Dimensionamento de estoques de peças de reposição com uso de simulação de Monte Carlo

Abstract

The adoption of a correct spare parts inventory management policy MRO (maintenance, repair and operation), can bring several benefits to the companies, such as assets availability increase and operational costs reduction. Current techniques of inventory sizing do not take into account the individual characteristics of this type of parts, which are, low or very low turnover, high lead time and high cost, not presenting a demand in the normally distributed response time. For this type of inventory the great organizations challenge is the definition of how much to order, when to order and how many to keep in stock. This work proposes a model for spare parts sizing, whose objective is to find the best relation between the number of parts in stock and the assets availability at lowest final cost. The study includes the collection and analysis of historical data from time to failure, time to repair and time to replacement of each piece studied individually in order to model the probability distributions that will represent these characteristics. The study was applied at ArcelorMittal Vega, a large steel company located in the south of country. The model applied in this work uses reliability and simulation tools with software by the Monte Carlo method that can be applied to any probability distribution. The proposed model is applied in two spare parts, one electrical and one mechanical type. This method was adherent to the proposed objective and effective on definition of spare parts number in stock, being easy to apply and understand. The simulation results are shown at the end of this study and a comparison is made with the current stock of these spare parts.