Data-Driven Decision-Making for Flexible Natural Gas Allocation Under Uncertainties: An Agent-Based Modelling Approach

• Fast-paced changes in final energy markets require dynamic assessments for informed decision-making in mature natural gas supply chains. • Agent-based modelling aids quick, informed decisions under market uncertainties. • Agent-based modelling acts as a dynamic decision-making tool for deciding on optimal natural gas allocation to commodities under economic and environmental constraints. • The results of the study inform policymakers about market uncertainties and support strategic resource allocation. Despite the anticipated growth in the global demand for energy commodities, the frequently changing market dynamics imposed by environmental regulations and sanctions create end-user demand uncertainties. This imposes the need for prompt quantitative decision-making approaches to understand how different market structures (i.e., market scenarios) affect the planning of current natural gas projects. Agent-based modelling (ABM) emerges as a powerful approach, facilitating expedited and well-informed decisions amidst limited timeframes. This analysis employs agent-based modelling to investigate natural gas allocation across various utilisation routes under diverse economic and environmental scenarios. Results from four main cases and two sub-scenarios imply that the allocation strategy is driven by utilisation routes considered in each case, followed by the allocation target (i.e., economic or environmental) and the operational bounds. The results revealed that cases prioritising natural gas monetisation for export outperform those meeting power requirements in average annual profitability. In case 4, considering a full network with power, the average annual profitability in the economic scenario dropped by approximately 47% compared to case 3, representing the optimal network configuration with $5.22 billion in average annual profitability. However, the economic scenario of case 3 had the second-highest emissions (0.66 CO 2 -eq t/y), following the hydrogen-rich process routes in case 2. Overall, this work presents an innovative data-driven framework for enhancing strategic resource allocation in dynamic business environments. By integrating empirical evidence and technical data with an advanced technical tool (i.e., ABM), the framework provides decision-makers and policymakers with valuable insights for managing uncertainties and shifts in market structures, particularly in existing natural gas projects.