This paper explores the trade-offs among various objectives in determining optimal interport connections and the modal mix for inland transportation. The study is based on precise data on specific export and import ports, the American Northeastern rail network, vessel types, mode-specific costs, transit times, and CO2 equivalent (CO \({}_{2}e\) ) emissions. We employ integer linear programming to address a multimodal, multi-objective, and multi-period transportation problem concerning unidirectional flows, from Western Europe to customers along the Ontario-Quebec trade corridor. Our findings suggest that reducing land-based transportation costs outweighs the benefits of achieving economies of scale at sea. The results emphasize the importance of the geographic proximity of import ports to customers in the hinterland, in terms of determining optimal traffic flows, making the nearest import port to customers the most effective in traffic capture. This observation aligns with arguments in the literature on the limited effectiveness of large vessels due to factors beyond vessel size—such as port capacity and cargohandling efficiency among other logistical considerations. We tested the practicality of this approach through sensitivity analyses, considering scenarios such as a carbon tax regime and increased rail capacities.
