Preliminary interpretation of P-P and P-S seismic data from the Blackfoot broad-band survey

Susan L. M. Miller, Evsen Ö. Aydemir, Gary F. Margrave

A 4.0 km 3C-2D broad-band seismic line was acquired over the Blackfoot Field near Strathmore, Alberta in the summer of 1995. The target rocks are incised valley fill clastic sediments within the Glauconite Formation of the Lower Cretaceous period. The exploration objectives of the survey were to determine if, through coupled analysis of P-P and P-S seismic data, clean channel sands could be distinguished from shale-plugged channels and regional non-channel sediments.

Preliminary modelling indicates that the channels can be defined seismically by character changes on both the P-P and P-S sections. Vp/Vs analysis of the synthetic P-P and P-S cross sections showed a clear distinction between sandstone and shale lithologies within the zone of interest. The average interval Vp/Vs was lower at the sand channel (1.85) than at either a shale-plugged channel or the regional section (1.93).

The vertical and radial components of the 10 Hz seismic data were correlated using P-P and P-S offset synthetic seismograms generated from dipole sonic logs. Corresponding horizons were interpreted on the migrated P-P and P-S sections. In general, the models matched the data quite well, and were useful in developing the interpretation. Both the P-P and P-S sections showed character changes in the interpreted sand-channel facies which were consistent with the models.

Interval Vp/Vs values were calculated from P-P and P-S isochrons across intervals containing the zone of interest. The P-P section is interpreted to have a time-structural low at the Mississippian unconformity which partially coincides with the projected location of the channel. This time structure is much less evident on the P-S section. As a result, the interval Vp/Vs value is anomalously low at this location, about 1.75 compared to an average value of about 2.00. The trend to a lower Vp/Vs value in the channel is in agreement with the model response, although the amplitude of the anomaly is larger than the model predicts. The reason for the large difference in time structure on the Mississippian unconformity is presently unclear. Further processing and modelling are planned.