Frequencies of leaks increase with aging pipelines and environment factors. Consequences of leaks can be severe because population densities near pipeline routes have increased and also because of environmental concerns. Thus, the uncertainty concerning when and where leaks would occur becomes a major consideration in identifying the leak-prone segments. This paper presents a statistical method based on the Bayes? theorem, to assess the probability of leaks for different portions of a pipeline system as a function of significant characteristics. The illustrated case study, in which this method was applied successfully, describes a Bayesian diagnostic model and its application to a cast iron gas pipeline system.
A Bayesian diagnostic model for pipeline leak prediction
Diagnosemodell auf Basis der Bayes Regel fuer die Vorhersage von Leckagen in Rohrleitungen
Pipeline Engineering Symp. 1987, Petroleum Div. of the ASME ; PD-6 , Feb ; 21-28
1987
8 Seiten, 3 Bilder, 4 Tabellen
Conference paper
English
Model based pipeline monitoring with leak detection
Tema Archive | 2009
|Model-based pipeline leak detection and localization
Tema Archive | 2003
|Reconfigurable Model for Multi-leak Location in a Pipeline
Tema Archive | 2003
|Post-earthquake pipeline leak detection technologies
Tema Archive | 2008
|Canadian pipeline installs leak-detection system
Tema Archive | 1988
|