but not the P-SV, data. It is this polarity reversal which causes the poor amplitude response of the P-P data at the location of the Cardium conglomerate. This arises from the destructive interference of the near- and far-offsets of the P-P data upon stacking, whereas the radial-component offsets add constructively, producing a higher amplitude P-SV Cardium event. The P-P and P-SV data were also found to be sensitive to variation in conglomerate thickness over the 0-1000 m and 500-2000 m offset ranges respectively. These results indicate that better imaging of the Cardium conglomerate could therefore be achieved by either limiting the offset range during acquisition or by generating offset-range stacks. The generation of a P-P 0-1000 m offset-range stack, in fact, showed a significant increase in the relative amplitude of the Cardium event at the location of the conglomerate deposits. A conventional seismic data set (far-offsets of 1200 m), also acquired in the Carrot Creek area but in a different location than the multicomponent data, showed no indication of thick conglomerate deposits.

It was also found that Poisson's ratio for specific seismic intervals could be derived from the P-P and P-SV traveltimes. Since the Cardium conglomerate possesses a low Poisson'sratio (0.18-0.22) relative to the surrounding shales (0.31), this interval analysis was capable of identifying variations in conglomerate thickness by the presence of lows in the calculated Poisson's ratio. Two such lows were determined using the multicomponent data, which in turn correlate well with the location of known conglomerate deposits.