? EAF The Evolution of the Consteel
C. Giavani - Tenova SpA
E. Malfa, V. Battaglia - Centro Sviluppo Materiali SpA
After more than 20 years from its first commercial installation, with 41 references worldwide (8 of ?which are on the way), Tenova’s Consteel EAF has become a proven steelmaking technology,
appreciated for its efficient use of energy and raw materials, operational and maintenance ease, and environmental friendliness. ?The experience made throughout these years, on Consteel EAFs running in quite different
scenarios, and suggestions coming from the end users have given Tenova’s process engineers
several ideas on possible improvement of this technology. Some of these improvements had to be investigated with a deeper look into the complex scrap heating phenomenon taking place inside the ?Consteel system; this was done with the help of the Centro Sviluppo Materiali (CSM), by means of laboratory trials and an extensive use of Computational Fluid Dynamics (CFD) analysis.
This paper describes part of this development work, in which experience, passion and engineering ?have given birth to a new generation of furnaces, the Consteel Evolution?, which will fully ?express the potential of John Vallomy’s great idea: the Consteel system.
?Consteel, EAF, steelmaking, CFD, burners
?The changes introduced with the Consteel Evolution? span on the entire system: from the furnace
up to the heating tunnel section.
The most noteworthy changes in the furnace are a greater freeboard, to allow a better slag retention practice, a revised configuration of the injection system and a larger hot heel , but is definitely in the heating tunnel were the departure from the previous configuration is more evident. The main driver for the whole development has been the reduction the electrical energy consumption for the process by means of an improved scrap heating inside the tunnel, also considering the use of burners, similarly to what is normally done by all modern conventional top-charge EAFs to help the initial melting of the charge.
Substituting electric energy with chemical energy provided by burners can reduce the operating costs – depending on the local cost of electric energy and natural gas – and can improve the overall
efficiency and carbon footprint of the melting process in countries were electric energy is mostly produced by thermal power plants. ?Simply putting burners inside a conventional Consteel heating tunnel has proven to be an
inefficient and ineffective solution (unless someone is looking to heat the fumes and not the scrap) because the burners flames, unless of an unrealistically high power, will get carried away by the main flow of primary process fumes, without reaching the scrap.
The solution to this problem was found in the separation of scrap heating by burners and by process off-gas, introducing a dedicated scrap heating tunnel by burners, the so-called “tunnel B”.
This tunnel is refractory lined and provided with air/natural gas burners located on the roof, at a relatively short distance from the scrap surface. In such configuration the burner flames impinges, relatively undisturbed, on the scrap layer with sufficient momentum to penetrate its cavities, heating it more uniformly.
After exiting from the “tunnel B”, the scrap will continue his travel towards the furnace inside the so-called “tunnel A” where it continues to be heated by process off-gas, as in the standard ?; a common fumes extraction point is provided between the two sections. The following Consteel?Figure 1 reports the conceptual configuration of a Consteel Evolution? EAF and clarifies the
arrangement of the two heating tunnels.
?Figure 1 – Conceptual configuration of the Consteel Evolution? EAF
The configuration of the “tunnel B” was studied with the help of Centro Studi Materiali (CSM) by
means of a synergic approach between a series of laboratory trials, to investigate the effects of an impinging flame on the heating process of scrap located inside a conveyor, and CFD simulations based on an original scrap model validated by experimental data.
MODELLING HEAT EXCHANGE IN THE SCRAP DUE TO FLAME IMPINGEMENT
Flame impingement heating of solids has been used for many years to enhance the convective heat transfer from combustion products to the charge. Some typical applications include melting of scrap metal, shaping glass, heating metal bars, metal fabrication and assembly including soldering, brazing, cutting and welding . In Electric Arc Furnaces (EAF) steelmaking is common practice to use oxygen/fuel burners in order to achieve a faster and more uniform meltdown of the charge, avoiding cold spots ; in such case, the main goal is to achieve a fast scrap meltdown in a specific zone. On the contrary, in the envisioned application of burners to a continuous charging process, the burners must heat an advancing scrap bed - with a speed normally ranging from 1.8m/min to 4m/min - through a dedicated tunnel in which any significant melting must be avoided; the uniformity of the heat flux is, therefore, an important feature for this type of scrap heating process. For this specific application of burners there is very limited data available in literature, therefore, physical and mathematical models have been set up by CSM to evaluate the effects of the various parameters affecting the heating phenomenon due to flame impingement processes, mainly: the position of the burner in respect of the scrap bed, the burner operating conditions, the different shape and layering of the scrap pieces. In the following part of the paper will be presented the approach adopted and some of the results obtained during the R&D work for one type of scrap.
