Biobjective Optimization of Intermodal Networks under Supply Uncertainty: Logistics Costs and CO2 Emissions

This paper proposes a biobjective mixed-integer linear programming model for sustainable intermodal network design under supply uncertainty, modeled via chance constraints. The two objectives minimize, respectively, logistics costs (transportation and terminal opening) and CO₂  emissions across road, rail, and maritime modes. The network is structured as a layered graph that endogenously determines terminal opening. Two scalarization methods generate the Pareto frontier: the global criterion method and the augmented Tchebycheff metric. Computational experiments on instances built with real geographic data from Brazil, the USA, and China show that both methods solve each scalarization in a few tens of seconds, and that raising the confidence level in the chance constraints shifts the Pareto frontier, with proportionally larger impact on emissions than on costs.