Pipeline leak detection system in a palm oil fractionation plant using artificial neural network

Abstract

A leak detection system for pipelines is designed and tested. Detection of leak in pipelines is an important task for economical and safety operation, loss prevention and environmental protection. Therefore, a leak detection of pipelines plays an important role in the plant safety operation. In this paper, a neural network based detection scheme integrating a neural Elman network dynamic predictor and a feedforward neural network fault classifier is proposed to overcome the problem of leak detection. The scheme was implemented to detect leakage in a palm oil fractionation process. To generate the required simulation data, Hysys.Plant dynamic process simulator was employed. The use of the output prediction error, between a neural network model and a non-linear dynamic process, as a residual for detecting leakage faults is analysed. A second neural network classifier is developed to detect the leak from the residuals generated, and results are presented to demonstrate the satisfactory detection of leakage achieved using this scheme that can detect leak as small as 0.1%.