Multi-parameter Analysis Based Research In Oil Pipeline Leak Detection And Location

In view of the advantages of cheapness, convenience and safety, pipelinetransportation has been widely used in the oil pipeline transportation. With the servicelife increasing, natural corrosion, and the damage of human factors, pipeline leakaccidents occur frequently. Pipeline leak interferes with normal production and causesa great deal waste of resources, serious environmental pollution and huge loss of lifeand properties. It is the sudden large leak which is still the most harmful toenvironment and causes the most wasteful resources and most serious economic lossin modern pipeline leak accidents. The pipeline leak detection and location technologyhas become an important technology of pipeline fault diagnosis.On the basis of previous study results, the aging pipeline at home is studies.Based on the mature Supervising Control and Data Acquisition (SCADA) system, amulti-parameter oil pipeline leak detection and location method are proposed, inwhich the sudden leaks caused by corrosion, theft and natural disasters, etc. areanalyzed and the better results are achieved. The research works are accomplished asfollows:(1) Aiming at the changes of flow, temperature and pressure in the condition ofpipe line leak, this paper introduced the principle for ultrasonic flow detection method,temperature gradient method and negative pressure wave method in oil pipeline leakdetection and location. On the basis of the existed SCADA system, a multi-parameteranalysis scheme for oil pipeline leak detection and location is proposed, by whichaccurate detection and location can be achieved.(2) Based on the ultrasonic flow detection method in the oil pipeline leakdetection scheme, an improved method of generalized harmonic wavelet packetdecomposition to extract the major frequency signal and chaos oscillators array todetect the frequency are proposed. The simulation results of the ultrasonic signalcollected in the laboratory prove that the major frequency can be detected effectivelyand pipe line flow can be calculated exactly, and then oil pipeline leak can be detected.It brings new ideas to oil pipeline detection.(3) In light of the pressure signal detection and location method in oil pipelineleak detection scheme, singular value decomposition is first adopted to extract thenegative pressure character signal collected from SCADA system to judge the oil pipeline leak; After confirmed pipeline leak occurring with other methods, recurrenceplots in pattern recognition with average grayscale method is adopted to locate oilpipeline leak, and contrasted with wavelet modulus great value method, the feasibilityof the scheme is verified, which provides a new way to oil pipeline leak detection.(4) It is proved by the comprehensive analysis of simulation results oftemperature, flow and pressure that the scheme of multi-parameter analysis of oilpipeline leak detection and location is feasibly, which provides a new scheme to oilpipeline leak detection and location.