| dc.description.abstract | Traveltime difference between the seismic surveys is a crucial factor for the repeatability in time lapse seismic, and much of the observed timeshift has its origin from the water layer (Grude, 2009). This master thesis evaluates the effect of diffracted energy from the sea bed, with the aim to see how this influences the water bottom timeshift and the impact on the 4D quality at reservoir level. The influence from mis-positioning and velocity changes between the seismic surveys on the water layer timeshift and hence the 4D quality is observed as well, and the effect from smoothening of the sea bed timeshift, multiple energy and migration on the 4D quality at reservoir level. This is done through estimations and evaluation of synthetic seismic, first for a simplified sea bed model to see the small scale effects, and then for more complex models.
The synthetic seismic is made from 2D finite difference modelling, followed by standard processing in ProMAX. The results are compared to real seismic from the Norne Field to see if they can explain some of the observed variations in the timeshift. Some weaknesses are observed in the use of 2D finite difference modelling. Among others grid effects, dispersion effects and numerical instabilities. These have resulted in a modelled water layer timeshift that is smaller than expected from estimations.
Mathematical estimations and modeling of synthetic seismic are used to see the impact on the water layer timeshift from one single diffractor. The derived equastion is an approximation valid for small offsets, and is made from cross correlation of the seismic signals from base and monitor survey in time lapse seismic (Grude, 2009). The approximation depends on the offset from diffraction apex, the strength of the diffractor and mis-positioning and tidal changes between the seismic surveys. Both the modelling and mathematical estimations have shown that the timeshift variation from one single diffractor without tidal changes and mis-positioning are little influenced by the tuning effect between the sea bottom and the diffractor, and can be neglected. Diffractors combined with acquisition mis-positioning however create additional timeshift close to the diffractor apex point which is not negligible.
The complex topography of the sea bed from the Norne Field is used in combination with a simple reservoir below, to see the impact on the 4D quality at reservoir level from the seabed diffractors created by the irregular sea bottom. From this modelling one observe that the water layer timeshift caused by an irregular sea bottom is much stronger than from a single diffractor. This create a variation to the timeshift that is no longer negligible, and increases with acquisition mis-positioning. The strongest error in the timeshift seems to be related to the areas of steepest seabed profile; from this one can evaluate the seabed topography to see where one can expect good quality on the water layer timeshift correlation factor.
The 4D quality at reservoir level is measured through difference in amplitude in the reservoir and timeshift variation through the reservoir. The timeshift variation is measured from cross correlation with a window length of 30ms at diffractor apex for the point diffractors below the reservoir. The amplitude analysis in the reservoir is performed in an analysis window between 900 and 1020ms, covering the entire reservoir.
The 4D differences are reduced by the complex sea bed, mis-positioning and velocity changes between the surveys, which are observed both through difference amplitudes and timeshift measurements. This is compensated by migration. Mis-positioning and velocity change between base and monitor survey show no degradation on the 4D difference amplitude and timeshift at reservoir level for a flat sea bottom, both pre and post migration. Water bottom multiples interfering with primary energy will increase the magnitude of the difference amplitude, and the timeshift at reservoir level becomes more unstable with multiple energy, both the difference amplitude and timeshift variations are improved by migration. Smoothening of the complex sea bed does not affect the 4D quality at reservoir level measured through difference amplitude, but it does not improve the irregularities induced by the complex sea bed with mis-positioning and velocity changes either. These variations in the difference amplitudes are improved by migration.
The seismic modelling performed here is a simplification of a complex problem observed in field data from the Norne Field, making the synthetic seismic a tool for observation of a bigger problem, and not directly comparable to the field data. Still they have shown to indicate that asymmetric diffractors explain the observed amplitude asymmetry on the field data from Norne, and that multiple energy disturb the 4D quality at reservoir level, and should be removed during processing of the data to ensure that the quality is not reduced. | nb_NO |
