Plasma liquid-phase reduction of iron from its oxides using gaseous reducers.
Deficiency of coke used in the blast furnace stipulated the need of developing a new technology with application of a less deficient reducer, in which byproducts of the production would be harmless for the environment.
Application of coal as a reducer causes contamination of metal with sulfur and phosphorus. That's why for direct production of high-purity metal it is advisable to use gaseous reducers, which do not contain these harmful impurities.
Increased degree of iron reduction and rate of the reaction proceeding enables temperature increase, which is the most efficiently ensured, when plasma heat sources are used [1, 2].
So, we investigated process of plasma liquid-phase reduction of iron from oxides with application of gaseous reducers.
In the monograph  system Fe--C--H--O is considered. One of important characteristics of complex gaseous mixtures is their theoretical work capacity in relation to oxygen (mass of oxygen absorbed or released by 1 [m.sup.3] of initial system in its transition into equilibrium state with condensed phases, containing oxygen), i.e. function of difference between equilibrium and actual degrees of oxidation, where degree of the latter is a ratio of the number of oxygen moles in a complex gas mixture to the number of oxygen moles in it in case of complete oxidation of all its components up to [H.sub.2]O and C[O.sub.2].
In practice hydrocarbons prior to their application as a reducer are subjected to the conversion: steam-and-water, oxygen, carbon dioxide or a combined one . However, in this case actual degree of oxidation of the gas increases and reduces its theoretical work capacity. That's why it is advisable at the first stage of investigations to determine reduction potential of the carbon black-hydrogen mixture, produced by pyrolysis of hydrocarbons in plasma.
According to , carbon black-hydrogen mixture is the heterophase one: in hydrogen volume pure finely dispersed carbon, not containing ash and sulfur and having size of particles about 1-[10.sup.-6] m, is uniformly distributed. Black carbon has high reaction capacity due to big specific surface (about 1-[10.sup.-5] [m.sup.2] per 1 kg of carbon). Gas part of the mixture is highly pure and active. Carbon is well wetted by the melt of iron oxides that ensures its good assimilation and high efficiency of use. This peculiarity of black carbon-hydrogen mixture imparts it with such properties, which do not have conventional gaseous reducers.
Decrease of concentration of wustite FeO in the melt in the course of the reduction process causes decrease of iron monoxide activity and, as a result, complication of the metal reduction from it and increase of the reducer consumption . So, in application of carbon monoxide for reduction of wustite at temperature 1600 [degrees]C not more than 16 % CO transform into carbon dioxide, while at FeO activity of 0.5--not more than 6 % (Figure 1) . Application of hydrogen under the same conditions gives 51 and 26 % at FeO activity 1.0 and 0.5, respectively.
Content of silica Si[O.sub.2] in the melt significantly influences activity of wustite. So, at silicon dioxide concentration 12 % not more than 8 % CO is used for reduction . The reason of this is binding of iron oxide by silicon dioxide with formation of iron meta(FeO-Si[O.sub.2]) or orthosilicate (2FeO-Si[O.sub.2]) . According to , reduction of iron from such melts can not be full. Depending upon content of Si[O.sub.2], maximum degree of reduction constitutes 86--97 %. As concentration of wustite decreases, lamination of the melt is possible with precipitation of solid silica Si[O.sub.2] at the temperature 1600--1700 [degrees]C (Figure 2) . It is advisable to improve reduction of iron by means of the slag basicity increase. Iron silicates may be destroyed in case of addition into the slag of calcium oxide, which binds Si[O.sub.2] into calcium meta- or orthosilicates (CaO-Si[O.sub.2] or 2CaO-Si[O.sub.2]) and thus increases activity of FeO . Maximum activity coefficient iron oxides achieve at slag basicity of 1.8--2.0, although at further increase of the latter and concentration of FeO about 100 % activity coefficient is close to one, i.e. the melt gets under Raoult's law.
[FIGURE 1 OMITTED]
Temperature exerts significant influence on rate and completeness of the reduction reaction proceeding. In Figure 3 [3, 8] dependence of the degree of reduction of a solid particle or a drop of the melt of wustite upon time at different temperatures is shown.
In reduction of iron by solid carbon from molten oxides, rate and completeness of reduction are positively influenced by temperature increase in the melting space, reduction of CO and C[O.sub.2] pressure, and mixing of the melt [3, 8].
