Detection and on-line prediction of leak magnitude in a gas pipeline using an acoustic method and neural network data processing

Abstract

-Considering the importance of monitoring pipeline systems, this work presents the development of a technique to detect gas leakage in pipelines, based on an acoustic method, and on-line prediction of leak magnitude using artificial neural networks. On-line audible noises generated by leakage were obtained with a microphone installed in a 60 m long pipeline. The sound noises were decomposed into sounds of different frequencies: 1 kHz, 5 kHz and 9 kHz. The dynamics of these noises in time were used as input to the neural model in order to determine the occurrence and the leak magnitude. The results indicated the great potential of the technique and of the developed neural network models. For all on-line tests, the models showed 100% accuracy in leak detection, except for a small orifice (1 mm) under 4 kgf/cm² of nominal pressure. Similarly, the neural network models could adequately predict the magnitude of the leakages.

Discrete event based hybrid framework for petroleum products pipeline activities classification

“…The system enhances learning from previous data, identification, extraction and classification of oil spillage patterns but lacks the capability for timing and automatic monitoring of events that led to the oil spillage on PPP. In Santos et al, [11] prediction of leak magnitude in a gas pipeline is presented using NN and acoustic sensors. The NN and acoustic sensors hybrid produced accurate predictions for high frequency signals of oil leakages but could not give correct predictions for low frequency signals even under occurrences of leakages.…”

“…[7] Different approaches have been used to evolve systems that monitor, detect, classify or respond to emergencies resulting from oil spillages and leakages. [8][9][10][11][12][13][14][15] These systems are limited by lack of a systematic way of tracking the time of activities, high probability of false detection and inefficient localization of detected activities due to non inclusion of intelligent tools for explicit timing of operations, pattern recognition and data imprecision handling. Discrete event system specification (DEVS) offers a plausible solution for the specification of timing and localization need of this problem, while adaptive neuro-fuzzy inference system (ANFIS) proffers solution for pattern recognition and data imprecision.…”

Modeling of gas pipeline in order to implement a leakage detection system using artificial neural networks based on instrumentation

“…Oil and natural gas are significant natural resources providing approximately 60% of world energy . These natural resources are mostly transported through a network of the insulated pipes, which is known as an economical way to transport large quantities of oil and gas.…”

“…It is worth mentioning that this strategy differs from simulation and statistical points of view. See the study of Molina‐Espinosa et al for more details. The second is the external implementation of hardware‐based implementation, such as sensors with impedance or capacitance changes, fiber optic cables, acoustic sensors, infrared rays for image processing, and video monitoring. The third solution, which combines the first and second methods, for example, acoustic analysis and pressure with mass and volume balance. Recently, the neural network has been given special attention for pipelines, with complex physical behavior . This method has some advantages and disadvantages that are as follows:…”

“…This paper, by using of gas flow pattern, proposed a novel neural network‐based fault detection method to detect the leakage in the gas pipeline. Compared with most of the recent works that only considered simulation models, in this paper, the acquired practical data from the real life gas pipeline are gathered and utilized for training a neural network to model the process. Some of the data are used for training set to adjust the neural network weights, and others are used to evaluate the performance of the neural network‐based fault detection system .…”