Spatially explicit supply chain for nationwide CO₂-to-fuel infrastructure: Data-driven optimization with Gaussian mixture model -based region screening and clustering

This study proposes a novel machine learning-based method for designing spatially explicit supply chains for establishing a nationwide CO₂ utilization strategy. The proposed method is capable of considering not only technical feature for CO₂-to-fuel conversion, but also the geographical features for facility allocation by integrating spatial data and various sociological information. To develop the proposed method, map polygonization and the ray-casting algorithm for precise spatial data generation are implemented, which identifies and screens out geographically and socially unfavorable terrains for facility allocation. The scenario evaluation method, which includes the Gaussian mixture model-based expectation–maximization algorithm and supply chain economics, such as transportation price and facility capital investment, are considered to calculate a large set of supply chain scenarios. As a case study, the proposed method is applied to determine the optimal supply chain configuration to secure economic feasibility in the Korean CO₂ capture and utilization project. The results identify the hydrogen price as the most critical factor in the supply chain economy and show the potential profitability in 2043–2045. This study can promote the nationwide CO₂ reduction strategy and the development of a sustainable energy supply system.