Monitoring and assessment of the operational and structural conditions in steel pipelines

Abstract

Oil and water transport pipeline systems require ongoing maintenance and inspection as they are susceptible to damage due to harsh environmental conditions and operational factors. In this study, the monitoring and assessment of pipelines is performed using a network of Fiber Bragg Grating (FBG) sensors mounted along the longitudinal and circumferential directions. The sensitivity of the measurements to assess pressure and flow variation in the pipe, in addition to leakage detection and localization were evaluated. This was validated experimentally in our laboratory on a six meters pipe. Water at a controlled pressure and a controlled flowrate was pumped into the pipe section. Leakage was simulated by opening one of the four designated valves installed on the pipe and spaced at 1.5 meters intervals. Experiments were performed for various levels of pressures ranging from 0.6 to 1 bar, and flowrates reaching 20 gallons per minute (GPM). Moreover, binary and multi-class Support Vector Machine algorithms (SVM) were implemented, using the collected data, to assist in the prediction of the operational and structural conditions of the pipe. The variation in the pressure inside the pipe, from 0.6 bar to 1 bar, highly impacted the amplitude of the measured strain increasing it significantly reaching 20percent. The change in the flow rate, however had an inverse effect, where a 5 GPM increase in the flowrate resulted in a 5percent decrease in the amplitude of the measured strain drop. The change of leakage hole size greatly influenced the measured signal, resulting in a 55percent change in amplitude between a 2 cm2 and a 5 cm2 leakage hole. As for the location of leakage, only the temporal aspects of the signal was affected resulting in a slight shift in the response time of sensors relative to each other. The developed SVM classifiers reached accuracies of 88percent for flowrate classification, higher than 95 percent for pressure classification, and 100percent for leakage size classification. On the other hand, the localization of leakage accuracies did n