Sammendrag
We describe and evaluate a methodology combining high-resolution seismic waveform tomography and rock physics inversion to monitor the CO 2 plume at Sleipner. For geophysical monitoring, we derive P-wave velocity by applying Full-Waveform Inversion (FWI) to both baseline and monitor data. We then infer selected rock physics parameters using a Rock Physics Inversion (RPI) methodology. We define the physical relations between rock physics properties and acoustic wave velocities in the Utsira unconsolidated sandstone using dynamic rock physics models based on Biot-Gassmann theories. For data prior to injection, we estimate rock frame properties and then use these rock frame properties as a priori to infer rock physics properties from the monitor data. The derived FWI results show nicely resolved thin layers of CO 2 -brine saturated sandstones and intrareservoir shale layers. The CO 2 saturation estimation is carried out by inserting effective fluid phase properties into the rock physics model. Calculating the effective fluid bulk modulus of the supercritical CO 2 -brine mixture is shown to be the key factor linking P-wave velocity to the spatial distribution of CO 2 saturation. The estimated velocities are compared with published work at Sleipner. It is shown that using multiparameter seismic inversion or multiphysics integration of data can lead to a better discrimination between the different effects and reduce uncertainties.
Vis fullstendig beskrivelse
