UNDERLYING THE PERFORMANCE OF REAL-TIME SOFTWARE-BASED PIPELINE LEAK-DETECTION SYSTEMS

This paper analyses the underlying approaches of two software methods to detect pipeline leaks and estimate their location. One method is based on real-time simulation, and the other is based on real-time statistical analysis. Almost any leak-detection system can detect pinhole leaks if given precise data, but in practice these data contain uncertainties complicated by random errors. If sensitivity is made too great, too many false alarm indications are given when in fact there is no leak. The paper discusses the underlying factors influencing the performance of leak-detection systems, and does not refer to commercial products, except where they are specifically mentioned. Before comparing the two software methods, the paper considers existing pipeline infrastructure, leak detectability importance and limits, and leak analysis and methods. Reference is made to WE Deming's concept of systems thinking and statistical process control in relation to quality assurance and management. Real-time simulation methods depend more on the absolute accuracy of the smooth SCADA information acquired, and need to perform extra data tests. Statistical leak-detection methods reduce complexity, thus making it progressively easier to identify leaks; they observe the real situation, and apply Deming's philosophy of total quality management.

• Corporate Authors:

  Scientific Surveys Limited

  P.O. Box 21
  Beaconsfield, Buckinghamshire,   England 
• Authors:
  • AL-RAFAI, W
  • BARNES, R J
• Publication Date: 1999

Language

• English

Media Info

• Features: References;
• Pagination: p. 44-51
• Serial:
  • PIPES & PIPELINES INTERNATIONAL
  • Volume: 44
  • Issue Number: 6
  • Publisher: Scientific Surveys Limited

Subject/Index Terms

• TRT Terms: Administration; Computer programs; Detection and identification; Detectors; Errors; Mathematical analysis; Methodology; Pipe; Quality assurance; Real time control; Real time data processing; Seepage; Sensitivity; Simulation; Statistics; Stochastic processes; Warning systems
• ITRD Terms: 155: Administration; 6471: Analysis (math); 9115: Detection; 6440: Error; 9102: Method; 3361: Pipe; 9123: Quality assurance; 8679: Real time; 4315: Seepage; 5928: Sensitivity; 9103: Simulation; 8645: Software; 6555: Statistics; 6574: Stochastic process; 9113: Warning
• Subject Areas: Administration and Management; Data and Information Technology; Hydraulics and Hydrology;

Filing Info

• Accession Number: 00790343
• Record Type: Publication
• Source Agency: Transport Research Laboratory
• Files: ITRD
• Created Date: Apr 11 2000 12:00AM