It has been proven that the common Arps empirical decline curve analysis (DCA) to be an extensively used tool for many years, however, it has much error in production forecasting. In the past, engineers forecasted the oil production rates without understanding of the reservoir engineering principals behind and without having the reservoir properties and operating conditions. These problems influences the quality of the production prediction by Arps model and decreases the validity of this technique. In addition, even by having those properties and conditions a reliable production prediction still cannot be promised without understanding of the theory that connects reservoir properties and operation to the production decline. Experience discovered that the production decline is completely complex that can't be handled just by a simple mathematical model such as Arp, so the lack of an efficient tool for DCA is still essential. This paper presents the simultaneous use of the conventional Arps method and ANN-PSO as forecasting tools to find the optimum decline curve model. Production data from Ramin oilfield in Iran is analyzed by using these two techniques to forecast future the production performance. Arps decline curve and ANN-PSO models are compared, and the most appropriate approach is introduced in this paper. Consequently it is found out that ANN-PSO model gives a better results and can be used as a novel tool for DCA.

Decline curve analysis (DCA) Arps model ANN-PSO Oil Production forecasting Ramin Oilfield.