Exploring risk-based portfolio selection with global sensitivity analysis to support strategic decisions in natural gas pipeline networks

Considering the dense network of natural gas pipelines, gas transportation companies face the dilemma of maintaining a profitable operation while ensuring efficient risk management. The inherent uncertainties in risk assessment and the need for well-founded strategic decisions further exacerbate this challenge. To address these issues, the paper proposes a risk-based portfolio selection approach, incorporating global sensitivity analysis, which helps prioritize high-risk areas for pipeline maintenance and integrity. As a result, this approach optimizes resource allocation and ensures better allocation to the most critical areas. The main innovation of the study lies in developing a robust methodology that guides actions in the natural gas transportation system, handling uncertainties, and considering both safety concerns and organizational constraints. The developed model is divided into three parts. The first part uses a multidimensional risk assessment methodology to evaluate the consequences across three dimensions: environmental, financial, and human. These dimensions are associated with potential accidents caused by pipeline leaks and their impacts. The second part employs a hybrid search algorithm to select the most critical sections of the pipeline portfolio. The last part uses Monte Carlo simulation to comprehensively analyze the uncertainties involved in risk assessment. A case study is presented to validate the proposed methodology. Based on the results, the approach offers more reliable decision-making when selecting the most critical zones, and decision sensitivity levels are statistically estimated for each pipeline section. This information guides managers in efficiently allocating resources to mitigate losses and maintain operational integrity. Thus, the proposed structured procedure decisively supports risk management in natural gas transportation and guides strategic decisions.