Experimental test rig set up at CSM Combustion Station, includes (Figure 2): ？ ceramic fiber lined furnace characterized by internal dimension of 2020x1740x1470mm; ？ scrap bucket (800x780x750mm), water cooled on two sides.
Figure 2 – Experimental test rig set up at CSM’s Combustion Station
Figure 3 – Temperature measurement for a bucket charged with shredded scrap
The tests have been performed using a commercial available 600kW Tenova’s THS burner ,
typically used in re-heating and treatment furnaces, with combustion air at ambient temperature.
The bucket, charged with about 500kg of scrap (layer of 600mm), has been instrumented with 75 thermocouples to monitor the temperature evolution during the heating process at different scrap height from the top (100/200/300/400/500 mm); other four thermocouples have been placed very close to the scrap surface (5mm), just to monitor the initial heating.
Figure 3 reports the temperature evolution with time for the thermocouple 8 (T8) placed at the center of the scrap bulk; this test was performed using shredded scrap, with three different level of the burner power: 200, 400 and 600 kW. In order to preserve the test setup from local scrap meltdowns it was chosen to stop the test at a temperature of about 1250?C on the top layer. The conditions for local meltdowns where tested last.
These first results have demonstrated the feasibility of flame impingement heating process in conditions similar to those that will be achieved inside a conveyor, giving useful information regarding the required distance between the burner and the scrap surface, the correct power density to be achieved inside an heating tunnel equipped with multiple burners and maximum amount of heat to the charge before the occurrence of a significant superficial meltdown.
One of the goals of the scrap heating trial was, also, to provide data for the tuning of a mathematical model of the scrap, to be used in a complete Computational Fluid Mechanics (CFD) simulation of scrap heating inside the tunnel equipped with burners. The CFD model was targeted, also, at the definition of the proper distance between the scrap and the burner, to achieve the most uniform heating, taking into account the interaction between several burners and include the effects of radiation in the evaluation of the scrap heating phenomenon, since it is known that radiation plays an important role in the case of a complete tunnel; the mathematical model has been developed inside AnsysFluent? CFD code.
The task has involved the development of a comprehensive model of the fluid dynamic, mixing and reaction of the gaseous species, heat transfer between the tunnel and the moving scrap. Towards this end, several sub-models have been necessary, including:
; proper turbulence model to represent the burner flame;
; reaction mechanism of natural gas;
; interaction between turbulence and chemistry;
; radiation from the flame to wall and from wall to scrap;
; solid scrap movement;
; flame penetration inside the scrap charge cavities.
The main effort has been the achievement of a suitable scrap representation, being the other sub-model already set-up and validated at CSM for the modeling of burners and re-heating furnaces . We had to deal with a multiphase transport phenomena in a porous media and most literature on the matter approach this problem assuming local thermal equilibrium between the solid and the fluid phase, an assumption that is not realistic in our case. Attempts have been done by McMaster University to model the scrap melting process in EAF by oxygen/fuel burners and electric arcs without this assumption , however, a validated model that includes, also, the effects of radiation within the scrap pieces has not been available in literature. Therefore, due to the difficulty of developing a comprehensive model for the complex heating phenomena taking place in such heterogeneous material as the scrap, an original “simplified” model has been set up by CSM.
Pin-line geometry defined to
obtain the target A/V ratio
hot gasScrap level
Jet penetationJet penetation
Figure 4 - Scrap representation developed by CSM
First, the penetration of the flame inside the scrap cavities has been obtained considering the scrap as a porous media, using the Brinkman-Forschheimer extended Darcy Model . The porosity and permeability of the various scrap types have been defined by using the McMaster University formulation and parameters (void/filled volume ratio, scrap characteristic lenght) for scrap characterization . To take into account the convective heat transfer due to penetration of the flame and, at the same time, the heat transfer by radiation, the scrap has been represented with a “groove geometry”: the depth of the groves has been assumed equal to the flame penetration length calculated in the previous step, whilst the groove’s width has been selected in order to maintain the
same ratio between area and volume (A/V) of the scrap being considered. Area (A) and volume (V) have been calculated according to the porosity and characteristic dimension of voids considered in the previous step. Then the scrap has been represented as a solid material with equivalent density (？) and equivalent conductivity (k) calculated as follows: equiv equiv
where m is the mass flow of scrap inside the conveyor, S is the cross section of the scrap chargecharge？layer, V is the transport velocity of scrap inside the conveyor, k is the conductivity of pure eqscrapFe%；iron, k is the conductivity of the air, % is the ratio between equivalent density of the charge (？)airequiv ？feand density of iron (？). Fe
The accuracy of the model has been verified comparing the results of the experimental trials with shredded scrap. The test rig, including the furnace, the burner and the bucket have been modeled first for the heating test with the burner set at 200kW of power, in order to reach a steady state condition. The comparison between measured and computed temperature is reported in figure 5 and, as it can be seen, the scrap model has proven to be quite satisfactory.
