Abstract
Liquefied natural gas (LNG) is natural gas transformed into liquid state for the purpose of transportation mainly by specially built LNG vessels. We study a problem arising in the LNG business where an LNG producer is responsible for transportation of LNG from the production site to customers located all over the world. The LNG producer need to develop an annual delivery program (ADP) listing the shipments to the customers during the year. The aim is to create an ADP with routes and schedules for the vessel fleet that are more robust with respect to uncertainty such as in sailing times due to changing weather conditions. We propose a solution method and several robustness strategies that are tested on problem instances with time horizons of 3 to 12 months. The resulting solutions are evaluated by a simulation model with a recourse optimization procedure. The results show that there is a significant improvement potential by adding the proposed robustness approaches. Our proposed solution strategies in combination with the simulation-optimization framework for evaluating solutions form a good foundation for a complete decision support system that will support both the initial planning process and the re-planning activities.
