Investigating the performance of transient model based leak detection system for liquid pipelines

点击翻译按钮获取中文摘要

Canada's pipeline network is massive with approximately 3 million barrels of crude oil transported every day. Even with extensive federal and provincial regulation, and pipeline companies undertaking many measures to maintain and monitor their pipelines to ensure safe operations, there is always a possibility that a pipeline leak could occur. Oil pipeline leaks can result in significant environmental impacts and huge economic losses, which highlights the need for effective leak detection to facilitate fast response to a leak event. The occurrence of a leak causes a transient event, in the form of pressure waves that propagate in both directions along the pipe. The presence of a leak can be detected by analyzing the hydraulic behaviour of the pipeline system using a computer model. Computer model based leak detection techniques are extensively used in the energy pipeline industry but their effectiveness is often limited by the ability of the computer models to accurately reproduce the complexities of real-world pipeline systems, and by inaccurate and insufficient measured data.