The major differences are for the thermocouples located close to the water-cooled walls of the bucket, due to the perfect coupling between scrap and walls assumed by the model and, also, due the simplified representation of the scrap cavities; a situation that is not happening in reality.
T11T5 CFD Thermocouples 5-11T8 CFD Thermocouple 8
grid400200200scrap height [mm]scrap height [mm]500100100Scrap height [mm] 000200400600800100002004006008001000otemperature [?C]coolingTemperature [C] temperature [?C]ocoolingTemperature [C] T14
T2T14CFD Thermocouple 2-14gridT7T9CFDThermocouple 7-9
200200400scrap height [mm]scrap height [mm]100100500
00Scrap height [mm] 0200400600800100002004006008001000temperature [?C]otemperature [?C]Temperature [C] oTemperature [C]
Figure 5 – Scrap model validation: 200kW burner power on shredded scrap at steady state
conditions, experimental vs. CFD results
Due to the good results for the steady state case, the 600kW has been also considered. In this condition, since the steady condition could not be reached experimentally, a translation velocity has been imposed on the scrap layer, to simulate the proper residence time under the burner. The quality of the comparison between measured and calculated temperature is very similar to the previous case. In the following Figure 6 is reported the temperature map for the vertical plane, just under the burner axis, for the physical test and CFD simulation.
Figure 6 – 600kW burner on a moving shredded scrap, experimental vs. CFD results
Measurement and calculations using different scrap mixes have confirmed the capability of the approach to predict the flame penetration and the effects of flame impingement heating, including the representation of radiation within the scrap.
MODELLING OF SCRAP HEATING IN THE CONSTEEL TUNNELS
The scrap model developed for the representation of the scrap heating with burners (inside “tunnel
B”) has been also applied at the simulation of the classic scrap heating by means of the melting
process off-gas (inside “tunnel A”), considering the effects of draft air intake and the combustion of CO inside this tunnel . Figure 7 reports the comparison of the heating curve in a conventional tunnel (20000x2000x2000mm), for the same working condition of the EAF, in the case of scrap modeled as a moving solid (no porosity) and with the new model, still considering the use of shredded scrap. With the new scrap model, a higher average temperature of the scrap is achieved towards the connecting car zone (the last part of the conveyor that discharges scrap into the furnace) where the draft air ingress produces a vortex structure that attaches the CO combustion to the scrap surface, generating a velocity field characterized by significant vertical component towards the scrap. These results coming from CFD simulations have been confirmed by field observations that oCscrap filledhave pointed out a better heating of the charge inside the last portion of the tunnel when operating a ? with a porous scrap charge. Consteel
scrap pin linesgroovessolidfilledpin lines 350
300groovessolidscrap pin linesscrap filled
50average scrap temperature [?C]
axial length [m]
Figure 7 – Scrap heating curve for different scrap types: solid and porous (shredded)
After achieving a satisfactory scrap model, it was possible to continue the CFD study for different configurations of the tunnel implementing the flame impingement heating concept. The main goal for this work has been the definition of design guidelines for this tunnel in order to achieve the best possible heat transfer efficiency, with a natural gas consumptions similar to those used in 3conventional Electric Arc Furnace (<9 Nm/t). Soon it was discovered that a modular approach,
such as the one shown in Figure 8, was the way to go.
Scrap - Shredded 3 Equiv. density:0.8 t/m Jet penetration: 62% Water cooling system 2 T =300K - h=5100 W/m K water
1.35 m M 2 M 3 M 1 M 4 M 5 Recovery tunnel
Figure 8 – Example of a heating tunnel equipped with burners: 5 zones, each made by 5 burners
In Figure 9 are reported some results of CFD simulations for this tunnel configuration, in which the
tunnel length is maintained fixed (12m) whilst the burner zones are switched on to maintain a 3/t) at a different scrap feeding rate and, consequently, constant specific consumption (5.7 Nm
different scrap velocity: zone 1 and 2 for 1.76 t/min, zone 1 to 3 for 2.64 t/min, zone 1 to 4 for 3.53
t/min and, finally, all the five zones for 4.4 t/min. Scrap heating curveScrap heating curveScrap heating curveScrap heating curve3333FR NG: 5.7 NmFR NG: 5.7 NmFR NG: 5.7 NmFR NG: 5.7 Nm/t - Energy input: 56.8 kWh/t /t - Energy input: 56.8 kWh/t /t - Energy input: 56.8 kWh/t /t - Energy input: 56.8 kWh/t 300300300300
2002002002004.4 t/min4.4 t/min3.52 t/min4.4 t/min3.52 t/min4.4 t/min1501501501502.64 t/min3.52 t/min2.64 t/min1.76 t/min100100100100temperature [?C]temperature [?C]temperature [?C]temperature [?C]
50505050Scrap heating curve
00003024681012024681012024681012024681012FR NG: 5.7 Nm/t - Energy input: 56.8 kWh/t axial lenght [m]axial lenght [m]axial lenght [m]axial lenght [m]4060Module 538
3654Efficiency1.76 t/minModule 434
32482.64 t/minModule 3303.52 t/min2842Efficiency %264.4 t/minModule 2kWh/ton to scraptotal kWh/t to the scrap2436
2030Module 1Axial lenght1.001.502.002.503.003.504.004.505.00
Figure 9 – Example of CFD simulation results for tunnel equipped with burners
According to these findings it has been possible to conclude that the value of the specific consumption and calculated efficiency are in the range of those typically recognized for wall mounted oxy/fuel burners used in conventional top-charge EAFs; in fact, considering that combustion is performed with gas and cold air and extrapolating the cold air/fuel burner to an equivalent oxy/fuel burner, an average efficiency higher than 60% is expected, indicating the effectiveness of the flame impingement heating technology also in this type of application.
