Ironmaking
New Method for Preparation of Coke Analogues and Comparability for Industrial Metallurgical Coke
2020 Volume 60 Issue 9 Pages 1918-1923
Details
2020 Volume 60 Issue 9 Pages 1918-1923
To better understand metallurgical coke behavior in blast furnace, the preparation of coke analogues was improved by using demineralized coke powder. Scanning electron microscope, Raman spectroscopy, mercury intrusion method, and CO2 gasification reactivity test were used to establish the representativeness relation between different coke analogues and metallurgical cokes. The results show that the coke analogues prepared by graphite powder and demineralized metallurgical coke powder were in general representative of industrial coke. With the controlled and reproducible pore characteristic, the average pore diameter of coke analogues is smaller and the average pore area is bigger, while the pore connectivity of metallurgical coke is stronger. Analogue prepared from demineralized coke has a slight superiority over that prepared from graphite in carbon structure and gasification reaction with CO2. In general, the coke analogue made from demineralized coke has higher comparability to industrial metallurgical coke and are more suitable for laboratory research.
Metallurgical coke is the essential reagent in blast furnace (BF) iron-making.1) It is the main source of fuel and carbon monoxide gas in the high-temperature process, which are used to reduce iron oxide.2) And the most important function of coke is to support the column and provide a channel for gas flow upward. The reaction behavior of coke in BF is mainly related to its thermal strength and reactivity.3) A series of multiphase complex reaction of coke in BF have been studied in detail by scholars and researchers, including composition, structure, performance after its interaction between gas, liquid and solid.4) Due to the gradual shortage of resources, the main problems that coking faces are the diversity of coal selected for metallurgical coke, the complexity of coal types themselves, and the variability of coal blending schemes.5) So it is impossible to determine the results caused by a single influencing factor when studying the strength and reactivity of coke in cold and hot states.6) Metallurgical coke experiences a series of physical and chemical changes from top to bottom in the blast furnace, such as the violent collisions in lump zone, the dramatic gasification in soft zone and drip zone, the rapid combustion in tuyere zone, as well as the slag-iron erosion in hearth zone.1) The influence of mineral matter and metal in coke from the parent coal is identified as a key parameter for the variation of carbon morphology, distribution of pore structure; however a detailed mechanism could not be established.7,8,9)
To exclude the complex effects of various minerals on coke properties in heterogeneous reactions, the research group directed by Brian J. Monaghan developed a coke analogue.10,11,12,13,14) These coke analogues are mainly made from carbonaceous materials, specific minerals and metallic element, of which the mineralogy and porosity can be controlled. Similar to metallurgical coke, two temperature zones showing the different kinetic were identified in the coke analogues, chemical reaction controlled and mixed controlled respectively.12) Fe increased the sp2 graphite bonding character of carbon near the Fe-carbon interface. The effects of Fe added gave a higher reactivity of the coke analogue by increased the number of active carbon sites (Cf) and promoting oxygen delivery to the active carbon sites (Cf), with the activation energy values not changed significantly. The addition of quartz (SiO2) to the coke analogue had no clear different effects on the carbon bonding type but decreased its reactivity.15) The CO2 gasification reactivity was greatly increased by CaO, due to the developed active carbon sites by the significant absorption of CO2 gas molecules on CaO.16)
In general, some results have been obtained with the aid of coke analogues, and the effect of single mineral on carbon structure and reactivity has been obtained. However, the raw material of those coke analogues is graphite powder, which is completely free from the complex carbon type and structure formed by coal during the metamorphism process a few years ago.17) It is very necessary to take the representative of the complex reaction of metallurgical coke into consideration, especially in the diverse blast furnace environment. Based on this, the graphite used in the preparation of coke analogues was replaced with demineralization coke powder in compared with the research of Brian J. Monaghan, which could better represent metallurgical coke and eliminate the influence of minerals. The carbon morphology, pore structure distribution and reactivity of the two coke analogues were compared with two common metallurgical coke to representing their representativeness.
