The production of steel using Direct Reduction and Electric Arc Furnace (DR-EAF) has increased in the last decade and shows no sign of slowing down. This technology is the main alternative for the well-established Blast Furnace - Basic Oxygen Furnace (BF-BOF) route. In its favor, the DR-EAF route drastically reduces the consumption of coal and CO2 emissions while maintaining productivity at the same pace as BF-BOF. However, regardless of the technology used, the steel produced must be competitive in a global multi-player market.
Steelmakers face many challenges when it comes to the strategic decision-making process. There are many levers throughout the production chain, which have significant consequences in the process as a whole. Before taking action, the business is constantly trying to weigh the trade-offs, searching for a way to produce steel for the lowest production cost. To find it, there are 5 main concepts that decision-makers need to consider and they are described in the following topics.
No matter the design of the steelmaking plant, the proportion of raw material costs in the production cost of steel is huge. Therefore, it is crucial to correctly select suppliers and specific consumptions to stay competitive. However, the implementation of cost-saving measures focused on finding new cheaper suppliers can quickly backfire.
The reason is simple: the quality of raw materials directly impacts the process. High-quality inputs can raise productivity, decrease the consumption of reduction gas, and reduce energy requirements, which can compensate for their higher prices. Low-quality materials are cheaper, but they can increase the presence of undesired chemical elements in the process chain and, since steel has specifications to meet, the plant will experience further costs to remove that extra mass of unwanted material. The best procurement strategy will certainly be a mix of those suppliers on diametrically opposed sides of quality and cost characterization.
As seen in the previous concept, the production of steel involves a wide variety of raw materials. Some of them are so important that steelmaking companies have decided to incorporate them into their production chain. The idea behind this centralization is to better coordinate the generation and quality of intermediate products, aiming at lower production costs for their steel. For instance, in steel plants based on electric arc furnaces it is common to have direct reduction reactors, briquetting machines, and even pelletizing plants.
Within this long process chain, a single decision over the quality of an intermediate product can affect many steps. A good example is the carbon content demanded by the EAF. This element has a major role in the thermal balance of the furnace and a high concentration of carbon in sponge iron can significantly decrease the consumption of electricity. Nonetheless, the DR plant has limited control over the sponge iron's carbon proportion. Usually, by decreasing the reactors' productivity it is possible to increase the carbon content. Another possibility is for the pelletizing plants to produce pellets with different physical properties, which are known to affect the presence of carbon in the sponge iron.
Definitely, it is not trivial to know where to make the changes needed to update the carbon content. Should it be in the DR reactor, in the pelletizing plant, or in both? Should the plant reconsider the EAF's carbon needs?
To avoid friction between production processes, it is common to adopt product specifications for each interface. There are agreed ranges for physical and chemical properties for intermediate products, which serve as a practical way to coordinate cross-department decisions. This system is ideal for contexts where there are minimum and maximum suitable requirements for process capacities and a stable raw material supply.
Since the raw material is the main driver of product quality and process performance, the revision of specifications should vary according to market conditions; not only considering the availability of the material, but also its price. Steelmakers must consider the alternatives of the main course to adapt to new scenarios. For instance, if there is a shortage of the main manganese ferroalloys' supplier, the plant should reconsider its specifications at the crude steel, sponge iron, and even at their pellet level.
A written-in-stone specification can be the blind spot of a process.
Process engineers must also consider how to operate their equipment to increase the benefits for the whole plant. The impact of process related decisions is clear regarding the treatment of the gases at the DR plant and the charging strategy in the EAF.
The material charged in the DR reactor chemically interacts with the bustle gas, generating the sponge iron and the off gas which is recaptured. The gas system is designed to maximize the recirculation, and guarantee its reduction capability. To achieve this goal, the off gas is treated to control its chemical composition, mostly regarding: CO2, H2, CH4, and CO. The efficiency of CO2 removal is subjected to a certain capacity, while the increase of hydrocarbon compounds requires the addition of reformed gas or natural gas to the system. The composition of the former is a process lever, which can be changed considering operational factors of the reformer, such as pressure and moisture. Additionally, oxygen is injected prior to the entrance of the reactor, resulting in partial combustion reactions that increase the gas temperature and allow for a greater productivity.
All decisions have major implications in the others. The reduction demand of the raw material, the CO2 removal, the reforming process, and the oxygen injection have a delicate balance, which is up to the engineers to define the path to follow.
Another important aspect of the steelmaking industry is production flexibility. There are many types of raw materials that can be used along the production chain. In the DR reactors, lump ores can substitute pellets. In the EAF, scraps can substitute sponge irons. The tricky part is to know when to change the input mix.
For EAF's process engineers, it is natural to determine the mix by comparing the production cost of sponge iron with the market price of scraps. Once the market is able to supply a scrap with a better cost-benefit ratio, it is logical to shift the blend. However, due to the complexity of the process, decision-makers usually consider the average production cost of sponge iron as their index to take action. Yet, this economic concept does not fit the analysis to be performed. In its place, consider the marginal production cost, which determines the production cost of the next tonne of DRI produced.
That microeconomic concept is crucial. If it is known that the next tonne of DRI is much more expensive than the alternative scrap supplier, it is trivial to purchase the external material. However, if this concept is not clear, the plant will observe only an increase in the production cost. The extra cost is diluted in the total average production cost and it may take many DRI tonnes to realize it was time to start substituting sponge iron for scraps.
This delicate marginal production cost function is hard to track in this industry. Every process has its own operational constraints beyond its product specifications, and all changes in raw material consumption or operating factors can affect the marginal production cost. Furthermore, its complexity increases because bottlenecks in one process can potentially affect all the others.
Finally, in such an integrated process, the only way to make the right decisions considering all delicate balances and concepts is by applying a mathematical model. It will consider all mass, chemical, thermal and energetic balances while tracking the revenues and costs of every possible solution. Additionally, optimization algorithms are suitable to find the best solution among all possibilities, given a certain objective. By combining them it is possible to know what the plant should do to minimize costs or maximize profit.
After the successful implementation of the BF-BOF model in Gerdau, CSP, and Usiminas, among others, the DR-EAF model is running over Ternium's plants in Mexico.
Author: Guilherme de Castro Martino - Senior Consultant at Cassotis Consulting
Co-authors: Fabio Silva - Senior Manager at Cassotis Consulting
Emmanuel Marchal - Managing Partner at Cassotis Consulting
R. da Paisagem, 220, sala 11S, Vila da Serra, Nova Lima - MG, Brasil