直接还原铁

Thermochimica Acta 410 (2004) 133–140 Thermal investigations of direct iron ore reduction with coal Gui-su Liu a, Vladimir Strezov b,∗, John A. Lucas c, Louis J. Wibberley d a Niksa Energy Associates, 1745 Terrace Drive, Belmont, CA 94002, USA b Newbolds Applied Research, The University of Newcastle, Cnr Frith and Gavey Sts, Mayfield, NSW 2304, Australia c Discipline of Chemical Engineering, Faculty of Engineering and Built Environment, The University of Newcastle, University Drive, Callaghan 2308, NSW, Australia d BHP Billiton Newcastle Technology Centre, Off Vale Street, Shortland 2307, NSW, Australia Received 8 April 2003; received in revised form 16 June 2003; accepted 25 July 2003 Abstract In this paper, fundamental mechanisms for iron ore reduction in coal–ore mixtures have been investigated using several advanced experi- mental techniques. Firstly, the thermal properties of coal–ore mixtures were studied and apparent specific heat of coal–ore mixtures against temperature was obtained at a heating rate of 10 ◦C/min. Several exothermic and endothermic peaks were observed which were related to the decomposition reactions and reduction. The flue gases from the mixture were analysed using a mass spectrometer. Secondly, the X-ray diffrac- tion (XRD) and the iron phase analytical techniques were applied to identify the iron phase changes with the temperature. It has been found that coal devolatilisation and iron oxides reduction occur simultaneously during the heating of the mixture. H2 and CO gases produced from coal devolatilisation and char gasification were responsible for the reduction of iron oxides at these temperatures. Iron oxides undergo step-wise reduction over the whole process. The results in this work provide a fundamental understanding for the direct reduced ironmaking processes. © 2003 Elsevier B.V. All rights reserved. Keywords: Coal; Iron ore reduction; Direct reduced iron; Thermal property 1. Introduction Substantial developments in direct reduction ironmaking (DRI) technologies have been recently conducted providing sustainable mean, for metallurgical operations. The largest advantage of the DRI technologies relies on the fact that DRI does not require cokemaking and sintering. Both coke- making and sintering, being at the front end of the conven- tional blast furnace ironmaking technology, are considered as costly for the new process construction and are consis- tently causing environmental concerns. The DRI process, on the other hand, consists of carbothermic reduction of iron oxide directly with the volatiles liberated during coal de- volatilisation as well as the carbon monoxide regenerated from coal char. This process provides an advanced utilisa- tion opportunity for the high volatile coals, which were oth- erwise unusable in the steel industry. There has been extensive work performed on iron ore re- duction of coal–ore mixtures and its kinetics [1–20]. Opti- ∗ Corresponding author. Tel.: +61-2-4968-6771; fax: +61-2-4968-6777. E-mail address: vladimir.strezov@newcastle.edu.au (V. Strezov). misation of the DRI process, however, requires knowledge of the thermal properties of the coal–ore mixtures and mech- anism of the reduction reactions, which have still not been well understood. It is therefore necessary to have an insight into fundamental mechanisms for these complex reactions. The aim of this work is to investigate coal–ore reactions of their mixtures during heating, using several advanced mea- suring techniques. Thermal property, gas analysis, and iron phase changes during heating of the mixtures were obtained, through which fundamental insights into coal and iron ore behaviour and their interactions are provided. 2. Experimental Coal C1 and iron ore O1 both from Australian origin were used in this work. The coal has 33.5% volatile matter and 6.5% ash in air-dry base, and has 85.1% C, 5.5% H, 2.1% N, 0.7% S and 6.6% O in daf base. The major components of the iron ore are 62.1% Fe, 4.33% SiO2 and 1.98% Al2O3. Both samples were dried under vacuum at 80 ◦C for 2 h before proceeding with experiments. The mean particle size of coal and iron ore was −80 and −50�m, respectively. 0040-6031/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0040-6031(03)00398-8 134 G.