The risk of hydrocarbon and toxic spills increases with the aging of oil and chemical plant related infrastructure. The need for early detection of hydrocarbon and toxic chemical pollution is paramount, particularly in view of potential environmental damage, cleanup costs, and the loss of public confidence in industry’s ability to quickly respond to leaks.
Rigorous right-of-way monitoring, control of third party activities within proximity of pipelines, in conjunction with a robust preventative maintenance program is key to leak prevention. The first line of defense in the event of a leak is early detection and operational response to limit product loss from the pipe. Sophisticated mathematical modeling of flow regimes coupled with multiple pressure sensory relay devices in pipelines has increased the sensitivity of these leak detection technologies. However, despite these technological improvements significant leaks have occurred recently on major pipeline systems with damaging consequences.
Operators are challenged to interpret and respond to leak alarms in the absence of corroborating information. Frequent false or ambivalent “indications” can foster complacency, and worse, inaction.
The authors contend that reliance on a single technology for detecting leaks is imprudent and unacceptable in certain environments given the consequences of a late response to a loss of product from the line. Leak detection can be significantly enhanced by the application of molecular sensory technology in tandem with real time pipeline monitoring systems. The systems are synergistic and do not compete with each other.
This paper describes the development of the LEOS® leak detection system, its application and operational experience in high sensitivity locations. Four distinct applications are described, including: an Arctic subsea pipeline, an arctic above ground pipeline, a river crossing, and a buried onshore pipeline right of way. In the latter, a situation is described where a hydrocarbon leak was discovered on an adjacent third party pipeline not directly monitored by the system.