In order to prevent major plant accidents, a domestically unique technology has been developed that can detect pipe leaks from a distance and identify the location of the leak in real time. It is expected to contribute not only to improving the accident response capabilities of large plants, but also to ensuring the safety of workers.

The research team led by Senior Researcher Woo Jeong-han of the Korea Institute of Machinery and Materials (President Ryu Seok-hyun, hereinafter referred to as KIMM) under the Ministry of Science and ICT announced on the 28th that it has developed a technology that can detect the location of a pipe leak in real time from a distance using an acoustic sensor for the first time in Korea.

The research team built a system that can immediately estimate the location of a pipe leak by optimally distributing 4-channel acoustic sensor modules capable of direction detection. By checking the location information estimated from the 3D virtual space and CCTV within the system, workers can quickly identify the location and respond within the golden time in the event of a plant leak accident.

Previously, only acoustic pressure, which indicates the size of sound, was mainly used, and the exact location could only be identified by approaching the leak point from a close distance, which reduced worker safety. It also required the use of expensive equipment that utilized about 100 channels of acoustic sensors per module.

The core of the technology is an acoustic intensity-based signal processing technique that can minimize detection errors that occur in environments with significant reflection or reverberation effects through an automatic abnormal data removal algorithm.

It can detect with an error of less than 1m (estimated error of 3°) and confirm the location in a 3D virtual space within the system, allowing for a quick response to accidents caused by leakage.

Compared to existing equipment, the number of acoustic sensors per module has been reduced from 100 to at least 4 channels, miniaturizing the device and reducing the price by more than 30%. In addition, since leak detection results can be checked in real time from a distance, it is possible to detect and respond immediately to accidents in the early stages.

Senior Researcher Woo Jung-han of the Korea Institute of Machinery and Materials said, “3D acoustic intensity-based leak detection technology can easily identify noise sources occurring in various fields and analyze their characteristics, making it easy to identify the cause of accidents based on noise.” He added, “It can be expanded and applied to various industrial fields, such as preemptive response to incidents and accidents or other crime prevention alarm surveillance systems.”

This study was conducted with the support of the basic project of the Korea Institute of Machinery and Materials, ‘Development of AI-based machine system prediction diagnosis and accident response technology.’