Detecting leaks in a liquid pipeline is not the most difficult task for a leak detection system (LDS); detecting leaks without giving false leak alarms is the main challenge. An operator will have trouble identifying a real leak if he has to sift through many false alarms. Therefore pipeline leak trials should test the reliability (number of false alarms) of a leak detection system as well as its ability to detect real leaks. This paper reviews how a number of pipeline operators tested their leak detection systems with simulated leaks, verifying the reliability as well as the sensitivity of their new leak detection systems. These simulated leaks were introduced by removing product from the pipeline by bleeding. The paper also outlines a simple table based on the API 1155 guidelines to evaluate software based leak detection systems that can be used as part of the bid evaluation process to hold the leak detection vendor accountable to deliver the performance promised in his bid proposal. This paper high-lights some of the performance limitations to watch for when selecting and testing an LDS, For example; will a pipeline leak detection system detect the quoted minimum leak if the normal operations include transients? Does the system block leak alarms to reduce frequent false alarms? Are the leak detection times based on the time it takes to declare a “Leak Warning” or on the time it takes to declare a “Leak Alarm”? Finally, the paper discusses how to perform more realistic leak tests.

This content is only available via PDF.
You do not currently have access to this content.