We address a problem of inventory management of spare parts in the context of a large energy company, producer of oil and gas. Spare parts are critical for assuring operational conditions in offshore platforms. About 200,000 different items are held in several inventory plants. The inventory system implemented at the company corresponds to a min-max system. The control parameters are decided based mainly on the expert judgment of the planners. Also, though the inventory plants can in practice be supplied from each other, the inventory planning is performed separately by the plant planners. This is because of different ownership structures where the studied company has the operative responsibility. The company is pursuing a system in which all planners conform to the same inventory management approach and evaluation, as well as being more cost efficient. Our work focuses on supporting this goal. We apply methods to decide the inventory control parameters for this system under a service level constraint. The methodology we use distinguishes unit-size and lot-size demand cases. We perform computational experiments to find control parameters for a sample of items. After the control parameters are found, we use them to explore the impact of risk pooling among the plants and inaccuracy arising from duplicate item codes. Key words:Email addresses: mario.guajardo@nhh.no corresponding author tel.: +47 55959834 fax: +47 55959650 (Mario Guajardo), mikael.ronnqvist@nhh.no (Mikael Rönnqvist), amaha@statoil.com (Ann Mari Halvorsen), svkal@statoil.com (Svein Inge Kallevik) Preprint submitted to NHH Discussion Series June 6, 2012Inventory of spare parts; risk pooling; duplicates; inventory inaccuracy; OR in the oil and gas sector.