A Comparison of Dynamic Lot Sizing Rules for Use in a Single Stage MRP System with Demand Uncertainty

Much has been published in the production and operations management literature regarding the performance of lot sizing rules in Material Requirements Planning (MRP). Most of this literature has been concerned with performance of these rules in a deterministic environment. In this research, the authors add uncertain demand and compare the performance of three lot sizing techniques. These techniques are economic order quantity, part‐period balancing and Wagner‐Whitin. Performance is compared using total inventory cost and computer time as criteria. In this research with stochastic demand, part‐period balancing proves to be the best lot sizing rule, overall. In many of the 128 problems generated, part‐period balancing and Wagner‐Whitin are very close to equal on total cost, but part‐period balancing uses far less computer time. Also included in the results is a discussion of particular situations where each of the lot sizing rules works best.