A NEW GENERATION OF FURNACES, THE CONSTEEL EVOLUTION?
The results shown here are relevant to the very first stage of the research that has led to the configuration of the new Consteel? Evolution? system and they have been largely superseded by
the latest refinements. Nevertheless, this first part of the work has been of fundamental importance in indicating the main areas of improvement.
Thanks to this work, it has been discovered a greater than expected potential for scrap heating by convection, even if inside a conveyor, and this has brought to the definition of design criteria for “tunnel B”, but it has, also, brought to a new design for “tunnel “A”, introducing solutions that
increase the turbulence in the primary off-gas stream and, hence, the heating of scrap; even if with a relatively shorter “tunnel A”.
These changes have, also, required a revision of the criteria for charge management and layering inside the conveyor: now the goal is to achieve a porous layer of scrap at the top and the placement of pig iron (due to its lower melting point) and other denser materials in the lower layers of the charge.
Improvements have been made also in the configuration of the burners, in close cooperation with Tenova LOI Italimpianti that provides them: the burners to be used are rated for a power input of 700-800 kW and their flame characteristics has been optimized for the specific application. The control system for the burner section, and for the entire system, will take advantage of Tenova Goodfellow’s EFSOP technology, to achieve a dynamic optimization of the operational parameters.
? Evolution? furnace. Table 1 reports an expected performance level achievable with a Consteel
These benefits would be further enhanced combining this type of furnace with a Tenova ReEnergy Evaporative Cooling System (ECS), to recover most of the heat lost by the process for the ?production of steam; the truly continuous process carrier out in the Consteel Evolution? makes it
a very steady source heat, thus allowing an optimal sizing of the heat recovery unit.
Heat size 100 tls
Power on 33 min
Power off 7 min
Electric energy 297 kWh/tls 3Oxygen 33 Nm/tls 3Natural gas 8.5 Nm/tls
Coal 20 kg/tls
Electrode 1 kg/tls ?Table 1 – Performance achievable with Consteel Evolution? EAF technology
The authors wish to thank colleagues N. Monti (Tenova Melt Shops), for his fundamental support in the beginning of this research, M. Fantuzzi (Tenova LOI Italimpianti) for his work on the burner, C.Bressani, A.Landi and U.Zanusso (CSM) for the set-up of the test rig and the execution of the laboratory tests at CSM Combustion Station.
 F. Memoli, C. Giavani, M. Guzzon, "The evolution of preheating and the importance of hot
heel in supersized Consteel? system", AISTech 2011 Proceedings - Volume I - pag. 823.  S. Chander, A. Ray “Flame impingement heat transfer: A review”, Energy Conversion and
Management 46 (2005) 2803–2837.
 “KT Injection System: the key for chemical energy in high performance Electric Arc Furnace”,
Millennium Steel 2001.
 THS: High Turbulence Tight Flame Burner, Tenova LOI Italimpianti’s burners catalogue.
 V.Battaglia et al., “CFD simulation of Combustion System for Steel Reheating Furnace”, 16?
IFRF Members’ Conference, Boston, USA, June 08-10, 2009.
 E.Malfa et al., “Application of CFD at the EAF Process Simulation”, Innovation in EAF and
Steelmaking Process, Milan, Italy, 27-28 May 2009.
 K. Mandal and G.A. Irons, Numerical modelling of scrap heating by burners, AISTech 2010
Conference, Pittsburgh, USA, May 3-6 2010.
 Kladias N. and Prasad V., 1991, Journal of Thermophysics, 1991, V5, pp560-576.