Demand uncertainty is a concern and difficulty of primary importance for companies. In particular, it has an important impact on the optimal design of a supply chain network. For example, demand uncertainty forces to store products in warehouses close to customers (safety stocks), in order to react quickly to variations and meet customer expectations. In this work, we integrate the impact of demand variability in a location-inventory model. Extending previous results, we propose a new mathematical formulation to consider safety stocks and study their impact on the supply chain network design.