The steel industry is responsible for 5% of total energy consumption and contributes 6% of CO₂ emissions worldwide [1]. Brazil produced 31,869 million tons of steel in 2023. The industrial park has 31 plants, 15 of which are integrated, 7 of which are coking plants, meaning that they consume around 74% of all the coal imported by the country [2].

Mineral coal is the main source of energy still in use in modern society. This input is exported by several countries around the world, such as Australia and the USA, of which Japan, India and China are major importers of the main input for reducing iron ore. A reactor is used to make steel and is called a blast furnace [3].

In the steel chain, 40 to 50% of the cost of steel is in the coal to be used in the coking plant and the constant search to optimize this commodity directly reflects on the competitiveness of the business, which is why various parameters are evaluated in the composition of the coal and subsequently the properties of the coke [4-5].

The main objective of this work is to present a proposed predictive model using combinatorial mathematical analysis and probability, where particles from different coals interact within the coke oven. The model shows promise for the proposed mixtures and could be used as a tool to improve the quality of the coke produced.