Adnan Husakovic, Primetals Technologies Austria, Austria

Co-Author:
Klaus Grasserbauer, Primetals Technologies Austria
Andreas Melcher, Primetals Technologies Austria
Dietmar Hofer, voestalpine AG
Wolfgang Höfer, voestalpine AG
Anton Tushev, Primetals Technologies Germany
Ali Abbas, Primetals Technologies Austria

Abstract:
Green steel production requires a better knowledge of processes and their optimization. The advances in data science and hardware capabilities within recent decade provided new digitalization methods. Two of the most prominent areas recently discussed in science are computer vision and data-driven process modelling. The scrap consistency and EAF-process model gain additional insights and potentials towards EAF-process stabilization and improvement. Within the scope of this paper, we show the importance of Computer Vision and data-based modelling on two use cases: scrap composition monitoring and EAF-process modelling. It summarizes the benefits for the steel producer.

Matthias Werner, VDEh-Betriebsforschungsinstitut GmbH, Germany

Co-Author:
Martin Hubrich, VDEh-Betriebsforschungsinstitut GmbH
Matthias Kozariszczuk, VDEh-Betriebsforschungsinstitut GmbH

Abstract:
Due to climate change, water shortages are developing worldwide, which amplifies the existing water stress and is now also occurring regionally in European countries. This is accompanied by a change in the water composition, such as the increase in mean conductivity by a factor of 2.5 and in the chloride content by a factor of 7 in the period from 2013 to 2019 in an exemplary surface water body. The effects for the iron and steel industry, in which water is an essential process medium, are an increased demand for freshwater due to earlier reaching of the limit conductivity and an increase in corrosion in the pipe systems due to increased chloride contents. This results in the need for the development of a prognosis tool to predict emerging bottlenecks and the development of suitable methods for tapping alternative water sources to ensure the water supply. In the application case, an integrated iron and steel works consisting of the units blast furnace, hot rolling and cold rolling mill as well as its central waste water treatment plant (CWTP) is considered in more detail. The focus of BFI-work is the desalination of the CWTP effluent using membrane-based capacitive deionization (MCDI). Furthermore, as part of the project work, the digital representation of the entire water management system with more than 20 circuits and a wide variety of water compositions was created in the SIMBA# software environment. This forms the basis for the development of the forecast simulation tool. Based on the results of the sampling and tests, various scenarios relating to the chloride content in the surface water and the effects of returning the treated effluent from the CWTP were examined using simulation technology. Further work focuses on the integration of geodata in the developed simulation model for forecasting water bottlenecks.

Yale Zhang, Hatch Ltd. , Canada

Co-Author:
Sa Ge, Hatch Ltd.
Nooshin Nekoiemehr, Hatch Ltd.

Abstract:
Climate change has quickly become the focal point of both challenges and opportunities for today’s iron and steel industry. The industry-wide focuses have shifted from pursuing mainly economic objectives to balancing with environment-centric KPIs such as GHG emissions, energy and raw material efficiency. Many steel producers are facing great challenges on transitioning towards green steel, including: (1) undertaking serious commitments on carbon emission reduction; (2) responding to stakeholder’s pressure on visibility of product carbon footprint; and (3) developing optimal strategy to deal with the impact of carbon pricing. Among many emerging disruptive technologies of Industry 4.0, digital twins have received a high adoption rate due to the steadily growing maturity of IoT, cloud computing and increased number of successful applications. This paper focuses on how digital twin technology can help in driving and optimizing steel producers’ decarbonization strategy. A novel concept of “Carbon Twin” is proposed from the following three perspectives: Product carbon footprint tracking, to monitor GHG emissions in near real-time and calculate product carbon footprint across entire value chain to create a reliable and transparent GHG emission disclosure to customers and stakeholders. Carbon reduction scenario analysis to conduct what-if analysis to assess the impact of different decarbonization technologies, flowsheet variations, process and operation decisions, raw material selection on GHG emissions reduction, to support decarbonization roadmap development and project implementation priorities. Integrated cost and emissions optimization to introduce carbon taxes and/or green premiums into the integrated steel value chain cost modeling, offering the opportunity to optimize overall profitability by considering process, logistics, and emission constraints, and determine the best strategy to achieve balance between production cost and carbon emissions. Several industrial project examples will be discussed in the present paper to demonstrate significant benefits of carbon twin.

