University of Illinois Chicago
Browse

Reliable Monitoring of Leak in Gas Pipelines Using Acoustic Emission Method

Download (2.01 MB)
thesis
posted on 2013-06-28, 00:00 authored by Hazim Yalcinkaya
Acoustic emission (AE) method can be used to detect and locate small leaks on buried and unburied pipelines in 1D and 2D. A 152 cm long, 11.43 cm diameter steel pipeline is built in the laboratory to study the leak variables for the Acoustic Emission (AE) method. Two thickness modes AE sensors (R6) that are coupled to the pipeline are used to understand the leak variables: internal pressure, leak size and earth pressure. Increasing leak size and internal pressure results in an increase in leak rate, therefore, a higher voltage output from the AE sensors. However, the presence of earth pressure increases the attenuation on the pipe resulting in lower sensor output and closer sensor spacing. Accuracy in leak location is also compared for different variables. Attenuation profiles for two specific frequencies (10 kHz and 60 kHz) are studied numerically to determine the spacing of the AE sensors sensitive to displacements in the radial and tangential directions. To increase the sensor spacing and the accuracy in leak localization, a new shear mode sensor is designed. Admittance curves of the new shear mode sensor that are found numerically and experimentally are compared. The voltage outputs of the two sensors, thickness mode and shear mode, are compared numerically for 60 kHz loading in the radial and tangential directions. The differences in polarization directions and linearity in sensor response for the two sensors are proven experimentally. Leak localization accuracy of the shear mode sensor is compared with that of the thickness mode sensor. To demonstrate an absolute calibration method utilizing the laser beam as an AE source, angular dependence of the sensors are shown in polar coordinates.

History

Advisor

Ozevin, Didem

Department

Civil and Materials Engineering

Degree Grantor

University of Illinois at Chicago

Degree Level

  • Masters

Committee Member

McNallan, Michael Foster, Craig

Submitted date

2013-05

Language

  • en

Issue date

2013-06-28

Usage metrics

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC