Start-Up Experience and Results of Consteel at the SOVEL Meltshop

ThIS ARTICLE IS AVAILAbLE OnLInE AT www.AIST.ORg fOR 30 dAyS fOLLOwIng pUbLICATIOn.

he SOVEL steel plant, which is part of the SIDENOR group, began production in 2001. It is located at a seacoast area of Almyros next to the city of Volos in the middle of Greece. The initial installation followed
In 2006, SOVEL decided to convert to a Consteel operation in order to increase production capacity and decrease electrical energy consumption. Installation experience and production results are presented in this paper.

the modern standards of the time, with the principle of attaining the maximum possible productivity of rebar steel under the tightest technical specifications. The steel plant, built by Danieli, consists of a meltshop and a rolling mill. The key meltshop production stations include a 100-tonne EAF with power supplied by a 100-MVA Tamini transformer, an LF station, and a 6-strand (five in operation) CCM, casting mostly billets of 140 x 140 mm2. Within a couple years, the annual production reached the level of 600,000 tonnes of good products, and it was decided to increase productivity. A major arc furnace revamping was carried out by increasing the shell dimensions in order to tap a maximum steel quantity of 130 tonnes. Figure 1 depicts, in schematic form, the new EAF design, intended to contain the required tapped weight of 130 tonnes liquid steel. Table 1 presents the main features of the EAF equipment. Apart from the installation of a new 120-MVA Tamini transformer, a major upgrade of the scrap buckets and ladles was also designed and implemented.

The improvement in productivity followed not only the meltshop revamping, but also the personnel learning trends. When SOVEL founded the steel plant in the area, it was understood that converting local farmers into specialized industrial workers would take some time. However, due to the improved quality of life that the plant offered to the workers, the job satisfaction and inspiration, together with respect for the growing SIDENOR group, they effectively succeeded in their goals. It was a real case in which a companys success resulted directly from the success of a team of workers. Within the next three years, in which continuous improvements were made, the annual production leveled off to values around 770,000 tonnes. These production results were attained during night and weekend operations (around 5,000 production hours per year), due mainly to domestic electric energy pricing. Then, a top-management decision was made that production should be increased even further. The selection of the Consteel process was not an overnight decision. From the primary meltshop installation, it was the Consteel process that was envisaged as the future key method of production. Even the tilting EAF platform was selected in such a way that the foreseen production method would fit, with minor changes. In addition to this, space in the scrap yard was the most necessary part for a successful implementation of the new system, and this was improvised by the team of the main technical office of the SIDENOR group, at the proper time.

ConsteelInstallation

The installation of the new system took place in four stages in order to minimize as much as

Authors
GeorgeBouganosopoulos(left),meltshop technical manager, SOVEL SA, Volos, reece (gogano sovel.vionet.gr); Vasilis Papantoniou, chief operating officer, SIDENOR SA, hessalonii, reece (vpapasienor.vionet.gr); an Panagiotis Sismanis (right), meltshops irector, SIDENOR SA, vpapasienor.vionet.gr); Athens, reece (psismanissienor.vionet.gr) 38 Iron&SteelTechnology

figure 1

Schematic (not etaile) esign for an EB EAF with a capacity of 130 tonnes liqi steel.

possible any unsteady meltshop production. In fact, the following stages were carried out: Construction of a new section in the scrap yard. This consisted of two cranes, each supplied with a 10-tonne-capacity double-magnet system for the loading of the 60-m-long x 2,400-mm-wide conveyor belt for feeding scrap into the furnace. In addition, a new duct for the removal of exit gases from the pre-heater, along with a new dust-settling chamber, were designed and built. This stage did not influence production at all. Continuation of the necessary construction inside the EAF building. This stage gave rise to only minor production delays. Construction of some necessary parts that created some daily production cutbacks. In the final 15 days, production was completely stopped in order to attach the connecting car that feeds scrap into the furnace. This included modifications in the lower part of the furnace shell that had been designed and executed in advance. Furthermore, a new EAF roof was placed with no fourth hole for the removal of exit gases, as the gases in a Consteel plant are removed from the scrap pre-heating section. It should be noted that the pre-heater was constructed to be partly water-cooled (about 8 m long) and partly refractory protected (about 32 m long). During this stage, strand No. 6 in the CCM was put into effect in order for the caster to be able to match the higher furnace productivity. In general terms, it took about one month of ceasing meltshop operations (from stopping production until hot commissioning), for the installation to be completed.

