| 1. |
- Carlsson, Leo, 1992-
(författare)
-
Applied Machine Learning in Steel Process Engineering : Using Supervised Machine Learning Models to Predict the Electrical Energy Consumption of Electric Arc Furnaces
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The steel industry is in constant need of improving its production processes. This is partly due to increasing competition and partly due to environmental concerns. One commonly used method for improving these processes is through the act of modeling. Models are representations of the reality that can be used to study and test new processes and strategies without costly interventions. In recent years, Machine Learning (ML) has emerged as a promising modeling approach for the steel industry. This has partly been driven by the Industry 4.0 development, which highlights ML as one of the key technologies for its realization. However, these models are often difficult to interpret, which makes it impractical to validate if the model accurately represents reality. This can lead to a lack of trust in ML models by domain practitioners in the steel industry. Thus, the present work investigates the practical usefulness of ML models in the context steel process engineering. The chosen application to answer this research question is the prediction of the Electrical Energy (EE) consumption of Electric Arc Furnaces (EAF). The EAF process was chosen due to its widespread use in the steel industry and due to the difficulty to accurately model the EE consumption using physical modeling. In the present literature, the use of linear statistical models are commonly used even though the EE consumption is non-linearly dependant on multiple important EAF process variables. In addition, the literature does neither investigate the correlations between input variables nor attempts to find the most optimal model with respect to model complexity, predictive performance, stability, and generalizability. Furthermore, a consistent reporting of predictive performance metrics and interpreting the non-transparent models is lacking. These shortcomings motivated the development of a Model Construction methodology and a Model Evaluation methodology that eliminate these shortcomings by considering both the domain-specific (metallurgical) aspects as well as the challenges imposed by ML modeling. By using the developed methodologies, several important findings originated from the resulting ML models predicting the EE consumption of two disparate EAF. A high model complexity, governed by an elevated number of input variables and model coefficients, is not necessary to achieve a state-of-the-art predictive performance on test data. This was confirmed both by the extensive number of produced models and by the comparison of the selected models with the models reported in the literature. To improve the predictive performance of the models, the main focus should instead be on data quality improvements. Experts in both process metallurgy and the specific process under study must be utilized when developing practically useful ML models. They support both in the selection of input variables and in the evaluation of the contribution of the input variables on the EE consumption prediction in relation to established physico-chemical laws and experiences with the specific EAF under study. In addition, a data cleaning strategy performed by an expert at one of the two EAF resulted in the best performing model. The scrap melting process in the EAF is complex and therefore challenging to accurately model using physico-chemical modeling. Using ML modeling, it was demonstrated that a scrap categorization based on the surface-area-to-volume ratio of scrap produced ML models with the highest predictive performance. This agrees well with the physico-chemical phenomena that governs the melting of scrap; temperature gradients, alloying gradients, stirring velocity, and the freezing effect. Multiple different practical use cases of ML models were exemplified in the present work, since the model evaluation methodology demonstrated the possibility to reveal the true contributions by each input variable on the EE consumption. The most prominent example was the analysis of the contribution by various scrap categories on the EE consumption. Three of these scrap categories were confirmed by the steel plant engineers to be accurately interpreted by the model. However, to be able to draw specific conclusions, a higher model predictive performance is required. This can only be realized after significant data quality improvements. Lastly, the developed methodology is not limited to the case used in the present work. It can be used to develop supervised ML models for other processes in the steel industry. This is valuable for the steel industry moving forward in the Industry 4.0 development.
