Domino effect analysis using Bayesian networks.

Area of Science:

• Risk Analysis
• Probabilistic Modeling
• System Safety Engineering

Background:

• Domino effects in processing facilities pose significant risks.
• Existing models often struggle to capture complex interactions and uncertainties.
• A probabilistic framework is needed for accurate domino effect analysis.

Purpose of the Study:

• To introduce a novel Bayesian network methodology for modeling domino effect propagation.
• To estimate domino effect probabilities at various levels.
• To analyze domino effects considering synergistic effects, noisy probabilities, and common cause failures.

Main Methods:

• Utilizing Bayesian networks for their flexible structure and unique modeling capabilities.
• Developing a probabilistic framework to analyze domino effects.
• Incorporating techniques to capture uncertainties and complex interactions among components.
• Implementing probability updating based on new information.

Main Results:

• The Bayesian network effectively models domino effect propagation patterns.
• The methodology accurately estimates domino effect probabilities.
• Complex interactions, synergistic effects, and common cause failures are successfully captured.
• Probability updating allows for qualitative and quantitative model refinement.

Conclusions:

• Bayesian networks provide an effective tool for modeling and analyzing domino effects in processing facilities.
• The proposed methodology enhances the understanding and prediction of domino effect scenarios.
• This approach offers a robust framework for improving system safety and risk management.