The importance of PVT properties, such as the bubblepoint pressure, so
lution gas-oil ratio and oil formation volume factor, makes their accu
rate determination necessary for reservoir performance calculations. A
n enormous amount of PVT data has been collected and correlated over m
any years for different types of hydrocarbon systems. Almost all of th
ese correlations were developed with linear or nonlinear multiple regr
ession or graphical techniques that may not lead to the highest accura
cy. Artificial neural networks, on the other hand, once successfully t
rained, can be excellent, reliable predictive tools for the determinat
ion of crude oil PVT properties. In this study, we present neural-netw
ork-based models for the prediction of PVT properties of crude oils fr
om the Middle East. Several neural-network architectures using back-pr
opagation with momentum for error minimization were investigated to ob
tain the most accurate PVT correlations. The data on which the network
was trained contain 498 experimentally obtained data sets of differen
t crude-oil and gas mixtures from the Middle East region. This represe
nts the largest data set ever collected to be used in developing PVT m
odels for Middle East crude oils. The neural-network model is able to
predict the bubblepoint pressure and the oil formation volume factor a
s a function of the solution gas-oil ratio, the gas relative density,
the oil specific gravity, and the temperature. A detailed comparison b
etween the results predicted by the neural-network models and those pr
edicted by other correlations are presented for these Middle East crud
e-oil samples.