Reduction kinetics of carbon containing pellets made from metallurgical dust

Abstract

Reduction experiments of carbon containing pellets made from metallurgical dust were conducted under a weak oxidising atmosphere in the temperature range of 1348–1573 K. Analysis of kinetics and the reduction mechanism revealed that the rate determining step of the reduction of the pellets is the interfacial or local reaction with the activation energy 111·66 kJ mol⁻¹. The reduction rate can be expressed by the McKewan equation 1−(1−R)^1/3 = kt. In addition, temperature is an important factor influencing the reaction rate as dezincification and metallisation increase with the increased temperature. The amount of dezincification and metallisation could be up to 97·8 and 79·9% respectively at 1573 K compared to a minimum of 75·3 and 60·2% at 1348 K.

Introduction

With the rapid development of the steel industry in China, treatment of metallurgical dust from iron and steel plants as a secondary resource is of increasing interest. Treatments for metallurgical dust include sintering and direct reduction.1 As dust contains a high content of harmful elements such as zinc, potassium and sodium as fine particles, this adversely affects the sintering process. Of the direct reduction processes, the rotary hearth furnace (RHF) has been well developed in recent years and has been applied in some Chinese companies such as Ma Steel and Rizhao Steel. The process is shown in Fig. 1. The advantages are the full utilisation of carbon and effective recycling of zinc while removing potassium, sodium and other harmful elements.

Figure 1.

Rotary hearth furnace process

Reduction studies of carbon containing pellets by previous researchers indicated that the process is by indirect reduction, with the reduction rate being controlled by the interfacial reaction or local reaction. Temperature and the flow speed of the non-reducing gas have a significant influence on the reduction rate of iron oxide in the pellets. The reduction rate of zinc oxide is mainly controlled by the diffusion of the gas phase, and the atmosphere has little influence on the volatilising rate of zinc and lead in the pellets.

There are several kinds of dust with a high content of carbon in steel plants, such as BF gas mud (dust collected from the BF off-gas wet Venturi dust trap) and the dust collected from mine and coke sieves (dust from the dust collecting fans on the top of mine and coke sieves), which can be directly reduced by the carbon contained in the dust without adding additional carbon. In previous studies, attention was mainly paid to the metallic elements in the dust and the use of additional coal or coke powders. There are limited studies on the reactions of carbon in the dust, and no studies have been reported about the reduction mechanism of carbon containing pellets with the dust as the only raw materials. In addition, the heating source of the reduction reaction of carbon containing pellets is mostly from the combustion of high temperature gas, resulting in a weak oxidising atmosphere after combustion. Inert atmosphere conditions were mainly studied in the previous research.2 ^– 6

In this paper, the gas composition for the experiments was selected according to the industrial air/fuel ratio for gas combustion in an RHF. Investigations on the reduction kinetics of the carbon containing dust pellets were carried out under a weak oxidising atmosphere. By analysing the weight reduction and the degree of reduction of the pellets, the possible rate determining steps (RDSs), including gasification of carbon, diffusion of the gas phase, interfacial reactions and local reactions, etc., were investigated in order to determine the reduction mechanism of carbon containing dust pellets in the weak oxidising atmosphere.

Material properties

The raw materials for the experiments were provided by Ma Steel. They are BF gas mud, BF dust 1 (collected from mine and coke sieves), sintered dust, BF gravity dust, BOF mud and BF dust 2 (BF cast house dust). Their chemical compositions are shown in Table 1. It is shown that the contents of iron (TFe) of BF dust 2, sinter dust and BOF off-gas dust are relatively high, and the contents of carbon (C) of BF gas mud and BF dust 1 are high. By adjusting the ratio of raw materials with a high content of iron and carbon respectively, an appropriate ratio of the content of iron and carbon in the pellets can be obtained according to the requirements for raw materials of RHF to achieve a complete reduction of iron oxides with high valence in the pellets.

Table 1.

