We believe the best end to end test of a leak detection solution is through fluid withdrawal tests. Through our continued partnership with Bridger Pipeline, we have been given unparalleled access to conduct dozens of such tests on a variety of pipelines and under various scenarios.
This has allowed for verification and validation of the Flowstate LDS solution as well as provided valuable data for research and development.
Test have included:
Line Types
Transmission Lines
Gathering Lines
Crude & Diesel lines
Leak Detection Methods
Statistical Volume Balance (SPRT)
Leak Signature Recognition
Rupture Recognition
Hydraulic Conditions
Packing/unpacking
Flow rate changes
Pump/Set-Point Changes
The image below shows some results of testing for different types of lines – illustrating our achieved Time to Detection for various leak sizes:
Our testing activities include tests designed to stress the system or verify success in challenging conditions. Here are a couple examples of some of the stress tests we’ve run.
In this example, oil is withdrawn at just 40 bph to simulate a 1% leak. Although the leak is small, the Flowstate Signature Recognition model is raised an alarm in just 8 1/2 minutes. The Statistical Volume Balance model raised an alarm in 30 minutes.
In this example, a leak starts prior to flow initiating on the line. Flow begins 20 minutes into the leak and takes some time to “pack” the line. Despite these conditions, the leak is detected in less than 2 1/2 hours.
In this example, a gathering line with 4 intermittently active inputs experiences a leak with many units coming on and off. Despite these conditions, the leak is detected in under an hour (50 min).
Jump back to the top and download several examples in the Case Study.
