Dynamic risk analysis of hydrogen gas leakage using Bow-tie technique and Bayesian network

Using fuzzy cognitive map in bow tie method for dynamic risk assessment of spherical storage tanks: A case study

“…Aliabadi et al [ 20 ] in their LPG release study using BT identified the causes leading to TE and Zhao et al [ 62 ] demonstrated failures of valve sealing, flange sealing, and tanks as the main causes of LPG leakage when assessing the hazards of explosion and fire incidents in spherical storage tanks by the BT method. The relationship and effect of the factors on each other were identified using the opinions of experts and the FCM method [ 20 , 25 , [25] , [25] , [26] , [27] , 30 , [33] , [34] , [35] ], which provided an understanding of the factors influencing LPG release and the relations between different barriers, helping decision-makers determine the important nodes in preventing the LPG incident. Utilizing the FCM to determine the critical nodes in the LPG release is a helpful method for developing dynamic approaches based on computer models, which has been approved by other studies [ 63 , 64 ].…”

“…These limitations have led to more advanced safety assessment methods in recent years that are dynamic. For instance, Khakzad et al [ 25 ] used Bayes' theorem for dynamic risk analysis in the BT method; Ferdous et al [ 19 ] developed a dynamic safety and risk assessment under uncertainty using a bow-tie diagram; Abimbola et al [ 26 ] proposed a dynamic safety assessment approach based on the BT and real time barriers failure probability assessment; Xin et al [ 6 ] utilized a dynamic methodology to identify hazards and scenarios using a Bayesian network; Chang et al [ 27 ] performed a dynamic risk analysis based on the mapping of the BT into a Bayesian network; Li et al [ 28 ] showed a risk assessment based on the fuzzy Bayesian network for mine ignition sources; Wu et al [ 29 ] presented a dynamic Bayesian risk analysis for modelling the influence of uncertainty in accident probability and consequences, which is reflected in failure rates; Borgheipour et al [ 30 ] determined the type and relationship between the effective causes of catastrophic accidents by integrating the BT technique and Bayesian network; Pang et al [ 31 ] also used a dynamic Bayesian network-based reliability in safety assessment of subsea Christmas tree; and Sarvestani et al [ 32 ] developed a dynamic Bayesian risk assessment for propane storage tanks. In this regard, Zarei et al [ 33 ] proposed a dynamic approach for analyzing the event of hydrogen release scenarios, which used the BT technique and Bayesian network in integration with improved D Numbers Theory and Best-Worst Method.…”

Evaluating and optimizing performance of public–private partnership projects using copula Bayesian network

“…It was used for reliability evaluation of tunnels (Liu et al, 2020), and complexity management in 2020 (Luo et al, 2020). The combination of BNs with XGBoost and machine learning methods emerged in 2021 in oil and gas projects (Borgheipour et al, 2021;Hao et al, 2021;Paul et al, 2021). In 2022, the BNs have been used for construction risk management (Arabi et al, 2022;Ji et al, 2022;Khanh et al, 2022).…”

“…, 2020). The combination of BNs with XGBoost and machine learning methods emerged in 2021 in oil and gas projects (Borgheipour et al. , 2021; Hao et al.…”

Risk analysis of waterlogging in a big city based on a bow-tie Bayesian network model, using the megacity of Wuhan as an example

“…Combining the bow-tie model with Bayesian network analysis can clearly express the basic event that caused the accident, analyze its basic event importance, and propose measures to prevent the occurrence of that accident and emergency protection measures to prevent injuries caused by the accident. Borgheipour et al (2021) used the hydrogen leakage accident of a chlorination plant as a research use case and determined the relationship between the type of catastrophic accident and the actual cause using the bow-tie technique and Bayesian network analysis, suggesting that the use of Bayesian networks can update the probabilities to reduce the uncertainty of the parameters. Bilal et al (2017) used the fault tree method, bow-tie, and Bayesian networks to quantitatively analyze the causes and consequences of accidents, estimate the probability of occurrence of undesirable events, and use fuzzy logic to evaluate the occurrence of each basic event in the fault tree to develop a complete and effective assessment model for pipeline fire and explosion risk accidents.…”