An intelligent gamma-ray technique for determining wax thickness in pipelines

Area of Science:

• Nuclear Engineering
• Petroleum Engineering
• Artificial Intelligence

Background:

• Wax deposition in oil and gas pipelines impedes flow and requires accurate measurement for effective management.
• Existing methods for wax thickness measurement can be invasive or lack precision.

Purpose of the Study:

• To develop and validate a novel, non-invasive method for measuring wax thickness in pipelines.
• To utilize artificial neural networks (ANNs) combined with gamma-ray backscattering for wax thickness determination.

Main Methods:

• Simulating a gamma-ray backscattering system using MCNPX code to analyze wax thickness in pipes of varying diameters.
• Optimizing the system design and training a multilayer perceptron ANN model using simulated data from Cesium-137 and Cobalt-60 radiation sources.
• Validating simulation results with experimental data from a real-world setup.

Main Results:

• The simulation and experimental results showed excellent agreement, with a root mean square error below 1%.
• The ANN model demonstrated capability in accurately predicting wax thickness across different pipe sizes (2-4.5 inches).
• Simultaneous use of two radiation sources yielded the best wax thickness prediction accuracy.

Conclusions:

• The proposed artificial neural network-based gamma-ray backscattering method is a highly accurate and viable technique for measuring wax deposition in pipelines.
• This non-invasive approach can significantly improve flow control and operational efficiency in the oil, gas, and petrochemical industries.