Optimal dry port configuration for container terminals: A non-linear model for sustainable decision making

In recent years the maritime freight transport has rapidly increased, causing congestion in many port areas. In some cases, in order to improve the capacity and the reliability of the temporary storage, a solution, recommended by industry officials, is the expansion of the terminal capacity. When this solution is not available, the ‘dry port’ area represents an effective alternative. The adoption of a dry port, if on one hand leads to benefits on terminal congestion, on the other hand requires resources and investments due to the transport of the container from port to dry port and vice versa. In the evaluation of the strategy to be adopted different aspects shall be evaluated to estimate time required for the container handling inside and outside the terminal on the basis of the congestion degree. In this paper, to support decision makers in identifying the best strategy to be adopted, a mathematical model allowing to identify the number of containers to be stocked in port and/or in dry port is defined considering the intra-/inter-terminal handling of the containers, in order to minimize the overall running costs and of the carbon footprint. The model, based on a computational algorithm for non-linear programming, is able to provide the number of containers to be stocked in port and/or in dry port, ensuring an effective strategy dependent on ‘road’ and ‘non-road' material handling equipment adopted, on the number and size of containers, as well as on the distance from port to dry port. Results obtained from numerical experiments show that, on the basis of the running cost and the carbon footprint of the container handling activities, it is possible to identify the most economic and eco-friendly container handling configuration. The case study of the Port of Bari (Italy) is investigated. In this case, given the overall number of containers to be stocked and the distance between port and dry port, the solutions found by the model identify a configuration able to ensure a reduction of 7% and 11% of the running cost and of the carbon footprint, respectively, when compared to the configuration in which all containers are stored in the port.