|
|
| 2. |
- El-Tawil, Asmaa
(författare)
-
Bio-coal as an alternative reducing agent in the blast furnace
- 2020
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The steel industry is aiming to reduce CO2 emissions by different means; in the short-term, by replacing fossil coal with highly reactive carbonaceous material like bio-coal (pretreated biomass) and, in the longer term, by using hydrogen. The use of bio-coal as part of top charged briquettes also containing iron oxide has the potential to lower the thermal reserve zone temperature of the Blast furnace (BF) and, due to improved gas efficiency, thereby give a high replacement ratio to coke.In order to select a suitable bio-coal to be contained in agglomerates with iron oxide, the current study aims at investigating the devolatilization behavior and related kinetics of different types of bio-coals. In addition, the aim is to investigate the self-reduction behavior of bio-coal-containing iron ore composite under inert condition and simulated blast furnace thermal profile.In the BF the temperature of the top-charged material will increase rather quickly during the descent in the upper part. Ideally, all the carbon and hydrogen contained in the top-charged bio-coal should contribute to the reduction. The devolatilization of bio-coal is thus important to understand and to compare between different types of bio-coal.To explore the devolatilization behavior for different materials, a thermogravimetric analyzer equipped with a quadrupole mass spectrometer was used to monitor the weight loss and off-gases during non-isothermal tests for bio-coals having different contents of volatile matter. The samples were heated in an inert atmosphere up to 1200°C at three different heating rates: 5, 10 and 15°C/min. The thermogravimetric data were evaluated by using the Kissinger–Akahira–Sonuse (KAS) iso-conversational model and the activation energy was determined as a function of the conversion degree. Bio-coals with both low and high content of volatile matter can produce reducing gases that can contribute to the reduction of iron oxide in bio-agglomerates. Bio-coals containing a higher content of catalyzing components such as CaO and K2O will enhance the devolatilization and release of volatile matter at a lower temperature. The self–reduction of composites was investigated by thermogravimetric analyses in argon atmosphere up to 1100°C and evolved gases were monitored by means of quadrupole mass spectroscopy. Composites with and without 10% bio-coal and sufficient coke breeze to keep the C/O molar ratio equal to one were mixed and Portland cement was used as a binder. To explore the effect of added bio-coals, interrupted vertical tube furnace tests were conducted in a nitrogen atmosphere at temperatures selected based on thermogravimetric results, using a similar thermal profile as for the thermogravimetric analyzer. The variation between fixed carbon, volatile matter contents and ash composition for different types of bio-coal influences the reduction of iron oxide.The results showed that the self-reduction proceeds more rapidly in the bio-coal-containing composite and that the volatile matter could have contributed to the reduction. The self-reduction of bio-coal-containing composites started at 500°C, while it started at 740°C with coke as the only carbon source. The hematite was successfully reduced to metallic iron at 850°C with bio-coal present as a reducing agent, but not until 1100°C when using coke.Use of bio-coal with high content of volatile matter but low content of catalyzing elements as potassium, sodium and calcium in bio-agglomerates for the BF can be recommended because it enhances the self-reduction of iron oxide, e.g., wustite was detected by XRD analysis in samples treated up to 680°C. Bio-coal with low content of volatile matter, low alkalis, low phosphorous and high content of fixed carbon will also be suitable to use in the BF.
|
|
| 3. |
- Hessling, Oscar, 1988-
(författare)
-
Some aspects of hydrogen reduction of iron ore
- 2024
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Fines of hematite and magnetite were studied in the fluidized bed using a pure hydrogen atmosphere in the temperature range of 768 to 888 K. Hematite pellets were studied based on Thermogravimetric Analysis (TGA) experiments using hydrogen atmospheres containing 0–15 % pH2O, in the temperature range of 873 to 1173 K. Thermocouples in contact with the pellet’s surface and embedded in the pellet’s center recorded the thermal gradient in the pellet during reduction. The fluidized bed and the TGA experiments used an alternative method to start the reaction. The furnace was pre-heated with a reactive atmosphere. After this, the reaction was initiated by introducing the sample to the hot furnace, to eliminate the effect of gas dilution during gas switching. After the experiments, Scanning Electron Microscopy (SEM)analyses were employed to study the reduction microstructures. Both types of fines showed similar reduction rates. Fines and pellets showed high initial reduction rates, which increased with increasing temperatures. The reduction rate in the last reduction stage was low for both fines and pellets. An increasing pH2O content in the atmosphere lowered the reaction rate, and theeffect decreased with increased temperature. A difference between the pellet's surface and center temperatures was observed during reduction. The pellet'smacro-pore structure was seen to be unaltered by changes in temperature or atmosphere. However, at 873 K, the iron product microstructure was found to be highly porous. Furthermore, increasing temperatures caused dense iron to form. In addition, when porous iron or iron oxides were observed, increasing pH2O contentsincreased the pore diameter but decreased the pore amount. Pellet properties with varied pellet compositions were also investigated using Cold Crushing Strength (CCS), reduction in a TGA setup, and melting experiments. The composition was not found to influence the mechanical or reduction properties but significantly affected the phosphorus refining during melting.The results showed that a mixed reaction rate control occurred during the early reduction stage for both the fines and the pellets. The temperature differences observed during this reduction stage resulted from a combined effect of heat transfer and an endothermic chemical reaction. The impact of water in the atmosphere influenced the reaction rate through the backward reaction and mass transfer. At 873 K, the retarding effect is mainly caused by the backwardreaction. The results show the late stage of reduction to be primarily diffusioncontrolled. In addition, it should be possible to alter the pellet composition while maintaining pellet properties to increase the usefulness of the pellet.