Table 1
features of the first-Time-Renovated EAf
3-phaseACEAFEBT Shell iameter Max primary Max seconary Electroe iameter Pitch-circle iameter hree oxyfel rners hree caron moles 120-MVAtransformer 7,000 mm 2.3 A, 30 V 63 A, 1,200 V 600 mm 1,250 mm 4 MW each 15 g/minte each

figure 2

he SOVEL plant in Almyros, reece. he photo was taen efore the final installation of the Consteel process.

February2009 39

Table 2
production data before and After Consteel Implementation
ap-tap time (min) Power-On time (min) Power-Off time (min) Proctivity (tph) Energy consmption (Wh/t) Oxygen consmption (Nm3/t) Natral gas consmption (Nm3/t) Caron (g/t) Yiel (%) FeO (%) Average power (MW) Electroe consmption (g/t) Refractory elta-roof life (heats) Mar point (t/h/MW) Before 56 41 15 138 430 32 4 10 87 37 81 1.77 300 1.68 After 48 40 8 160 395 32 0 18 88 23 77 1.75 200 2.14

ExperienceWiththeNewSystem

Start-up took place at the end of August 2007, and proved to be smoother than expected. A better working environment and a more predictable operation enabled personnel to adapt easily to the new facility. Over-voltages were reduced, primary voltage variations were minimized, energy consumption was decreased, arcing with the water-cooled roof was minimized, and productivity increased. Table 2 presents some of the production data that were improved. Scrap mixtures were selected mainly upon 50% shredded scrap, 45% HM 1 and 2, and HI up to 5%. 1 2, There is a length limit of about 2 m for the scrap; overweight pieces of scrap are avoided. HM 1 and 2 scrap is screened through a sieve (manufactured by LAEA Ingeniera y Servicios, Spain) that results in less gangue scrap. Generally, at start-up some problems are expected on the new technology applied. At SOVEL, this was also the case, but any incoming problem was tackled with relatively minor efforts. As far as the actual process is concerned, the slag basicity is kept at 2.22.3, with a specific consumption of 47 kg lime/tonne and 4 kg olivine/tonne fed through the roof fourth hole. Olivine is supplied by local mines as a main source of magnesia; it actually contains about 55% MgO and is cheaper than local caustic magnesia or imported dolomitic slag. Refractory lining thickness is 400 mm; most erosion is encountered at the slag door area due to the continuous efflux of slag. However,
40 Iron&SteelTechnology

at the scrap feed area, the refractory erosion measured was less than 70 mm after a campaign of 500 heats. Anthracite is injected through the modules in order to keep foaming slag conditions for most of the time. The reduction of FeO content in the slag, the better slag removal and the slag-free tapping have extended ladle life by 15 heats, bringing the average up to 105. Problems were encountered at the delta roof refractory due to the steady use of arc. There is some inclination toward testing monolithic refractory systems. Although the load cells that were supposed to measure the furnace weight, and hence the steel content at any time, malfunctioned within the first month after start-up, the furnace operators managed to control the scrap feeding schedule by adjusting the electric power with the scrap feedrate, as the scrap feedrate is correlated with the belt speed. Generally, the scrap feedrate is kept within 3.24.5 tonnes/minute at power levels around 6278 MW. Liquid heel is kept after each heat, and the target quantity is about 50 tonnes; it is indispensable for such a way of introducing scrap into the melting furnace. Finally, dust production was reduced, as expected, by a value close to 33%.

InsightsforImprovement

Since Consteel implementation, the team engaged in the project has continued to strive for improvement in energy consumption and productivity. It is known that slag conditioning and specifically foaming slag is of paramount importance in a Consteel practice. In their monumental work, Pretorius et al1 discussed the role of MgO into the foaming slag process. A remarkable achievement of this theory in the field of Consteel practice has been recently published2 with excellent overall results. At SOVEL, the tendency is to keep an attainable basicity of 2 = 2.22.3, or equivalently, 3 = 1.651.75. The target value for MgO in the slag is around 10%. In this aspect, the gangue part of the scrap is kept apart as much as possible, although scrap densities may be as low3 as 0.3 tonne/m3. Improvements in productivity, mark point, and specific energy consumption are highly related to the proper exploitation of chemical energy in the furnace and in the preheater. In 1995, Lahita4 started the discussion about the role of a proper system for the CO post-combustion inside the pre-heat conveyor. He claimed that unsuitable instrumentation and lack of previous practice were responsible for the inadequacy to burn off the excess CO efficiently inside the preheater. He also added that an optimized post-combustion system is necessary not only to further improve energy consumption, but

