Arash Tahmasebi, The University of Newcastle, Australia

Co-Author:
Arash Tahmasebi, The University of Newcastle
Brody Brooks, The University of Newcastle
Kim Hockings, BHP

Abstract:
The plastic layer permeability was investigated by using an integrated permeability/dilatation measurement rig and an in-situ permeability rig. The integrated measurement rig enabled the synchronized measurements of permeability and dilatation of coal samples during heating. The in-situ permeability measurement was fitted in the 4kg lab-scale double heated wall coke oven to measure the plastic layer permeability under thermal gradient induced coking conditions. To interpret the measurement data, the microstructure transitions across the plastic layer samples obtained from the plastic layer sampling technique fitted in the 4kg coke oven were analyzed by Synchrotron micro-CT. The permeability results and pore structure parameters derived from those analyses were correlated to better understand the mechanism of plastic layer permeability. Among all samples tested, the high-rank coking coal with low fluidity showed the lowest permeable peak at the later thermoplastic stage, corresponding to the high internal gas pressure (IGP). The microstructure in the resolidified layer of the high-rank coal was characterized by the lowest number of isolated pores and the larger volume of open pores within a larger size range of 50-100 µm, compared to those of the lower rank coals. In addition, its pore wall structure was highly compacted due to the high IGP. Based on these observations, a likely explanation of the lowest permeability of the high-rank coal is that the lower deformability of the pore wall structure at the later thermoplastic stage may have prevented additional pore growth, which prevented pore interconnectivity and decreasing permeability. For the high-pressure coal, there was a slight decrease in porosity in the initial softening layer. It is possible that the formation of the low permeable barrier redirected the plastic mass toward the loose coal side, thus increasing the fluidity of the region and filling the interparticle voids. These results suggest that the plastic layer permeability

Brody Brooks, The University of Newcastle, Australia

Co-Author:
Stephen Brant, BHP
Kim Hockings, BHP
Soonho Lee, University of Newcastle
Brody Brooks, University of Newcastle

Abstract:
Maceral composition is one of the key parameters used to assess coals and to predict coke quality. However, coal grains are often a mixture of different macerals and mineral matters. The development of fluidity in coal grains depends on grain size and the degree of maceral association. It is important to determine the composition of coal grains to understand the fluidity drivers. Coal Grain Analysis (CGA) was used in this study to determine the maceral compositional information of individual coal grains for 4 metallurgical coals varying in rank and maceral composition. The thermoplastic behaviour of coal samples was tested using a Gieseler plastometer and the custom-designed permeability/dilatation testing apparatus and a 4kg dual-heated coke oven. The experimental results showed that coal inertinite content and the degree of association between inertinite (and minerals) and vitrinite greatly influence coal thermoplasticity. Coals with a lower vitrinite particle size and a higher degree of association of vitrinite with inertinite and minerals showed lower dilatation and higher permeability. It was postulated that coals with close maceral association are more prone to volatile gas escape during the plastic phase, effectively hindering bubble formation and growth and thermoswelling while also increasing the viscosity, leading to decreased Gieseler fluidity measurements. However, coals with different degrees of maceral associations showed similar internal gas pressures (IGP) in the 4kg dual heated coke oven and made strong cokes. The liquid phase present in the melt provided sufficient binder material to form strong cokes. Results suggest that some coals possess a higher “effective fluidity” than reported from standard Gieseler testing. On the other hand, coals that display limited maceral association and relatively high proportions of larger vitrinite prolific grains benefit from enhanced bubble growth and coalescence. The apparent fluidities of such coals are more accurately represented by standard Gieseler testing.

