|
Evaluation of Post-stack Seismic Inversion Techniques for Detection of Sands in Gulf of Thailand |
|---|---|
| รหัสดีโอไอ | |
| Creator | Wiphawan Purisa |
| Title | Evaluation of Post-stack Seismic Inversion Techniques for Detection of Sands in Gulf of Thailand |
| Contributor | - |
| Publisher | Department of Geology, Chulalongkorn University |
| Publication Year | 2557 |
| Journal Title | Bulletin of Earth Sciences of Thailand (BEST) |
| Journal Vol. | 7 |
| Journal No. | 2-Jan |
| Page no. | 186-192 |
| Keyword | Post-stack Seismic Inversion, Pattani Basin, Gulf of Thailand |
| URL Website | https://www.bestjournal.org/ |
| Website title | Bulletin of Earth Sciences of Thailand |
| ISSN | 1906-280X |
| Abstract | The majority of petroleum production in Thailand comes from the Pattani Basin of the Gulf of Thailand. The fluvial sands are the target reservoirs in this area. To predict these reservoirs is not always easy using conventional seismic amplitudes because of rapid lateral and vertical lithological variations due to the fluvial depositional environment. Rock physics analysis and different post-stack seismic inversion techniques were applied to detect sands within the study area. Moreover, the study also evaluates different post-stack seismic inversion techniques for accurate prediction of the sand reservoirs. According to rock physics analysis, P-impedance can differentiate sands from shales in the shallow (from Middle Miocene Unconformity to Lower Middle Miocene) and deeper (below the Upper Lower Miocene) stratigraphic intervals, but not for the middle section in between that. Therefore, post-stack inversion solving for P-impedance is only suitable for shallow and deeper sections not for middle section. Model Based, Band limited and Sparesspike (Linear programming and Maximum likelihood) inversions show reasonable correlation at well locations whereas Colored inversion shows lowest correlation coefficient. Blind test results shows that Model Based inversion is more effective in area close to used wells for initial guess model, but Colored inversion performed better in the area away from the used wells for initial model. Model Based inversion scales P-impedance by using well data, whereas, Colored inversion is mostly seismic data driven. Therefore, in the case of good well control Model Based inversion works better, but in the case of poor well constraint, Colored inversion can provide reasonable prediction for P-impedance. The predicted P-impedances are low for sands in shallower and deeper sections. In the middle section, it is not easy to isolate sands from shales. In deeper section, sands with thickness less than 30 m cannot be detected on inverted volumes. This may be due to tuning phenomena. Sand bodies can be extracted along horizon slice by applying cutoff for Pimpedance. Cutoff values for sand can be computed from cross-plots. The extracted sand bodies along horizon slices are promising targets for future exploration and development programs. |
