3D SRME and Interbed Multiple Attenuation on Fully Regularized Common Offset Vector COV WAZ Data from Kuwait Single Sensor Survey
- Moosa Al Jahdhami (Petroleum Development Oman) | Jan Willem de Maag (Shell Global Solutions International B. V.) | Alexander Mueller (Shell Global Solutions International B. V.) | Srinivasa Rao Narhari (Kuwait Oil Company) | Olusegun Kolawole (Kuwait Oil Company) | Vijaya Kumar Kidambi (Kuwait Oil Company) | Bashar Al-Qadeeri (Kuwait Oil Company) | Qasem Dashti (Kuwait Oil Company)
- Document ID
- Society of Petroleum Engineers
- Abu Dhabi International Petroleum Exhibition & Conference, 13-16 November , Abu Dhabi, UAE
- Publication Date
- Document Type
- Conference Paper
- 2017. Society of Petroleum Engineers
- Multiple Attenuation, SRME, Interbed de-multiple, Regularization and interpolation
- 0 in the last 30 days
- 77 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 28.00|
Seismic data in Raudhatain field in Kuwait is strongly contaminated with multiples that impair the image of reservoir reflectors and challenge the structural as well as the quantitative interpretation. A reliable reservoir interpretation depends on an optimal attenuation of multiples (free surface and interbed). In this project, we demonstrated that a significant amount of multiple energy was attenuated by implementing 3D SRME together with data and model driven interbed de-multiple using two strong shallow reflectors as the dominant multiple generators.
The project was run with a primary objective of a successful multiple attenuation at the Jurassic level. Initially, 3D Fourier interpolation was applied to regularize and infill shots and receiver independently, to prepare the data for 3D multiple attenuation. However, the survey in some areas was undershot due to agricultural activities, resulting in massive data gaps which presented challenges for data regularization/interpolation.
Since the 3D data interpolation results demonstrated that this approach was suboptimal and inadequate to infill all data gaps, especially the large ones, a new strategy of incorporating legacy data (from the same area), was followed to infill these gaps in a more appropriate way.
The 3D merged dataset (vintage and new data) was used as input for 5D interpolation software which regularizes and infills the data in 5D senses producing fully regular Common Offset Vector (COV). The 5D interpolated COV panels show a significant improvement to the 3D interpolation results. With the incorporation of vintage data and implementation of 5D interpolation, the multiple predictions were improved.
After applying 3D SREM on the data, 3D interbed demultiple was implemented assuming that the dominant multiple periodicity at the target level is related to two shallow reflectors which have approximately 138ms TWT separation. As the real shallow data was too poor to use, full waveform inversion velocities were used to model the reflectivity of the shallow overburden, which used as input for 3D interbed multiples prediction together with the actual data. This 3D data and model driven approach successfully predicted the 1st order interbed multiples bouncing between these two reflectors.
This advanced 3D de-multiple approach has significantly attenuated multiples at target level allowing more reliable interpretation of the reservoir sections. Robust data interpolation and optimal implementation of novel multiple attenuation techniques were the key elements to the success of this project. Following the same processing approach on surveys or data with similar issues and challenges will help to address the seismic multiples and improve the reliability and accuracy of reservoir interpretation.
|File Size||2 MB||Number of Pages||10|