Damien Streiff, Zentralkokerei Saar GmbH, Germany

Abstract:
123123

Zion Guetta, thyssenkrupp Industrial Solutions AG, Germany

Abstract:
The Gas Treatment Plants (GTP) of Acciaierie d'Italia has been just upgraded with new Claus plant, ramp up H2S/NH3 Desorption plant capacities and new chiller plant. The Arcerlormittal Gent (Belgium) has just started construction of new H2S/NH3 Desorption plant and new Claus plant. Both GTPs have a common goal, which is to reduce to low as possible the downtime of the desulfurization process by installing a 1+1 setups. This is state of the art of GTPs today. After the modification, Acciaierie d'Italia possess at least one standby unit in all plants of the GTP from Primary Gas Cooler (PGC) and up to Desulfurization. Acciaierie d'Italia has made itself completely weather independent and it is able to run on a hot south Italian summer a constant PGC outlet temperature. The GTP in Gent will continue to keep a clean COG even during boiler inspections or cleaning of internals owning to the new installation. The upgrading of an existing GTP (brown field installation) is counted to be one of the most difficult work in chemical engineering. That kind of work begins with a study to identify targets, concepts and last but not least budgets. At the pre-engineering phase, the tie-in points are investigated and identified by the engineers. The beigest challenge of engineering is to adapt a know process to the existing installation considering for example: flows, pressure and temperatures, accessibility, process control, local standards, modern standards, as well as minimize spare parts by trying to use the same wear parts as in the existing installation. The paper will present the arrangements of the GTPs and examples of engineers challenges, who developed solutions for some given specific difficulties with regard to brown field installations.

Michael Cremer, thyssenkrupp Steel Europe AG, Germany

Abstract:
Coke plant Schwelgern, commissioned in March 2003, is still running the biggest coke ovens worldwide with a coke output of more than 56 tons per oven. The so-called Coke Stabilizing Quenching, the state of the art regarding wet quenching processes, realizes the procedure of coke quenching. The two quenching towers are the highest industrial wood constructions worldwide and hardly stressed by the coke quenching process itself. The presentation illustrates the quenching process itself and shows the different steps of maintenance optimization in a battlefield of abrasion. The presentation gives also an impression about the actual project of demolishing and new erection of Quenching Tower No. 2.

Volodymyr Kuskov, Paul Wurth S.A., Luxembourg

Co-Author:
Christian Dengler, Paul Wurth S.A.

Abstract:
As the equipment designer, Paul Wurth has an intimate knowledge of the WHY’S of the equipment design as well as an in-depth knowledge of process and equipment logic and operations of the latest technology. Growing global competition and increasing pressure to maximise production efficiency are key forces acting on the adoption of new digital tools within the iron and steel industry. Organisations recognise the importance of extracting and analysing information from data, as well as persisting and amplifying knowledge accumulated during years of experience in the types of equipment used in the primary iron making industry. The following will focus data acquisition and on knowledge-based condition monitoring systems and their practical application in maintenance, specifically for the Paul Wurth coke oven machines and in particular for our Stamping Charging Pushing Machines. Monitoring and evaluating the condition of complex equipment requires analysis of everything from simple variable data to recognition of trends and complex phenomena resulting from multiple factors. Effective evaluation of critical events in the operation of complex equipment may rely on knowledge acquired from both human and machine. A secure platform and tools that allow integration of human and machine learning can enable continuous growth of baseline knowledge used to operate and monitor critical assets. Digitally captured knowledge from process and equipment experts with an abundance of real-time data adds the potential to predict and prevent asset failure by identifying and preventing the re-occurrence of detected failures and recognize repetitive events. Enabling engineers with tools omitting the need for programming or data science experience makes this possible. Rule engines driven by human knowledge reduce the potential for overwhelming operators faced with continuous alerts requiring no immediate remedy or action. THE COMPLETE ABSTRACT IS ATTACHED.

Bert Gols, Tata Steel Europe, Netherlands

Abstract:
The integrated steel site in IJmuiden has two coke plants to supply coke to the blast furnaces. Major rebuilds are not economically feasible. So the challenge is to keep these coke plants in operation for as long as possible to remain at least self-sufficient. Especially Coke Plant # 2 is vulnerable for high coking pressures due to its age (45+ years), weaker design and higher oven walls. To control coking pressure, pilot test facilities are used to measure the gas pressure. A good relation between gas pressure and pushing force of a coke oven was found. Limiting the amount of power required to push an oven is believed to prevent damage to an oven wall. In the pilot oven the gas pressure of all blending piles entering the coke plant is measured as well as all individual coal shipments and several trial blends to investigate intended blend changes. The dataset of all the pilot oven tests is used to develop predictive models for gas pressure. This model requires the data of the individual coal test to widen the data range to give the model predictive power outside the range of blends used for the coke plant. Even though a model like this can predict the general direction and variation between coal types quite well, it fails to predict the difference between various shipments of the same coal. A further investigation into this revealed that variations in moisture and particle size of the same coal type have influenced the gas pressure measurement. These variations do not show up in analytic models, because the influence of these factors is coal type dependent. When the limitations and constraints of these models are understood they can be a useful tool in the optimisation of the coal blends and protection of our assets.

