Long offset time-lapse seismic and monitoring of unconventional reservoirs

Abstract

Time-lapse seismic is highly accepted as a reservoir management tool. However, it has severe challenges related to monitoring of stiff rock reservoirs due to relatively small seismic property changes for such reservoirs. This thesis aims to investigate monitoring of stiff rock reservoirs. Long offset time-lapse data is used as the key method for investigation. Amplitude variation versus offset is used to find the offset that has the maximum amplitude beyond critical offset.

The maximum amplitude offset has the potential to monitor velocity changes. A timeshift of head waves associated with velocity change of the stiff rock is also used for monitoring. The main goals of thesis are (i) to interpret the maximum amplitude changes versus offset for time-lapse seismic PP and PS data, (ii) investigate the sensitivity of the maximum amplitude offset with respect to changes in P-wave velocity, S-wave velocity and density, (iii) investigate the timeshift associated with velocity changes on long offset time-lapse seismic data and (iv) study the challenges associated with long offset time-lapse seismic. When possible, I have tried to test the proposed methods on field data. However, for some cases, there were no field data available. In such cases I used two-dimensional finite difference forward modeling to generate synthetic seismic data to evaluate the sensitivity of the method.

The maximum amplitude offset is observable in both PP and PS data and has the potential to improve monitoring of small velocity changes in a high velocity background medium, e.g. in stiff rock reservoirs. In case of Valhall field, the predicted velocity change using this method matches qualitatively the velocity changes predicted from independent acoustic impedance and synthetic study.

The long offset time-lapse method is sensitive to velocity changes and practically independent of density changes.

The timeshift observed on head waves matches with synthetic study for PP and PS data. This timeshift helps to monitor the change in velocity of reservoir or overburden.

The main challenge associated with long offset time-lapse seismic is overburden noise. The overburden noise may be attenuated by processing filters or during acquisition by the environmental friendly method of using super long source arrays. It is shown synthetically that super long source arrays attenuate the overburden noise by a factor of 3-4