Paula Pomaro, Vetta, Brazil

Co-Author:
Lis Soares, Vetta
Thiago Maia, Vetta

Abstract:
Seeking to meet the environmental agreements and climate conventions, a lot of effort and resources have been made to decarbonize industrial sectors. The steel sector alone is responsible for 7% of global emissions and 5% of emissions in Europe. However, besides the energy crisis, base industries in Europe are facing unfair competition with industries from countries that do not have the same level of ambition in climate matters. Despite the huge challenge, key factors are helping to drive the transition towards carbon-friendly steel products and maybe the most powerful and effective aspect is the change in customer requirements. There is an increasing trend in the steel-consuming industries, such as the automotive industry, of eliminating carbon emissions from their entire value chains and these customers are even willing to pay more for the steel produced under reduced emissions. In order to help steel industries to manage their emissions and create added value by differentiating their products, Vetta®, an SMS group company, has developed a digital solution for the management of emissions. This work presents Viridis Carbon, a digital solution for the management of gaseous and particulate emissions applied to the steel industry, from scopes 1, 2, and 3. The tool is able to issue reports on CO2 and GHG emissions in real time, generating transparency for stakeholders, policymakers, and competent authorities for taxation. It is also capable of tracking the emissions generated by each billet of steel produced, issuing emission certification seals, which makes it possible to differentiate products and increase added value for an increasingly demanding market with environmental commitment.

Andreas Wolff, VDEh-Betriebsforschungsinstitut GmbH, Germany

Co-Author:
Stefano Dettori, Scuola superiore di studi universitari e di perfezionamento Sant'Anna

Abstract:
As a result of strong globalization and rapid scientific progress in almost all subfields relevant to industry, large industrial companies increasingly embrace the advantages accompanying the use of digital twins’ automation, and of analysis and optimization of previous process structures. In addition to these changes, the industry faces challenges such as reducing CO2 emissions, efficiency and optimization processes across the entire value chain driven by global competition, and the minimization of waste products. This is particularly relevant in the steel industry, where the identification of optimal process parameters for complex control models is often a limiting factor for improving efficiency and reducing CO2 emissions. To achieve a carbon-free steel industry and further improve their gas and steam networks and management practices, established approaches need to be reconsidered to reduce CO2 emissions and waste and increase overall efficiency. In this paper, we propose an approach that leverages Machine Learning and the Koopmann approach to improve the accuracy of digital twins. By demonstrating this approach using case studies from recent projects within the Research Fund for Coal and Steel (RFCS).

Bertrand Orsal, Dassault Systèmes, France

Abstract:
The Steel industry is at a tipping point where sustainability and greenhouse gas emissions are top items on CEOs’ agendas. The old philosophy of engineering and production for profit is giving way to engineering and production for sustainability. Steel producers are adapting their business models, changing their priorities, and shifting their strategies as a response to the shifting market conditions. We know that this involves a transformation towards new ways of production limiting emissions like the direct reduction, but also capturing the CO2 emissions and optimizing the whole production chain. Dassault Systèmes provides a collaborative work environment (3DExperience) and comprehensive set of solutions that focus on green steel to enable and help accelerate this transformation. Solutions range from design, engineering and industrial process simulation capabilities to environmental assessment tools to calculate the impact of a plant over the course of its lifecycle. It includes a set of solutions to operate facilities such as integrating planning & scheduling. It provides capabilities for certification of green steel production through tracking of all plant data in a single source of truth platform. Find out how Dassault Systèmes and the 3DExperience platform make it possible to manage all industrial projects for green steel in the same collaborative space, for all roles involved, from design to production.