Michael Specht, Refratechnik Steel GmbH, Germany

Co-Author:
Michael Cremer, thyssenkrupp Steeleurope

Abstract:
The requirements for the door plug building blocks of the Schwelgern coking plant are particularly challenging due to the door height of 8.43 meters. In order to ensure the tightness of the doors, the durability of the insulating bricks of the coke oven doors, the so-called door plug components, is essential for a safe oven door repair process. In the course of the continuous improvement of processes (CIP), the proven door plugs modules of the Schwelgern coking plant were further improved. The focus of further improvement was on reducing the expansion coefficient of the components in order to achieve a longer service life and less susceptibility to cracking. Initial tests in the laboratory and in practice showed positive results and confirmed the theoretical approach. In addition to the newly developed material grade, the prefabricated components were reinforced with a special steel fiber to achieve another positive effect. The operational experience of the last six years, which is shown in this presentation, shows that the newly developed components outperform the conventional parts in terms of service life. The consumption of door plug modules could be reduced significantly.

Ko-ichiro Ohno, Kyushu University, Japan

Co-Author:
Kazuya Kunitomo, Kyushu University
Takayuki Maeda, Kyushu University
Shungo Natsui, Tohoku University
Shingo Ishihara, Tohoku University
Takashi Watanabe, Tokyo Institute of Technology
Hirokazu Konishi, Osaka University

Abstract:
Correct knowledge about softening behaviour of iron burden in cohesive layer is becoming more and more important for the minimum carbon blast furnace operation. In order to precisely understand cohesive behaviour, a unique softening simulator under rapid heating and quenching conditions was applied. Outcomes from normal softening simulator are very complex because they depend on reduction degree, chemical composition, basicity, melts physical property, etc. One of the reasons to make complex this test is sample for this test provided as particle packed bed. In order to comprehend softening behaviour of cohesive zone, focus on behaviour of single particle as a component of particle packed bed in this study. For simplification, a pre-reduced simulant sinter iron ore was prepared as single particle sample. The samples were made from mixture of reagent oxides and they were pre-reduced to several reduction levels. The sample was rapidly heated up to 900℃, and then gradually heated up with 10℃/min under inert gas atmosphere and 0.1MPa load. Shrinkage degree of the sample particle was measured during the softening test, and quenched sample was made at certain temperature when the sample shows a characteristic tendency. Cross-sectioning observation of the quenched sample provided mineral phases distribution in the sample particle. The observation indicated that peripheral structure has a significant effect on deformation resistance. Especially in there, existence of molten slag phase could make easy to deform the sample particle shape, and metallic solid Fe phase helped to strengthen the particle's deformation resistance.

James Small, Tata Steel, Netherlands

Co-Author:
Chloe Buytendijk, Tata Steel
Stefan Melzer, Tata Steel
Sieger van der Laan, Tata Steel
Kooij Chris, Tata Steel
Enno Zinngrebe, Tata Steel
Yanping Xiao, Tata Steel
Magdalini Ntouma, Tata Steel
Frank van der Does, Tata Steel
Mary Wijngaarden-Kroft, Tata Steel

