Template Matching Technique for Unobstrusive Leak Event Detection in Oil and Gas Pipelines

Oil & Gas is the most common source to meet the ever-increasing energy demand in the world; thus, the pipelines run inter-country to inter-continents. One of the biggest challenges faced by the Oil & Gas companies is to monitor such long pipelines for leak events as unattended leaks may cause a catastrophe (environmental hazards, life loss, economic loss etc.). Negative Pressure Waves (NPW) can monitor leaks from a considerable distance. These pressure waves can travel a very long distance (similar to 10 km) and can be detected since low frequencies have significantly less attenuation in fluids. Most existing NPW techniques rely on invasive pressure sensors for their enhanced reliability. However, for existing industrial scenarios drilling pipelines poses a severe hindrance. Accelerometer-based NPW techniques, although more attractive to industrial use-cases due to their unobtrusive, easy-to-install nature and maintainability, come with their challenges. These sensors are susceptible to environmental noises and are generally likely to generate false leak event alarms during routine pipe maintenance jobs. This paper investigates the template matching technique for unobtrusive accelerometer-based leak event detection to isolate NPW signatures from other vibrational signatures on the pipe surface. More particularly, we setup a small pipeline with an air compressor to create leaks in a pressurized pipe and exploited an accelerometer sensor to pick up the NPW signatures on the pipe surface. Thus, the proposed approach offers a complete non-invasive yet robust solution for capturing low-frequency NPW waves.