Font Size: a A A

Research On The Signal Propagation Characteristics Of Acoustic Emission And Localization Of The Gas Pipe Network Leak Based On Multi-level Method

Posted on:2018-04-21Degree:MasterType:Thesis
Country:ChinaCandidate:P GongFull Text:PDF
GTID:2381330596468660Subject:Safety science and engineering
Abstract/Summary:PDF Full Text Request
Pipeline network is the most important oil and gas transmission element in the petroleum industry.More and more countries pay attention to the establish of pipeline network system in the period of city planning.With the increase of the usage of pipeline network,many factors like aging,bad sealed node connection,corrosion,environmental corrosion,wear,fatigue and defects would lead to pipeline leakage during operation.Acoustic emission(AE)technique can achieve early detection,the leakage warning,and pinpoint the leakage source promptly,which could prevent accidents in advance and reduce the losses caused by leakage.In conclusion,AE technique can provide instrumentalizaion and technicalization support in the oil and gas pipeline network system.COMSOL software is used to simulate the transmission characteristics of leakage signals under different process parameters and medium types by building the three-dimensional model of pipeline network.We can study the factors which would affect the leakage signals like different collecting locations,pressures,leakage source locations and medium types.The results of COMSOL simulation could provide guidance for pipeline network experimental design.Based on the gas and liquid leak and diffusion large-scale experimental system of Safety Engineering in China University of Petroleum(East China)with advantages of adjustable gas parameters,readable flow and so on,we carried on a series of experiments to collect AE leakage signals and study the transmission characteristics of leakage signals.At first,according to the characteristics of normal and leakage signals,we can estimate the pipeline leaks or not.Combined with signal processing methods,we can achieve the early detection of small leakage.Simultaneously,the influence of factors like medium types,flow and pipeline network components on signal amplitude attenuation is studied.At last,the results of experimental data are compared with COMSOL simulation results and a good agreement is achieved.Finally,a method called multi-level leak source localization was presented.This method contains three steps: the first step called Modified Subtract on Negative Add on Positive(MSNAP)is to achieve the coarse localization of region.This method can transform the large,complex spatial topological pipeline network into the localization of small plane pipeline network.Based on the construction of likelihood matrix,the maximum value of the likelihood matrix shows the leakage source in the plane pipeline network;The second step called multi-information fusion technique can achieve the accurate localization of region.It can transform the localization problem of small plane pipeline network into the localization of straight pipeline;The last step is called precise localization,which contains single sensor based localization method and modified long-range pipeline localization method.Single sensor based localization method is suitable for the localization of small branch.Based on modal acoustic emission theory,we obtain the frequency dispersion curve of guided waves,then combined with empirical mode decomposition-Hilbert Huang(EMD-HT)and Gabor wavelet methods,the velocities and arrival times of guided waves are calculated.A maximum relative localization error of 6% is achieved;The modified long-range pipeline localization method makes use of the attenuation characteristics of leakage signals to determine the region of leakage source.And then combined with wavelet packet decomposition method and cross-correlation function,we can calculate the source-sensor distance.A maximum relative localization error of 5.3% is achieved.
Keywords/Search Tags:Acoustic emission(AE), Pipeline network, leakage, COMSOL, transmission characteristics, Multi-level leak source localization
PDF Full Text Request
Related items