Abstract:
Here we discuss the influence of chemistry, mineralogy and microstructure of iron ore sinter on its reduction behavior under conditions relevant to the upper shaft and the reserve zone of the blast furnace process. Sinter material was produced in a sintering pot varying only limestone flux- and coke addition rate with an otherwise constant sinter mixture, to yield variation in basicity (CaO/SiO2: 1.7 – 4.1) and FeO (6 - 19 wt%). This material was coarsely ground and screened to 250-500 µm and used as starting material for isothermal reduction treatments in lab-experimental gas-mixing furnaces at the following conditions: (1) 750 °C with XCO (= CO/[CO+CO2]) of 0.55; and 950 °C with XCO of (2, 3 & 4 respectively) 0.65, 0.76, 0.90. The experimental starting materials thus retain the mineralogical and microstructural variations of quasi-industrial sinter, while excluding mesostructural variations (e.g. sinter size). Long duration experiments were performed to compare the resultant net reduction degree and mineralogy with (Factsage-)predicted thermodynamic equilibrium. Excellent agreement was found between the predicted and measured reduction degree of samples from long-duration experiments within the wustite stability field (conditions (1) & (2)), with broad agreement too in terms of mineralogical composition. Samples were consistently less reduced than predicted with conditions (3) & (4), reflecting kinetic limitations. For (4), these are the formation of dense metallic layers around wustite domains in the case of lower-basicity sinters, and the sluggish reduction of metastably formed brownmillerite (C2(A,F)) to metallic Fe, lime and Ca-aluminate in higher basicity sinters. Shorter duration time-series experiments performed at conditions (1) and (2) show clearly the passage of sequential concentric reaction(-zone) fronts through the particles. Categorical differences are seen in the initial reduction behavior of Si-Al-rich SFCA versus Fe-rich SFCA-I, while all SFCA types remain initially unreacted as the hematite-magnetite reaction front passes.

Olli Mattila, SSAB Europe Oy, Finland

Co-Author:
Timo Paananen, SSAB Europe Oy
Lauri Halonen, Sapotech Oy
Juha Roininen, Sapotech Oy
Iikka Salmela, Sapotech Oy
Henrik Saxen, Åbo Akademi

Abstract:
Ironmaking pellets transported overseas are usually moist to prevent dust emissions during vessel loading. Dust generation cannot be avoided when pellets collide on conveyor system parts and when they move in contact with each other starting from the pellet incineration line and ending in the blast furnace (BF). Due to restrictions in available field area or construction costs, large pellet storage silos are not optimized with respect to fines flow behavior. Their operation often leads to the accumulation of fines and the discharge of sudden pulses of fines. Pellet fines pulses together with moisture create sieving problems prior to the BF and in the BF process itself by impairing burden permeability and, in consequence, BF stability. If the share of fines in the feed is constant, it is possible to fine-tune the sieving practice and charging pattern in the BF to maintain a proper permeability distribution. To tackle the uneven fines discharge-related problem, the behavior of fines in silos during sequential charge and discharge cycles was modeled with the Discrete Element Method (DEM). Based on the modeling results, a new systematic practice to fill and discharge the storage silos was applied in the automation system. To measure the amount of fines in the pellet stream discharged from large storage silos, an optical image analysis measurement on the bottom of the conveyor line was developed. In addition, the separated fines originating from multiple simultaneously operating sieving machines were analyzed with respect to pellet quality with a newly installed conveyor balance and simple delay model. By combining information from the extracted flow of fines with measurements of pellet moisture, the fines flow to the BF is estimated, and the information can be used for taking appropriate actions to guarantee BF operation stability.

Igor Shepetovsky, JSC "Kosaya Gora Iron Works" , Russian Federation

Co-Author:
Ivan Chmerenko, Kosaya Gora Iron Works
Andrew Shalygin, Kosaya Gora Iron Works
Rustam Nuriev, Kosaya Gora Iron Works
Konstantin Myasnikov, Kosaya Gora Iron Works
Alla Shepetovskaya, Kosaya Gora Iron Works

Abstract:
Stringent requirements of contemporary production of high-duty castings and special grades of steel lead to a new approach specific to basic burden material for foundry and special steelmaking which tends to use of virgin ferrous material with minimum impurities. Over a long period Kosaya Gora Iron Works (Tula, Russia) develops blast furnace technology for large scale manufacturing of pig iron with extended set of quality requirements in accordance with the above mentioned new approach. Along with the typical task to minimize content of conventionally detrimental sulfur and phosphorus in pig iron, technologists of Kosaya Gora must minimize content of manganese, titanium, vanadium and other tramp elements (up to 10 elements in all) in produced pig iron. Partially this task is solved due to application of ferrous burden with 100% pellets which are chemically purer than sinter. This paper describes peculiarities of production of high quality pig iron with precise qualitative properties through blast furnace route using state-of-the-art technologies and equipment.