figure 3 also to reduce to the minimum the NOx and CO concentrations at the baghouse. It was as early as 1992 that Vallomy et al.5 discussed the important role of the energy liberated by the CO combustion in the pre-heating of scrap. They had presented plant results from Kyeoi Steel, Japan, and with the use of a thermodynamics tool, they had claimed that the proper use of oxygen and carbon in the furnace can create sufficient quantities of CO for post-combustion in the pre-heater in a controlled and not random manner. In 2001, Vallomy6 introduced the practical idea of charging the furnace with hot metal, as well. The remarkable notion was the relatively constant values of carbon in the bath, even during the hot metal introduction period. Manenti et al.78 strengthened the idea based on plant results of worldwide Consteel installations. They also commented on the necessary progressive and controlled post-combustion of CO in the pre-heating section, by the means of an automatically controlled injection of air. However, no CO will be burned off in the pre-heater if it is not there in sufficient quantities to react. The ultimate answer came in 2002, in an experimental work in the O.R.I. Martin Consteel plant. Di Donato et al.9 identified the various parameters by which sufficient quantities of CO are generated in the furnace and combusted in the pre-heater in order to pre-heat the fed scrap to a desired temperature. They did not neglect, however, that the exploitation of a part of the chemical energy of CO directly in the EAF can give rise to a higher overall benefit. Their findings were statistically treated by the present authors, and the results are summarized and depicted in Figure 3. Post-combustion ratio, or PCR, is a useful parameter10 for this kind of study, defined as:

Post-comstion ratio (PCR) as a fnction of specific oxygen an caron consmption in the EAF.

PCR =

CO2 CO + CO2
(Eq. 1)

PCR is a number between 0 and 1, and it is usually expressed as a percentage. It is realized that the larger the CO concentration, the lower the PCR value becomes. From the shape of the curves in Figure 3, it seems that there is an upper limit for the carbon charge in EAF, insofar as an acceptable value for PCR is obtained for specific oxygen consumption. Most likely, charge values for carbon, injected or in the scrap, above 23 kg/tonne may be a waste of material in order to attain the sufficient quantities of CO for post-combustion in the pre-heater. Although carbon oxidation phenomena require specific oxygen consumptions in the

range of 1020 Nm3/tonne, it is understood that some more oxygen will be required for the rest of the oxidation (i.e., oxidation of Si, P, Mn, etc., that provide extra heat into the EAF). A homemade thermodynamics tool, developed into an Excel spreadsheet and capable of performing mass and energy balance1113 computations for the EAF and the pre-heater, is used for decision making. In this work, it was used in order to deduce the effects of PCR and pre-heater oxygen upon the entry scrap temperature and energy savings, or specific energy consumption. Figure 4 depicts a screenshot with most of the data involved in a heat. Uncertainties related to scrap or liquid-steel chemical compositions from one heat to another are compensated for by trends and average values. Figures 58 present computed results for the heat data presented in Figure 4. Figure 9 illustrates the energy balances in the EAF and the scrap pre-heater, for a pre-heater efficiency of 70%. For a specific value of electrical energy consumption and pre-heater efficiency, the model computes the scrap entry temperature and matches the energy balances. Graphs from this kind of analysis with pre-heater efficiencies of 50%, 70% and 100% are presented in Figures 1012. It is interesting to note that, depending on the pre-heater efficiency, the scrap entry temperatures fall in the range of 200600C values that had been mentioned by Vallomy years ago. Furthermore, scrap pre-heat temperatures in the range of 400600C relating
February2009 41

figure 4

Inpt ata for a heat se for the mass an energy alance calclations.

figure 5

Reslts compte with respect to the heat ata presente in Figre 4.

42 Iron&SteelTechnology

figure 6

Mass alance reslts compte with respect to the heat ata presente in Figre 4.

figure 7

EAF energy alance reslts compte with respect to the heat ata presente in Figre 4.

February2009 43

figure 8

Pre-heater energy alance reslts compte with respect to the heat ata presente in Figre 4.

figure 9

Compte reslts in graphical form for the energy alances in the EAF an the scrap pre-heater, for a pre-heater efficiency of 70%.

