Print this pageReservoir Characterization and Modeling of from Isolated Platform Reservoir, Tengiz Field, Kazakhstan
An improved reservoir model for the Tengiz Field, Caspian Basin of Kazakhstan, will support future reservoir development including a potential miscible gas injection project, deemed the Future Growth Project (FGP). Since the generation of the previous reservoir model, stratigraphic research advanced in two directions: 1) the improvement of the sequence stratigraphic framework (SSF) and, 2) the application of Multiple Point Statistics and Facies Distribution Modeling (MPS/FDM) to incorporate diagenetic concepts and trends.
Reservoir quality in the Late Visean to Bashkirian (Unit 1) platform portion of the reservoir is controlled by a combination of depositional and diagenetic processes. Depositional elements include a highly cyclic succession of generally shoaling lithofacies overlying a sharp base, where there is evidence for subaerial exposure and/or flooding. This initial depositional fabric was strongly overprinted by late burial diagenesis, which increased porosity by dissolution in the center of the platform and reduced porosity through compaction, calcite cementation and pyrobitumen cementation towards the margins. The SSF (3rd, 4th and 5/6th order) is anchored on the stacking of ash beds (gamma ray spikes) and low porosity intervals, which are generally easy to recognize from the Tengiz well logs.
Petrophysical Rock Types (PRTs) formed the basis for the MPS/FDM simulation in the Unit 1 platform. The PRTs are defined from current reservoir properties resulting from the original depositional processes overprinted by: 1) the combined effect of late burial calcite cementation and (later) corrosion, 2) late burial pyrobitumen cementation. Trends of corrosion/cementation, pyrobitumen effect and distribution of volcanic ash were captured for each individual 3rd and 4th order sequence and used to generate the facies probability cube for MPS/FDM. Finally, diagenetic concepts and dimensional data inferred from well data were used to create training images, specifying the PRT spatial juxtaposition rules for each of the 3rd and 4th order sequences; these training images were then convolved with the probability cubes to generate the PRT realizations.
The revised SSF and integration of novel concepts in modeling the diagenetic overprint resulted in a refined understanding of the Tengiz platform. In addition, the extensive use of MPS/FDM to integrate depositional and diagenetic trends has resulted in a more realistic model of the Unit 1 platform.
AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009