Xue Feng Dong, School of Mechanical, Materials, Mechatronic and Biomedical Engineering, University of Wollongong, Australia

Co-Author:
Apsara Jayasekara, School of Mechanical, Materials, Mechatronic and Biomedical Engineering, University of Wollongong
Dominique Sert, ArcelorMittal Maizieres Research SA
Rodolfo Paulo Santos Ferreira, ArcelorMittal Maizieres Research SA
Pascal Gardin, ArcelorMittal Maizieres Research SA
Brian Monaghan, School of Mechanical, Materials, Mechatronic and Biomedical Engineering, University of Wollongong
Sheng Chew, Coke & Ironmaking Technology, BlueScope Ltd
David Pinson, Coke & Ironmaking Technology, BlueScope Ltd
Paul Zulli, School of Mechanical, Materials, Mechatronic and Biomedical Engineering, University of Wollongong

Abstract:
Molten slag, which is primarily generated in the blast furnace (BF) cohesive zone, trickles down through the coke packed bed in the form of films, rivulets or droplets in the lower zone of the BF. During its downward flow, there are significant interactions occurring between slag and other phases. In terms of these interactions, slag flow behaviour can greatly affect BF productivity and be associated with furnace irregularities. Hence, understanding the interactions between phases is useful to maximizing BF efficiency. In the current study, a Volume of Fluid (VOF) modelling technique was applied to track the movement of individual slag droplets at a mesoscopic level. The numerical modelling was firstly developed and validated through comparisons between simulation and experimental results via qualitative and quantitative methods. The VOF approach was then used to investigate the slag flow behaviour in the packed bed, considering various bed permeabilities, more wide-ranging slag properties, and strong interactions between slag and other phases. Results demonstrate the significant role of modelling at a mesoscopic level in understanding macroscopic slag flow behaviour. Modelling work helps to visualize the trickling behaviour of slag droplets in more realistic and complex conditions representing a BF, and clarify the mechanisms of the different flow patterns generated for variations in operating conditions. As examples, flow characteristics such as localised slag flooding and gas channelling caused by the strong interaction between gas and slag were uniquely identified and analysed. As the hydrostatic pressure dominates the flow of rivulets downwards, slag and liquid iron enhance their respective flows through the bed rather than retard each other. The current modelling proved to be a valuable tool to provide a foundation for better understanding the slag flow behaviour and its interactions with other phases in the BF lower zone.

Rob van Opbergen, Danieli Corus, Netherlands

Abstract:
Title: Channelling in the blast furnace Oscar Lingiardi (1), Rob van Opbergen (2) and Maarten Geerdes(3) 1) Ternium, Argentina, olingiardi@gmail.com 2) Danieli-Corus, The Netherlands, Rob.van-Opbergen@danieli-corus.com 3) Geerdes Advies, The Netherlands, geerdes@mgeerdes.nl Channelling in a blast furnace is a local, preferential gas flow through an area with the lowest gas resistance. Channelling occurs frequently and has major impact in loss of chemical and physical energy of the gas. Most of the channelling starts along the wall of the furnace. Because of the inverse conical shape of the shaft a gap is easily formed when burden descends. Channels have a self-enhancing effect: a strong gas flow blows away materials, preventing materials collapse in the channel. Therefore channels are observable from strong decrease of the gas utilization, locally high heat losses and/or stave temperatures as well as pressure taps. Experience shows, that channels can be caused by irregular burden descent, because fast descent creates a gap between wall and burden. Irregular burden descent is caused by uneven melting of the cohesive zone. This can be caused by - Concentration of fines. - Drainage of primary slag because of poor melting. - Unstable burden descent. Remedial actions for channelling are the following. - In the very short run an operator will try to have the channels collapse by checking and operate at a lower blast volume. - In the long run, the occurrence of channels has to be prevented by having stable burden descent. The paper will show, that effects on ETA CO, stave temperatures, heat losses and pressure taps indicate the same channelling phenomenon. Remedial actions are discussed and operational examples will be given.

