Research On Detection And Location Method Of Large Diameter Gas Pipeline Leakage

Large diameter gas pipelines transmission is one major way of long-distance natural gas transportation, and leakage is one of the main risks of safety in the operation of pipeline, the natural gas pipeline leakage may cause serious accident such as poisoning, fire disaster and explosion. So effective security detection method is needed in natural gas pipeline detection. The change of the parameters such as flow state, pressure have been used in gas pipeline leakage detection at present, those methods is easy to be affected by the condition of pipeline transportation, and could not alarm before abnormal situation of the pipeline.This paper studied both the acoustic method and distributed optical fiber detection method of large diameter pipelines leak detection, combining the advantages of two leak detection methods to reduce the rate of false alarm. Identifying and controlling the leakage risk, reducing the occurrence of safety accidents, which is the purpose of this research subject.This paper mainly studies the following aspects:1. Analysis the acoustic transmission mechanism caused by pipeline leakage,acoustic mainly produced by the interaction of fluid and solid in theory. Simulating the gas pipeline leakage by FLUENT, from the velocity and pressure distribution nearby leakage, we obtained that there must produce acoustic wave signals.2. Based on the research of Mach-Zehnder distributed optical fiber leak detection method, put forward a kind of linear Mach-Zehnder distributed optical fiber detection and location method, setting up test platform based on linear Mach-Zehnder distributed optical fiber structure, the test platform convert the detecting light signal to electrical signal and then transmit to the Labview platform for demodulation processing. According to the test results, when there have vibrations caused by outside interference, the system can quickly detect the changes, and be able to response to the abnormal situation of the pipeline in advance.3. To analyze leakage signal processing by the wavelet analysis method,according to the non-stationary and random character of leakage signal, we use threshold value, wavelet packet decomposition and Multi-scale cross-correlationalgorithms to process the leakage signal respectively, compared denoising result by the three algorithm, found that the multi-scale cross-correlation algorithm, not only eliminate the background noise effectively, also can reduce noise signal strength at mutation, which can effectively extract leakage signal at mutation.