|
|
| 4. |
- Swartling, Maria, 1983-
(författare)
-
An Experimental and Numerical Study of the Heat Flow in the Blast Furnace Hearth
- 2008
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- This study has focused on determining the heat flows in a production blast furnace hearth. This part of the blast furnace is exposed to high temperatures. In order to increase the campaign length of the lining an improved knowledge of heat flows are necessary. Thus, it has been studied both experimentally and numerically by heat transfer modeling. Measurements of outer surface temperatures in the lower part of a production blast furnace were carried out. In the experimental study, relations were established between lining temperatures and outer surface temperatures. These relations were used as boundary conditions in a mathematical model, in which the temperature profiles in the hearth lining are calculated. The predictions show that the corner between the wall and the bottom is the most sensitive part of the hearth. Furthermore, the predictions show that no studied part of the lining had an inner temperature higher than the critical temperature 1150°C, where the iron melt can be in contact with the lining.
|
|
| 5. |
- Vesterlund, Mattias, 1980-
(författare)
-
District heating system analysis and challenges within the urban transformation of Kiruna
- 2017
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- There is currently an ongoing urban transformation in a small Swedish town named Kiruna, it is located in the very north of Sweden well above the Arctic Circle in a sub-arctic climate. Large part of the town will be relocated due to the ground deformation that is caused by the progressing iron ore mining activity and it is affecting all infrastructures of the town. This thesis aims to accomplish a holistic analysis on the district heating (DH) system for the town of Kiruna and its future challenges.Energy companies with a DH system recognize the importance in having a good understanding about the network characteristics, for obtaining an efficient and stabile heat delivery to the end-users. In this thesis, a method for modeling and simulation of meshed DH networks is described, that makes it possible to study and analyze the flow pattern in order to locate non-obvious paths, bottlenecks and overloaded pipes.For carrying out the DH simulations a fundamental input is to set the thermal losses for each pipe segment in the model, a fictitious series with all pipe diameters is created which corresponds to the annual losses in the real network. In comparison with the pipes series manufactured today the created one is best described by the series with least insulation and highest thermal losses. The studied network has its origin in the 60th and is the sum of the different piping technics that has been valid over time, this mixture is positioning the thermal performance as a close to a worst-case scenario.To the meshed DH network a number of heat production sites are connected for delivering the thermal requested by the end-users, each site consisting of several boilers and using different resources. A hybrid evolutionary-Mixed Integer Linear Programing (MILP) optimization approach is developed and applied for finding the cost-optimal heat production for three scenarios in combination of two heating demand levels. It is stated that no matter the geographically location of the site the cheapest resource should always be favorable as fuel, in the case when the same resource is viable at different sites a differentiated heat production is obtained. The supply temperature from each site is found to be the one lowest possible in order to serve all site-concerned end-users with a temperature level high enough for hot water production. The findings recommend a network temperatures reduction with the consequence in higher cost related to pumping work, but is lower than the savings due to the reduction in thermal losses.In order to provide the relocated part of the town with DH the hybrid evolutionary-MILP optimization routine is reshaped for finding different alternatives for network expansion layouts. The result is presented as a multi objective analysis between the operation cost and installation cost, showing the complete spectra of all optimal possible solutions and how the different cost correlate to each other. In this way, the outcome can be used for support in decision making, helping network owners is their planning and pipe sizing for new areas.For constructing the buildings that will populate the new city-area the Swedish government has stated a number of recommendations for achieving livable thermal indoor climate. An investigation is carried out analyzing the impact from the usage of three different heating system; air/air heat pump, air heating and floor heating in a low energy family house, where the first two system are aimed