Research On New Method And Key Technology Of Leak Detection For Crude Oil Pipelines

Pipeline leak detection technologies have been playing an important role in protecting the safety of pipeline transportation. In recent years, new requirements are proposed in the pipeline transportation field, such as improving location precision, detecting small leaks and reducing false alarm rate. On the one hand, the real-time model method based on transient flow is the focus of international research on pipeline leak detection methods in recent years, because even small leaks will produce a change of the pressure waveform in a pipeline transient event. Compared with other methods, the method mentioned above could get a better result of leak location. On the other hand, the negative pressure wave (NPW) method which does not require a complicated mathematical model, with the advantages of small construction quantity, low cost and convenient maintenance, has been widely used in the field of leak detection of long-distance pipelines. At present the length of crude oil pipelines installed with leak detection system based on the NPW method has already surpassed 10,000 km in China, which has played a significant role in the pipeline running safety. However, the NPW method has some shortcomings in reducing false alarm rate and improving location precison.Aiming at solving the problems existed in present leak detection technologies of crude oil pipelines, a new pipeline leak detection method based on an additional dynamic micro-pressure excitation is put forward, and a number of key technologies of leak detection for crude oil pipelines are studied thoroughly. The dissertation mainly covers the following aspects:1. A new pipeline leak detection method based on an additional dynamic micro-pressure excitation is proposed, particularly suitable for the detection of small leaks, which can detect the smallest amount of 0.1% of total flow.2. A high sensitive dynamic pressure transmitter is used to capture the transient signal caused by a leak, of which the performance is verified through field experiments.3. The signal feature extraction methods based on wavelet package entropy and the signal feature extraction method based on chaotic characteristics are applied to extracting the feature of the dynamic pressure signal along the pipelines.4. Aiming at the leak detection based on the NPW method, the signal processing method based on high-order statistic analysis and the signal processing method based on unscented kalman filter (UKF) are used to improve the location precison.5. A number of key technologies on combination of pipeline leak detection system with SCADA (Supervisory Control and Data Acquisition) system are investigated, and a leak detection system based on SCADA is designed, of which the reliability and accuracy are validated through field experiments and practical applications.