Datensatz

Zusammenfassung

At the Ketzin pilot site, CO2 is being injected since 2008 in a sandstone aquifer at depths of approx. 630-650 m. As a part of a comprehensive geophysical monitoring program, 4D seismic surveys were acquired. The baseline measurement was carried out in 2005 and the 1st repeat in 2009, after 22000 tons of CO2 were injected in the target reservoir. The difference of the surveys revealed a time-lapse signature associated with the injected CO2, which matches the expected distribution of the plume derived from reservoir simulations. In scope of long-term low-cost monitoring, the possible sparsity of geometries for 2D seismic surveys, which still would enable imaging of the CO2 distribution, were investigated. The baseline and repeat seismic surveys were modeled for a set of sparse geometries and time-lapse effects were analyzed regarding the visibility of the CO2 signature, which is the focus of this presentation. For building 3D elastic models of the subsurface, the results of the acoustic impedance (AI) inversion of baseline and 1st repeat data sets, obtained in a previous study , were used. The P-velocity and density distributions were calculated from the AI-cubes with the Gardner’s equation. Since no S-velocity cube was available, the Vp-cube and the Castagna’s relationship obtained from logs were used for calculation of the required Vs-model. The same Vs-model was used for modelling of baseline and repeat data, as no changes in the S-velocities due to the CO2 injection are expected. The difference cubes, both from measured data and from inverted impedances, show besides the CO2 signature also a noisy background and some features owing mostly to the processing artifacts and acquisition. Within the modelling, these differences can approximate noise level which can be expected on the real processed data.