09 Damage Prevention 2.0 : Analysis of operational data from an automated ROW airborne visual inspection of seven pipelines: crude oil leak detection, 3rd party encroachment detection and advanced image documentation.

Abstract

For years, pipeline operators have relied on human observers flying at low altitude to perform the mandated visual inspections of their right-of-way (ROW), often completed in single pilot-observer configuration, without any automatic documentation or detection system. This paper presents the first real-life results of a test campaign performed by Flyscan Systems over ROW’s of seven operators in seven US states and two Canadian provinces. Capabilities developed by Flyscan in collaboration with Enbridge include real-time hyperspectral leak detection and location and reporting of threats in the right-of-way. The paper will discuss the technology and system capabilities with details outlining detection of unauthorized 3rd party activity (machinery, abnormal construction activities), as well as generation of high definition 2D orthomosaics and 3D point cloud, vegetation analysis and digital surface ground mapping of the entire length of a 500-meter-wide ROW.

The paper will also review how new technologies can provide operators with a complementary tool that provides consistent, repeatable, and automated detection of high-priority threats to the integrity of pipelines. Statistics on detection performance will be presented, including unplanned threats and simulated (hidden) leaks. Real-life operational results will be covered including latency in detection, volume of data to be manipulated, cloud computing aspects as well as operational “up time” that can be expected. A development roadmap will be reviewed illustrating the path to fully automating all functions performed by human observers, including detection of ground movement, riverbank erosion, marker counting and localisation, exposure of pipelines after serious weather events, and automation of class location determination.

Read/Download PaperGradeAvailable after the event