-s. Liu et al. / Thermochimica Acta 410 (2004) 133–140 Fig. 1. The cross-sectional diagram of the instrument used for thermal property measurements. A mixture of coal and iron ore particles at a mass ratio of 20:80 was used throughout the study. Several measuring techniques were used for investigating the reactions between coal and ore in the study. The thermal property measurements were performed using the computer aided thermal analysis technique [21,22], with the instru- ment shown in Fig. 1. This method provides an opportunity for dynamic thermal analysis by heating the sample with controlled heat flux while continuously monitoring boundary temperature conditions within the sample. The heat transfer is, then, calculated using inverse numerical modelling tech- nique. There was a substantial effort in the past [21–27] to apply inverse methods in a variety of different thermal stud- ies. It was shown that, with controlled heating conditions it is possible to obtain accurate thermal characterisation of the materials. In this work, the sample with 30 mm in length and 10.6 mm in diameter was packed in a silica glass tube to the density of 1860 kg/m3, and heated by radiation from a surrounding graphite cylinder. The heating rate of the furnace was typically maintained at 10 ◦C/min, controlled through a type K thermocouple embedded in the graphite. The sample was maintained under inert atmosphere with an argon flow of 5 ml/min through the glass sample tube. Sample temperatures were continuously measured by ther- mocouples positioned at the surface and in the centre of the sample. The measured data were acquired at a frequency of 1 Hz, and the typical temperatures for one experimental run for coal–ore mixture in a ratio of 20:80 are shown in Fig. 2. The heat flux at the surface of the sample was calculated by assuming heat transfer from the graphite heating element to the sample was performed predominantly by radiation and was estimated using Eq. (1). For this purpose and to ensure uniform emissivity of the glass, the outside surface of the sample glass tube was coated with a thin layer of carbon soot, prior to positioning the sample tube centrically to the graphite heating cylinder. Q = F1−2σ(T 4g − T 4s ) (1) 0 100 200 300 400 500 600 700 800 900 1000 0 1000 2000 3000 4000 5000 6000 Time (s) Te m pe ra tu re (º C) graphite sample surface sample centre Fig. 2. Time–temperature history during heating of coal–ore mixture at a mass ratio of 20:80 and heating rate of 10 ◦C/min. where Q is the heat flux (W/m2), F1−2 the radiation shape factor, σ the Stefan–Boltzmann constant (5.67 × 10−8 W/m2 K4), and Tg and Ts are the temperatures of the graphite and the surface of the sample (K), respectively. The one-dimensional heat conduction in the sample is given by the following expression: ρCp ∂T ∂t = k ∂ ∂r ( r ∂T ∂r ) (2) where ρ is the density (kg/m3), Cp the specific heat (J/kg K), k the thermal conductivity (W/m K), T the temperature (K), t the time (s) and r the radius (m). The sample was numerically divided into a number of nodes (n) across the radius. For each node an estimate was made based on the heat balance principle (i.e. heat accu- mulated in the node equals to the difference between the incoming and outgoing thermal energies). The boundary conditions of the system were the temperatures measured at the centre and surface of the sample, and the heat flux calculated according to Eq. (1). A computational matrix was then generated using Eq. (3) to estimate the volumetric specific heat based on the initial mass of the heated sam- ple. For closer familiarising with the equation evaluation procedure the reader is referred to [21–24]. ρCp = 2πn�xQ(t) (�x2π/4�t)(T t0 − T t−10 )+ (�x2π/�t)(n− (1/4))(T tn − T t−1n )+ ∑n−1 i=1 (2π�x2i/�t)(T t i − T t−1i ) (3) G.-s. Liu et al. / Thermochimica Acta 410 (2004) 133–140 135 where ρCp is the volumetric specific heat (J/m3 K), n the number of nodes, T ti the temperature expressed in K of the node i for the time t (s) and Q(t) the heat flux expressed in W/m2 for time t. The volumetric specific heat in this case is apparent, which means that it includes heats developed due to decomposi- tion, transformation and reaction. If, for instance, endother- mic heat is developed in the sample, it results in increase of the apparent specific heat, hence the heat is consumed by the sample. On the contrary, if an exothermic reaction is developed during heating, the apparent specific heat will show decrease of its values. The performance of the above- mentioned measurement method was tested previously [22] on a range of different materials and accuracy of approxi- mately ±2% was found in the measurement data. Maximum temperature was limited to 1000 ◦C, which was the maxi- mum limit of the furnace used. Results are normally plotted against average of the two sample temperatures. Thermal studies of direct iron ore reduction were, in this work, incorporated with analysis of the gaseous products us- ing a mass spectrometer Prima 600 connected to the gas out- let of the apparatus described above. Argon gas was flown at a rate of 100 ml/min across the sample for the purpose of these measurements. The water vapour and larger molecu- lar weight volatiles generated during heating of the samples were condensed at the outlet of the furnace prior to the gas analyser. The volume percentages of gases as a function of sample temperature were then obtained. The coal–ore sam- ples, which were heated up to certain temperatures in the above furnace, were further collected and analysed using an X-ray diffraction (XRD), and an iron phase analysing tech- nique. The XRD uses the “fingerprint” of a crystalline mate- rial to allow identification of unknown phases in a mixture. Rapid identification of unknown phases can be possible us- ing search/match software available at the unit. The samples were also sent to chemical laboratory for carbon and iron phase analysis. 