Author/Authors :
Kiumars ،kamalyar Petroleum University of Technology, Ahvaz, Iran , Morteza ،Yeganeh Petroleum University of Technology, Ahvaz, Iran
چكيده لاتين :
Knowledge of reservoir fluid properties is very important in various reservoir engineering computations such as
material balance calculations, well testing, reserve estimating, and numerical reservoir simulations. Ideally, those data
should be obtained experimentally. On some occasions, these data are not either available or reliable; then, empirically
derived correlations are used to predict PVT properties. An enormous amount of PVT data has been collected and
correlated over many years for different types of hydrocarbon systems. Almost all of these correlations were developed
by linear or nonlinear multiple regression or graphical techniques. Artificial neural networks (ANN), once successfully
trained, offer an alternative way to obtain reliable results for the determination of crude oil PVT properties.
In the present study, a new ANN is developed to predict the bubble point pressure. 200 data sets were collected from
more than 70 different reservoirs in Middle East and North Sea. total solution gas-oil ratio (Rs),gas specific gravity (γg),
oil specific gravity (γo), Temperature (T), were used as the input data and bubble point pressure was as target data,60%
of these data points were used for training and the remaining for predicting the Pb (validation and test).An ANN was
developed and a correlation coefficient (R2) of 0.977 and Absolute Average relative error (%) of 4.49% were obtained
by comparing Pb predictions and the actual measurements.
