Risk assessment of syngas pipeline leakage based on a Dynamic Bayesian Network

The syngas pipeline serves as the primary carrier for syngas exported from the coal gasification furnace, is vulnerable to corrosion and erosion from the transported medium, and is prone to leakage due to long-term high-pressure operation. Moreover, due to its compositional characteristics, syngas pose flammability, explosive, and toxicity risks, which can potentially lead to severe accidents if a leak occurs. Therefore, it is essential to conduct a risk assessment of syngas pipeline leakage. This study proposes a risk assessment approach for syngas pipeline leakage in coal gasification using Dynamic Bayesian Network (DBN). First, the risk identification model is built using Bow-Tie (BT) analysis and then mapped into DBN using a mapping algorithm. Expert evaluation, improved similarity aggregation methods, and fuzzy set theory are employed to quantify prior probabilities. To address the uncertainty of the DBN model, a Leak Noisy-or gate model is introduced. Time series are added to predict the dynamic probabilities. Nine key hazard events, six highly sensitive factors, and the maximum causal chain are identified, and predict the dynamic probability of syngas pipeline leakage and potential consequences. This study provides a theoretical basis for routine maintenance and risk assessment of syngas pipelines.