AI-driven methane monitoring for LNG assets

OPTICAL GAS imaging (OGI) cameras paired with AI plume detection can complement traditional sensor systems, helping raise the standard for visual methane leak detection. 

In 2025, a public corporation in South Korea responsible for domestic LNG operations introduced a Flir OGI solution to strengthen methane leak monitoring at Boryeong LNG Terminal’s pressure regulation facilities.

The system combined real-time OGI video with an AI-based plume detection engine to support safer, more effective monitoring, particularly in outdoor, unmanned environments where conventional sensors can be less reliable.

Before installing the system at the LNG pressure regulation site, the team conducted an on-site test by intentionally releasing a sample of methane gas to validate performance. The Flir GF77a, optimised specifically for methane detection, enabled the operator to observe the gas release pattern immediately on a monitor. Following the successful trial, the customer opted to deploy the solution.

Integrating visual and AI-based detection

The system was installed to provide a field of view covering the entire outdoor pressure regulation facility. The camera was deployed using a network-based configuration similar to standard CCTV installations, allowing the operator to leverage existing site infrastructure rather than introducing additional mounting frameworks.

In operation, the GF77a transmits real-time optical gas imaging video to an AI analysis engine, which automatically identifies leak and plume patterns. When a potential leak is detected, the system triggers alarms and logs the event for traceability. Operators then respond in accordance with established site procedures and the specific circumstances of the incident, enabling a structured and timely intervention process.

Compared with conventional sensor-only approaches, the new system enabled visualisation-based monitoring over a broader area, supporting earlier awareness of potential risk indicators. As an AI-powered, image-based monitoring layer that complements existing sensors, it improved surveillance capability for outdoor, unmanned LNG pressure regulation facilities.

A roadmap for continuous improvement

The facility manager affirmed that although the installation is in its early stages, "we expect it to help prevent safety accidents before they occur". "While there are areas for improvement, such as sensitivity adjustment and environmental impact correction in the method of detecting gas leaks and visualising their patterns, we clearly see the advantage of enhancing detection performance while supplementing existing measurement methods."

Looking ahead, the operator plans to further refine the AI models underpinning the system, tailoring them more closely to site-specific environmental conditions and operational characteristics. There are also plans to extend visual monitoring across a wider area of the terminal to support more integrated, facility-wide oversight, as well as expand AI-assisted visual detection to additional categories of gas equipment.

For more information: 

www.flir.com

Tel: +44 (0)1732 220 011