44 Iron&SteelTechnology

to efficiencies in the range of 7080% and in agreement with present findings have already been published.8 In the calculations performed, the oxygen used in the pre-heater was considered to be from air drafted inside the pre-heating section by appropriate flaps that exist in specific locations at the pre-heater. As an example, for the case of the 50% preheater efficiency, a decrease in the PCR value by 7.4%, that is from 54% (at CO = 22.92%, CO2 = 27.05%) to 50% (at CO = 26.16%, CO2 = 25.93%) would shift the data presented in Figure 10 to the following new results: airdrafted oxygen from 10.8 to 12.0 Nm3/tonne (line 1), entry scrap temperature from 295 to 305C (line 2), energy savings from 45 to 48 kWh/tonne (line 3), and specific electric energy consumption from 394 to 391 kWh/ tonne (line 4). In a recent publication,14 Ferri et al. mentioned that, in the O.R.I. Martin Consteel installation, the CO post-combustion was actually performed by the injection of a controlled quantity of air into the pre-heater. They also considered it advantageous to add carbon lumps together with scrap in the conveyor.

figure 10

Conclusions

Compte vales for the specific energy consmption an the scrap entry temperatre with respect to PCR an oxygen se in the preheater, for a 50% pre-heater efficiency.

For SOVEL, the road, albeit on the right track, will still be long. Although the proper instrumentation for the air injection in the pre-heater has not yet been installed, the various parameters that influence efficient furnace operation have been identified. Scrap entry temperatures are on the order of 320C, and substantial energy savings have already been obtained. Everyday plant data verify the selected improvement guidelines. References
1. E.B. Pretoris an R.C. Carlisle, Foamy Slag Fnamentals an heir Practical Application to Electric Frnace Steelmaing, 56th Electric Furnace Conference Proceedings, New Orleans, 1998, pp. 275292. 2. H.M. Ajei-Sarpong, M. Fox, B. rmle an J. Powers, Optimization of Ncor SteelHertfor Contys Consteel an DC EAF Operations, Iron & Steel Technology, Vol. 3, No. 2, 2006, pp. 6574. 3. J.C. Simmons, K.J. Shoop an J.M. McClellan, FASEELM: he Hot Metal of FASMEL an the Continos Scrap Feeing of CONSEEL, Electric Furnace Conference Proceedings, San Antonio, 2002, pp. 231240. 4. J.A. Lahita, he CONSEELM Process in Operation at New Jersey Steel Corporation, METEC, Paris, 1995, pp. 524534. 5. J.A. Vallomy, . Fse an S. Naamra, Sccess of CONSEELM Process in USA Leas to Start-Up of 120-mt/hr Unit in Japan, Electric Furnace Conference Proceedings, Atlanta, 1992, pp. 309313.

figure 11

Compte vales for the specific energy consmption an the scrap entry temperatre with respect to PCR an oxygen se in the preheater, for a 70% pre-heater efficiency.

February2009 45

figure 12

Compte vales for the specific energy consmption an the scrap entry temperatre with respect to PCR an oxygen se in the preheater, for a 100% pre-heater efficiency.

6. J.A. Vallomy, Continos Charging of Hot Metal an Preheate Scrap in the EAF, Electric Furnace Conference Proceedings, Phoenix, 2001, pp. 787 795. 7. A.A. Manenti, EAF Meltshop with Consteel an Hot Metal Charging, Electric Furnace Conference Proceedings, San Antonio, 2002, pp. 251260. 8. J. Simmons, A. Manenti an K. Shoop, CONSEEL to FASEEL: he Ftre of Steelmaing, Electric Furnace Conference Proceedings, San Antonio, 2002, pp. 377385. 9. A. Di Donato, V. Volponi, U. De Mirana an P. Argenta, Development of Flexile Operating Practices to Proce Steel With Consteel-EAF Process in ORI Martin Plant, METEC, Venice, 2002. 10. R.J. Frehan, The Making, Shaping and Treating of Steel, AIS, 1998, pp. 605615. 11. H.A. Fine an .H. eiger, Handbook on Material and Energy Balance Calculations in Metallurgical Processes, he Metallrgical Society of AIME, 1979. 12. C.H.P. Lpis, Chemical Thermodynamics of Materials, Elsevier Science Plishing Co., 1983. 13. C.E. Sims, Electric Furnace Steelmaking Volume II: Theory and Fundamentals, Ch.19, Energy Balance in the Arc Frnace, Interscience Plishers, New Yor, 1963, pp. 283296. 14. M.B. Ferri, . Cozzi an E. Lomari, Energy an raphite Electroes Saving in ORI Martin, Brescia, Italy, ISSTech 2003 Conference Proceedings, pp. 11011110.

This paper was presented at AISTech 2008 The Iron & Steel Technology Conference and Exposition, Pittsburgh, Pa., and published in the Conference Proceedings.

Di yo fin this article to e of significant relevance to the avancement of steel technology? If so, please consier nominating it for the AIS Hnt-Kelly Otstaning Paper Awar at www.aist.org/hntelly.

46 Iron&SteelTechnology