Título

Optimization Models of Coking Coals Blends

Optimization Models of Coking Coals Blends

Autoria

Echeverría, Manuel Sirgado

Echeverría, Manuel Sirgado

Eu sou esse autor

DOI

10.5151/1TH-MEETING-COAL-1987-1987044

Downloads

Resumo

With the introduction in 1976 of the Coal Park of Abono, it was necessary for Ensidesa to use a coal mixing optimization model in order to produce coke. In fact, the storage installations, grinding, dosification, and mixing at the Coal Park of Abono allow us to work simultaneously with a great number of coals in the coal blend. This number is normally greater than 18 coals, and in many cases, it reaches up to 22. In addition, the need to use a maximum of national coals (which currently makes up 25% of the blend) requires that widely separated coals in the range scale be used, resulting in very different properties. In the coking coal blend of Ensidesa, we can find a semianthracitic infusible coal with 11% volatile materials and 2.0% reflectance, as well as a coal with 35% volatile matters, 0.7% or 0.8% reflectance, and a maximum Gieseler fluidity greater than 60,000. Within these two types of coals, which may be considered extremes, there is a wide variety with intermediate properties, whether national, Polish, Australian, or North American, and high, medium, or low in volatiles; even in additives, such as petroleum coke. Given the complexity of the raw materials used, it is indispensable to utilize an optimization model. Otherwise, the rigorous resolution of the problem would be practically impossible, especially if we consider that, when calculating the mixture for each month, we must take into account the situation that will arise several months later. The model not only provides the composition of the coal paste that will be consumed in the actual month but must also be useful for planning the procurement of the necessary coals in the near future, while naturally respecting the stocks that will be in the ports, the supply, and the long-term contracts that we may have.

With the introduction in 1976 of the Coal Park of Abono, it was necessary for Ensidesa to use a coal mixing optimization model in order to produce coke. In fact, the storage installations, grinding, dosification, and mixing at the Coal Park of Abono allow us to work simultaneously with a great number of coals in the coal blend. This number is normally greater than 18 coals, and in many cases, it reaches up to 22. In addition, the need to use a maximum of national coals (which currently makes up 25% of the blend) requires that widely separated coals in the range scale be used, resulting in very different properties. In the coking coal blend of Ensidesa, we can find a semianthracitic infusible coal with 11% volatile materials and 2.0% reflectance, as well as a coal with 35% volatile matters, 0.7% or 0.8% reflectance, and a maximum Gieseler fluidity greater than 60,000. Within these two types of coals, which may be considered extremes, there is a wide variety with intermediate properties, whether national, Polish, Australian, or North American, and high, medium, or low in volatiles; even in additives, such as petroleum coke. Given the complexity of the raw materials used, it is indispensable to utilize an optimization model. Otherwise, the rigorous resolution of the problem would be practically impossible, especially if we consider that, when calculating the mixture for each month, we must take into account the situation that will arise several months later. The model not only provides the composition of the coal paste that will be consumed in the actual month but must also be useful for planning the procurement of the necessary coals in the near future, while naturally respecting the stocks that will be in the ports, the supply, and the long-term contracts that we may have.

Palavras-chave

-

-