Pipeline Leakage Detection Based On Acoustic Emission Techniques

As the energy crisis and environmental issues are becoming more and more serious, Carbon Capture and Storage (CCS) technology has been considered to be one of the most promising options to deal with the problem of global warming. For this technology, long distance transmission pipeline is often used for transporting captured CO2 to the storage site. Therefore, the research on leakage detection of the pipeline applied in CO2 conveying is of great significant. To master the CO2 gas leakage and diffusion can be useful to accurately predicted, evaluation and control it, reducing gas pipe accidents.In this paper, a model of gas continuous leakage and diffusion of pipeline is established. Theoretical analysis, numerical simulation and experimental demonstration are used in this paper.First, AE signals are analysed both in time and frequency domain in this paper. In time domain, the characteristics of burst and continuous signals are extracted. Meanwhile, the effects of material type of the pipeline are explored as well. In frequency domain, the distribution of the acoustic signals and noise signals are illustrated based on the principle of power spectral density.Second, the gas leakage diffusion turbulence model is established based on the principle of fluid jet and the simulation results are analyzed. Using Fluent to simulate the leakage of gas diffusion under different line pressure, leakage aperture and conditions of obstacles, analysis the influence of different factors on natural gas leakage diffusion. Experimental tests are condcuted to determine the spectrum of AE signals and type of sensors and capture device.Finally, in order to eliminate the background noises and extract the interesting characteristics of the AE signals, wavelet transform (WT) and empirical mode decomposition (EMD) algorithms are applied to reconstruct the signals. The traditional location method makes use of the phenomenon that Acoustic Emission (AE) signals propagate along the pipe when CO2 leakage takes place. Acoustic sensors mounted on the both sides of the pipe are used to sense the AE signals. Due to the limitations of the conventional method, the measurement results have large errors and randomness. Therefore, this paper focuses on the optimization of sensors installation and location algorithms.Based on the principle of cross-correlation, the time difference between signals from the two sensors can be calculated, then the leakage location can obtained through calibrating the propagation speed of the signals passing along the pipeline. In view of the frequency dispersion, phase shift and attenuation of AE signals, one option is proposed to improve location accuracy. Multi data fusion algorithm is used to combine the information of the signals from upstream and downstream sensors, which increase the fault tolerance performance of the measurement system. The measurement results suggest the maximum error in the estimation of the leak location is less than 5%, multi-sensors detecting is a good method to solve the leakage problem and make the system effectively. The localization accuracy is improved by data aggregation in multi-levels.