Parallel Self Organizing Neural Network (PSONN) Prediction of Water Saturation in Carbonate Reservoirs

Abstract Carbonate reservoir rocks are considered heterogeneous and the distribution of reservoir quality in carbonates depends primarily upon how diagenetic processes are modifying the rock microstructure, leading to significant petrophysical heterogeneity and anisotropy. Water saturation determination in carbonate reservoirs is crucial parameter to determine initial reserve of given an oil field. Water saturation determination using electrical measurements is based on Archie’s formula. Consequently, accuracy of Archie’s formula parameters affects seriously water saturation values. Determination of Archie’s parameters (a, m and n) is proceeded by three techniques conventional, CAPE and 3-D. This work introduces the hybrid system of parallel self-organizing neural network (PSONN) targeting an accepted value of Archie’s parameters and consequently reliable water saturation values. This work focuses on calculation of water saturation using Archie’s formula. Different determination techniques of Archie’s parameters such as conventional technique, CAPE technique and 3-D technique have been tested and then water saturation was calculated using Archie’s formula with the calculated parameters (a, m and n). This study introduced parallel self-organizing neural network (PSONN) algorithm predict Archie’s parameters and determination of water saturation. Results have shown that predicted Arche’s parameters (a, m and n) are highly accepted with statistical analysis lower statistical error and higher correlation coefficient than conventional determination techniques. The developed PSONN algorithm used big number of measurement points from core plugs of carbonate reservoir rocks. PSONN algorithm provided reliable water saturation values. We believe that PSONN can be supplement or even replacement of the conventional techniques to determine Archie’s parameters and then water saturation of carbonate reservoirs.