Universal neural-network-based model for estimating the PVT properties of crude oil systems

This study presents a universal neural-network-based model for the predicti on of PVT properties of crude oil samples obtained from all over the world. The data, on which the network was trained, contains 5200 experimentally o btained PVT data sets of different crude oil and gas mixtures from all over the world. They were collected from major-producing oil fields in North an d South America, the North Sea, Southeast Asia, the Middle East, and Africa . This represents the largest data set ever collected to be used in develop ing PVT models. An additional 234 PVT data sets were used to investigate th e effectiveness of the neural-network models to predict outputs from inputs that were not used during the training process. The neural network model i s able to predict the solution gas-oil ratio and the oil formation volume f actor as a function of the bubble-point pressure, the gas relative density, the oil specific gravity, and the reservoir temperature. The neural-networ k models were developed using back-propagation with momentum for error mini mization to obtain the most accurate PVT models. A detailed comparison betw een the results predicted by the neural-network models and those predicted by other correlations are presented for these crude oil samples. This study shows that artificial neural networks, once successfully trained, are exce llent reliable predictive tools for estimating crude oil PVT properties bet ter than available correlations. These neural-network PVT models can be eas ily incorporated into reservoir simulators and production optimization soft ware.