INTERPRETING MULTICOMPONENT (3C) SEISMIC DATA:EXAMPLES OF SAND CHANNEL IDENTIFICATION

摘要

We analyse two multicomponent (3C) seismic data sets in the quest to delineate sand reservoirs. The first case is a 3C-3D seismic dataset over the Ross Lake oilfield in southwestern Saskatchewan, Canada. We find a reasonable correlation among associated logs, synthetic seismograms, VSP, and surface seismic volumes. The PS synthetic seismogram increases the confidence of PS seismic event identification and provides an essential guide to pick PS horizons. Combining the PP and PS time-thickness maps yields a V_p/V_ s map between horizons bounding the reservoir. Relatively low V_p/V_s values are interpreted as a sand indicator. A break in the mapped low V_p/V_s values suggests that there is a shale-cut or shaly part within this target sand body. This interpretation is supported by a subsequent gamma ray log from a horizontal well drilled through the sand and shaly region. PP and PS impedance inversions are performed on the post-stack PP and PS data. A V_p/V_s map, resulting from the ratio of the P and S impedances, also shows anomalies roughly correlated with the time-thickness results. The second example is from the Pikes Peak heavy oilfield, also in Saskatchewan. In this 2D-3C case, we follow a similar procedure as above to outline the productive sand intervals. In addition, we use geostatistical methods, which combine multicomponent seismic data and well logs, to estimate density and porosity along the line. Again, known sand areas are indicated on the PP/PS results as well as several new, yet untested, anomalies.