Chemical composition of raw materials (wt-%)

Raw materials	TFe	SiO₂	CaO	MgO	Al₂O₃	Zn	S	C
BF gas mud	37·98	5·35	3·17	0·86	2·28	4·71	0·70	23·58
BF dust 1*	38·45	5·57	6·02	1·11	2·28	1·20	0·44	19·22
Sinter dust	50·27	5·72	13·00	3·12	2·09	0·05	0·12	1·42
Gravity dust†	46·49	10·82	10·46	1·79	6·34	1·28	0·51	7·57
BOF dust	56·40	1·57	13·19	0·86	0·30	0·80	0·14	2·47
BF dust 2‡	64·99	1·51	0·61	0·08	0·44	0·65	0·52	2·16

*Collected from mine and coke sieves.

†BF casthouse dust.

‡Gravity dust comes from BF off-gas collected by gravity dust separator.

The pellets were prepared by mixing the selected raw materials in predetermined amounts, drying and artificially pelletising with additional water. The pelletiser is of disc type with a diameter of 1000 mm, a side height of 200 mm, a declining angle of ∼45° and a rotation rate of 17–18 rev min⁻¹. The water content in the green pellets was ∼8%. The total time for pelletising was 12 min including 2 min for the formation of mother pellets, 8 min for growing them and 2 min for compression. The diameter of the pellets was 12–15 mm, and the chemical composition is shown in Table 2.

Table 2.

Chemical composition of green carbon containing pellets (wt-%)

Components	TFe	SiO₂	CaO	MgO	Al₂O₃	Zn	S	C
Content	49·10	4·50	7·15	1·23	1·88	1·81	0·42	11·64

Experimental

The experimental device is shown in Fig. 2. A SiC resistance furnace was used for heating with a maximum working temperature of 1673±2 K. The accuracy of the TGA electronic balance is 0·1 mg. Data were collected every 6 s by computer. The reactor is made up of two parts. The gas enters the reactor via the external part, and then after being preheated, the gas flows into the reactor from the bottom of the inner part.

Figure 2.

Schematic of experimental apparatus

The normal working temperature of the RHF is 1473 K; therefore, the chosen temperatures in the experiments are 1348, 1423, 1498 and 1573 K. According to the theoretical calculation of the air/fuel ratio, the volume ratio of N₂ and CO₂ is ∼93/7, and their flowrates are 14·0 and 1·0 L min⁻¹ respectively.

After the temperature in the reactor reached the preset value, the mixed gas of N₂ and CO₂ was introduced, and the temperature was kept constant for 30 min. Then, the basket was hung under the electronic balance and the balance reset. Two pellets were put into the basket and moved into the reactor, and the data collection procedure started. When the weight of the pellets stabilised, it was considered that the reduction reactions were completed and the test was stopped.

The pellets were removed, and the contents of the metal iron (MFe), total iron and zinc were determined by chemical analysis. The phase composition of the pellets was analysed by scanning electron microscope (JSM-6510 A) and X-ray energy dispersive spectrometer analyser (JSM-6510 LA).

Results

Reduction

During the reduction of carbon containing dust pellets, reduction of zinc oxide also occurred; thus, zinc could be removed from the dust as it is easy for the produced zinc to be gasified. The main composition of reduced carbon containing dust pellets was analysed by chemical analysis methods. The degrees of metallisation (R _m) and dezincification (R _Zn) of the pellets were calculated according to (1) (2) where MFe is the content of metal Fe (%) of the carbon containing pellets after reaction, TFe is the total amount of Fe (%) of the pellets after reaction, m _Zn is the total mass (g) of zinc after reaction and M _Zn is the total mass (g) before reaction. The calculated results are shown in Table 3. It can be inferred that based on an appropriate ratio of dust with high contents of carbon and TFe respectively, the carbon containing pellets can be self-reduced without adding extra carbon. The metallisation and dezincification reached the highest values of 79·9 and 97·8% at 1573 K respectively, while they decreased to the lowest values of only 60·2 and 75·3% at 1384 K, which indicates that temperature is an important factor for metallisation and dezincification. With the increase in temperature, the metallisation and dezincification degrees increase.