Bernd Koch, Matplus GmbH, Germany

Co-Author:
Alex Miron, Matplus GmbH
Igor Alperovich, Matplus GmbH

Abstract:
For structural applications, steel already has the best overall recycling rate of all materials - however, steel for structural components, automotive chassis and bodywork is today largely produced via the classic blast furnace route, which is still associated with high CO2 emissions for the foreseeable future. A key aspect in reducing the overall life cycle balance is the step towards the "circular economy", in which largely complete recycling is to be achieved. To this end, a holistic view of the production, usage and end-of-life processes is required. For steelmaking this task is faced with a “scale” issue – the actual processes are highly sophisticated and detailed down to atomistic chemical reaction scale, but the overarching material and energy flows are in the kiloton / GWh regime. While for both extremes process and simulation tools are available – classical LCA for the macroscale for instance – a universal modelling tool is missing, which bridges between the detail of process simulation and measurement and the macroflows, allowing to balance all efforts and outputs in a mathematically sound and cohesive manner. In this paper an innovative approach is presented, integrating well-established flow analysis methodology in a state-of-the-art materials knowledge system. Concrete approaches for mapping and modelling EAF steelmaking processes in terms of material and energy flows from input scraps to steel and recyclate output are presented, and concepts for bridging the scale gap are developed. Finally, the vision of a universal communication platform based on this model and data structure for all concerned parties, from waste managers over steelmakers to producers, is established.

Uwe Diekmann, Matplus GmbH, Germany

Co-Author:
Petra Becker, Matplus GmbH
Dominik Zuegner, voestalpine Forschungsservicegesellschaft Donawitz GmbH
Sabine Zamberger, voestalpine Forschungsservicegesellschaft Donawitz GmbH

Abstract:
A key challenge for steel manufacturers and users is the rapid creation and use of new technologies with improved environmental performance as a product of lightweighting and circularity. The acceleration of development processes can be achieved through the use of modern simulation techniques that also draw on extended knowledge bases. One element of the extended knowledge base is VDEh's Stahldat SX. Beyond standard data the Stahldat SX knowledge base includes research reports of applied steel research (FOSTA) and a growing number of models for transformation kinetics (TTT, Jominy), plasticity (flow curves) and compliance information (CAS-numbers). In practical application the information is extended by own results from material qualification projects along the process chain as well as specific CAE interfaces. For green steel technologies the mass- and energy flows as time series data along the process chain are subject of evaluations. Practical results with regard to accelerated development of new steel technologies are presented. This includes the consideration of energy demands along the process chain and comparisons to LCA data.

Didier Morice, AST Technology S.A.S. France, France

Co-Author:
Fabrizio Masia, AST Technology
Said Alameddine, AST Technology
Geraldo Ferreira, AST Technology

Abstract:
The quest for global optimization of Industrial processes to tackle environmental management recently took a new path thanks to the advent of big data analysis, IIoT, machine learning and artificial intelligence. During the last three decades, the industry in general has accumulated vast amounts of data but only a very small amount was generally explored and exploited. The new techniques mentioned, however, will enable the maximum benefit from all the accumulated data. The authors explain through tangible examples how supply chain and trade flows as well as available environment open-source databases can be pooled in the same data lake along with process data and explored with powerful analysis tools to twin not only the industrial process chain but also its footprint and impact on the environment. The authors conclude with a new emerging paradigm called “The Green AI”. This new paradigm will be used more and more extensively in the next decade to the benefit of Environmental Management. The obvious payback of this paradigm will be accompanied by major improvements of process and environmental factors.

Ali Vazirizadeh, Aisimpro Inc, Canada

Co-Author:
Roholah Abbaszadeh, Gol Gohar Sirjan F.C.
Ladan Foroughi, Aisimpro inc
Mehdi Azizkarimi, Gol Gohar Sirjan F.C.
Reza Khaksarpour, Gol Gohar Sirjan F.C.

Abstract:
Advances in data science, machine learning, and artificial intelligence are transforming mining and metals and making them more algorithm intensive. The paradigm is shifting from one of detection and control to one of prediction and optimization where advanced soft sensors take an essential position. While conventional soft sensors were mainly based on linear and physical models, modern machine learning techniques bring the opportunity to improving the commonly used soft sensors and provide new types of soft sensors that couldn’t be developed through conventional methods. In this paper, two industrial case studies on advanced soft sensor applications in a pelletizing plant will be discussed. The first one is a soft sensor estimating the cold crushing strength (CCS) index of pellets which helps operators to make better decisions in real-time and improve pellet quality. The value propositions of the application of CCS soft sensor from economical and environmental perspectives have been studied. It was demonstrated that a reduction of pellet rejects and consequently CO2 emission can be achieved in the Direct Reduction Iron (DRI) plant by improving pellet CCS in the pelletizing plant. The second soft sensor predicts the amount of fired pellet FeO% which associates penalty if it goes beyond the target. An economical analysis demonstrates the potential for revenue improvement if FeO% soft sensor is implemented in a 5 MTPA pellet plant.

