A Success Story in Utilizing Big Data in Reporting Tier 3 Process Safety Incidents and Future Forward in Utilizing this Tool with AI to Save +10 Million Dollar Per Annum

Objectives/Scope This paper introduces a novel approach to increased tier-3 incident reporting by tenfold. Methods, Procedures, Process This paper presents a transformative approach to process safety management through the integration of big data analytics and predictive modeling, with a focus on Tier 3 incident detection. The study outlines the development and deployment of a digital tool, capable of real-time monitoring and automated identification of safety system demands, particularly on Pressure Safety Valves (PSVs) and Safety Instrumented Functions (SIFs). The tool uncovered significant underreporting of Tier 3 incidents, revealing latent risks and enabling targeted maintenance interventions aligned with API RP 576 standards. The developed tool is equipped with process parameter tags and their corresponding safe operating limits (SOLs) to detect any exceedance, demand on safety systems, or relief devices. This paper outlines the methodology for implementing this big data detection system, as applied in a real-world case study. It highlights the challenges encountered during deployment and the valuable results achieved in enhancing plant safety and operational reliability. Results, Observations, Conclusions Prior to the implementation of this tool, less than 10% of Tier-3 process safety incidents were identified despite field operator rounds, operational monitoring, and trend analysis. This left 90% of incidents undetected, many of which recurred and posed a high risk of escalating into Tier-1 incidents. The future direction involves integrating Artificial Intelligence (AI) to proactively analyze process parameters, identify potential activations of critical Emergency Shutdown (ESD) loops, and recommend operator actions. Operators will retain control by modifying these recommendations to ensure smooth operations. This predictive approach is expected to address 30% to 60% of plant trip scenarios, potentially saving over $10 million annually. Novel/Additive Information This research presents a significant advancement in the detection and reporting of Tier-3 process safety incidents. The novel tool introduced in this study has successfully increased incident reporting by tenfold, addressing a critical gap where 90% of incidents previously went undetected. Unlike other research that primarily focuses on preventing spurious trips, this study emphasizes a proactive approach by utilizing AI to prevent demands on safety systems. The proactive analysis of process parameters and identification of potential activations of critical ESD loops represent a forward-thinking strategy that significantly reduces the risk of plant trips. This innovative approach is expected to address 30% to 60% of plant trip scenarios, potentially saving over $10 million annually.