Lijia Wu, AG der Dillinger Hüttenwerke, Germany

Co-Author:
Rongshan Lin, AG der Dliinger Hüttenwerke
Dieter Senk, IEHK RWTH Aachen University
Alexander Babich, IEHK RWTH Aachen University
Stephan Hojda, AG der Dliinger Hüttenwerke
Pierre Van Dorpe, PAUL WURTH S.A.
Juraj Micák, PAUL WURTH S.A.

Abstract:
The blast furnace is nowadays still the main facility for metallic iron production in steel making. Because of the current economic and ecological situation, a stable operation aiming at lowing fuel consumption is strongly required. Thus, reliable thermal state prediction plays an important role in thermal control. Compared to the silicon content in hot metal, the hot metal temperature is considered to be more representative of the current thermal state. Hence, for model fine-tuning, many thermal state prediction models especially machine learning models require not only high measurement accuracy of hot metal temperature measurement but also enough datasets. This paper introduces first a short overview of methods of hot metal temperature measurement and presents secondly the model-based thermal control system applied at ROGESA. To further improve this system, continuous hot metal temperature measurements are installed for each of the two tap holes on blast furnace No.5. The results are compared with the conventional hot metal temperature measurements that are taken by immersion thermocouple probes and integrated into the current thermal control system for further improvement of thermal state prediction supported by machine learning. Key words: thermal control; hot metal temperature prediction; continuous hot metal temperature measurement; machine learning models

Eric Schaub, Paul Wurth Deutschland GmbH, Germany

Co-Author:
Ralf Allmannsdörfer, Paul Wurth Deutschland GmbH
Markus Bierod, Paul Wurth Deutschland GmbH
Volker Dulz , Paul Wurth Deutschland GmbH
Rudolf Hebel, Paul Wurth Deutschland GmbH
Martina Engelmann, Paul Wurth Deutschland GmbH
Stephan Bamberg, Paul Wurth Deutschland GmbH

Abstract:
Modern blast furnace (BF) hearth linings can reach life times of 15 years and more. During such a long BF campaigns, a number of lining properties and characteristics (e.g. thermal conductivity (TC)) can change unnoticed, if alkalis form potassic feldspars, especially in carbon hearth lining layers. Besides this, other phenomena like “heat resistances” can create unnoticed changes in the hearth side wall lining, too. One has to distinguish between positive heat resistances like “skulls” and heat resistances with a negative effect on the BF hearth linings like cracks, gaps, or partly destroyed lining structures, like “brittle layers”. It is hard to detect such changes of important lining properties with classical observation measurement techniques. Without identifying such changes, it is very difficult to calculate the important “true and real” hearth lining wear profiles in different measuring sections and measuring levels with a 2D- or 3D- mathematical FE model. Paul Wurth Deutschland GmbH (Paul Wurth) has developed so-called multipoint thermocouple sensor probes (MTP sensor probes) in combination with necessary heat flux probes, to detect such kind of unnoticed changes of BF hearth lining properties on the one hand and “heat resistances” in the BF hearth linings on the other. In order to detect such phenomena, the multipoint MTP sensor probes are installed deep in the BF hearth lining wall and work reliably throughout the whole BF campaign. The MTP sensor probes provide “online” the necessary calibration and reference points for the true and real lining wear calculation with a special tailor-made mathematical FE model. Aiming at monitoring the condition and extending the service live of modern blast furnaces, especially under today’s harsh technical and economic conditions, Paul Wurth presents in this paper examples of typical installations, recent experience of detected abnormalities as well as application of MTP sensor probe technology and mathematical models.

Yakov Gordon, Hatch, Canada

Abstract:
Beginning of development of gas based direct reduction processes goes back to the end of 19 century. The first industrial application was Wiberg-Soderfors DRI shaft furnace process. The history of development and implementation of various gas based direct reduction processes was studied. The strength and week features of this processes were evaluated. It was shown that economics of the direct reduction process depends on cost of raw material and reducing agent, reliability of equipment and strict application of direct reduction principles. Deviation from these main principles leads to the failure of the new DRI technologies. Results of evaluation allowed to show, why Midrex and HYL are the only successful DRI gas-based processes, while other lost their competitive position and stop operation.

