科研之友
微信
新浪微博
Facebook
分享链接
摘要
Three-dimensional (3D) reflection seismic data were recorded as part of a pilot scale carbon dioxide (CO2) geological storage project (CO2 SINK) with the aim of mapping the structural geometry of the site and providing a 3D baseline prior to CO2 injection. Standard processing originally focused on the storage target, a saline aquifer at 500-700 m depth and successful imaged coherent reflections from 150 ms to 900 ms in the seismic volume. However, the relatively sparse distribution of sources and receivers, the frequency content of the data and artefacts of the processing resulted in the uppermost 150 ms being poorly resolved. This depth range contains caprock, shallow faults and an aquifer system. Thus, characterizing the shallow subsurface is important in terms of site delineation of potential leakage paths and monitoring after CO2 injection. In order to study the potential of mapping the uppermost reflectors and shallow structures associated with major fault zones, a comprehensive reprocessing effort on a subset of the 3D data was performed. The challenge in imaging shallow reflections is dependent upon the separation of ground roll and refracted energy from the reflected energy, as well as compensation for time shifts due to statics. Among the processing sequences, refraction static corrections, careful muting and filtering, velocity analysis and 3D time migration were key steps for enhancing the resolution and coherency of shallow seismic reflections. This study images a previously unmapped horizon, close to the Quaternary-Tertiary boundary, at about 95-120 ms (similar to 65-90 m depth). Correlation of lateral variations in reflectivity along this boundary, lateral velocity variations in the tomographic image and the seismic signature in modelling studies suggest an aquifer/aquitard complex and variable lithology with associated localized silty or clayey sediments, overlying the Tertiary Rupelton clay unit. In the previous processing it was not clear if the deeper faults imaged on the 3D seismic survey extended to shallower levels than the base Tertiary. Thus, a comprehensive fault detection technique, multi-attributes and neural networks analysis, was employed in this study to allow a more reliable fault geometry to be interpreted. Tracking of faults in the seismic image and comparisons with a tomography study indicate that some deeper faults may penetrate into the overlying Tertiary unit. These findings are important for understanding potentially risky areas and can be used as a database for future monitoring programmes at the site.
- 出版日期2009-4
