Supply Chain Design under Uncertainty: Locating LNG distribution centers in an emerging market with uncertain demand

Abstract

According to the International Energy Agency, the global use of natural gas will increase dramatically in the next two to three decades. Due to factors ranging from national and international energy market regulations to availability of energy and economical growth, there is a high degree of uncertainty in these predictions concerning how the natural gas demand will develop in the future.This thesis looks specifically at how to optimize the profit of a gas distribution company through development of distribution centers along the Norwegian coastline, given different scenarios for future demand. Both the amount of distribution centers to be constructed, their locations and capacity are considered. The distribution methods are limited to shipping between liquefaction plants and distribution centers, and subsequent truck transportation to end-customers. A deterministic model with one aggregated demand scenario and a stochastic model with three different scenarios are presented, implemented and compared. Due to the high flexibility in the problem, where it is possible to expand and construct new distribution centers throughout the lifetime of the project, it is found that the difference in achieved profit between the stochastic and deterministic solution is insignificant in most cases. Only when the low or high demand scenarios are heavily weighted in the probability distribution does the use of a stochastic model become valuable in certain cases. Tests show that the usefulness of the stochastic model, compared with the deterministic model, increases when the flexibility decreases, and vice versa.