Anomaly detection using flow and pressure

• Main Author: Kachireddy Lakshmi Prasanna
• Other Authors: Justin Dauwels
• Format: Theses and Dissertations
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/72563
• Institution: Nanyang Technological University

Pipelines are the major media for transport of oil and gas from a source to the target location in oil and gas industries. As pipeline networks are very complex, small changes in pressure and velocity values due to leaks and water ingression affects the total network. Leakages in a gas pipeline network are mostly due to inevitable aging, erosion, abrasion, sabotage and pipeline joints. The existing water ingression and leak detection techniques are manual which requires a lot of manpower, time taking and also results in the interruption of the normal supply of gas through pipelines. This demands software based real-time leak and water ingress detection techniques. Following the infiltration of water into the pipe, there is a mixed flow of water and gas which results in different flow regimes. In such cases, the real-time leak detection helps in detecting minor leakages, back pressure and other disturbances like ‘water ingression' in the gas pipeline network. The project not only aims at exploring feasible monitoring technology to identify, analyze and localize anomalies, such as gas leakage, water ingression or deformation but also to find out the potential failure prone zones and then to carry out preventive overhaul work. Studies are carried on different multi-phase flow regimes and are approximated into a single phase flow. These approximated single phase flows are simulated by using computational fluid dynamics simulation tool and compared with experimental results. Simulations were also carried out for different cases in the approximated flow regimes.