Abstract
The definition of Reservoir Rock Types (RRT) is a key challenge in the evaluation and characterization of carbonate reservoirs, and this step is critical as the RRT's define the building blocks for constructing 3D models, as RRT definition links to static and dynamic reservoir properties. This paper describes an innovative and synergetic rock typing process linking geology and petrophysical properties, with a customization of the Flow Zone Indicator (FZI) method to identify RRT's and characterize the heterogeneous oil-bearing Pre-salt carbonates of the Santos Basin, Brazil offshore.
A data set of 448 MICP from the Pre-Salt carbonates of Barra Velha Formation was used to build the FZI-RRT model. The optimal number of RRTs, five in total, is determined by using an unsupervised neural network with capillary pressure parameters as inputs, permeability, effective porosity and water saturation. The five classes are delineated by FZI values at 10% porosity and key permeability values, chosen for reasons due flow properties at the core and log scale and suitability in EOR treatments. The five RRTs define a unique permeability/porosity equation that can be propagated to the full core dataset and to the log domain. An ID card for each RRT is then created with specific static and dynamic properties (porosity, permeability, water saturation, relative permeability) that can be used for 3D reservoir modeling.