Shawn Siroka, ITR | Industrial Technology Research , United States

Co-Author:
Borui Li, Industrial Technology Research
Andrew Lauden, Industrial Technology Research
Jonathan Davis, Industrial Technology Research
Klaus Stohl, Primetals Technologies Austria

Abstract:
"Bearing defects of low-speed bearings are difficult to detect with traditional vibration analysis because the energy generated from these defects is often insufficient to show a measurable change in the acceleration signature. Therefore, these applications require special techniques such as the shock pulse method (SPM) or ultrasonic resonance excitation (URE) to accurately quantify the health status of an asset. These techniques rely on the impacting nature of bearing defects to excite an accelerometer at resonance which amplifies the measurability of the defect. Conventionally, SPM and URE techniques have utilized piezoelectric accelerometers. However, MEMS accelerometers warrant investigation for their recent advances in noise performance and impact detection at low frequencies. This paper compares MEMS and piezoelectric accelerometers for low-speed bearing applications. First, the impulse response of both sensor types is characterized. Then, the performance of each sensor is experimentally quantified by measuring known bearing faults. Finally, sensors are used in a real-world application on a ladle turret to classify the operating condition of the asset. Overall, this work shows how MEMS accelerometers can increase the detectability of bearing defects in critical low-speed applications. "

Robert Brunnbauer, S+P Samson GmbH, Germany

Abstract:
In the steel industry, things can get pretty hot in a rough and dirty environment. To ensure reliable material tracking, optimum warehouse organization and smooth shipping logistics under these adverse conditions, modern and efficient identification solutions are required - ideally automated. This is exactly where S+P Samson has its roots and where it has acquired over 40 years of experience. The robust GRAPHIPLAST® data carriers have been customized for the requirements of the steel industry and can be used either as a self-adhesive label or a hangtag. The new ID4track® RFID High Temperature hangtag is a polymer film with a 1-side heat-resistant coating. The material withstands a continuous load up to 360 °C / 680 °F, for a short time up to max. 450 °C / 842 °F and is weather, aging and scratch resistant. Its resistance to ultraviolet radiation makes it suitable for outdoor use. A protective shield over the RFID inlay ensures trouble-free use in drying processes with high temperatures. The tags are attached with metal clips. Read range at non-metals is up to 8 m and up to 5 m at metals. The new ID4track® RFID High Temperature Tag consists of three main components: 1. tag material: GRAPHIPLAST® 7377.140 with a tear-out strength of 140N 2. transponder: bonded on PI substrate and works in the UHF frequency band of 860 to 960 MHz (ETSI and FCC, ISO/IEC 18000-6C). It is based on the chip NXP UCODE 7xm (EPC: max. 448 bit, TID: 96 bit factory locked, User memory: 2048 bit) 3. protective heat-shield: also GRAPHIPLAST® 7377.140 The RFID solution for your processes in the steel industry and its difficult industrial identification requirements.

Dongsong Li, RWTH Aachen University, Germany

Abstract:
Thickness requirements must be met when determining toughness properties in typical Charpy or fracture mechanics testing. As a result, the characterization of toughness qualities is a challenge in thin-walled constructions. This problem can be solved by replacing Charpy impact toughness testing with impact notch tensile testing. Toughness requirements, however, are still specified in terms of conventional test findings. As a result, a framework for translating these standard test criteria into impact-notch tensile test requirements is proposed here. The proposed framework is based on numerical simulations with a phenomenological damage mechanics model, which predicts local damage and global fracture using state-of-stress-dependent, strain-based criteria. This model considers the impacts of non-proportional strain routes and uses different cleavage and ductile fracture criteria to accurately anticipate the activation of cleavage and ductile fracture mechanisms in the associated numerical simulations.