Two kinds of metallurgical coke commonly used in industry were selected for comparison, marked as 1# metallurgical coke (CRI = 21.80, CSR = 70.96) and 2# metallurgical coke (CRI = 33.43, CSR = 60.15) respectively. Their components and properties are shown in Table 1. In order to eliminate the complex influence of minerals in mother coal of coke, 2# metallurgical coke powder (<74 μm) was demineralized by 6 vol% HCl and 40 wt% HF aqua solution, the specific operation can refer to our previous research results.18)
| Industrial analysis of metallurgical coke powder before and after demineralization (air dried basis, wt%) | |||
|---|---|---|---|
| Item | Ash | Volatile | Fixed carbon |
| 1# metallurgical coke | 12.10 | 1.24 | 86.66 |
| 2# metallurgical coke before demineralization | 12.772 | 1.342 | 85.886 |
| 2# metallurgical coke after demineralization | 1.235 | 2.41 | 96.355 |
| Ash composition analysis of metallurgical coke powder before and after demineralization(air dried basis, wt%) | ||||||
|---|---|---|---|---|---|---|
| Item | SiO2 | Al2O3 | Fe2O3 | CaO | TiO2 | Other |
| 1# metallurgical coke | 50.43 | 32.6 | 5.26 | 3.337 | 1.99 | 6.383 |
| 2# metallurgical coke before demineralization | 44.723 | 31.782 | 8.984 | 6.114 | 1.492 | 6.905 |
| 2# metallurgical coke after demineralization | 43.096 | 32.627 | 4.494 | 4.236 | 8.857 | 6.634 |
| 1# Coke analogue | 2# Coke analogue | ||
|---|---|---|---|
| Materials | Proportional addition | Materials | Proportional addition |
| Graphite Powder | 28% (120–160 μm) + 28% (<45 μm) of dry mass | Demineralized coke powder | 56% (<74 μm) of dry mass |
| Phenolic resin (Bakelite) | 40% (<105 μm) of dry mass | Phenolic resin (Bakelite) | 40% (<105 μm) of dry mass |
| Phenolic resin (Novolac) | 4% of wet basis | Phenolic resin (Novolac) | 4% of wet basis |
The preparation scheme of the analogue in this paper referred to the methods of Aladejebi O A,19) but the difference was that two coke analogues were prepared by using graphite powder and metallurgical coke powder after demineralization respectively. Phenolic resin (Novolac) was used as a binder, gluing graphite powder/demineralization coke powder and Phenolic resin (Bakelite) powder together. 5 g samples were pressed under 40 pa pressure to make a cylinder (φ20 mm×17 mm). The pressed samples were placed at 378 K, and dried for 24 hours.
The dried samples were refined at different temperatures and atmospheres in a high temperature tube furnace to obtain the final coke analogue, and the specific refining temperature and atmosphere settings is shown in the Fig. 1.
The refining temperature and atmosphere settings for coke analogue production. (Online version in color.)
Scanning electron microscope (SEM) was used to determine the microstructure of metallurgical cokes and coke analogues. The pore diameter distribution and porosity in coke were measured by Mercury injection method. Raman spectrum was used to compare the bonding types of carbon in coke. In addition, the microcomputer thermal gravity balance was selected to perform the reactivity test of coke. The specific test method and operation process can refer to our previous research.20)
Figure 2 shows the microstructure characteristics of metallurgical coke and coke analogues. On the whole, the pore size of metallurgical coke is relatively uniform, and the pore size is much smaller, around 100 μm. But the connectivity between pores is strong, especially the 2# metallurgical coke. The pore distribution of coke analogues is obviously different, with some larger pore more than 200 μm. In addition, there are a large number of micropore (< 10 μm) in the 2# coke analogue, which are distributed around the macropore.
Morphologic characteristics of metallurgical coke and coke analogues: (a) 1# metallurgical coke; (b) 2# metallurgical coke; (c) 1# coke analogue; (d) 2# coke analogue. (Online version in color.)
