Multi-dimensional classification of wind turbine spare parts in a multi-echelon inventory system

Optimising inventory management of spare parts is important for wind power companies to reduce operation and maintenance (O&M) costs. We summarise the indicators for wind turbine spare parts classification by analysing O&M management characteristics of the wind power industry. Spare parts of wind turbines are classified based on three dimensions: value, demand, and importance. The existing multi-dimensional spare parts classification methods discretise the indicator on each dimension. However, we use the K-means algorithm to classify spare parts based on normalised indicators. The proposed classification method significantly decreases the reliance on expertise and information loss caused by indicator discretisation. The proposed multi-dimensional classification method is validated using a practical case study of wind turbine spare parts classification, demonstrating that the proposed method can obtain reasonable classification that simultaneously stabilises the service level and reduces inventory costs.