Laterally and temporally variable amplitude anomalies are observed on 3-D time lapse seismic volumes over a heavy oil field which is undergoing cyclical steam stimulation. These time lapse seismic data were acquired during production and steam injection cycles. The amplitude anomalies seen in the stacked data are interpreted to be caused by gas-saturated zones in the 1990 data (production) and by steamed zones in the 1992 data (injection). Forward modelling predicts that a measurable increase in amplitude should occur in data acquired in steam- or gas-saturated zones. The modelling also shows that the farthest source-receiver offset should be restricted to less than 500 m and that, to avoid ambiguous results, offsets in the 1990 data, where partially gas-saturated zones are under analysis, should be restricted to less than 250 m.
Amplitude anomalies were selected from the two stacked 3-D data volumes to investigate intervals interpreted to be steam- or gas-saturated. For the majority of the supergathers examined, the AVO values obtained agree with the interpretation of the stacked data amplitude anomalies. Analysis of selected supergathers along a 3-C P-P line, to detect partial gas-saturation at the top of the reservoir, resulted in negative AVO values for all the gathers away from the injection wells and positive AVO values for most of the gathers at the well locations. Crossplots of AVO value against Vp/Vs from the 3-C line and against scaled average amplitude from the corresponding 3-D line indicate that a negative AVO value with Vp/Vs over 2.2 or with negative average amplitude can be used to classify that point as lying in cold reservoir with a high degree of confidence.
View full article as PDF (0.74 Mb)