In production of cast iron in the blast furnace distribution of sulfur between the slag and the metal is affected, first of all, by basicity of the slag: as basicity increases less sulfur transits into the metal. However, increase of basicity enables increase of the slag amount and consumption of coke . Optimum basicity of the blast furnace slag is considered 1.0--1.3 . Phosphorus transits in blast furnace melting almost completely from the charge into cast iron .
The investigations were carried out on developed in the E.O. Paton EWI of NASU pilot installation (Figure 4).
The installation was powered by single-phase alternative current from transformer OP 108 through regulator A1474. Current was adjusted within 500-2200 A, open-circuit voltage was 380 V, arc voltage drop was 30--180 V, maximum power was 400 kW.
Before beginning of the process, a hollow electrode was short-circuited on the metal ingot with known mass and chemical composition in the lower part of the mould. Into space between the electrode of 80 [micro]m diameter and the mould wall with internal diameter 200 [micro]m iron ore pellets of the following composition were poured, wt.%: 9.18 Si[O.sub.2]; 0.32 [Al.sub.2][O.sub.3]; 0.87 CaO; 0.37 MgO; 0.039 S; 0.007 P; 0.062 C; [Fe.sub.2][O.sub.3]--the base (Figure 5).
[FIGURE 2 OMITTED]
Melting chamber was blown by argon, which was supplied through the electrode cavity. After switching of the power source, electric arc was excited by lifting the electrode above the ingot. As iron ore feedstock melted, the electrode was lifted above surface of the melt with preservation of distance from end of the electrode to surface of the melt.
After complete melting of the feedstock hydrocarbons were fed into the cavity and supply of argon was stopped. In plasma of the arc in reduction zone pyrolysis of hydrocarbons took place. Produced hydrogen and carbon interacted with oxides and reduced the metal, which mixed with molten layer of the ingot and accumulated above it. Formed water vapor, carbon mono- and dioxide, and unreacted hydrogen and pyrocarbon were withdrawn from the melting space.
After termination of the process the melting chamber was blown by argon, which was supplied through the electrode cavity. In the course of plasma liquid-phase reduction of iron from oxides, flow of supplied through the electrode cavity reduction gas, length of the arc, and electric parameters of the arc burning were controlled. Then material balance was calculated and chemical analysis of produced metal and slag was made.
A series of experiments was carried out for determining influence of composition and flow of the gaseous reducer on rate and completeness of proceeding of plasma liquid-phase reduction reaction.
[FIGURE 3 OMITTED]
Comparison of experimental data in use of methane C[H.sub.4] and propane-butane mixture [C.sub.3][H.sub.8]--C[H.sub.4][H.sub.10] (6 % of butane) proves that ratio of the amount of metal, produced with application of each of the reducers under the same conditions, as a whole is close to the ratio of their theoretical work capacities in relation to oxygen. So, one may draw conclusion that results of the experiments with application of one kind of a reducer may be transferred on other reducers, taking into account their theoretical work capacities in relation to oxygen and ratio of content in them of carbon and hydrogen.
[FIGURE 4 OMITTED]
A series of experiments was carried out, in which at the same flow of the propane-butane mixture (35 and 15 l/min) and basicity (0.1) duration of the gaseous reducer supply and duration of reduction varied.
Dependences of the degree of iron reduction upon duration of gas supply and its volume are presented in Figure 6.
Presented dependences make it possible to track thermodynamics and kinetics of the process. During the first minutes transition [Fe.sub.2][O.sub.3][right arrow][Fe.sub.3][O.sub.4][right arrow]FeO takes place, which is proved by absence of a renewed metal. Further amount of produced iron quickly in creases, but then reduction process decelerates, which may be caused by reduction of wustite activity.
[FIGURE 5 OMITTED]
It should be noted that at the beginning of the process reduction rate insignificantly depends upon flow of the reducer, i.e. 35 l/min flow is excessive. But at a lower flow the process decelerates earlier because of decrease of iron oxide activity and increased need in the reducer.
So, for full recovery of the metal, saving of the reducer, and acceleration of the process it is necessary to increase flow of gaseous reducer in the course of the process.
For increasing activity of iron oxide, basicity of the slag melt was increased by addition of CaO. Basicity of the melt was determined by formula
B = (CaO) + 1.5(MgO)/ (Si[O.sub.2]) + 0.6([Al.sub.2][O.sub.3]V
where (CaO), (MgO), (Si[O.sub.2]), and ([Al.sub.2][O.sub.3]) are the contents of respective oxides in the slag, wt.%.
