<p>Risk analysis in gas processing facilities is essential for preventing unwanted events that could lead to catastrophic accidents and significant financial losses. Traditional failure assessment techniques, such as Event Tree (ET) analysis, have been widely used to identify potential consequences and to propose preventive and protective barriers. However, ET models are limited by their static structure and inability to handle uncertainties effectively. To address these limitations, this paper introduces a Bayesian Belief Network (BBN) model, which allows for the representation of event dependencies, multi-state variables, and the dynamic updating of probabilities, capabilities that are not supported by conventional ET methods. Furthermore, a Dynamic Bayesian Network (DBN) is employed to model the temporal evolution of sequenced events and to integrate safety barriers within a unified framework. This dynamic approach enhances the depth and flexibility of the analysis. The methodology is demonstrated through a case study involving the release of Liquefied Petroleum Gas (LPG), showcasing the advantages of the proposed model in capturing complex accident scenarios and improving risk-informed decision-making.</p>