Table 3.

Metallisation (R _m) and dezincification (R _Zn) of pellets after reduction (%)

Temp/K	Time/s	Total Fe (TFe)	Metallic Fe (MFe)	Zn	R _m	R _Zn
1348	1500	65·9	39·7	0·25	60·2	75·3
1423	1000	68·3	49·0	0·083	72·0	91·8
1498	875	69·2	51·0	0·027	73·6	97·3
1573	600	69·1	55·2	0·022	79·9	97·8

Phase analysis

After the reaction, the pellets were mounted in resin and polished to expose the circular section across the centre of the sphere and analysed by scanning electron microscope (SEM) and X-ray for the energy spectrum analysis. As shown in Fig. 3, at 1348 K, there are many coral reef-like holes in the microstructure of the pellets, which are caused by the volatilisation of volatile components and the gasification of carbon. Owing to the low temperature, metal Fe cannot be observed, but it has actually formed according to the chemical composition shown in Table 3. The amount of voids is much less at 1423 K than at 1348 K, and metallic Fe has appeared but is not easy to observe. At 1498 and 1573 K, white, off white and greyish black phases appeared. Six points were selected from the figures of 1498 and 1573 K for the energy spectrum analysis, and the results are shown in Fig. 4. It can be inferred that the white area in Fig. 3 is the metallic Fe. The greyish black area has a complicated composition that is the slag phase. The off white area is mainly composed of elements Fe, O and low level Mg. Therefore, this off white area is mainly wüstite. As the temperature increases, FeO is reduced to metallic Fe, which enters the white area. Therefore, the off white area is much smaller at 1573 K than at 1498 K, and the white and greyish black areas are larger at 1573 K than at 1498 K.

Figure 3.

Micrographs of pellets after reaction

Figure 4.

Energy spectra of six positions from Fig. 3c and d

It is concluded that temperature is an important factor for the growth of metallic Fe in the pellets. It is very hard to observe Fe with the electron microscope at 1348 and 1423 K; however, at 1498 K, large areas of sheet or strip shaped crystals of Fe began to appear. At 1573 K, the areas continued to grow to form a large area. In addition, no Zn was discovered by the energy spectrum analysis, which is consistent with the results shown in Table 3, indicating that Zn has almost been removed at 1498 and 1573 K.

Weight loss rate analysis

Figure 5 shows the weight loss and weight loss rate of pellets with temperature. Weight loss was rapid at first and then declined as the reaction proceeded. It can be inferred from Fig. 5 that the reaction times of the pellets are approximately 600, 875, 1000 and 1500 s at 1573, 1498, 1423 and 1348 K respectively, which indicates that temperature has a key influence on the reaction time. Figure 5a is different from the others. After 1500 s, the weight loss rate was almost stable, caused by the low temperature and the slow rate of reaction. At 1500 s, the reduction and reoxidation of the pellets were almost in equilibrium, and the weight loss is mainly due to the volatilisation of volatiles, so the reaction time is ∼1500 s at 1348 K.

Figure 5.

Weight loss of pellets during reduction

The main reaction in carbon containing pellets at high temperatures is the reduction reaction, and the reacted fraction R is defined as follows (3) where m is the mass (g) of pellets before the reaction, m _t is the instantaneous mass (g) of pellets at time t, m _o is the mass of oxygen in the iron and zinc oxides in the pellets and m _c is the carbon content (g) of the pellets. The relationship between the reacted fraction R and time t can be obtained by calculation according to this formula with data from the experiment and is shown in Fig. 6. With the increase in temperature, the changing rate of the reaction fraction increased. The time for R to reach its maximum value was the shortest and longest at 1573 and 1348 K with reaction times of around 600 and 1500 s respectively. This shows that temperature is an important factor for the rate of reaction. When the temperature is higher, the reaction is faster.

Figure 6.

