Abstract

The recognition of reservoir facies variation based on seismic wave shape has become an increasingly important part of seismic interpretation. The seismic wave shape is a resultant effect of amplitude, frequency and phase and is sensitive parameters for changing lithology, porosity, bed thickness, and fluid content etc. In the present study, seismic facies analysis is based on pattern recognition of wave shape of seismic traces within a specified window from reservoir top i.e. LIII of Mumbai High North. Seismic facies analysis has been carried out in two ways i.e., supervised classification and unsupervised classification of seismic traces. The supervised classification is aimed to predict good or bad reservoir facies based on porosity, lithology and shale volume directly integrating petrophysical data with seismic. From the result of direct interpretation based on supervised classification of seismic facies, some aspect of depositional process and environment can be predicted indirectly. Unsupervised classification is carried out without considering the well data to assess the maximum variability or heterogeneity of the facies from reservoir to non-reservoir part and un-biased classification of seismic facies with some limitation. In this paper we have restricted our discussion mainly to direct prediction of reservoir facies of LIII layer of Mumbai High North based on supervised classification. The 3-D seismic data acquired with OBC technique was used for classification. Seismic facies maps prepared for Upper stack and Lower stack of LIII carbonate reservoir (Early Miocene) layers, in combination with other seismic attributes and petrophysical property maps, indicate lateral variation of facies. It can be used for sector wise analysis of seismic and reservoir data in future.