Safety stock allocation in a two-echelon distribution system

In this paper, an analytical method for determining the optimal system stock in a two-level arborescent inventory system subject to management-specified customer service constraints is presented. The system examined consists of one central warehouse and N not necessarily identical regional depots all operating under daily review (lot-size, reorder-point)-policies and facing stochastic demands. Transshipments between depots are not allowed. Analytical approximations of the warehouse demand process are derived and tested via computer simulation. It is shown that the probability distribution of the warehouse service time (the lead time seen by a depot replenishment order, which may include a transshipment time) plays a key role in the optimal allocation of system safety stock. Based on the author’s earlier work an analytical approximation of the warehouse service time probability distribution is used to solve the problem.