Study On A Real-Time Computer Leak Monitoring System For Oil Pipelines

Oil pipeline leak detection and location is an important topic for reducing oil loss and environment protection. Pipeline leak has been a big problem for many oil fields in our country due to pipeline aging and local destructors who steal oil from oil pipelines. Thus there is an urgent need for oil pipeline leak detection and location system in our country. Actually, research work on leak detection methods has been undertaken domestically and abroad for several decades. However, there are still not many real systems that have been successfully used. Especially that, since there is no flow meter in most stations in our country, many leak detection methods cannot be put to use. In view of all these conditions, we developed a "real-time computer leak monitoring system" that is especially applicable in our country.Centered on the development of this leak detection system, the first part of this thesis explores leak detection and location methods adopted by the system: Correlation Analysis and Wavelet Transform. Discussion covers the essential ideas of these methods and their computer realization. In order to improve leak detection sensitivity, a practical adaptive filtering algorithm is proposed for filtering pressure signal; in order to reduce false alarm, this thesis also discusses/proposes several methods for disturbance recognition/elimination, hi the second part of the thesis, in-depth description of the hardware and software structure, as well as the software design and function has been given. The results of two field experiments are shown.Test run and field experiments prove that, with microwave telecommunication device, this system on-line monitors leaks of oil pipeline; it can recognize operating disturbances automatically; this system detects, locates leak point within 3 minutes after its occurrence, geographical information is shown on the computer screen and alarm is running when a leak is detected; this system adopts Correlation Analysis and Wavelet Transform to detect leak simultaneously; with only two pressure signal, it reaches a satisfying level of leak detection sensitivity and reliability; the minimal amount of leak that can be detected is 5m3/h, approximately 0.6% of the total flow rate, location error is about 2% of the total length of the pipeline. This system is taking the lead in our country and reaches the advanced level of similar systems abroad. It is going to be installed on several other major pipeline stations in Sheng Li Oil Field.