Relationship between time and temperature during reduction

Dynamics analysis

Basic principles

The main reactions in the carbon containing pellets under the weak oxidising atmosphere are as follows (4) (5) (6) (7) Reaction (4) demonstrates the indirect reduction of iron oxide. At the same time, the oxidation of iron oxide (i.e. reoxidation) by CO₂, the main component in the weak oxidising atmosphere, also occurs according to reaction (5). In addition, the reduction of zinc oxide (reaction (6)) and the gasification of carbon (reaction (7)) also take place. Du and Du7 studied the steps of reduction of pellets under an oxidising atmosphere. During the early period of the reaction, the main reaction is the self-reduction, that is, the rate of reaction (4) is larger than that of reaction (5); during the middle period, the rate of reduction and reoxidation of pellets is almost the same, that is, the rate of reaction (4) is equal to that of reaction (5); during the late period, the reoxidation of pellets is the primary process, that is, the rate of reaction (4) is smaller than that of reaction (5).

Assumptions on the reduction process of carbon containing dust pellets in a weak oxidising atmosphere are as follows. First, the effect of the diffusion of volatiles in carbon containing dust pellets on the reduction process is ignored; second, the effect of reduction reaction of iron and zinc oxides on the whole reduction process is considered, and the other reactions are ignored; and third, there is a uniform distribution of particles of chemicals inside the carbon containing pellets.

The reduction process of carbon containing dust pellets is a gas–solid reaction and a two-way gas–solid reaction, that is, the simultaneous occurrence of the internal gas–solid reaction and the gas–solid reaction between pellet and external gas.8, 9 The main reaction process includes:10 ^– 13

the external gas diffuses through the boundary layer of gas–solid to the particle surface

ii.

the gas diffuses through the porous solid product layer of iron to the boundary layer of Fe_xO_y–ZnO

iii.

in the reaction interface, CO gas reacts with Fe_xO_y and ZnO, and CO₂ is generated

iv.

the reoxidation of Fe_xO_y−1 by CO₂ occurs

CO₂ reacts to the solid carbon with the production of CO

vi.

CO₂ goes through the porous solid product layer of iron to the surface

vii.

CO₂ diffuses through the gas–solid boundary layer to the gas phase.

The corresponding process is shown in Fig. 7.

Figure 7.

Schematic of reduction mechanism of carbon containing pellets in weak oxidising atmosphere

Based on the above analysis, it is suggested that the possible RDSs during the reduction process of carbon containing dust pellets include:

gasification of carbon

ii.

diffusion of gas between the solid particles

iii.

the interfacial or local reaction.

Analysis of dynamics under possible RDSs

The experimental data were analysed by different models with the assumed RDS.14 If the gasification of carbon is the RDS, the rate of reduction reaction can be expressed following the rate of gasification of carbon as follows (8) where n is the coefficient with its value close to 1·0, and k is the rate constant of the reduction reaction of carbon containing pellets. Guo et al.4 studied the gasification of carbon in the pure carbon containing pellets of zinc oxide and obtained a value of 0·67 for n according to the molar mass of zinc and CO. In this study, the reduction of iron oxide is the dominant process due to the low content of zinc, and CO is also dominated in the gas phase with the assumption of n = 0·95. According to the above equation, the data from experiments at 1348–1573 K were analysed, with the results shown in Fig. 8. It shows a good linear relationship between −ln (1−nR) and t indicating that carbon gasification could be the RDS.

Figure 8.

Kinetic analysis with carbon gasification as RDS

If the reduction of carbon containing dust pellets is controlled by the interfacial reaction or local reaction, the rate of reduction can be expressed by the corresponding McKewan equation (9) The experimental data at 1348–1573 K were analysed according to the above equation, with the results shown in Fig. 9. This also shows a linear relationship between 1−(1−R)^1/3 and t before the end of the reaction. Therefore, the interfacial reaction or local reaction could also be the RDS.

Figure 9.