The reason for the macropore in coke analogues may be the direct use of Phenolic resin (Novolac) as binder in the preparation process. Because of the viscosity of the binder itself, the non-uniformity mixing of sample preparation resulted in a large number of gases generated locally in the pyrolysis process. However, the presence of micropore around the macropore in the 2# coke analogue may be a result of some macromolecules in the coke powder after demineralization, which are decomposed to form small molecules and released in the gaseous form during heating.21)
To compare the mineral distribution of prepared coke analogues with metallurgical coke, the back scattering SEM and EDS analysis were used, as shown in Fig. 3. From Figs. 3(a) and 3(b), the minerals in the coke are more dispersed. And the EDS map scanning analysis results show that the element is most abundant of Al, Ca, Si and Mg, also contains a small amount of K and Na, which have different effects on coke behavior in blast furnace. However, the content of minerals in 2# coke analogue refined with demineralized coke powder is significantly reduced, and the distribution is also limited. Therefore, it has certain advantages for the research of its application in the laboratory.
SEM micrographs and EDS mapping showing mineral distribution in coke: (a) and (b) 2# metallurgical coke; (c) and (d) 2# coke analogue. (Online version in color.)
The pore size distribution and specific surface area were measured by mercury injection, and the data obtained are shown in Fig. 4. As can be seen, the measurement results of pore size distribution are quite different from that of morphology analysis by SEM. The average pore diameter of the two coke analogues is smaller than that of industrial metallurgical coke. The main reason for this result was that the porosity connectivity could not be distinguished in the mercury injection method, so the result represents more macropores. Although the pores of coke analogues exist alone, their pores are larger. The porosity of coke analogues prepared from demineralized coke powder is similar to that of industrial metallurgical coke, but the average pore area is bigger, especially 2# coke analogue. The main reason is the large number of micropores formed by the decomposition and escape of macromolecules in coke powder.
Pore distribution obtained by mercury intrusion method: (a) 1# metallurgical coke; (b) 2#metallurgical coke; (c) 1# coke analogue; (d) 2# coke analogue. (Online version in color.)
The Raman spectrums of carbon in different metallurgical cokes and coke analogues are shown in Fig. 5. The traditionally D band and G band in disordered carbonaceous materials were observed at around 1350 and 1600 cm−1, showing the disordered structure and crystal symmetry graphite structure, respectively.22) Deconvolution method was adopted to obtain structural information hidden in the overlapped area between the D band and G band.23,24,25) Peak intensity ratio I(V)/I(G) and I(D)/I(G) was also calculated to characterize the structural information. The value of I(V)/I(G) indicates the structural defect degree of disordered sp2-sp3 binding in coke, which shows irregular changes among metallurgical coke and coke analogue. The range of I(V)/I (G) value and I(D)/I (G) value of industrial metallurgical coke are 0.3–0.5 and 0.9–1.4 respectively, and the carbon structure of 2# coke analogue is located in the middle of this region, while the carbon structure of the other analogue is the opposite. The main reason is 1# coke analogue chose highly ordered graphite as the raw material, which has great differences in the structure of coal samples used in coke refining.
Deconvolution peak analysis of Raman spectrum: (a) 1# metallurgical coke; (b) 2# metallurgical coke; (c) 1# coke analogue; (d) 2# coke analogue. (Online version in color.)
In order to study the accuracy of coke analogues in simulating metallurgical coke gasification reaction in CO2, the gasification of four coke was tested under three heating rates (5 K·min−1, 10 K·min−1, 15 K·min−1). Figure 6(a) shows the results of non-isothermal gasification reaction at 5 K·min−1. It can be clearly seen that the reactions of 2# coke analogue and two metallurgical coke are relatively uniform in the whole gasification process, except for the 1# coke analogue. The reactivity of 2# coke analogues is enhanced relative to industrial metallurgical coke, which may be related to the removal of ash, and the specific causes of minerals catalyze need further study. The whole reaction process of 1# coke analogue is divided into two stages, which is mainly because the structural difference of carbon types. One is the highly ordered graphite carbon, and the other is the irregular carbon structure generated by the pyrolysis of phenolic resin (Bakelite). According to previous research results, the higher the degree of graphitization of carbon, the more difficult it reacts with CO2, resulting in the formation of two-stage gasification.26,27) However, the demineralized coke powder was selected as raw material for 2# coke analogue, and their carbon structure was similar to that of Bakelite after pyrolysis, so there was no significant segmenting gasification.
