Pipeline parameter identification and leak localization using experimental data

Kamal Moustafa, Yousef Haik

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Pipeline systems are important in many fields of real life to distribute fluids from one location to another. Leakage from such pipeline systems poses serious problems from the technical, environmental and economic points of view. Early leak detection and localization is, therefore, important for real life applications. In this paper, an experimental study is conducted to collect the pressure head measurements at a number of nodes along a pipeline carrying oil for both the healthy and leaky cases. The experimentally measured data are utilized to identify the leak factor and coefficient of friction of the considered pipeline. The identified parameters are utilized by the propose localization scheme to determine the leak location. The identification is implemented by a window marching technique that uses the collected pressure head measurements and seeking the minimum objective function that represents the mismatch between the measured and numerically modeled pipeline variables. Monte Carlo simulation results are reported to demonstrate the effectiveness of the proposed parameter identification and leak localization techniques.

Original languageEnglish
Title of host publicationOperations, Applications and Components
PublisherAmerican Society of Mechanical Engineers (ASME)
ISBN (Electronic)9780791846063
Publication statusPublished - 2014
EventASME 2014 Pressure Vessels and Piping Conference, PVP 2014 - Anaheim, United States
Duration: Jul 20 2014Jul 24 2014

Publication series

NameAmerican Society of Mechanical Engineers, Pressure Vessels and Piping Division (Publication) PVP
ISSN (Print)0277-027X


OtherASME 2014 Pressure Vessels and Piping Conference, PVP 2014
Country/TerritoryUnited States

ASJC Scopus subject areas

  • Mechanical Engineering


Dive into the research topics of 'Pipeline parameter identification and leak localization using experimental data'. Together they form a unique fingerprint.

Cite this