Density and velocity determination for single-phase flow based on radiotracer technique and neural networks

Measuring the density and velocity of fluids is one of the important tasks in oil and petroleum industries. The article demonstrates the measurements of these parameters precisely for different fluids and various diameters of pipes by using radiotracer injection and Artificial Neural Network (ANN). The required data for training and testing the ANN model were obtained by the MCNPX code simulations. Before using the simulation results for training the ANN, simulation geometry was validated with an experimental setup. The experimental setup consists of two 2-inch NaI(Tl) detectors that are positioned in distance of 120 mm from each other and one Ba-133 radioactive source as a tracer. It is shown that the estimated Mean Relative Error (MRE) of the density determination in presented system was less than 0.9%. The relative combined standard uncertainty of the fluid velocity measurement did not exceed 0.5%.