Hesham Ahmed, Lulea University of Technology, Sweden

Co-Author:
Bo Björkman, Luleå university of Technology
Johanna Alatalo, johanna.alatalo@afry.com
T K Sandeep Kumar, LKAB
Hesham Ahmed, Lulea University of Technology

Abstract:
Iron and steel industries are major contributors to the greenhouse gas emissions. The majority of these emissions are linked to the use of fossil-based carbon. Hence, decarbonization of the steel industry is one of the pathways toward a fossil-free environment. One of the decarbonization strategies is to replace the fossil-based reductant by green hydrogen. However, the produced zero-carbon DRI will cause the refining and steelmaking operations to have a starting point far from today’s operational practices. The furnace feedstock melting point will be increased by several hundred degrees, melting will take place over a very short melting interval, gas formation in the melt will be very limited, and slag foaming will be difficult to achieve without introducing additional carbon. Therefore, in hydrogen-based iron making process, a way to introduce carbon in a controlled manner is essential. In the present study, attempt have been made to investigate the carburization of hydrogen-reduced DRI at the single pellet scale. The factors affecting the carburization process, such as gas composition, temperature and duration were investigated thoroughly in the temperature range of 600 to 800oC. Further, the carbon introduced into the hydrogen-reduced DRI can either be in the form of cementite, graphite (free-carbon) or mixture of both. Therefore, the effect of carbon content and the form in hydrogen-based DRI for their respective melting characteristics have also been investigated, and presented here.

Nella Janakova, Progres Ekotech, s.r.o., Czech Republic

Co-Author:
Nella Janakova, Progres Ekotech, s.r.o.

Abstract:
The paper presents a cold technology to produce a low-carbon BF feedstock which is used to supplement the sintering process of iron ore as it is currently known. The implementation of this production process of BF feedstock causes lower energy consumption compared to the high-temperature process of sintering. The result is quite a reduction of energy and significantly lower carbon footprint. A key factor in reducing the carbon footprint is the cold production process, as there are no direct CO2 emissions. In addition, the technology contributes to lower consumption of iron ore and coke and to efficient re-utilization of by-products from metallurgical processes like scale and fractions of dust and slag. The BF burden produced by The EcolBriq® process shows mechanical product characteristics, very favorable mechanical and physical properties, such as: high strength 6-12 MP, which makes it easily ready as BF charge. An additional benefit of EcolBriq® is the possibility of year-round outdoor storage without changing the briquette properties. A single production line with a capacity of 400KT / year can be utilized for a wide variety of materials with different structures, eg: powder, oxide, metal with sizes 0-200 mm. In contrast with sinter the briquette´s shape and size can be adjusted within a wide range to optimize the BF operation. This flexibility offers the possibility of creating an optimal low-carbon charge tailored to optimum blast furnace operation and quality of the hot metal. The presentation will show the advantages by two case studies, which quantify the results arising from: i. Reduction of CO2 emissions for production of the EcolBriq® feedstock compared to sinter ii. Reducing the consumption of natural resources for EcolBriq® iii. Dust reduction for EcolBriq® compared to sinter plant iv. Maximizing the efficiency of the circular economy in BF operations v. Reducing energy intensity

Thomas Baumann, AKW Apparate + Verfahren GmbH, Germany

Co-Author:
Thibaut Richard, AKW Apparate + Verfahren GmbH

Abstract:
Almost all the iron and steel manufactured in the world is made from pig iron produced by the blast furnace process (BF). However, the dust and especially the sludge represent a great challenge to improve the overall sustainability of this highly material and energy intensive industry. During the production of pig iron in blast furnace, a Zn- and Pb-containing sludge is generated in the air pollution control system. More than ½ of the mass input becomes outputs in form of off-gas and solid wastes/by-products. This toxic waste can be landfilled after dewatering and pretreatment, which is very costly. The sludge particles contain large amounts of Fe and C that could be recycled in the furnace. However, the Zn content of the sludge is high, and the Zn input to the blast furnace must be limited, so Zn has to be removed (major portion concentrated in the <20 µm). There are no standard plant concepts for a BF-sludge treatment plant. The process design and plant arrangement will primarily depend on the nature of the feed sludge and therefore will be based on pilot test work that is being performed in AKW Equipment + Process Design technical laboratory. On basis of the test results, the suitable and customized process solution will be developed, discussed and later on engineered and executed by AKW Equipment + Process Design. This unique process concept is presented in the following paper: multi-stages hydrocycloning, combined with thickening and filter pressing.

