Degree of water saturation using intelligent method (ANFIS) in one of the hydrocarbon reservoirs in southwestern Iran(North Azadegan oil field)

Degree of water saturation is one of the key parameters in petroleum geological engineering to calculate the volume of reservoir hydrocarbons and also reduce economic risk in the development of oil fields. The first attempts to estimate the degree of water saturation of the formation were made by Archi in the data of well drilling diagrams in clean sandstone reservoirs, the results of which were generalized as an experimental relation for carbonate reservoirs. In carbonate reservoirs, assuming these parameters to be constant due to severe heterogeneity is associated with a large error. On the other hand, it is not possible to calculate these parameters due to the time and cost and loss of part of the core for the entire length of the well. Artificial intelligence methods are new, low-cost and accurate methods that can indirectly estimate the degree of saturation of the reservoir water in the shortest possible time using well data. Therefore, in this study, using different wells and intelligent method of adaptive neural-fuzzy inference system (ANFIS-FCM), the degree of water saturation in one of the hydrocarbon reservoirs in southwestern Iran (North Azadegan Square) has been indirectly estimated. To use this artificial intelligence method, the database consisted of 2511 well data, which were divided into two parts: training data (1758 data) and test data to evaluate the models (753 data). The results show the very good performance of the adaptive neural-fuzzy inference system (ANFIS-FCM) method of estimating the degree of water saturation. Therefore, the adaptive neural-fuzzy inference system (ANFIS-FCM) model can be used as a powerful, fast and accurate method for indirectly estimating the degree of water saturation in reservoirs where the degree of water saturation is not measured through the core.