Integrated steel plants are full of opportunities to implement solutions that allow them to enter the so-called Industry 4.0. People may think about many cost reduction initiatives at different stages of the process from an operational standpoint. However, we should not forget that Advanced Analytics can also be applied at a more strategic level, where decisions are worth millions of dollars!
Just think about how using a certain raw material impacts the processing cost and the production capacity and also changes the overall energetic balance of the entire plant. Any strategic decision has so many repercussions in the whole production chain that it becomes impossible for a human, even assisted by some spreadsheets, to be sure of its effectiveness.
In an integrated steel plant, the cost of raw materials responds for 75% of the production cost. With such a proportion, purchasing the cheapest materials seems a rational choice. However, the material highly impacts the process itself (performance, energy requirements), and may generate an increase in future costs for removing undesired chemical elements, for instance. A unique particularity of integrated steel plants is that the quality of materials impacts the entire production chain. We may think about:
Additionally, some material properties, together with operational decisions, determine the quality of intermediate products, which directly impacts subsequent processes.
Due to the fact that steelmaking is a chain with so many processes (including the coking plant, sintering plant, blast furnace, primary and secondary refining), most of the decisions have to be agreed upon between different departments.
Here are a few examples:
Chemical or physical specifications are a practical way to coordinate these decisions regarding the quality of intermediate products. However, they usually come up from a supposedly fair compromise, not a real quantitative analysis. Furthermore, the quality of intermediate products greatly depends on raw materials, which price can vary a lot. Thus, a revision in specifications should go along with price variations!
People used to compare the production cost with the purchase price to determine the “optimal” production level. It sounds logical, doesn’t it? The problem starts when this comparison is made based on the “average production cost”, while it should have been done with marginal production cost (much harder to estimate when you don’t have any mathematical model…)
Why are average and marginal costs so different in iron and steel making? Here are some examples:
Marginal cost is also the indicator used to make the right decision regarding purchasing alternative materials, like pellets, as a substitute to sinter, or scraps as substitutes to hot metal. Similarly, coke can be produced in loco or purchased from suppliers.
Decision-makers need to be supported by trustworthy data and tools. Given such complexity, the only way to make the right decisions regarding all the topics presented above is a mathematical model. It should contain all the empirical or theoretical relations between quality, production, fuel requirement, and emission of pollutants (that can influence the purchase of some materials). Cost and profit are usually indicators that are respectively minimized and maximized.
Many integrated steel plants are already using such a model: CSN, Gerdau, Usiminas, Ternium, CSP, among others. All their strategic decisions are being optimized by a tailored model, adapted to each plant. They have achieved tremendous, ongoing cost savings, which remains the main goal of implementing such technologies.
Author: Emmanuel Marchal - Managing Partner at Cassotis Consulting
R. da Paisagem, 220, sala 11S, Vila da Serra, Nova Lima - MG, Brasil