Consistent porosity–permeability modeling, reservoir rock typing and hydraulic flow unitization in a giant carbonate reservoir

Abstract

Abstract Proper porosity–permeability modeling, reservoir rock typing and incorporation of hydraulic-flow-units are crucial parts of an integrated reservoir characterization/modeling and dynamic simulation study. These control the quality of a reservoir model with effect on wells production behavior and prediction performance. In this study, the routine/SCAL core data and petrophysical/production log with prior sequence stratigraphy study of an Iranian gas carbonate reservoir were examined for proper multi-scale permeability modeling, consistent selection of reservoir rock-type classes and profile of hydraulic flow units. The considered reservoir has a very heterogeneous nature with Dykstra-Parsons coefficient of 0.98. The permeability was correlated to the porosity data using the Discrete Rock Type (DRT) concept. The DRT was used just for correlation purpose and was not considered as Reservoir Rock Type. The DRT itself was calculated from well log data (for permeability determination) by applying an artificial intelligence technique known as Local Linear Neuro-Fuzzy Model (LLNFM). The model was trained a priori. The reservoir rock-type classes were selected using routine and special core data complemented with thin section analysis results. A scheme mainly based on rock hydraulic concept (derived from the hydraulic-pore-throat size obtained from core mercury injection tests) proved practical by identifying the Petrophysical Rock Type class (PRT). Each PRT class may be divided to Reservoir Rock Type (RRT) classes by considering the effect of rock wettability. However, the PRT classes are essentially the RRT classes in gas reservoirs since the rock is water wet. The formation layering was carried out using the concept of sequence stratigraphy and the layering was subdivided into hydraulic flow units (HFU). The distribution of PRT classes across the reservoir sections was also investigated. This was carried out based on the core permeability (near wellbore) and the permeability derived from Production Logging and DST data (well drainage volume). Boost factors were obtained and applied to match the near wellbore core-based permeability to that of the DST/PLT results. This proved to minimize the mismatch between the simulation results and production history.