3. Results 3.1. Apparent specific heat The apparent specific heats for single coal C1 and iron ore O1 were measured at a heating rate of 10 ◦C/min under Ar atmosphere, as shown in Fig. 3. Coking coal C1 ex- hibits a rapid and significant exothermic reaction occurring between 420 and 460 ◦C, which is thought to be caused by both physical and chemical related changes in the coal plastic region, i.e. tar formation, tar vaporisation and reso- lidification. Following the tar formation, the secondary devolatilisation and hydrogen release occurred over a tem- perature range of 500–1000 ◦C. Previous work [28] has shown that the evolution of CO for C1 initiated at 450 ◦C and completed at about 950 ◦C, with a maximum evolution rate at 720 ◦C, whereas H2 started from 495 ◦C and contin- 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 0 100 200 300 400 500 600 700 800 900 1000 Temperature (ºC) Sp ec ifi c H ea t ( J/m 3 K ) Mixture of C1+O1 Single O1 Single C1 Fig. 3. Specific heats of single coal C1, single ore O1, and their mixture at 20:80 mass ratio. ued to evolve at 1000 ◦C, with a maximum rate at 785 ◦C. The specific heat of O1 iron ore was also measured, which exhibited several sharp peaks. The first peak at 120 ◦C represents the endothermic water vaporisation, and the sec- ond at 340 ◦C was caused by dehydroxylation of goethite (FeO(OH), hydrated iron oxide) with formation of hematite (Fe2O3). The third endothermic peak at around 600 ◦C was due to breakdown of kaolinite (Si4Al4O10(OH)8) [29]. The fourth peak, caused by magnetic transformation of Fe2O3 [30], appeared at 685 ◦C, whereas the exothermic trough at 843 ◦C was most likely caused by the partial reduction of hematite from the carbon present in the ore. The specific heat of the mixture at a mass ratio of 20:80 exhibits similarities with the pure iron ore, prior to 600 ◦C. Dehydroxylation and decomposition of kaolinite are the dominant reactions for the mixture, with peaks being similar to those of single iron ore. Primary coal devolatilisation has initiated below 600 ◦C as shown for single C1 coal, however the rate is reduced at the presence of iron oxides [3,4]. Above 600 ◦C, the curve was substantially different from that of either the single iron ore or coal. An exothermic reaction occurred at around 690 ◦C, followed by two strong endother- mic reactions. These reactions are most likely due to the iron ore step-wise reduction, i.e. the reduction of hematite (Fe2O3), magnetite (Fe3O4) and wustite (FeO), respectively. Fig. 4 shows the specific heats of iron ore reduction of mixtures of C1–O1 and coke–O1 at a mass ratio of 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 7.E+06 8.E+06 0 100 200 300 400 500 600 700 800 900 1000 Temperature (ºC) Sp ec ifi c H ea t ( J/m 3 K ) C1+O1 Coke+O1 Fig. 4. Specific heats of mixtures of C1–O1 and coke–O1 at a 20:80 mass ratio. 136 G.-s. Liu et al. / Thermochimica Acta 410 (2004) 133–140 20:80. Similar to that of C1–O1 mixture, the specific heat of coke–O1 mixture only exhibits two endothermic peaks at 340 and 595 ◦C, respectively. These reactions are of the same origin as the dehydroxylation and decomposition reac- tions of the pure iron ore in the same temperature range. The intensity of the peak at 595 ◦C in coke–O1 is smaller than C1–O1 mixture. The reason is that the peak for the latter is a sum of the endothermic reactions of iron ore decomposition and the secondary devolatilisation of coal. 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 0 200 400 600 800 1000 Temperature (ºC) Temperature (ºC) Temperature (ºC) Temperature (ºC) Sp ec ifi c H ea t ( J/m 3 K ) 0.E+00 1.E-02 2.E-02 3.E-02 4.E-02 5.E-02 G a s Co m po sit io n (% wt ) specific heat CH4 of C1 CH4 of C1+O1 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 0 200 400 600 8001 000 Sp ec ifi c H ea t ( J/m 3 K ) Sp ec ifi c H ea t ( J/m 3 K ) Sp ec ifi c H ea t ( J/m 3 K ) 0.E+00 1.E-03 2.E-03 3.E-03 4.E-03 5.E-03 6.E-03 7.E-03 8.E-03 G as C om po sit io n (% w t) specific heat H2 of C1 H2 of C1+O1 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 0 200 400 600 800 1000 0 0.05 0.1 0.15 0.2 0.25 0.3 G as C om po sit io n (% w t) specific heat CO of C1 CO of C1+O1 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 0 200 400 600 800 1000 0 0.05 0.1 0.15 0.2 0.25 G as C om po sit io n (% w t) specific heat CO2 of C1 CO2 of C1+O1 Fig. 5. Mass spectrometric analysis of gas evolution of C1 + O1 mixture during heating. Unlike that of C1–O1 mixture, at higher temperatures no significant