Kinetic analysis with interfacial reaction or local reaction as RDS

If the reduction of zinc oxide in the carbon containing pellets is considered to be controlled by the diffusion of gas phase and the diffusion is subject to Fick’s law, the rate of reduction can be expressed following the diffusion controlling equations of Jande and Ginstling-Brounshtein (10) (11) The experimental data at 1349–1573 K were analysed according to the above equations, with the results shown in Fig. 10. It does not show a good linear relationship in both cases because it is easy for the volatiles in the pellets to volatilise at high temperature, resulting in the formation of large holes, beneficial to the diffusion of gas. In addition, at high temperature, it is also easy for metallic Fe to form, resulting in an increase of the distance between iron atoms, in favour of the diffusion of gas. Therefore, the diffusion of the gas phase could not be the RDS.

Figure 10.

Kinetic analysis with gas diffusion as RDS

It is concluded that the RDSs for the reduction reaction of carbon containing pellets are the gasification of carbon and the interfacial or local reaction.

Calculation of reaction activation energy of carbon containing pellets

Given the rate constant k for the reduction reaction at different temperatures, the activation energy is calculated according to the Arrhenius equation as follows (12) (13) where E is the activation energy (kJ mol⁻¹), A is the frequency factor and T is the temperature (K). The value of k at different temperatures can be obtained according to the slope of lines, as shown in Figs. 8 and 9. The relationship between −ln k and T ⁻¹ is shown in Fig. 11. E is determined according to the slope of the line, with the results shown in Table 4. When the gasification of carbon is the RDS, the activation energy of the reduction of pellets is 87·52 kJ mol⁻¹, and when the interfacial or local reaction is the RDS, the activation energy is 111·66 kJ mol⁻¹. The former value is lower than the latter one, indicating that the gasification of carbon has less influence on the reduction of pellets, because the occurrence of the produced CO and CO₂ as well as the original CO₂ from the weak oxidising atmosphere could speed up the gasification of carbon. In addition, there is CO and CO₂ at the surface of the iron oxide particles, and both reduction (reaction (4)) and reoxidation (reaction (5)) of iron oxides take place in the interfaces, which make the interfacial reaction hard to continue. Ding et al.3 have measured that the activation energy of metallurgical wet dust (mud) collected from the ironmaking and steelmaking processes of pellets from a converter in an inert atmosphere is 77·48 kJ mol⁻¹, which is lower than the activation energy in this research. It can be concluded that the reoxidation of iron oxides in a weak oxidising atmosphere opposes the interfacial reaction of pellets. The zinc content of the mud used in the experiment is higher than that of converter mud. Wang et al.15 reported that under the coexistence of zinc, lead and iron, the RDS for zinc volatilisation is the reduction of ZnO by CO with the activation energy of 79·42 kJ mol⁻¹. This is lower than the activation energy of the interfacial reaction in this research, which shows that the volatilisation of zinc does not have much influence on the reduction reaction of iron oxides in the dust pellets.

Figure 11.

Relationship between ln k and T ⁻¹ under different rate determining steps

Table 4.

Activation energy of carbon containing dust pellets

Reaction rate equation	Regression equation of ln k and T ⁻¹	Activation energy E/kJ mol⁻¹
1−(1−R)^1/3 = kt	ln k = −13 431·1T ⁻¹+3·0686	111·66
ln (1−R) = −kt	ln k = −10 526·6T ⁻¹−0·25171	87·52

Conclusions

Carbon containing pellets can be self-reduced without adding extra carbon based on an appropriate ratio of dust and high contents of carbon and TFe.

Temperature is an important factor for the growth of metallic Fe in the carbon containing pellets. It is very difficult to observe metallic Fe with an electron microscope at 1348 and 1423 K, but at 1498 K, large areas of sheet or strip shaped metallic Fe began to appear. At 1573 K, the areas continued to grow to form a large area.

Metallisation and dezincification increase with the increase in the temperature and reached a minimum of 79·9 and 97·8% at 1573 K respectively.

