News and events

Back to News & Events
Posted: 24 Nov, 2017

Four Reliable Ways to Detect Pipeline Ruptures

Rupture alarms require more urgent action than a leak alarm. Therefore, rupture detection must be highly reliable and give zero or few leak alarms to earn the total confidence of the pipeline operations team. This should ensure their immediate action to shut down the pipeline as soon as a rupture alarm is generated.

Using leak detection systems alone to monitor for ruptures can leave operators more exposed to the risk of missing a pipeline rupture since when tuning these systems to detect small leaks some reliability is sacrificed. Most leak detection systems cannot achieve the high-reliability level of few false alarms per year desired by the best engineering practice in rupture detection. This emphasizes the importance of deploying a separate rupture detection system as an extra layer of protection, thus ensuring that the pipeline controller has sufficient confidence in the rupture detection system to act swiftly in shutting down a pipeline. There is an increasing pressure to automate the response to rupture events as much as possible, even to initiate an automatic pipeline shutdown triggered by a rupture detection system. In this regard, a 2015 paper by Martin Di Blasi, Norense Okungbowa, and Zhan Li of Enbridge Pipelines Incorporated (EPI) presented at the Pipeline Simulation Interest Group (PSIG) 2015 conference in New Orleans1, US stated:

"These types of performance expectation are challenging to achieve with conventional, model-based, leak-detection systems (i.e. CPM-RTTMs) as the reliability measured in terms of the false alarm rate is typically too high to initiate automatic pipeline shutdown."

These authors confirm that although leak detection systems can detect ruptures, the generally low reliability of these systems (e.g. 1 or 2 false alarms a day for long, complicated pipelines) may have played a role in some past rupture incidents. The frequency of these false alarms may have caused pipeline controllers to ignore them, or not pay them enough attention.

The Atmos Rupture Detector is a preferred and proven choice that requires minimal changes to a SCADA system and uses comprehensive detection algorithms to detect ruptures occurring everywhere along a pipeline (i.e. without the 20-km limitation of most SCADA-based rupture detection systems).

Atmos Rupture Detector uses four separate algorithms to detect ruptures with high reliability. A rupture occurrence not covered by one algorithm will be detected by another of the four algorithms.

  1. Low KL method detects ruptures near pump stations where the product is easier to pump
  2. Inventory method detects ruptures on pipelines that can go slack - what comes out is less than what goes in
  3. Dynamic model analysis method uses pattern recognition to detect ruptures in all areas of a pipeline
  4. Pressure difference analysis method compares intermediate pressures along then pipeline to detect ruptures in all areas of a pipeline


1 Di Blasi M., Okungbowa N., Li Z. "Overview of Enbridge's rupture recognition program". Pipeline Simulation Interest Group Conference. New Orleans, USA. 2015.