In Figure 7 dependences of the metal reduction degree and content of iron oxide in the slag upon basicity in case of flow of propane-butane mixture 35 l/min and duration of its supply 14 min are given.
As it follows from the data of Figure 7, increase of basicity from 0.1 to 1.2 enables completeness of iron reduction.
Concentration of carbon in the metal is low--approximately 0.02 wt.%. Content of sulfur in the produced metal is 0.007 %, which is 5.5 times lower than in the charge, and does not depend upon basicity within the range 0.1--1.2 because of insufficiently high basicity or impossibility of producing metal with lower content of sulfur according to the proposed method. Contents of phosphorus in the metal and in the charge do not differ (about 0.007 wt.%) and do not depend upon basicity within mentioned range. Iron is not contaminated by impurities from the reducer, their content is rather insignificant and meets requirements established for majority of steels.
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
Content of nitrogen in the produced metal is about 0.0075, hydrogen--0.0005 wt.%, which corresponds to their usual content in steel; content of oxygen is 0.075 wt.%, like in a rimming steel.
In determination of the reducer use degree actual level of oxidation of the reaction products was taken into account. In Figure 8 dependence of completeness of use upon duration of gas supply at B = 0.1 is shown.
Presented dependence allows tracking kinetics of the process and confirms analysis data of Figure 6. Use of the gaseous reducer approaches 30 % at the flow of 15 l/min, which is close to its use in solid-phase reduction. Optimization of the process should enable more complete use of the reducer.
1. It is established that process of plasma liquid-phase iron reduction with application of gaseous reducers ensures production of a high-purity metal.
2. Main regularities of influence of flow and composition of gaseous reducer and basicity of slag on degree of iron reduction and use of the reducer are determined.
[1.] Dembovsky, V. (1981) Plasma metallurgy. Moscow: Metal lurgiya.
[2.] Lakomsky, V.I. (1974) Plasma-arc remelting. Ed. by B.E. Paton. Kiev: Tekhnika.
[3.] Bondarenko, B.I., Shapovalov, V.A., Garmash, N.I. (2003) Theory and technology of coke-free metallurgy. Kiev: Naukova Dumka.
[4.] Kartavtsev, S.V. (2000) Natural gas in reduction melting: Monography. Magnitogorsk: MGTU.
[5.] Kozhevnikov, I.Yu. (1970) Coke-free metallurgy of iron. Moscow: Metallurgiya.
[6.] Ajzatulov, R.S., Kharlashin, P.S., Protopopov, E.V. et al. (2002) Theoretical principles of steel-making processes: Tutorial for institutes of higher education. Moscow: MISiS.
[7.] (1965) Thermodynamical properties of inorganic materials: Refer. Book. Moscow: Atomizdat.
[8.] Yusfin, Yu.S., Gimmelfarb, A.A., Pashkov, N.F. (1994) New processes of metal producing (iron metallurgy): Manual for institutes of higher education. Moscow: Metallurgiya.
[9.] Tovarovsky, I.G., Lyalyuk, V.P. (2001) Evolution of blast furnace melting: Monography. Dnepropetrovsk: Porogi.
[10.] Kurunov, I.F., Savchuk, N.A. (2002) State-of-the-art and prospects of blast-furnace-free metallurgy of iron. Moscow: Chermetinformatsiya.
M.L. ZHADKEVICH, V.A. SHAPOVALOV, G.A. MELNIK, D.M. ZHIROV, A.A. ZHDANOVSKY,
K.A. TSYKULENKO and M.S. PRIKHODKO
E.O. Paton Electric Welding Institute, NASU, Kiev, Ukraine
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||PLASMA-ARC TECHNOLOGY|
|Author:||Zhadkevich, M.L.; Shapovalov, V.A.; Melnik, G.A.; Zhirov, D.M.; Zhdanovsky, A.A.; Tsykulenko, K.A.;|
|Publication:||Advances in Electrometallurgy|
|Date:||Apr 1, 2006|
|Previous Article:||Composite materials on base of copper and molybdenum, condensed from vapor phase, for electric contacts. Structure, properties, technology: Part 2....|
|Next Article:||Effect of allyying on composition, structure and properties of powders of alloy AlCuFe, containing quasi-crystalline PHASE.|