Service levels in single-stage MRP systems with demand uncertainty

In this paper the authors consider the problem of predicting customer service levels in a single-stage MRP environment. In the case of quantity uncertainty and a constant master schedule, the MRP system implements generalised periodic review (s,S) policies. The authors establish the relationship between the parameters of these policies and those of the original MRP systems for two general lot-sizing rules. They then derive analytic expressions for different measures of service levels using extreme rules for allocating priorities between the current period demand and the existing backlog. The results of this analysis are validated using a simulation model.