The used metering technique in this paper is based on the multienergy (at least dual) gamma-ray attenuation. The aim of
the current study is investigation of different combinations of sources in order to find the best combination for precise
metering gas, oil and water percentages in annular three-phase flows. The required data were generated numerically using
Monte Carlo N Particle extended (MCNPX) code. As a matter of fact, the current investigation devotes to predict the
volume fractions in the annular three-phase flow, on the basis of a multienergy metering system including different
radiation sources and one sodium iodide detector, using the hybrid model. Since the summation of volume fractions is
constant, a constraint modeling problem exists, meaning that the hybrid model must predict only two volume fractions. Six
hybrid models associated with the number of applied radiation sources are employed. The models are applied to predict the
oil and gas volume fractions. For the next step, the hybrid models are trained based on numerically obtained data from the
MCNPX code. The results show that the best prediction results are obtained for the oil and gas volume fractions of a system
with the (241Am & 137Cs) radiation sources.