Study On The Leakage Detection And Pre-warning Techniques Based On The Distributed Optical Fiber For The Long-distance Oil And Gas Pipelines

The sustained high-speed national economic development makes the demand of energy supplies especially oil and gas resources increasingly urgent. By virtue of a great many advantages, pipelines have become the principal means of oil and gas transportation. Sometimes, however, pipeline leakage takes place due to some natural or artificial damages. Leakage accidents may cause loss of life and properties along with environmental pollution. Presently the prevalent methods to judge leakage are evaluation of parameters such as pressure in pipeline, flow rate, and temperature obtained by common oil and gas pipeline leakage detecting devices to infer whether a leakage has occurred. These methods are subject to the quality of the material transported, transportation process employed and other factors. In addition, warning of leakage exclusively arrives after the leakage.Based on the previous team work, this dissertation involves the in-depth study on the leakage detection and pre-warning technique based on distributed optical fiber for the long oil and gas pipeline. It also constructs a leakage detection and pre-warning system based on distributed optical fiber for the oil and gas pipeline. Aside from detecting and locating the leakage points, the system can also perform pre-warning monitoring and locating when the abnormal events dangerous for the oil and gas pipeline take place.The major study of this dissertation covers the following aspects:1. It elaborates the principle of distributed optical fiber vibration sensor, pressure sensitivity, and factors affecting pressure sensitivity. It also analyzes to what extent the instability of polarization state of light wave has impact on the sensor in the single-mode optical fiber, and compares several approaches to the this issue.2. It applies the"energy- pattern"signal feature extraction method based on the wavelet package decomposition, the EMD (Empirical Mode Decomposition) based signal feature extraction method, and the signal feature extraction method based on the analysis of multi-dimensional chaotic characteristics to extracting the eigenvectors of the detected signals along pipelines.3. It applies the RBF (Radial Basis Function) neural network approach and SVM (Support Vector Machine) approach to recognizing the types of the detected signals along pipelines. 4. A set of experimental devices are designed to verify the performance of the pattern recognition and locating techniques in leakage detection and pre-warning system of the distributed optical fiber oil and gas pipeline through a great many field experiments. The experiment result demonstrates that the technique has a positive effect on monitoring and pre-warning as well as identifying and locating the abnormal events including leakage along pipelines.5. A leakage detection and pre-warning system based on distributed optical fiber for the oil and gas pipeline has been accomplished, which is being operated in a product oil pipeline. The results show that the system is able to detect the abnormal events in real-time and accurately recognize the types of them and locate the places where they take place.