Inversion of seismic data for assessing fluid replacement in the Nisku Formation

Patricia Esperanza Gavotti, Donald C. Lawton

The seismic response of the Nisku Formation in Alberta was evaluated in terms of its impedance variations with the surrounding formations. The methodology compares the effect of using a dataset processed with different approaches in an inversion study. These approaches are based on a previously conditioned dataset with a conventional processing sequence (a) versus a new specialized processing sequence (b) focused on attaining coherent noise without compromising low-frequency signal. A model-based inversion was performed with these datasets. In both cases, the inverted impedance showed good results at the well location and yielded a similar general trend and lateral variations. The inverted impedance in case b) showed a broadband result possibly related with the presence of more low frequency content in the seismic data. In case (a) the result yielded a cleaner section and the units look more continuous without much lateral variation.

A 2D seismic modelling was undertaken to simulate a CO₂ injection scenario in the Nisku Formation. The time-lapse study was performed by comparing seismic amplitudes and impedance changes before and after the CO₂ injection. The post-injection seismic section shows a time delay of 1.8 ms of the basal reservoir reflector and amplitude change of ~30% with respect to the baseline case. After performing the inversion, a decrease in the impedance values of ~7% is observed in the post-injection scenario.