Application of artificial neural networks for the prediction of volume fraction using spectra of gamma rays backscattered by three-phase flows

. The determination of the volume fraction percentage of the different phases flowing in vessels using transmission gamma rays is a conventional method in petroleum and oil industries. In some cases, with access only to the one side of the vessels, attention was drawn toward backscattered gamma rays...

Full description

Saved in:
Bibliographic Details
Published in:European physical journal plus 2017-12, Vol.132 (12), p.511, Article 511
Main Authors: Gholipour Peyvandi, R., Islami Rad, S. Z.
Format: Article
Language:English
Subjects:
Applied and Technical Physics
Artificial neural networks
Atomic
Backscattering
Cesium 137
Cesium isotopes
Cesium radioisotopes
Complex Systems
Condensed Matter Physics
Energy sources
Gamma rays
Gas oil
Mathematical and Computational Physics
Molecular
Multilayer perceptrons
Neural networks
Oil and gas industry
Optical and Plasma Physics
Petroleum industry
Physics
Physics and Astronomy
Radiation shielding
Regular Article
Theoretical
Vessels
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:. The determination of the volume fraction percentage of the different phases flowing in vessels using transmission gamma rays is a conventional method in petroleum and oil industries. In some cases, with access only to the one side of the vessels, attention was drawn toward backscattered gamma rays as a desirable choice. In this research, the volume fraction percentage was measured precisely in water-gasoil-air three-phase flows by using the backscatter gamma ray technique andthe multilayer perceptron (MLP) neural network. The volume fraction determination in three-phase flows requires two gamma radioactive sources or a dual-energy source (with different energies) while in this study, we used just a 137 Cs source (with the single energy) and a NaI detector to analyze backscattered gamma rays. The experimental set-up provides the required data for training and testing the network. Using the presented method, the volume fraction was predicted with a mean relative error percentage less than 6.47%. Also, the root mean square error was calculated as 1.60. The presented set-up is applicable in some industries with limited access. Also, using this technique, the cost, radiation safety and shielding requirements are minimized toward the other proposed methods.
ISSN:2190-5444
2190-5444
DOI:10.1140/epjp/i2017-11766-3