Pipeline monitoring and leak detection using system identification methods

点击翻译按钮获取中文摘要

We will use a distributed acoustic fibre-optic based sensor to monitor the condition of a pipeline and detect**leaks. We will develop data-driven modelling techniques that construct mathematical models of both the**propagation of acoustic signals along the pipeline, and the properties and locations of any acoustic signal**sources that affect the measurements. These acoustic sources will model the signals created by fluid flow in**the pipeline, pumps, vehicles travelling near the pipeline, leaks in the pipeline, and many other sources. Once**the sources have been separated, it will be possible to classify each as either a critical event, such as a leak, or a**non-critical or routine event. By monitoring the acoustic properties of the pipeline, it will be possible to infer**changes in its physical condition, and the locations of any such changes. The tools developed in this project**will allow for continuous monitoring and leak detection in pipelines that is precise and reliable.**The development of this technology has the potential to provide immediate social and environmental benefits**to Canadians. It will assist the Canadian government, the oil and gas industry, and pipeline operators in**obtaining the social license to better operate in all jurisdictions. The general public and specifically,**environmental groups and First Nations, will have the assurance that additional proactive measures are being**deployed to prevent pipeline leaks. More importantly, they will know that if a leak does occur, it will be**detected virtually as soon as it occurs, so that the size of the leak can be minimized and mitigation work can**begin almost immediately.