Quantitative time lapse seismic analysis - rock physics, repeatability and inversion aspects

Abstract

Quantitative prediction of pressure as well as saturation effects from time-lapse seismic data is one of the topics of interest for geophysical society since the last decade. Time-lapse amplitude variation with offset (AVO) could be used for this purpose. The conventional pressure-saturation discrimination method suffers due to leakage between estimated pressuresaturation changes. However, the method works reasonably well for deeper reservoirs where the P-to-S-wave velocity ratio (Vp/Vs) is around 2. Objective of this thesis is to investigate the applicability of the conventional inversion method for shallow unconsolidated reservoirs where Vp/Vs is often higher than 2. In addition, it is also important to investigate the inaccuracy in the estimates (of pressure and saturation changes) due to improper implementation of rock physics parameters as well as consideration of first order approximations in AVO gradient and intercept changes. I propose a new method, based on a stepwise linear approximation to the intercept and gradient reflectivity changes, to estimate pressure and saturation changes. Similar to the conventional method the new method utilizes the near- and far offset seismic surveys as two independent measurements, and then estimates the pore pressure and saturation changes from amplitude versus offset (AVO). As the conventional pressure-saturation discrimination method has inaccuracy mainly due to gradient reflectivity attribute, it is crucial to explore the applicability of other seismic attributes, such as PS-reflectivity changes and travel time shifts, to discriminate between pressure-saturation changes. I investigate the applicability of various combinations of seismic attributes for compacting reservoir scenario. The usefulness of time-lapse seismic data greatly depends on the repeatability of the data between different surveys. The more repeatable the data is, the more confidently and efficiently it may be used in reservoir management. As the refraction method is emerging as one of the promising and complementary 4D techniques, it is therefore necessary to perform repeatability analysis on refraction data as well. In addition, repeatability analysis on other seismic events (such as tank noise, normal modes, sea-bottom reflection) provide more insight on the efficient implementation of 4D seismic method. I study repeatability issues using two seismic data sets acquired at the permanent sea-bottom array that was installed at the Ekofisk field (North Sea) in 2010. I compare pre-stack repeatability of various seismic wave types, such as refraction, reflection, sea-bottom reflection, tank noise and water column noise. In carbonate reservoirs, implementation of a proper rock physics model is extremely difficult due the more complex pore-geometry in carbonate rocks. It is therefore important to investigate various rock physics models and their applicability for carbonate reservoir. I investigate the applicability of various rock physics models applied in three different wells at the Ekofisk Field, North Sea.