Application of Geophysical Strata Rating (GSR) in carbonate reservoir characterization

عنوان دوره: هجدهمین کنفرانس ژئوفیزیک ایران
کد مقاله : 1134-NIGS
نویسندگان
چکیده
The Geophysical Strata Rating (GSR), which is introduced in this study for carbonate reservoirs, is an empirical strength rating of rocks. It provides ratings from 10 to 100 where the lower values correspond to rocks such as shales which are weak from a borehole stability point of view and also the porous, permeable reservoir rocks. In comparison, the higher values of GSR are associated with intact rocks with few defects in the form of fractures and discontinuities and low porosity. In this study, GSR is calculated from petrophysical data using the equations developed in clastic rocks. The region investigated is the South Pars gas field where the Permo-Triassic Dalan and Kangan reservoirs host the largest accumulations of gas in the world. A 3D GSR model is then estimated from 3D post-stack seismic data of the South Pars gas field by using a probabilistic neural network model. Strong correlations between neural network predictions and actual GSR data at unseen borehole locations proved the validity of the intelligent model in GSR estimation. This 3D GSR cube can be utilized for construction of geomechanical models over the South Pars gas field.
کلیدواژه ها
 
Title
Application of Geophysical Strata Rating (GSR) in carbonate reservoir characterization
Authors
Abstract
The Geophysical Strata Rating (GSR), which is introduced in this study for carbonate reservoirs, is an empirical strength rating of rocks. It provides ratings from 10 to 100 where the lower values correspond to rocks such as shales which are weak from a borehole stability point of view and also the porous, permeable reservoir rocks. In comparison, the higher values of GSR are associated with intact rocks with few defects in the form of fractures and discontinuities and low porosity. In this study, GSR is calculated from petrophysical data using the equations developed in clastic rocks. The region investigated is the South Pars gas field where the Permo-Triassic Dalan and Kangan reservoirs host the largest accumulations of gas in the world. A 3D GSR model is then estimated from 3D post-stack seismic data of the South Pars gas field by using a probabilistic neural network model. Strong correlations between neural network predictions and actual GSR data at unseen borehole locations proved the validity of the intelligent model in GSR estimation. This 3D GSR cube can be utilized for construction of geomechanical models over the South Pars gas field.
Keywords
GSR, South Pars, well logs, Probabilistic Neural Network, geomechanical model