Permeability Modelling using Rock Fabric Number and Neural Networks in Carbonate Reservoirs

Determination of permeability is vital in the static and dynamic modelling of reservoirs. Carbonate reservoirs pose a unique challenge due to heterogeneity caused by primary and diagenetic processes. Generalized porosity and permeability relationships were developed by linking permeability and porosity with rock fabric classifications. Applications of Lucia Rock Fabric Number (RFN) method to estimate permeability is possible where specific data is available. It is postulated that spatial distribution of RFN within a field may be estimated based on Neural Networks. By combining well-based RFN classifications and Neural Network Modelling, a 3D profile of permeability and rock fabric can be obtained.