A stochastic model for facility locations using the priority of fuzzy AHP

In this study, a production-distribution network system for a company, which is active in producing bottled natural spring water, was established. For this, one of the seven provinces in the Aegean Region was selected as a pilot region. According to the new supply chain network to be established in the region facility location selection model (FLS) decisions were optimized. The production of 0,5 lt bottled water is considered because it is the primary market target of the company. In order to meet the customer demand with a minimum production and distribution cost a network was established with optimum factory capacities and warehouse locations. Accordingly, this study consists of two stages; in the first stage a Fuzzy Analytic Hierarchy Process (FAHP) method was used to choose the most appropriate province. In the second stage the network was established and optimized under stochastic parameters by developing optimization models. The first method, FAHP, is fuzzy, and the second method, stochastic mixed-integer programming (MIP), is also modeled under uncertainty. Thus, the results are realistic and in a range that the company also accepts.