Quantitative Seismic Reservoir Characterization of Mehar Block, Lower Indus Basin, Pakistan

Abstract

The exploitation of a reservoir can be enhanced significantly by means of reservoir characterization. The objective of this dissertation is to use the seismic and well log data of Mehar Block, Lower Indus Basin, Pakistan in order to demarcate and subsequently characterize the reservoir formation (i.e. Ranikot). The techniques used for this analysis includes seismic structural interpretation, wireline log analysis, seismic inversion analysis (P-impedance and S impedance), and LMR attributes (lambda-rho (lr) and mu-rho (µr)). Furthermore, porosity sections have been generated using cross plot analysis and machine learning approach, namely, Probabilistic Neural Networking (PNN). To analyze the spatial distribution of water saturation, Deep Feed-forward Neural Networking (DFFN) approach has been used. The structural seismic interpretation reveals that the regime of the Mehar block is compressional in nature as confirmed by the presence of thrust faulting. The wireline log analysis of the wells Mehar-01, 02 and 03 indicates that the effective porosity (PHIE) in the zone of interest in Mehar-01 is 7.3%, total porosity (PHIT) is 8.1%, shale volume (Vsh) is 10%, and hydrocarbon potential (HS) is 58.2%. In case of Mehar-02, the PHIE in the zone of interest has been estimated to be 7.6%, Vsh is 25%, PHIT is 13% and HS is 68%. Two zones have been demarcated in Mehar 03 well indicating PHIE values of 5.3% and 11.9%, total porosity is 7.8% and 17.9%, shale volume is 11.9% and 12%, and hydrocarbon saturation is 51.2% and 72% respectively. The values from the model-based inversion (MBI), maximum likelihood, and bandlimited inversion indicates low impedance in the zone of interest within the Ranikot Formation. This spatially identified zone is further confirmed by the low values of the lr sections (derived from each inversion model) indicating presence of hydrocarbons while the moderate values in the same zone from the µr sections depict the presence of sand. The spatial distribution of effective porosity using PNN and cross plots depicts high values in the reservoir formation. Furthermore, the DFFN based water saturation values indicate relatively low saturation values in the zone of interest