BAYESIAN MULTICRITERIA RISK ANALYSIS APPLIED TO HYDROGEN DOPING IN NATURAL GAS PIPELINES

The research addresses hydrogen doping in natural gas pipelines as a strategy to promote the energy transition in the face of environmental and geopolitical pressures. An integrated model is proposed that combines Decision Theory, MAUT and Bayesian Networks to assess risks in a multidimensional way, considering operational, structural, environmental and economic criteria. Decision Theory, associated with MAUT, translates the decision-maker's preferences into an aggregate utility function, while Bayesian Networks allow modeling uncertainties through subjective probabilities. The results indicate that mixtures with up to 10% hydrogen present admissible risks, and the sensitivity analysis, with 10,000 iterations, confirms the robustness of the model. From a managerial point of view, the approach helps in the definition of operational limits and in the prioritization of investments, while, from a socio-environmental perspective, it contributes to the reduction of dependence on fossil fuels and to the advancement of energy autonomy.