Olli Vitikka, University of Oulu, Finland

Co-Author:
Andrey Malkki, Kivisampo Oy
Illia Tkalenko, AMCOM GROUP LLC
Nikita Koriuchev, AMCOM GROUP LLC
Timo Fabritius, University of Oulu
Anne Hietava, University of Oulu
Mikko Iljana, University of Oulu

Abstract:
One way to improve the recycling of by-products from iron and steel production is briquetting, a process in which fine materials unsuitable for use as such are agglomerated to achieve a larger particle size. This work is about the high-temperature properties of auger pressing briquettes mainly consisting of blast furnace sludge and mill scale. The aim was to determine the suitability of the briquettes for blast furnace (BF) ironmaking by studying the reduction, swelling and cracking behavior using blast furnace simulator (BFS) furnace. The BFS, able to perform non-isothermal reduction experiments with changing gas atmospheres, was used to simulate the reducing conditions in a BF. In the BFS experiments, different stages of reduction up to 1100 °C were simulated for the auger pressing briquettes. A commercial olivine pellet and a differently agglomerated industrial blast furnace briquette were used as reference samples. The sample weight losses were monitored by thermogravimetry, swelling as a change in the external dimensions, and cracking by visual inspection. TG-MS analysis was carried out to determine the presence of potentially harmful volatiles. The samples were analyzed using LOM and FESEM to study the phase transformations. The auger pressing briquettes proved to be a promising raw material for BF use. They were of self-reducing type due to their carbon content, and they were reduced to metallic iron faster compared to the reference briquettes. The swelling was slight, and despite minor cracking the auger pressing briquettes did not degrade. No harmful volatile substances were found but their existence cannot be completely ruled out.

Stephan Höntsch, TU Bergakademie Freiberg, Germany

Co-Author:
Andreas Siegfried Braeuer, TU Bergakademie Freiberg
Christopher Harris, voestalpine Stahl GmbH
Christoph Thaler, voestalpine Stahl GmbH
Laura Lohmeier, TU Bergakademie Freiberg
Ralf Wollenberg, TU Bergakademie Freiberg
Hans-Werner Schröder, TU Bergakademie Freiberg
Volker Herdegen, TU Bergakademie Freiberg

Abstract:
Midrex process residues, such as screened oxide fines, hot briquetted iron (HBI) screened fines, dried sludge and process classifier dust contain iron and iron oxide that should be saved for better utilization of resources. The presentation is about basic research to achieve this by way of briquetting the residues with or without a binder for recycling in the reduction shaft furnace. After reduction the residue briquettes are converted into HBI by joint hot briquetting with the major proportion of direct reduced iron (DRI) pellets. Another way of utilizing the residues is adding them directly to the hot DRI pellets without prior reduction and integrating them finally also into the HBI. Process requirements for briquette production are described as well as quality parameters of the briquettes with special emphasis on mechanical strength, transport safety, thermal stability and reduction properties. REM and EDX analytical methods are used to get a deeper insight into the micro structure of the briquettes which helps understand the effect of process parameters on briquette quality. Finally, the results of a briquetting test with a roller briquetting machine in semi industrial scale gives prove of transferability of the results obtained in laboratory to industry.

Elsayed Mousa, Swerim, Sweden

Co-Author:
,
Hesham Ahmed, Luleå University of Technology
Mohamed El-Sadek, Central Metallurgical Research and Development Institute

Abstract:
The recycling of iron rich residues created by the different steel works has turn into an urgent issue due to ores depletion and strict environmental regulations. Mill scale is one of the iron rich residues generated during steelmaking, casting, and rolling processes. It is too fine to be charged directly to the iron and steel making units without pre-agglomeration. This paper demonstrates the feasibility of briquetting of the mill scale using organic binders to meet the requirements of hydrogen-based direct reduction. The influence of binder type, binder dosage, moisture content, compaction pressure on the briquetting process and the briquettes quality will be investigated. The briquettes quality will be characterized for their mechanical strength and reduction behaviour. The mechanical strength is measured by using a compressive tester machine and the reduction behviour is studied by thermogravimetric analysis (TGA).