Rupture Alarm Detection

Large-volume, uncontrolled releases of hazardous liquid. That’s how ruptures are described in PHMSA’s recently proposed rulemaking. And it describes a shared nightmare facing all pipelines.

Although pipelines remain the safest way to transport oil, they are inherently susceptible to rare but damaging events. Ruptures can be caused by line strikes, material fatigue or corrosion, or even natural occurrences such as landslides. If not detected promptly, these events can have catastrophic consequences.

Flowstate offers leak detection built to catch a variety of anomalies under a range of scenarios. But to help ensure a line rupture never goes undetected, we have added a Rupture Detection Model developed specifically to detect large pressure and flowrate changes that can be signs of large release. Utilizing only flow rate, pressures, and operational status (pump status, set point change), the model can detect the occurrence of a rupture in just minutes.

The proposed amendments to the hazardous liquid Mega Rule regarding rupture establish minimum standards for identification of ruptures and the initiation of mitigative actions.

The new rulemaking seeks to do these things:

1. Define a rupture
2. Require procedures for responding to a rupture
3. Require that a “Rupture” is declared within 10 minutes of notification or indication
4. Require immediate external notification
5. Require mitigating response within 40 min after Rupture identification

In other words, operators need to know fast and certain because they are going to be required to take extreme measures to mitigate the potential damage done by rupture-like releases.

As a leak detection provider, we understand this means something very important for our customers: a rupture alarm must be distinct and reliable; there is no room for false alarms.

Preceeding the proposed rulemaking from PHMSA, API and AOPL provided guidance on Liquid Pipeline Rupture Recognition and Response in a whitepaper. In the guidance, it is noted that the Rupture Recognition and Response Team focused on ideas to:

• Enable exiting SCADA systems to “Scream Rupture”
• Distinguish rupture from other alarms
• Enable rupture detection for “lightly instrumented” or legacy SCADA, without significant upgrades and investments

It is understood that these goals would promote and enable rupture detection in a broad swath of pipelines and elevate rupture alarms to a heightened priority of recognition and response. We have built the Flowstate LDS Rupture Alarm with these goals in mind.

Due to the severity of the event, it is expected that the Flowstate LDS signature recognition model will identify a rupture very quickly. However, guidance issued by the API/AOPL recommends that rupture alarms be distinguished due to their severity. And, to insure that there are no false alarms, the Rupture Detection model takes just a few short minutes to ensure that certainty.

The following shows data from a validation test performed using a commodity withdrawal test. Crude oil was withdrawn at a rate of 12% using a leak skid . It can be seen that a Leak Alarm was issued just 2 minutes after the leak began.  Because the Rupture Detection model is configured to ensure an imbalance persists – ruling out operational activities – the Rupture Alarm is not issued until 9 minutes after the leak began. At this point though, the combination of a large pressure/flow drop and growing/sustained imbalance is reliable evidence of an abrupt, large volume release. Under the Proposed Rulemaking, when the alarm is issued, the operator has 10 minutes to declare a Rupture and 40 minutes to shut it down.