Ali Gharbanian - Petroleum University of Technology, Ahvaz Faculty of Petroleum Engineering
Seyed Moein Hosseini - Petroleum University of Technology, Ahvaz Faculty of Petroleum Engineering
Misagh Mansoori Ghanavati - Petroleum University of Technology, Ahvaz Faculty of Petroleum Engineering
Saeed Ashtari Larki - Petroleum University of Technology, Ahvaz Faculty of Petroleum Engineering

Asphaltene precipitation is one of the major problems in the oil production and transportation. Change in pressure, temperature and composition of oil can lead to asphaltene precipitation. In the case of gas injection into oil reservoirs, the injected gas caused to change in oil composition and may lead to asphaltene precipitation. Accurate determination and prediction of the precipitated amount is vital, for this purpose there are several approaches such as experimental method, scaling equation, thermodynamics models and neural network as the most recent approach. In this paper we proposed new artificial neural network optimized by particle swarm optimization to predict the amount of asphaltene precipitation during the process of gas injection into oil reservoirs for EOR purposes. In the developed model oil composition, temperature, pressure, oil specific gravity, solvent mole percent, solvent molecular weight and asphaltene content considered as input parameters to neural network and the weight of asphaltene precipitation as output parameter. Comparison between the results of the proposed model in this work with Gaussian Process algorithm and previous researches shows that the predictive model is more accurate

Asphaltene precipitation, Gas injection, Artificial neural networks, Particle swarm optimization, Gaussian process