A Practical Implementation of Pressure Transient Analysis in Leak Localization in Pipelines

In the present study, a more realistic approach for using pressure transient analysis in leak detection and localization is proposed. In a previous publication [1] by the authors, the feasibility of using pressure transients, generated by full closure of a downstream solenoid control ball valve, in leak detection and localization is investigated. The main shortcoming of using the full closure of a downstream valve is the very high pressure rise that may reach 14 times the operating pressure. Also, full valve closure yields to discontinue the whole pipeline flow. In the present paper, a controlled partial downstream or upstream valve closure is used as a mean of generating pressure transients to overcome the above drawbacks. The percentage of the valve closure is controlled to reduce the pipeline flow rate by 20–80%. Pressure transients generated by a partial valve closure are investigated experimentally and numerically. The experimental setup consists of a 60 m long and 25.4 mm internal diameter PVC pipelines connecting two tanks. Leaks are simulated at different locations along the pipeline to investigate the effect of leak positions. The pressure time history is recorded using piezoelectric pressure transducers located at five equidistance points along the pipeline connected to a Data Acquisition System. Experiments are carried out for different leak quantities ranging from 2% to 20% of the pipe flow rate. The numerical model accounts for complex pipe characteristics, such as unsteady friction and viscoelastic behavior of pipe walls. The leak is treated as a flow through an orifice of prescribed size. The numerical model is experimentally verified to insure the capability of the model in accounting for unsteady and viscoelastic complex phenomena and efficiently simulating pressure transients in the presence of a leak.