繰り返し三次元地震探査による圧入CO₂および物性量の評価手法の開発について

抄録

To mitigate the global climate change due to excess carbon dioxide emission, geologic sequestration by carbon dioxide injection in the subsurface has been proposed. To monitor and verify the long-term safe storage in the subsurface, 4D or time-lapse 3D seismic technology is the most effective to spatially and efficiently detect the change of fluid saturation and pore pressure in the target aquifer and others. In RITE/METI Nagaoka carbon dioxide injection test field, 4D seismic survey was conducted to monitor injected carbon dioxide as a collaborative research of Japex and RITE/METI. It was the first 4D seismic survey for onshore aquifer injection monitoring in the world.
Target reservoir in saline aquifer is at approximately 1100m deep. Carbon dioxide saturation was observed with approximately 6m thick by logging survey. To estimate the saturation zone by 3D seismic surveys, I had to overcome the problems of data acquisition and processing, i.e., 1) irregular 3D geometry in the land 3D seismic surveys, and 2) to enhance data resolution to meet the thinner target reservoir.
In order to detect and highlight 4D anomaly zones, for relatively high noise content data, it is inadequate to simply subtract baseline and monitor seismic data as there are no ways to discriminate the noises from geological changes by this volume alone. It turns out that 4D anomaly detectability is improved if supervised pattern recognition technology is implemented in the multi-attributes approach. Estimated anomaly pattern is consistent with the observation of time-lapse wireline data and total amount of injected carbon dioxide incorporated with probability of distribution. It has improved the results of visual inspection of simple math difference of 3D seismic volume.
3D data were evaluated using well synthetics data and impedance inversion data to estimate several physical parameters such as porosity and permeability combined with the wireline data and geological constraints. Estimated high permeability zone by baseline data prior to the repeat survey showed a correlation with the detected 4D anomalies.
The first onshore 4D seismic monitoring of the injected carbon dioxide in aquifer was successfully conducted, though there were difficulties caused by the irregular 3D geometry and thinner target reservoir. It provides the prototype approach to the similar onshore carbon dioxide monitoring.