Reduction of industrial hematite pellets using hydrogen and some of its impacts on steelmaking
Time: Fri 2025-11-21 13.00
Location: B1, Brinellvägen 23, Stockholm
Video link: https://kth-se.zoom.us/j/68130437151
Language: English
Subject area: Materials Science and Engineering
Doctoral student: Julia Brännberg Fogelström , Processer
Opponent: Professor Kazuki Morita, The Department of Materials Engineering, The University of Tokyo
Supervisor: Professor Pär Jönsson, Processer; Prefessor Emeritus Du Sichen, Processer; Dr Johan Martinsson, Processer, Business Area Metallurgy, Swerim AB; Dr Niklas Kojola, Processer, Group of Research and Innovation SSAB AB, Klarabergsviadukten 70 D6, Box 70 101 21 Stockholm Sweden
Abstract
The transition to fossil-free steelmaking is one way to significantly reduce the anthropogenic carbon dioxide emissions and meet the climate targets. The thesis investigates the hydrogen-based reduction of hematite pellets and some impacts on the steelmaking. Gaining a better understanding of the reduction process and how it links together with the melting operation, opens the possibility for process optimizations of the new H-DRI- EAF route.
Since previous works have showed that the reduction reaction is greatly affected by the type of hematite sample used, this work uses commercial industrial hematite pellets to get relevant results that directly relate to pilot-scale processes in use today. The reduction behavior was studied under isothermal and non-isothermal conditions. During the isothermal reduction studies, the focus was directed to the effect of nitrogen dilution, since in some cases nitrogen gas is used in the inert gas seals that insert and extract the material from the reactor. The result showed that the nitrogen content, in the investigated range (0-30 vol%N2), reduced the reduction rate at all temperatures but the rate decreased more than expected at the temperatures of 600°C and 900°C. The rate decrease was mainly caused by the decreased driving force for gas diffusion and chemical reaction.The reduction temperature showed to have a large impact on the microstructure, however the N2 dilution showed no noticeable effect on the same.
To gain a deeper understanding of the reduction taking place inside a shaft furnace reactor, the reduction behaviors under non-isothermal conditions were studied. An extra focus was directed toward the effect of heating rate and water vapor contents in the reduction gas. Studying the reduction behavior under non-isothermal conditions is crucial when trying to optimize the reduction process taking place inside a shaft furnace reactor. The heating rate and water vapor content in the reduction gas had a significant impact on the rate of reduction, i.e., the rate increased with increasing heating rate and decreasing H2O content. The reduction was initiated at around 450°C when pure hydrogen gas was used for reduction. However, when the reduction gas contained 5-20vol%H2O, the reduction started at a temperature close to 525°C. The microstructure at the surface of the pellet showed had similar appearances, irrespective of the heating rateapplied. Furthermore, the main feature of the microstructure showed no signs of sintering when further heated after complete reduction. However, the main feature of the center microstructure varied depending on the heating rate. The appearance of the iron phase changed from porous to dense structure when the temperature exceeded 668°C. The formation of dense iron significantly decreased the rate of mass transfer of H2 and H2O to and away from the reaction site through the product layer- affecting the total reduction rate. When the reduction gas contained more than 5 vol%H2O, the reduction becamemore complex as a mechanism change occurred between 0.11-0.15 degrees of reduction. The mechanism change led to a low-rate stage, and it was related to the presence of FeO phase. Consequently, it is crucial to consider this mechanism change when modeling and optimizing the shaft furnace process. To summarize, the total reduction rate increased with faster heating rate due to increased rate of the chemical reaction, the mass transfer through the product layer, as well as it widened the reaction zone.
The melting behavior of H-DRI was investigated to gain insight into the preliminary impurity contents when melting H-DRI. This was done to better understand the need for optimization of the refining process when the new H-DRI-EAF process is in use. The work contained pilot-scale and laboratory samples to systematically study the pickup and release of hydrogen and nitrogen, and the formation of inclusions during melting of H-DRI. The result showed that significant amounts of the ingoing hydrogen and nitrogen in the H-DRI were adsorbed on the surface. During melting, the surface area significantly decreased, and the hydrogen and nitrogen contents almost decreased instantaneously. However, the final contents were around 4-36 ppm H and 20-30 ppm N, mainly due to the elements increased solubility in the crude steel and higher mobility at the higher temperature. The inclusions found inside the crude steel after melting 100% H-DRI were analyzed. First the main mechanism of formation of the inclusions were studied using laboratory work. The H-DRI was subjected to heat, and the residual oxides present inside the H-DRI after reduction was forced together when the porous structure sintered. The inclusions were of different sizes and composition depending on what type of oxides that merged. The merging of the oxidic particles seemed to be one of the main mechanisms of formation of the inclusions. The inclusions varied significantly in size and composition, but the majority was in the size range >5.6 to 22.4 µm. However, a considerable number of inclusions having a size above 22.4 µm were also found. In total three different types of inclusions formed, and they are denoted as Type I-1, Type I-2 and Type I-3. Type I-1 inclusions formed when the autogenous slag (O-2) merged with Mg-rich oxides (O-1) with high MgO contents. Type I-2 inclusions formed when large amounts of autogenous slag merged with only very small number of Mg-rich oxides. Type I-3 inclusions contained single Mg-rich oxides that had not merged with any other oxides.