The evaluation of multidimensional risk in natural gas pipelines is a topic of interest due to the possibility of human, environmental, and financial losses occurring. However, managing risks in pipeline systems might be a complex task due to the uncertainty inherent in the parameters of the model, being a factor that managers must take account. This requires more regular effort to be made to maintain safety conditions at stable levels, and for it to be borne in mind that pipeline sections at risk may be sensitive to both minor and major changes. Thus, this study undertakes a global sensitivity analysis based on a multidimensional assessment of risk, which uses ELECTRE TRI integrated into the Utility Theory approach, in order to classify sections of a natural gas pipeline into risk categories. For this analysis, parameters are set for operational, physical, and failure modes for different pipeline sections. The assessment is formulated to include preferences, graphical information and consequence modeling. Then, simulations are run to provide the decision-maker with relevant information, based on different settings of groups of parameters. From a graphical analysis appraisal, this paper gives the notion to the decision maker of sensitivity levels of each section and how sections are changing regarding their original risk class based on the variations tested. Thus, it is shown that, with the aid of procedures that evaluate risk under uncertainty, actions can be determined that lead to resources being applied better and more efficiently so as to guarantee acceptable safety levels for everyone and everything involved in a natural gas transportation system, thereby avoiding human, environmental, financial and company losses.