Developing a gamma ray fluid densitometer in petroleum products monitoring applications using Artificial Neural Network

Abstract

In gamma-ray densitometry the determination of density is performed by the use of calibration tables. Calibration and consequently the accuracy of the system can be influenced when a number of important system parameters such as pipe diameter, source to detector distance and so on are changed from one case to another. In this work an Artificial Neural Network model was proposed for developing a previously designed and constructed gamma ray densitometer in prediction of fluid density of different petroleum products. The required data for training and testing the ANN model has been obtained based on simulations using MCNP4C Monte Carlo code. Simulations for 4-inch polyethylene pipe had been validated with the experimental data previously. The Mean Relative Error (MRE) from ANN modeling was less than 0.5%. Results show that proposed ANN model represents a good estimation of the density in petroleum products monitoring application and can be used as a reliable and accurate tool.