Application of Mass Balance and Transient Flow Modeling for Leak Detection in Liquid Pipelines

Abstract Effective methodologies implementing proper flow hydraulics simulation approaches are presented for accurate leak detection in natural gas liquid (NGL) pipelines under steady- and transient-state flow conditions based on mass and force balance principles and head and flow measurements taken at pipeline end-points. Leak detection curves are generated using a process simulation package under steady-state operation conditions for systems with prescribed fluid compositions, pipeline characteristics, and instrumentation accuracy. Real-time transient-state leak detection along a pipeline is accomplished by simulation of pipeline fluid conditions with a proper flow model. Typical transient-state flow episodes are demonstrated for pump-start up and valve closing cases. Real pipeline data are facilitated in the leak detection model by implementing a time-marching algorithm. Deviations of patterns for measured inlet and outlet flow conditions and heads are facilitated for detection of occurrence and locations of leaks. It is shown that the response time depends on the leak location and has the lowest value for leaks occurring near the pipeline middle-point, leak location can be predicted based on the time required for detection of a discrepancy, and estimation of leak location is more difficult when the mass of pipeline fluid and hence resistance to flow increase. Proposed approaches can be instrumental for critical review and assessment of the physical characteristics, limitations, capabilities, and instrumentation adequacy of pipelines for safe operation and proper selection of effective leak detection technologies and instrumentation upgrades with improved response time and accuracy required for pinpointing of leak locations and estimation of leak sizes.