Batch Size and Stocking Levels in Multi-Echelon Repairable Systems

In multi-echelon repairable inventory systems with high set-up cost for order and/or high demand rates, the use of batch ordering may be more cost-effective than the common (S - 1, S) ordering policy. This paper addresses the issue of determining the optimal order batch size and stocking levels at the stocking locations in such a system. A power approximation is used to estimate the total system stock and backorder levels from which the optimal batch size can be readily determined. A search routine involving "one-pass" searches are then followed to obtain the stocking levels at the depot and the local sites of the system. This procedure has been evaluated using 900 test cases and has been found to be very effective. The power approximation approach also results in a simple analytical relationship to test whether or not (S - 1, S) is an optimal ordering policy for repairable items in a multi-echelon environment.