Repeatability of P-Cable 3D seismic data

Abstract

The use of 4D time-lapse seismic data is a viable technology for monitoring fluid changes in the subsurface using conventional data, e.g. reservoir exploitation and CO2 storage. In this master thesis, the potential for using high-resolution P-Cable 3D seismic data for 4D time-lapse studies has been evaluated. The P-Cable 3D seismic system is a versatile, lightweight system which employs 24 short streamers. Short offsets and close streamer spacing result in CMP coverage with small bin sizes. Together with air gun sources providing frequencies up to 400 Hz, P-Cable data has at least one magnitude better lateral resolution and several times better temporal resolution than conventional seismic data. Over the last 4 years, the Arctic University of Norway, in Tromsø has acquired 4D seismic data in 3 different areas, two in the Barents Sea; the Snøhvit and Lyngen areas and the third at the Vestnesa Ridge on the western continental margin of Svalbard. These 3 repeat surveys are processed for 4D in order to quantify the repeatability of high resolution 4D seismic. The repeatability is quantified using metrics like the normalized root mean square (NRMS) and predictability. A 4D processing software package from Hampson-Russel is used to calibrate, process and match the 4D data. The 4D processing includes correction for time-shifts, phase differences and amplitude variations between the base and monitor surveys. Repeatability measures like the NRMS, cross correlation, time shifts and predictability have greatly improved through the 4D processing A noise removal process in the GeoTeric software was also a part of the investigation of repeatability of the high frequency 4D data. Noise removal has proven to have a huge effect on the final repeatability of the data. In addition frequency filtering the surveys from one of the areas also contributed to increase the overall repeatability. Some geological features not associated with genuine 4D differences turned out to be particular challenging to remove by processing as they are strongly absorbing boundaries. It turned out that shallow targets where the geology is homogenous are the best targets for this type of seismic. In some shallow areas in the Lyngen repeat, containing stratified horizontal layers, the processing was able to result in an overall better match giving better repeatability. This support the assumption of P-Cable seismic being most suitable for shallow investigation. Some theoretical limits for the repeatability of P-Cable 4D data are reached with regards to the NRMS and Predictability for the repeat surveys, where a NRMS of around 0.4-0.5 and a predictability of 0.8-0.9 was reached. In conclusion, the P-Cable offers seismic data that is suitable for 4D if the acquisition and processing circumstances are similar for the base and monitor survey.