Seismic Methods for Heavy Oil Reservoir Monitoring

J. Helen Isaac

Two time-lapse three-dimensional (3-D) seismic surveys, provided by Imperial Oil from a heavy oil production area at Cold Lake, Alberta, were processed concurrently and a procedure was developed to minimise differences between them above the reservoir. Interpretation of the processed data reveals variations in interval traveltimes and reflectivity. These responses are mapped to present a spatial distribution of the interpreted steamed reservoir. Inversion of the data confirms that amplitude anomalies observed on the stacked data are caused by zones of low velocity.

Two experimental time-lapse multicomponent (3-C) surveys were also acquired. Modifications to the converted-wave processing routine included a revised method of determining the large shear-wave static corrections, iterative velocity analysis and deconvolution after normal moveout corrections. Analysis of Vp/Vs showed that this ratio is lower during steaming than during production and that, in general, lowest values are seen at the injection well locations. The average measured Vp/Vs in the cold reservoir and the lowest measured value in the heated reservoir are in excellent agreement with the theoretically calculated values. Areas of heated and cold reservoir mapped from the Vp/Vs analysis agree well with those determined from Imperial Oil's analysis of 3-D seismic data.

Selected amplitude anomalies from the two stacked 3-D seismic data sets were analysed by investigating amplitude variations with offset. In general, the AVO results confirm the interpretation of amplitude anomalies as due to low velocity zones. Analysis of data along a 3-C P-P line indicates a good correlation between positive AVO response and injection well locations.