Background and Objective: Safety issues of natural gas pipelines have become an important issue in recent years. In this regard, risk analysis of gas pipelines has received attention, and approaches related to the identification and determination of risks of natural gas pipelines, especially the assessment and quantitative analysis of gas pipeline risks, have become an important issue. Therefore, the present study aimed to model the risks of the gas industry distribution process using Bayesian networks.
Materials and Methods: This study was applied in terms of purpose and mixed-method in terms of data collection and analysis methods. The statistical population of the research consisted of experts and specialists of the Kermanshah Province Gas Company in Iran. Purposive sampling was used to select 30 participants. The required data were collected in the qualitative part using semi-structured individual interviews and in the quantitative part using a questionnaire. The thematic analysis method was used for qualitative data analysis and the Bayesian networks method was used for quantitative data analysis.
Results: Results of qualitative data analysis indicated that the risks in a gas industry distribution process can be classified and studied in three general categories, namely risks due to human error (consisting of 20 risks), organizational risks (including 11 risks), and technical risks (consisting of 4 risks). The results of sensitivity analysis also showed that gas pressure gauge, connections, quality of maintenance staff, filters, and maintenance strategy have the greatest impact on reliability, in that order.
Conclusion: Performance of a complex technological system depends on the interaction of technical, human, organizational, social, and environmental factors. The present study provided an integrated framework for the assessment of the potential reliability and risk in gas distribution units by considering technical, human, and organizational criteria.