An integrated, multi-disciplinary approach was developed to examine waterflooding processes in deepwater stacked turbidite reservoirs. Fluid flow in porous rocks was reviewed both at the pore and the reservoir section scales. The importance of a favourable mobility ratio for a stable oil/water displacement was highlighted. This information guided our choices for fluids characteristics to use in our fluid flow simulation models. A geological review of deepwater turbidite reservoirs provided a sound understanding of their geological characteristics, with a special emphasis on their impact on fluid flow within the reservoir. The vertical permeability distribution across the reservoir was identified as having a crucial impact on waterflooding efficiency in turbidite reservoirs. Permeability distribution within a channel element of a stacked turbidite reservoir follows three characteristic trends: homogeneous distribution, fining upward distribution, and coarsening upward distribution. A series of idealised reservoir models, representative of a single flow unit within a turbidite reservoir, was built. The idealised models represented the three different permeability distributions commonly found in turbidites and another model was added to simulate bottom drive waterflooding. Two scenarios were run on the models: water injection with pressure support and water injection where the pressure dropped by a maximum of 1500 psi. The change in Vp was around 4 to 5% regardless of the reservoir geology or the pressure variations. Pressure change has a global dimming effect on the P-wave velocity. It happens very briefly after the start of the simulation and spreads across the whole reservoir. Change in Vp due to pressure decline was around -0.5% and could not be detected on the synthetic seismic. The waterfront is easily interpreted both on 4D cross-section and 4D attribute maps. A realistic turbidite geological model based on the Ainsa II outcrop was built. The model was populated with rock characteristics of a turbidite reservoir on the West of Africa. The model was then up-scaled and fluid flow simulation was performed. Permeability values and NTG distribution played a major role in the advance of the waterfront inside the reservoir and controlled its shape and location. Petrophysical modelling showed that P-waves velocity would increase by up to 7% due to the substitution of oil by water and suggested that it can be extremely sensitive to water saturation changes. Even the smallest changes (less than 10%) would have a noticeable effect on Vp values, which is of crucial importance when time-lapse seismic is to be used in a quantitative way. Synthetic seismic was created using three different frequencies (35 Hz, 62 Hz, and 125 Hz). On 3D seismic sections, different channels within the reservoir were resolved separately on the high resolution seismic. Tuning phenomenon is observed for the three modelled frequencies due to the presence of very thin beds (1-2 meter thick). The interpretation of the OOWC or the MOWC on those sections is challenging because the reflections at the fluids front are obscured by reflections from geological interfaces. The complex geology of the reservoir resulted in 3 different RMS seismic amplitude maps showing an increasing degree of heterogeneity as the seismic dominant frequency increased. Interpretation of MOWC on time-lapse seismic cross-sections and maps is challenging and the inclusion of different attributes in the interpretation workflow might be necessary in order to assess the complexity of the waterflooding signature. Time-lapse seismic monitoring of waterflooding processes in deepwater turbidite reservoirs requires sound a-priori knowledge of the geology of the reservoir. On the other hand, an accurate interpretation of the time-lapse seismic signature of Waterflooding can improve our understanding of the reservoir characteristics. Therefore, the task should be performed by multi-disciplinary teams, where geologists, reservoir engineers, and geophysicists work closely together.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:547711 |
| Date | January 2010 |
| Creators | Kooli, Nader |
| Publisher | Heriot-Watt University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/10399/2364 |
Page generated in 0.0023 seconds