An analysis of leak indications, alarm management, alarm analysis and alarm reviews.
Analysis of a Leak Indication
Certain leak indications may require an immediate shutdown of the pipeline while other indications may dictate that the Pipeline Controller analyze the current pipeline operation or escalate the investigation. The magnitude of the leak and the level of risk involved may be the key factors in the decision.
Leak Indications Requiring Immediate Shutdown Response
Alarms that are clear and credible such as a clearly defined signature, or where a Pipeline Controller believes the leak indication or the indication is highly reliable.
Some operators may have the policy to shut down whatever the alarm.
Leak Indications Allowing Additional Analysis before Shutdown
An operator may have a second category of alarm response where alarms require timely investigation and preparation for shutdown. These instances include, among others, alarms not supported by the hydraulic conditions, loss of communications, alarms that coincide with start-up/shutdown or rate changes, and a number of others.
Validating the Leak Indication
Examples of leak validation include, but are not limited to; hydraulic calculations, pressure and flow trend monitoring (trending), visual inspection, and pressure testing where possible.
Alarm Management
Alarm management employs tuning and threshold setting methods driven by pipeline analysis, data collection, and review. Alarm management should aim at increasing Pipeline Controller responsiveness by increasing the reliability of alarming and maintaining the LDS performance.
Alarm review
Alarms should be analyzed with the goal of increasing the confidence of the alarms. Thresholds changes and tuning can be used to optimize sensitivity versus reliability.
Categorization
Alarm categories may be:
- Alarms that required immediate action to shut down the pipeline
- Alarms that required an immediate investigation and preparation to shut down or lower-credibility alarms
- Alarms that were proven to be non-leak alarms
Other possible alarm categories or subcategories may be cause-based, e.g.
- Data failure (e.g.) meter failure, communications failure to pressure transmitter, temperature out of range, or meter prove error.
- Operational issues such as filling a new spool piece on the manifold, new product type from supplier, fluid at much warmer temperatures than normal, or instrument calibration.
- Modeling/tuning issues where the non-leak alarms generated could be prevented by tuning or adjustment of the LDS.
Note: API 1130 divides alarms into three classes: data failure, irregular operating condition, and possible leak.
Alarm Review
The goal is to improve learning, to increase the number of clear and credible alarms and reduce uncertain alarms and overall improvement. Another objective to see if threshold changes are required.
The pipeline operator performs both short-term periodic reviews (daily, weekly, or monthly) of alarms and long-term periodic review (for example, a five-year cycle) of alarms. The evaluation may indicate what to do strategically in managing the types of alarms. This information then feeds into improvement planning.
Short-term Periodic Review (Daily, Weekly, Monthly)
May include
- Review of alarms, threshold trends, and imbalance trends
- Analysis of imbalances, threshold, line pack and meter over/short (flow balance) during pipeline start-ups and shutdowns, pump starts/stops, movement changes, valve close/open, column separation condition, process variable changes, etc.
- Analysis of measurement trends
- Review of response to the alarms and the procedures
Review may result in recommendations to improve the alarm response or repair instrument or equipment failures (e.g. PLC failures) etc.
Long-term Periodic Review
The time between the long term reviews should not exceed five years. The review should assess the alarm performance and thresholds from the perspective of sensitivity and reliability with respect to the KPIs and performance metrics.
In addition to the KPIs, the review may include accuracy, reliability, sensitivity, robustness), actual leaks, simulated leaks, equipment needs, changes operations need to help leak detection (eg. Avoid slack) and lessons learned.
Actual Leaks
The LDS response should be analyzed and documented. General classifications for review of actual leaks may within scope (the leak was big enough to be detected and was or was not detected) or out of scope (too small to be detected). Report the lessons learned. If the LDS failed to detect a leak that is within its scope of the LDS, investigate and identify corrective actions.
References
- American Petroleum Association. (2014). Recommended Practice 1175, Pipeline Leak Detection Program Management (2st ed.). Washington, DC: Author.
A Look Inside API Recommended Practice 1175 Series
A Look Inside API Recommended Practice 1175, Part 1
A Look Inside API Recommended Practice 1175, Part 2
A Look Inside API Recommended Practice 1175, Part 3
A Look Inside API Recommended Practice 1175, Part 5
A Look Inside API Recommended Practice 1175, Part 6
A Look Inside API Recommended Practice 1175, Part 7
A Look Inside API Recommended Practice 1175, Part 8