Machine learning and acoustic method applied to leak detection and location in low-pressure gas pipelines

The increase in pipeline safety would prevent incidents that can result in fatalities, environmental disasters, and economic losses. The present work proposes a technique that combines acoustic sensors and machine learning algorithms to identify and locate leakages in low-pressure gas pipelines. The patterns on the sound signal captured by microphones were used to accomplish those two tasks. The technique aims to solve two persistent problems, the detection of small leakages on pipelines operating under low pressures and the reduction of false alarms in the presence of external disturbances. The experimental results showed that the method identified 99.6% of the leakages and achieved a rate of false alarms of 0.3%, while the position of the leakages was estimated with a maximum location error of 4.31%. These results show that the technique proposed is an efficient and reliable alternative to monitor low-pressure pipelines. Graphic abstract