Models for multi-echelon repairable item inventory systems with limited repair capacity

The dominant models for inventory control of repairable items, both in the literature and in practical applications, are based on the assumption of ample repair capacity. This assumption can introduce a serious underestimation of the spare parts requirements in systems with high repair facility utilization, as is typical in industry. In this paper, we introduce approximations that can deal with limited repair facilities, under the scenarios of single-class exponentially distributed repair distributions, single-class general repair distribution, and multi-class general repair distributions. We provide numerical experiments that demonstrate how these models significantly outperform traditional models in the case of high repair facility utilization. Their ease of implementation is illustrated in a case study of the spare parts requirements at the Caracas subway system.