to use heat from the DH network. The analysis show that only the floor heating system satisfies the recommendations stated, but with carefully planning an air heating system could also fulfill the recommendations. Further, a techno-economical evaluation declares that the cheapest heating cost over 30 years is by using an air/air heat pump. In order to make DH competitive as heating source the needed price reduction is found for the hydronic floor and air heating system.Finally, three different building energy performance scenarios are studied in conjunction with the urban transformation in combination with the suggested energy measures from the Energy Performance certificates (EPC). In order to reach the national target entailing a reduction of 50% until 2050 all re-built buildings have to be built with passive standard and all advised measures in the EPC has to be carried out. Wort noticing is that the scenarios is analyzed as part of a 3-D City Model, which is found to be a worthwhile working tool for staff dealing with energy related issues.
|
|
| 6. |
- Brämming, Mats
(författare)
-
An Operational View on Foaming and Slopping Control in Top-blown BOS Vessels
- 2015
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Slag formation plays a decisive role in all steelmaking processes. In top-blowing Basic Oxygen Steelmaking (BOS) i.e., in the LD process, an emulsion consisting of liquid slag, dispersed metal droplets, undissolved particles and solid precipitates will, together with process gases, form an expanding foam. Extensive research has defined the parameters that govern the foaming characteristics of BOS slag-metal emulsions. It is a well-known fact that certain process conditions in the Basic Oxygen Furnace (BOF) will lead to excessive foam growth, forcing foam out through the vessel opening (mouth). This process event is commonly known as slopping. Slopping results in loss of valuable metal, equipment damage, lost production time, unsafe work environment and pollution. A literature survey covering the slopping phenomena has been carried out, as well as a deeper investigation into the causes behind slopping on the BOF type LD/LBE at SSAB Europe, Luleå, equipped with an automatic system for slopping registration using image analysis Good slag formation and foam-growth control in order to avoid slopping is primarily accomplished by taking preventive “static” measures. The most common pre-blowing operational conditions favouring foam growth and, hence, slopping were found to be linked to oxygen lance positioning, hot metal Si and Mn contents, scrap quality and large additions of iron oxide bearing materials. Improved slopping control may be achieved by developing oxygen lance control schemes with automatic adjustment of the distance between the lance tip and the metal bath (i.e., the lance gap) according to scrap quality and ore additions. If “static” measures cannot be effectuated, a set of in-blow slopping preventive measures is needed. For such “dynamic” measures to be effective, it is necessary to have a system for slopping prediction. Trials with vessel vibration measurements for indirect foam height estimation in industrial scale BOFs, type LD/LBE, have been carried out. FFT spectrum analysis was applied in order to find the frequency band with best correlation to an estimated foam height. The results show that there is a correlation between vessel vibration and foam height which can be used for dynamic foam level and slopping control, and this during the entire blow. The vessel vibration results have been tested against what is the perhaps most commonly implemented technique for dynamic foam height estimation and slopping control, the audiometric system. Parallel vibration and audio measurements have been carried out on 130-tonne as well as on 300-tonne BOFs. The results show that during stable process conditions there is good agreement between the two methods in regard to foam height estimation and that combining the two methods will provide a powerful slopping prediction and control system. A feasibility study has been carried out with the aim to describe the possibilities and limitations of multivariate data analysis, including batch analysis, for dynamic BOS process control, mainly in regard to slopping prevention. Two principal modelling approaches were tested.A central part of this PhD work is the performed emulsion characterisation and the subsequent investigation into the influence of emulsion mineralogy and morphology on slopping in the LD process. The results are based on the study of emulsion samples from trial heats conducted in a 6-tonne pilot plant LD vessel. The main emulsion slag phase mineral species identified were di-calcium silicate, monoxides (mainly FeO, MnO and MgO), calcium ferrites and late-appearing tri-calcium silicate. The study also show that the iron oxidation state has a large