Publication Date
Authored by
Mauricio Taborga, Tulio Castillo
Advisor(s): Jim Rice
Topic(s) Covered:
  • Network Design

To mitigate operational disruptions, oil and gas companies maintain high levels of Maintenance, Repair, and Operations (MRO) inventory. However, our sponsor company was found to have twice the nonmoving inventory value compared to its competitors, prompting an interest to reduce inventory holding costs — the cost associated with the storage of inventory, such as cost of capital, annual warehouse fees, annual taxes, and annual warehouse costs. This study aims to reduce such costs by segmenting 19,153 MRO SKUs based on their demand characteristics and building a Mixed Linear Integer Programming (MILP) model to redesign the network of warehouses and plants. By eliminating the 1:1 relationship between warehouses and plants, the warehouses can serve more than one plant and the sponsor can avoid individual inventory management for each plant. Through our MILP model, we investigated different levels of consolidation though scenario analysis. In the most conservative scenario without inventory systems integration, the new network design resulted in a 12% reduction in warehouse and transportation costs and a 22% reduction in safety stock holding costs. While full inventory system and legal entities integration led to 27% savings in both.