peaks were observed for coke–O1 mixture until the sample reached 920 ◦C. At temperatures above 920 ◦C, similar to the C1–O1, a rapid increase of specific heat was observed for coke–O1 mixture, due to char gasification with CO2. The above observation suggests that CO and H2 volatiles generated from coal decomposition are the predominant re- ductants for iron oxides at lower temperatures, whereas the iron ore reduction by char gasification dominates above 920 ◦C. 3.2. Product gas analysis The gas analysis was performed by connecting a mass spectrometer to the outlet of the heating furnace. Fig. 5 shows the weight percentages of product gases CH4, H2, CO and CO2 of the coal–ore mixture as a function of tem- perature, in comparison with the specific heat. The CH4 was predominantly produced from coal devolatilisation, and re- leased between 450 and 800 ◦C, with a maximum value at 600 ◦C. The H2 started to release at 500 ◦C, and the con- centration decreased with increasing the temperature from 530 ◦C, and disappeared from 600 to 700 ◦C, which indi- cated a great consumption of H2 gas due to the significant 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 7.E+06 0 200 400 600 800 1000 Temperature (ºC) Temperature (ºC) Sp ec ifi c H ea t ( J/m 3 K ) Sp ec ifi c H ea t ( J/m 3 K ) 0.E+00 1.E-02 2.E-02 3.E-02 4.E-02 5.E-02 G as C om po sit io n (% wt ) specific heat CO of coke+O1 0.E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 6.E+06 7.E+06 0 200 400 600 800 1000 0.0E+00 2.0E-02 4.0E-02 6.0E-02 8.0E-02 1.0E-01 1.2E-01 1.4E-01 G a s C o m po sit io n (% w t) specific heat CO2 of coke+O1 Fig. 6. Mass spectrometric analysis of gas evolution of coke + O1 mixture during heating. G.-s. Liu et al. / Thermochimica Acta 410 (2004) 133–140 137 reduction of the iron oxide. The H2 increased again from 700 ◦C, reaching a peak at 780 ◦C, and decreased again. Compared to the single coal C1, the H2 concentration of C1–O1 mixture is much lower due to the consumption of H2 as a reductant gas for iron oxides, with the peak temperature similar to each other, as shown in Fig. 5. The CO started to release through by the coal devolatilisation at 450 ◦C, and decreased with the temperature again from 530 to 700 ◦C, which was due to the consumption by reduction. The CO concentration increased rapidly from 700 ◦C, followed by a decrease at about 950 ◦C. The CO2 concentration exhib- ited an inverse trend of CO below 780 ◦C which was pro- Fig. 7. X-ray diffraction analysis of mixture of C1 + O1 heated at different temperatures. duced from iron ore reduction by CO predominantly from volatile, and increased dramatically at temperatures above 780 ◦C, indicating the carbon gasification has become sig- nificant within this temperature range. Rapid increasing of both CO and CO2 indicates that significant carbon consump- tion occurred due to gasification. Fig. 6 shows the mass spectrometric analysis of coke and O1 mixture at the same mass ratio. The CO and CO2 con- centrations were extremely low and remained constant be- low 700 ◦C. They started to increase at 700 ◦C, and became rapid from 800 ◦C. It is more than likely that direct reduc- tion of iron oxide by solid carbon occurred between 700 and 138 G.-s. Liu et al. / Thermochimica Acta 410 (2004) 133–140 800 ◦C with the CO2 as a primary product gas. Significant carbon gasification occurred at temperatures above 800 ◦C, and the reduction of iron oxide occurred through reducing gaseous phase, for example CO and H2 produced from car- bon gasification with CO2 and H2O, respectively. 3.3. XRD analysis Coal and ore samples were prepared in a furnace at vari- ous temperatures followed by cooling with a rate of approx- imately −100 ◦C/min, and were further analysed by XRD. Two of the samples heated at temperatures above 1000 ◦C were prepared separately in a muffle furnace. Fig. 7 shows the XRD analysis of coal–ore mixture made at different tem- peratures. At 25 ◦C (i.e. original coal and iron ore mixture), the sample was dominated by Fe2O3 and carbon, with mi- nor phases of goethite (FeO(OH)) and kaolinite. At 450 ◦C, only the disappearance of goethite was evident, due to the dehydroxylation reaction at 340 ◦C as indicated by the spe- cific heat. The first step reduction of Fe2O3 forming Fe3O4 was initiated below 530 ◦C, and completed at a temperature between 620 and 670 ◦C. The rate for the first step reduc- tion within 530–580 ◦C was much higher than that within 580–670 ◦C. It was evident that the kaolinite disappeared between 580 and 620 ◦C, due to the thermal decomposi- tion as found by the specific heat study presented above. The second step reduction of Fe3O4 to form FeO initi- ated between 670 and 740 ◦C, and completed below 870 ◦C. Rapid reduction rate was indicated between a temperature range of 800–870 ◦C. The third step reduction of FeO to form metallic Fe initiated from 870 ◦C. Small amount of metallic Fe