Gasification behavior analysis of metallurgical coke and coke analogue: (a) 5 K·min−1 non-isothermal gasification reaction curve; (b) apparent activation energy calculation results. (Online version in color.)
The apparent activation energy was calculated by the results of three different heating rates, as shown in Fig. 6(b). The activation energy of the two industrial metallurgical cokes is 70–70 kJ.mol–1, while coke analogue has a lower apparent activation energy than metallurgical coke, but within the reasonable range. The activation energy of the 1# coke analogue is 66.21 kJ.mol–1, and that of the 2# coke analogue is 56.32 kJ.mol–1. The main reason is that the minerals in industrial cokes have different positive catalytic and negative catalytic effects on the CO2 gasification, resulting in the higher activation energy. Because the catalysis of minerals is not the focus of this paper, no further study has been done here. The low activation energy of 2# coke analogs is due to the high disorder carbon structure of demineralized coke powder in the raw materials, which is more prone to gasification reaction. It is worth noting that the activation energy calculated in this paper is quite different from that of coke and its analogues in other literatures. The activation energy of coke analogues calculated by O. A. Aladejebi and A. S. Jayasekara in the chemical reaction control zone was 260–270 kJ.mol–1,11,19) and the activation energy of metallurgical coke calculated by G. W. Wang and K. J. Li was 140–266 kJ.mol–1.28,29) The main reason for the huge difference in the activation energy is that on the one hand, samples are selected differently in the experimental process, some are in powder form, some are in block form; on the other hand, the kinetic calculation models used in the activation energy calculation process are different.
Coke analogues has been developed to simulate the complex and inherent heterogeneous metallurgical coke for laboratory study on the basis of previous studies. The results of morphology analysis, Raman spectroscopy analysis, porosity characteristics analysis and gasification reaction characteristics analysis of the two different coke analogues were compared to show their respective advantages. According to the morphology and pore characteristics, the porosity of the two analogues is close to that of industrial metallurgical coke, and the porosity of 2# coke analogue is much lower. However, the pore distribution is quite different. Coke analogues do not show the strong connectivity similar to industrial coke, resulting in a smaller average pore size and larger average pore area. Although the porosity characteristics of coke analogues are quite different from those of industrial metallurgical coke, the characteristics of pores in coke are closely related to the selection of raw materials and its ratio, the parameter setting of coking process, which can be improved in the following experiments. From the perspective of carbon bond type and gasification, compared with the highly ordered carbon structure of graphite, 2# coke analogue prepared by demineralized coke powder shows disordered carbon structure and in-plane defects. And its reactivity with CO2 is slightly stronger than that of industrial coke, which shows uniform gasification compared to the two-stage reaction of 1# coke analogue.
In general, 2# coke analogue is more representative of metallurgical coke from the aspects of microstructure, porosity, carbon types, gasification reaction with CO2. Compared to coke analogue made from graphite, coke analogue made from demineralized coke powder is benefit by the coal rank with relatively simple structure and is controllable with mineral and metallic addition. Because the mineralogy and porosity of coke analogues can be realized by the operation of coking process, the behavior of special coke in blast furnace can be more scientifically simulated by using the demineralized powder of this coke as the raw material for preparing coke analogue.
This work was supported by the National Science Foundation of China (51774032); the National Science Foundation for Young Scientists of China (51804025); the National Key Research and Development Program of China (2017YFB0304300 & 2017YFB0304303); the Chinese Fundamental Research Funds for the Central Universities (FRF-TP-17-086A1).
