Abstract

The objective of the study is to decipher the sediment dispersal system and depositional facies, within a distal deepwater, Oligocene sequence in the KrishnaGodavari Basin. A work-flow is designed to integrate outputs from seismic facies classification, spectrally decomposed volumes and basic amplitude attributes to discern and map subtle variations of depositional facies in a basin floor channelized dispersal system. The work-flow was tested on +/- 4ms window along a mapped horizon, within a 3D PSTM volume and later applied to the studied interval. Integration of these outputs finally resulted in the facies characterization of the Oligocene sequence. A southwesterly flowing channel system is present in the area, with associated overbank facies. For most part, the channels are confined to single seismic cycles and hence their identification by a single method is constrained due to vertical resolution of the seismic trace. The channels are mostly sinuous and meandering, while some are relatively straight and shoe-string type. Spectrally decomposed volumes with central frequencies of 10, 20 & 30 Hz, individually bring out different features, depending upon the thickness of the depositional elements. Seismic facies classification on the other hand uses trace shape (frequency and phase being integral to it) to bring out broad facies classes, which can be attributed to lithologies, interpreted by RMS maps derived from spectral volumes. RGB blending of spectrally decomposed volumes on the other hand combines the response obtained from individual spectral bands to generate maps for characterizing the facies. Analysis of the resultant data set reveals the presence of broad meandering channels dominantly responding to low frequency bands, with variable fills of sand and clay, which are also deciphered in the seismic facies maps. The other set of channels which are relatively straight in their course and are resolved in higher frequencies are likely to be clay filled. These channels are especially better distinguished with seismic facies classification as compared to spectral decomposition or RGB blending.