influence on the emulsion mineralogy and morphology, as a higher Fe3+ content facilitates the precipitation of calcium ferrites, raising the emulsion apparent viscosity and, hence, the foam index. The same effect is caused by higher MgO contents (i.e., at saturation), resulting in the precipitation of monoxide phase. However, large volume fractions of emulsion precipitates will not always lead to slopping in the LD process. A second “requirement” for excessive foam growth is a simultaneously high gas generation rate. Vice versa; an LD heat may very well slop at low volume fractions of 2ndphase particles in the emulsion if the gas generation rate is sufficiently high. It is an indisputable fact that excessive foaming is one of the main features of the LD process, due to the practice of top-lance oxygen blowing, creating a highly oxidised slag, and heavy batch additions of basic slag formers, causing an initial formation of large quantities of precipitates. Therefore, preventing slopping is primarily a matter of tight process control, most importantly, control of the oxygen lance gap in order to reach a state of sufficiently high liquid MeO phase to minimise the emulsion apparent viscosity, but low enough to avoid over-oxidising and a high gas generation rate.
|
|
| 7. |
- Johansson, Maria, 1968-
(författare)
-
Improved Energy Efficiency and Fuel Substitution in the Iron and Steel Industry
- 2014
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- IPCC reported in its climate change report 2013 that the atmospheric concentrations of the greenhouse gases (GHG) carbon dioxide (CO2), methane, and nitrous oxide now have reached the highest levels in the past 800,000 years. CO2 concentration has increased by 40% since pre-industrial times and the primary source is fossil fuel combustion. It is vital to reduce anthropogenic emissions of GHGs in order to combat climate change. Industry accounts for 20% of global anthropogenic CO2 emissions and the iron and steel industry accounts for 30% of industrial emissions. The iron and steel industry is at date highly dependent on fossil fuels and electricity. Energy efficiency measures and substitution of fossil fuels with renewable energy would make an important contribution to the efforts to reduce emissions of GHGs.This thesis studies energy efficiency measures and fuel substitution in the iron and steel industry and focuses on recovery and utilisation of excess energy and substitution of fossil fuels with biomass. Energy systems analysis has been used to investigate how changes in the iron and steel industry’s energy system would affect the steel plant’s economy and global CO2 emissions. The thesis also studies energy management practices in the Swedish iron and steel industry with the focus on how energy managers think about why energy efficiency measures are implemented or why they are not implemented. In-depth interviews with energy managers at eleven Swedish steel plants were conducted to analyse energy management practices.In order to show some of the large untapped heat flows in industry, excess heat recovery potential in the industrial sector in Gävleborg County in Sweden was analysed. Under the assumptions made in this thesis, the recovery output would be more than three times higher if the excess heat is used in a district heating system than if electricity is generated. An economic evaluation was performed for three electricity generation technologies for the conversion of low-temperature industrial excess heat. The results show that electricity generation with organic Rankine cycles and phase change material engines could be profitable, but that thermoelectric generation of electricity from low-temperature industrial excess heat would not be profitable at the present stage of technology development. With regard to fossil fuels substituted with biomass, there are opportunities to substitute fossil coal with charcoal in the blast furnace and to substitute liquefied petroleum gas (LPG) with bio-syngas or bio synthetic natural gas (bio-SNG) as fuel in the steel industry’s reheating furnaces. However, in the energy market scenarios studied, substituting LPG with bio-SNG as fuel in reheating furnaces at the studied scrap-based steel plant would not be profitable without economic policy support. The development of the energy market is shown to play a vital role for the outcome of how different measures would affect global CO2 emissions.Results from the interviews show that Swedish steel companies regard improved energy efficiency as important. However, the majority of the interviewed energy managers only worked part-time with energy issues and they experienced that lack of time often was a barrier for successful energy management. More efforts could also be put into engaging and educating employees in order to introduce a common practice of improving energy efficiency at the company.
