One of the most important measurements of supply chain performance is inven-tory costs associated with the material and product flow within the whole supply chain. While there exists a large body of inventory management literature, the restriction imposed by a limited storage capacity has not been given due atten-tion. Limited storage capacity, coupled with a traditional warehousing approach, often makes it tough to match inventory with fluctuating demands....[ Read more ]

One of the most important measurements of supply chain performance is inven-tory costs associated with the material and product flow within the whole supply chain. While there exists a large body of inventory management literature, the restriction imposed by a limited storage capacity has not been given due atten-tion. Limited storage capacity, coupled with a traditional warehousing approach, often makes it tough to match inventory with fluctuating demands.

In this thesis, we build a stochastic inventory management model subject to storage capacity restrictions. By coordinating inbound and outbound flows, incoming goods are allocated to retail outlets before the leftover goods, if any, are stored to the warehouse. Our Markov Decision Process model differs from other existing work in that the storage capacity does not limit the ordering quantity. We derive optimal inventory policies under different cost structures. For different cases, managerial insights based on computational analysis are highlighted.

We also study a problem of water purchasing management under the context of limited storage space, which is motivated by Hong Kong's water purchasing practice from Guangdong. We restrict our attention to the importing side's opti-mal decisions. We build a regulated Brownian motion model to describe various issues of the problem. We provide analysis that helps optimize not only the annual purchasing amount, but also the storage capacity of the reservoir.