LINEAR PROGRAMMING APPLIED TO BLENDING IN A MINING COMPANY

This work addresses the optimization of iron ore blending in rail loading, relevant to achieve contractual specifications and maximize commercial margins. The objective of the research was to develop a mathematical model based on linear programming that assists in operational decision-making to carry out a loading batch that meets the needs of the process and product control specifications. The data and rules were obtained from a large mining company located in the iron quadrangle, one of the largest exporters of the commodity in Brazil. Currently, blending is created through load plans based on the technique of progressive adjustments, which makes the activity time-consuming and results in solutions with opportunities for improvement. The mathematical model developed proved to be reliable to the process, providing significant operational improvements and a reduction of approximately 4% in the silica content and promoting greater efficiency in the operation.