NEDJAR Yassine, MIDOUNI Khalil, ABBOU Abderrahmane Enc: BOUAFIA Abderraouf, BOUGOFA Mohammed2026-01-112026-01-112023-07-11http://dspace.univ-skikda.dz:4000/handle/123456789/5697This dissertation aims to evaluate the resilience of the pre-fractionation section in the Adrar Sbaa refinery by integrating the Functional Resonance Analysis Method (FRAM) and Bayesian Networks (BN). FRAM provides a comprehensive understanding of system functions, variability, and interdependencies, while the BN offers a probabilistic framework for capturing dependencies and making inferences. The research focuses on developing a FRAM model and then converting it to a Bayesian Network model to quantify the resilience of system reliability. The methodology includes data collection, analysis, and synthesis, with a sensitivity analysis to assess the models' robustness. The integrated approach allows for a holistic assessment of the section's adaptive capabilities and recovery potential, facilitating informed decision-making and resilience planning.enThesis