<![CDATA[Traditional risk assessment methods in chemical and process industries frequently fail to capture the dynamic nature of barrier degradation and hazard escalation. This study proposes a Dynamic Bayesian Network (DBN) framework integrating temporal indicators, including SIS unavailability and corrosion rate, to enhance the predictive accuracy of real-time risk management systems. The DBN was structured using nodes and dependencies derived from industrial scenarios and reliability parameters, then validated through two simulation cases — reactor runaway and corrosion-driven leak — utilizing real-time inputs of dT/dt and k_cor to dynamically update failure probabilities via MATLAB. Results demonstrate that barrier degradation significantly impacts risk profiles: escalation probability in the reactor runaway scenario increased from 0.10 to 0.45 as SIS unavailability rose, while leak probability reached severe consequences when barrier failure exceeded 60%. Compared to static models, the DBN approach more accurately captured emergent risks over time. The framework supports predictive maintenance, alarm prioritization, and human reliability modeling, establishing DBNs as valuable tools for transitioning toward intelligent, adaptive safety systems in high-risk industries.]]>
Copyrights © 2025
