Using Simulation to Evaluate a Continuous Review (r, Q) Two-echelon Inventory Model

The effective management of supply chains depends upon the ability of decision makers to control inventory while maintaining customer service at low cost. This paper examines an analytical model for the two-echelon one warehouse and multiple retailers system using (R, Q) inventory policies. A simulation model was developed to analyze the performance of the analytical model under conditions that violate the model’s fundamental modeling assumptions. The behavior of the multi-echelon inventory model under different demand (non-Poisson) and lead-time (non-constant) conditions were observed in an experimental design. The simulation results showed that there was significant error in estimating the total system cost (tends to be underestimated) when the model assumptions are violated. The cost relative error ranged between 34% and – 40% over all the experiments examined. Models of the one warehouse and multiple retailers two-echelon inventory system are significant because they can be embedded within aggregation/disaggregation techniques that allow the general approximation of supply chain networks. Practitioners looking to utilize these models will have a better understanding of when and when not to use these models in practice.