Depositional facies variability with a secondary overprint of Turonian palaeokarst controls the quality of the Cenomanian Mishrif reservoir, the main oil producer in the southeastern Persian Gulf. Subsequent drowning of the prominent area and successive deepening of the basin during the Coniacian and Santonian enabled the deposition of pelagic marls of the Coniacian Laffan Formation and the development of a carbonate turbidite system within the overlying Santonian Ilam Formation. The Ilam Formation occurs within slope and pelagic carbonates and consists of oil-bearing channel-reworked limestone facies.
Although high-quality 3D seismic data exist over the Sirri C/D oil fields studied in this paper, it is a real challenge to map the palaeokarst and turbidite deposits in 3D space using conventional seismic interpretation procedures.
This work describes a procedure using Paradigm's Seisfacies software for seismic facies classification and uses this to develop a volume-based interpretation of palaeokarst geobodies and sedimentary patterns of the carbonate turbidite. A hierarchical facies classification technique combined with principal component analysis (PCA) is used to analyse a set of seismic attribute volumes that capture the seismic stratigraphic patterns inherent in the data. PCA as a data reduction algorithm greatly optimized the analysis of the attribute volumes while preserving the essential features of seismic character. A hierarchical facies classifier recognized enough variability within the seismic data to reveal details of the associated geological features. This classification method uses multiple 3D volume attributes as input and generates a single 3D seismic facies volume (a synthesis of different attributes). Using this method, interpretative work can focus directly on geological features in 3D space. This study gives new insights into the internal variability of palaeokarst and carbonate turbidite systems in the Sirri C/D oil fields (SE Persian Gulf) that are essential for the estimation of reservoir volume, connectivity and variability.