Flare Stack Smoke and Flames… AI for Monitoring and Analysis to Optimize Operations
AI-Based Process Safety Management System: Development of Generative AI Services Using the ‘AX Platform’
S-OIL (CEO: Anwar Al Heezaji) is advancing its collaboration with LGCNS, a specialist in AI Transformation, to accelerate the construction of a Digital Refinery at its Ulsan plant.
S-OIL and LGCNS signed an MOU agreement at the LGCNS headquarters in Magok-dong, Seoul. The signing was attended by S-OIL’s COO Park Bong-su, LGCNS CEO Hyun Shin-kyun, and other key executives. The partnership aims to implement a digital refinery using breakthrough technologies like AI and big data.
Through this agreement, both companies will collaborate on developing an AI-based flare stack optimization system, a routine Process Safety Management (PSM) system, and the AX platform, aiming to improve operational efficiency and safety management at S-OIL’s Ulsan plant.
Flare Stacks are facilities that safely release combustible gas from refineries and petrochemical plants into the atmosphere. If the gas isn’t fully combusted, smoke and flames may occur, making real-time monitoring and swift responses essential.
The AI-based flare stack optimization system being considered by S-OIL and LGCNS uses video analytical AI technology to analyze the color of smoke and flame images from CCTV in real time, 24/7. Upon detecting anomalies, steam valves can be automatically optimized and controlled, significantly enhancing process stability. When implemented, this system is expected to reduce the burden on S-OIL employees, who currently have to check flare stacks manually every hour, offering more efficient plant operations.
S-OIL and LGCNS also plan to collaborate on enhancing the Process Safety Management (PSM) system.
PSM is a critical safety management protocol for businesses handling high-risk chemicals, covering 12 items including process safety information, work permits, process hazard assessments, change management, emergency action plans, and facilities management. This extensive reporting process takes at least 3 to 6 months and requires systematic document management.
To address this, the companies are considering the adoption of generative AI and voice recognition technology tailored for S-OIL, aiming to better assist employees with PSM tasks.
As part of the PSM process, they are actively developing an “AI Mock Interview” feature. This feature operates by having generative AI pose questions based on interview scenarios and environment/safety-related materials, providing AI feedback on employee responses.
This not only helps employees prepare more effectively for interviews but is also expected to enhance overall safety awareness.
Additionally, S-OIL and LGCNS plan to jointly promote the establishment of the ‘AX Platform’ at S-OIL’s Ulsan plant. For this, they intend to leverage LGCNS’s enterprise generative AI platform, ‘DAP Gen-AI Platform,’ along with its machine learning/deep learning-based AI/big data analytics platform, ‘DAP MLDL.’
Through this platform, S-OIL plant employees can directly develop various generative AI services and apply them in operations, facilitating the establishment of a rapid and accurate data-based decision-making system.
Consequently, S-OIL is set on accelerating Digital Refinery innovations and actively pursuing process optimization and work efficiency enhancements through AI technology.
S-OIL COO Park Bong-su stated, “S-OIL has significantly improved on-site safety and operational efficiency by actively adopting recent 4th Industrial Revolution technologies. We have high expectations for realizing a Digital Refinery through collaboration with LG CNS, which possesses both AI technology and smart factory operations know-how. We will continue to actively collaborate for the systematic establishment of a Digital Refinery from a long-term perspective.”
Through this agreement, S-OIL and LG CNS plan to accelerate AI and big data-driven factory innovations and continue to strengthen their close cooperation for realizing a Digital Refinery.