Estimation of permeability and effective porosity and determination of hydraulic flow units by using Artificial Neural Network method in Marun oil field

Permeability and effective porosity are the most important characteristics of a reservoir which can be used as input for creating petrophysical models of reservoir. The relationship between porosity and permeability in the form of hydraulic flow units can be used in describing heterogeneous reservoir rocks. Identifying hydraulic flow units can be used for evaluating reservoir quality based on relationship between porosity and permeability. Porosity and permeability are measured by injecting helium and air into the core samples respectively. In addition, these parameters can be measured by NMR well logging. Furthermore, well testing is another way for measuring permeability parameter. Although these measurement methods are accurate, they have some drawbacks. These methods are time-consuming and expensive.Therefore, they are used only occasionally. In this study, we estimated porosity and permeability by back propagation error Artificial Neural Network method (BP-ANN) using extensive dataset achieved by well logging in the field.Finally, after estimating parameters by using SSE method and K-means analysis, we improved the relationship between porosity and permeability in the Asmari Formation by dividing data from three studied wells into nine hydraulic flow units. Results showed that the Neural Network method predicted reservoir parameters successfully.