Adaptive liquefied natural gas supply chain planning under demand uncertainty

Global liquefied natural gas trade volumes have risen sharply as countries seek cleaner energy alternatives. As liquefied natural gas emerges as a primary energy source in many regions, securing it efficiently has become strategically critical. However, supply plans based on point forecasts often become infeasible or excessively costly when actual demand deviates from forecasts. To address this challenge, we propose a new optimization framework that explicitly accounts for demand uncertainty and adaptively updates decisions as real-world demand unfolds. We formulate the nationwide liquefied natural gas supply chain planning problem as a stochastic optimization model that jointly optimizes long-term contract selection, spot market procurement, volume allocation, and inventory control. To overcome the computational intractability of this formulation, we develop an adaptive robust optimization framework solved via a customized two-phase approach. Large-scale computational experiments calibrated with data from South Korea’s liquefied natural gas supply chain demonstrate that the proposed framework achieves an optimality gap within 5% relative to a perfect information lower bound, substantially outperforming the industry-standard deterministic benchmark. Furthermore, through extensive sensitivity analysis, we derive critical managerial insights for energy planners, identifying spot price as the dominant cost driver. The proposed framework offers national energy planners a practical and robust decision-support tool for liquefied natural gas supply chain planning under demand uncertainty.