A model which helps system designers to jointly evaluate a repairable equipment population system and its associated logistics support subsystem is developed. The modeled system consists of a deployed population of repairable equipment and a logistics support subsystem consisting of repair channels and an inventory of consumable components. In the system, a population of identical repairable equipment is procured and deployed to meet a known and constant demand. As the equipment fails, it is repaired and returned to service. Equipment repair generates secondary demands for consumable components and spares. Backorders within the inventory support subsystem add to the time failed equipment is not in service.

The model returns optimal values for the number of equipment units to deploy, the number of maintenance channels, the retirement age of deployed units, the procurement level, and the procurement quantity. These values are optimal in that they minimize the expected equivalent annual life-cycle cost of the system. The best system design may then be selected from among candidate designs on the basis of minimum life-cycle cost.

Previous attempts at integrated logistic system evaluation have led to suboptimal results because evaluation of the repairable equipment population system was disconnected from the evaluation of the inventory subsystem. Where inventory backorders are allowed, there are tradeoffs between system-level shortage costs and the cost of procuring and holding inventory.The backordered consumable components contribute to equipment repair times by increasing the time spent in repair. Furthermore, the procurement of additional repairable equipment to offset shortage penalties may result in increased logistics support costs which must be accounted for when evaluating system design alternatives.

This thesis presents a model for a repairable equipment population system with its associated logistics support subsystem. The evaluation function is developed along with assumptions and definitions of the cost elements. A numerical example which illustrates some basic aspects of the model is given. Finally, an optimization routine is presented which returns an optimal set of system design variables for a given set of design and source dependent parameters in the face of design and source independent parameters.