Performance evaluation of stochastic forward and reverse supply networks

The variability of the arrival and service processes has a strong impact on the performance of supply chain networks (SCNs), especially when the reverse flow to the manufacturer is considered. This paper proposes to use approximate analytical models to quickly evaluate the performance of SCN configurations during the design phase of the forward and reverse supply chains. The models are applied to the case of scrap-based steel production in which the role of the reverse flow is higher compared to the other reverse supply chains since a higher proportion of the raw material is provided by the reverse flow. As the solution methodology, an approach based on the queueing network model has been developed to represent general distributions of the stochastic parameters (i.e. arrival and processing rates). The accuracy of the proposed analytical models is assessed by comparing the results of numerical experiments against discrete-event simulations.