In the weak oxidising atmosphere, the activation energy of the interfacial or local reactions is 111·66 kJ mol⁻¹, while the activation energy of gas phase diffusion control is 87·52 kJ mol⁻¹. Thus, the reduction rate of carbon containing pellets is controlled by the interfacial or local reactions, and the reduction rate can be expressed according to the McKewan equation of 1−(1−R)^1/3 = kt.

References

She

Xue

Dong

: ‘Study on basic properties of typical industrial dust from iron and steel plant and analysis of its utilization’, Chin. J. Process Eng., 2009, 9, (Suppl. 1), 7–12.

Wang

Chen

Zhou

: ‘Reduction of carbon-bearing pellets made from metallurgical dust and the effect of atmosphere on reduction and volatilization’, Eng. Chem. Metall., 1996, 17, (4), 290–293.

Ding

Wang

Zeng

She

Yang

Xue

: ‘Research on reduction kinetics of carbon-bearing pellets of BOF dust and sludge’, Chin. J. Process Eng., 2010, 10, (Suppl. 1), 73–77.

Guo

Zhang

Yang

: ‘A kinetic of the reduction of ZnO pellets containing graphite’, J. Chongqing Univ., 2002, 25, (5), 86–88.

Bonalde

Henriquez

Manrique

: ‘Kinetic analysis of the iron oxide reduction using hydrogen-carbon monoxide mixtures as reducing agent’, ISIJ Int., 2005, 45, (9), 1255–1260.

Seaton

Foster

Velasco

: ‘Reduction kinetics of hematite and magnetite pellets containing coal char’, Trans. ISIJ, 1983, 23, (6), 490–496.

: ‘Application-basis study of key technologies in carbon-bearing pellets-iron bath smelting reduction process’, Acta Metall. Sin., 1997, 33, (7), 718–728.

Wang

Yang

Tian

Sun

: ‘Mechanisms of reduction in iron ore-coal composite pellet’, Ironmaking Steelmaking, 1997, 24, (6), 457–460.

Wang

Yang

Tian

: ‘Reduction kinetics of iron ore-coal pellet during fast heating’, Ironmaking Steelmaking, 1998, 25, (6), 443–447.

10.

Meng

Zhao

: ‘Effect of carbon gasification rate on melting behavior of iron ore/carbon composite pellet’, J. Iron Steel Res. Int., 2007, 19, (12), 5–9.

11.

Banerjee

Sen

Roy

: ‘Reduction kinetics of porous zinc oxide pellet with CO–N₂ gas mixture’, Miner. Process. Extr. Metall. Rev. (Trans. Inst. Min Metall. C), 2004, 117, (4), 221–230.

12.

Cheng

Chen

Zhou

: ‘Reduction kinetics mechanism of iron ore pellets containing graphite’, J. Chin. Rare Earth Soc., 2008, 26, 103–107.

13.

Duong

Johnston

: ‘Kinetics of solid state silica fluxed reduction of chromites with coal’, Ironmaking Steelmaking, 2000, 27, (3), 202–206.

14.

Coetsee

Pistorius

Villiers

: ‘Rate-determining steps for reduction in magnetite-coal pellets’, Miner. Eng., 2002, 15, (11), 919–929.

15.

Wang

Chen

Zhou

Lin

: ‘Direct reduction process of pellet containing carbon with addition of Zn–Pb-bearing iron and steel plant dust’, J. Univ. Sci. Technol. Beijing, 1997, 19, (2), 130–133.

Reduction kinetics of carbon containing pellets made from metallurgical dust

Abstract

Introduction

Material properties

Chemical composition of raw materials (wt-%)

Chemical composition of green carbon containing pellets (wt-%)

Experimental

Results

Reduction

Metallisation (R m) and dezincification (R Zn) of pellets after reduction (%)

Phase analysis

Weight loss rate analysis

Dynamics analysis

Basic principles

Analysis of dynamics under possible RDSs

Calculation of reaction activation energy of carbon containing pellets

Activation energy of carbon containing dust pellets

Conclusions

References

Metallisation (R _m) and dezincification (R _Zn) of pellets after reduction (%)