| There are two common types of supply uncertainties: the random yield of supply due to capacity and quality issues, and the disruption of supply caused by low likelihood events such as natural disasters or terrorism activities. In this dissertation, we study the impacts of these two types of uncertainties on supply chain design decisions using both analytical and numerical approaches. We also suggest the conditions under which significant cost savings can be achieved by considering supply uncertainties at the supply chain design phase.;We start our research from a simple EOQ-like inventory problem. To make this research extendable to more complicated integrated supply chain design problems, we assume both the retailer and the supplier are subject to random disruptions. By formulating and analyzing the total working inventory and recovery cost at the retailer, we conclude that the disruptions at the retailer have more significant impacts on the inventory decisions than the disruptions at the supplier. Since the cost function we set up at the retailer cannot be minimized in a closed form, we propose a tight approximation for this cost function.;We then extend this inventory problem to an integrated supply chain design problem with one supplier, one or more retailers and customers. Random disruptions at both the supplier and retailers are considered. Numerical experiments provide conditions under which cost savings from considering supply disruptions at the supply chain design phase are significant.;Moreover, we set up an integrated model for a three-tiered supply chain network consisting of one supplier, one or more facilities and retailers. The random yields of the supplier and facilities are considered, and the objective of the model is to maximize the total profit at the retailers and facilities. Our analytical study reveals the properties of the model based on which we propose an effective solution algorithm to solve this integrated model. We also analyze the optimality conditions of the solution to the integrated model to see the impacts of random yields of supply on the optimal decisions. |