|
|
| 8. |
- Muwanguzi, Abraham Judah Bumalirivu
(författare)
-
Investigating the parameters that influence the behaviour of natural iron ores during the iron production process
- 2013
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- In the iron production processes, sinters and pellets are mostly used as raw materials due to their consistency with respect to physical and chemical properties. However, natural iron ores, as mined, are rarely used directly as a feed material for iron processing. This is mainly due to the fact that they have small contents of iron and high concentration of impurities. Moreover, they swell and disintegrate during the descent in the furnace as well as due to low melting and softening temperatures. This work involves an investigation of the parameters that influence the use of natural iron ores as a direct feed material for iron production. Furthermore, it points out ways in which these can be mitigated so as to increase their direct use in iron production.Natural iron ore from Muko deposits in south-western Uganda was used in this study. Initially, characterisation of the physical and chemical properties was performed, to understand the natural composition of the ore. In addition, investigations were done to study the low temperature strength of the ore and its behaviour in the direct reduction zone. Also, simulations were performed with three models using the experimental data from the direct reduction experiments in order to determine the best model for predicting the direct reduction kinetics of natural iron ores.Chemical analyses showed that the Muko ore represents a high grade of hematite with an Fe content of 68% on average. The gangue content (SiO2+Al2O3) in 5 of the 6 investigated iron ore samples was < 4%, which is within the tolerable limits for the dominant iron production processes. The S and P contents were 0001-0.006% and 0.02-0.05% respectively. These can be reduced in the furnace without presenting major processing difficulties. With respect to the mechanical properties, the Muko ore was found to have a Tumble Index value of 88-93 wt%, an Abrasion Index value of 0.5-3.8 wt% and a Shatter Index value of 0.6-2.0 wt%. Therefore, the ore holds its form during the handling and charging processes.Under low temperature investigations, new parameters were discovered that influence the low temperature strength of iron oxides. It was discovered that the positioning of the samples in the reduction furnace together with the original weight (W0) of the samples, have a big influence on the low temperature strength of iron oxide. Higher mechanical degradation (MD) values were obtained in the top furnace reaction zone samples (3-25% at 500oC and 10-21% at 600oC). These were the samples that had the first contact with the reducing gas, as it was flowing through the furnace from top to bottom. Then, the MD values decreased till 5-16% at a 500oC temperature and 6-20% at a 600oC temperature in the middle and bottom reaction zones samples. It was found that the obtained difference between the MD values in the top and other zones can be more than 2 times, particularly at 500oC temperature. Furthermore, the MD values for samples with W0 < 5 g varied from 7-21% well as they decreased to 5-10% on average for samples with W0 ≥ 5 g. Moreover, the MD values for samples taken from the top reaction zone were larger than those from the middle and bottom zones.During direct reduction of the ores in a H2 and CO gas mixture with a ratio of 1.5 and a constant temperature, the reduction degree (RD) increased with a decreased flow rate until an optimum value was established. The RD also increased when the flow rate was kept constant and the temperature increased. An optimum range of 3-4g was found for natural iron ores, within which the highest RD values that are realised for all reduction conditions. In addition, the mechanical stability is greatly enhanced at RD values > 0.7. In the case of microstructure, it was observed that the original microstructure of the samples had no significant impact on the final RD value (only 2-4%). However, it significantly influenced the reduction rate and time of the DR process.The thermo-gravimetric data obtained from the reduction experiments was used to calculate the solid conversion rate. Three models: the Grain Model (GM), the Volumetric Model (VM) and the Random Pore Model (RPM), were used to estimate the reduction kinetics of natural iron ores. The random pore model (RPM) provided the best agreement with the obtained experimental results (r2 = 0.993-0.998). Furthermore, it gave a better prediction of the natural iron oxide conversion and thereby the reduction kinetics. The RPM model was used for the estimation of the effect of original microstructure and porosity of iron ore lumps on the parameters of the reduction process.
|
|
| 9. |
|
|
