Programação da produção de uma planta de fertilizantes Fosfatados por meio da representação discreta do tempo

Abstract

The fertilizer industry is strategically important to Brazil socioeconomic development, a country with intensive agricultural activity and soils needing of replenishment of nutrients. Phosphorus is one of the most essential nutrients for vegetables growth. The domestic production alone can not be satisfy by internal market, not only by low capacity, but also because the hard extracting deposits and also the industrial process are uncompetitive. In order to make the profitability of national facilities maximized, arises the production scheduling as one of the ways to achieve this goal. This work aims to maximize profit through the production scheduling of a multipurpose plant of phosphate fertilizers, considering the process, capacity, inventory and demand constraints. Propose a way to maximize the profitability of this type of operation is not a simple task, since all the products are obtained from the same main raw material, process recipes are differentes, the production units are shared and there are steps in nature continuous and batch connected in series. All mentioned characteristics raise the level combinatorial mathematical modeling of the problem and high complexity. A plant for the production of simple superphosphate, simple ammoniated, triple and phosphate monoammonium was built. Using all plant data a resolution algorithm with discrete representation of time was implemented in GAMS. The result of the production allocation quickly returned the production scheduling of higher profitability for 15 days, divided into 45 intervals of 8 hours each. All constraints were considered. Besides maximizing the profit generated, the program returned again: the changes in inventories, allocation of productive units, rates of production and consumption over time. Not only prediction of the best way to operate the plant, considering the maximization of profit, opposite a large number of possible combinations, other analyzes can be made from the allocation of production efforts over time, for example , which units are bottlenecks to satisfy the demands.