Volume 56

Experimental Study on Carbon Dioxide Pipeline Leak Detection Based on Distributed Acoustic Sensing Dianqiang Xu, Shuai Wang, Jiateng Duan, Kaixuan Feng, Yu Liu

Abstract

Carbon dioxide (CO2) is typically transported in liquid or supercritical states, which are highly corrosive, posing a risk of corrosion leakage in transport pipelines. The function of detection technologies lies in the ability to promptly identify pipeline leaks, thus reducing potential risks.
This study includes an experimental investigation into the acoustical technology for pipeline leak detection based on Distributed Acoustic Sensing (DAS). The experiments were carried out within a 210-meter-long pipeline with an inner diameter of 44 millimeters, equipped with electromagnetic valves and circulation pumps. In this experiment, designed to validate the feasibility of the detection technique, pipeline leakage detection with supercritical carbon dioxide was performed.
It was validated through detection experiments involving different phases and leak severities that the feasibility of pipeline leak detection technology based on DAS is established. The distinguishing properties of acoustic signals between leak and normal states were explored using the spectrum subtraction algorithm (SSA) and correlation analysis approaches to compare the DAS signals. The results indicate that significant signal intensity differences exist in specific frequency domains for pipeline leak signals. Signals remain reasonably constant during steady operation because flow rates are consistent. When a leak occurs, large changes in signals near the leak location are noted, allowing leak identification based on differences in signal response times. During continuous leakage, the energy of signals at specific frequencies around the leak site is significantly higher than in areas without leaks.
This study may provide novel methodologies for low signal-to-noise ratio processing and bad point signal identification in the DAS signals, and establish a theoretical foundation for the engineering application of DAS in CO2 pipeline leak detection.

Keywords Leak detection, Distributed Acoustic Sensing, Spectral Subtraction Algorithm, Correlation Analysis