Modelo para previsão de itens sobressalentes baseado em confiabilidade industrial: estudo desenvolvido em uma do setor petrolífero

Abstract

The ability to compete in global markets is a present qualifying in the current organizations. The management of productive spending (costs and investments) has a direct influence on the business decisions of term short and long. In this sense, the search for optimization of the resources, through reductions in the fixed capital, is a competitive factor sought by various companies. The inventory management involves conflicts of interests that can lead the manager to inconsistent decisions: stocks can absorb large amounts of resources or represent massive production losses (profit loss). Among the various types of stocks, immobilization of spare items applied to maintenance stands out for have some important features: the demand is irregular (lumpy), the investment for the acquisition of spare parts is high and the delivery time is long. This work aims to apply a method for spare inventory management based on industrial reliability with use of covariates. The model applies techniques related to probabilistic modeling and renewal process, in order to quantify the spare items. The analysis makes use of computer programs to manipulate of the reliability parameters (Weibull® and BlockSim® software - ReliaSoft) and R® software, necessary to define the covariates. The dimensioning of the spares is consolidated through the program developed in c ++ language, which aims to analyze the failures overlapping over time. The study is applied to the sizing of centrifugal pumps (BCS) used in the production of oil in offshore units. The tool search to structure the process of allocation of spare parts to be fixed in the oil sector, aiding decision making.