Perpustakaan Digital ITB
Pipelines are crucial tools in transporting fluids to designated locations. It is almost impossible that a pipeline
does not face any problems. One of the most common problems is a leak. Leak can be caused by corrosion, wax,
and third-party damage (illegal tapping). In field X, located in South Sumatra, 54 cases of illegal tapping were
found from 2014 to 2023. This condition cannot be left unresolved any longer because even a minor leak may
lead to environmental damage, risk to human safety, and economic loss. This raises the urgency of detecting a
leak for preventive actions to be implemented before a leak happens and fluid loss can be minimized. This study
develops a dynamic transient simulator-based model to address pipeline leak detection by simulating transient
flow behavior under both shipping and stop shipping conditions.
The methodology of this study includes three parts: preliminary study, simulation, and interpretation. A
literature review was conducted to define the study objectives. Data from field drain tests, including pressure,
flow rate, and temperature, were collected from a pipeline in South Sumatra under both shipping and stop
shipping conditions. These data were input into a dynamic transient simulator to build the model. Sensitivity
analyses were performed, and model predictions were compared to real leak data to assess accuracy. The results
were interpreted to refine the model and improve leak detection in pipeline systems.
A Leak was intentionally introduced with specific leak rates, releasing 300 liters of condensate followed by a
30-minute standby, then 600 liters with another 30-minute standby. Data was collected on October 24, 26, and
28 2024 under both shipping and stop shipping conditions. In shipping, fluid flow was driven by the pump,
while in stop shipping, the pump off and residual pressure caused minimal fluid movement.
At the end of this study, the model was successfully developed and validated with field drain test data, showing
accurate replication of pipeline behavior. The model achieved a pressure error of 1.92% under shipping
conditions and 11% under stop shipping conditions. Pressure trends during a leak showed distinct drops and
ripple. The model accurately predicted leak size and location with a maximum error of 5%, offering an effective
tool for pipeline leak detection.
Keywords: leakage, illegal tapping, shipping, stop shipping, simulator-based model