3.1. Coking Coals and Coal Blends Properties
Table 2,
Table 3,
Table 4,
Table 5,
Table 6,
Table 7,
Table 8,
Table 9 and
Table 10 present the compositions of coal blends and the characteristics of coals and blends. Blends are presented in rising order of R
om from 0.94 to 1.11 wt.%. The volatile matters of the components of coal blends vary between 17.4 and 37.6 wt.%. In turn, the values of the volatiles decrease from blend 1 to blend 9 and are in the range of 31.8–29.2 wt.%. Concerning the maceral contents of the coals, they display a moderate or high vitrinite content from 58.00 to 93.00 wt.%, low to moderate liptinite content from 0.2 to 5.4 wt.%, and a low or moderate amount of inertinite macerals from 7.0 to 41.0 wt.%. The maximum thickness of the plastic layer for coals ranges from 8.5 to 29.0 mm, and for blends from 14.5 to 16.3 mm.
Figure 1a shows the dependence between the R
om and VM of all coals used for coal blend preparation from which it follows that the values of VM decrease with increasing values of R
om. At the same time, the values of FC increase with increasing values of R
om, as follows from
Figure 1b. As the R
om increase of the coals, the VM decreases significantly (R
2 = 0.93), and the FC content increases significantly (R
2 = 0.92), consistent with the general law of the coalification [
58].
Figure 1a,b show a clear separation of the coals that are used to compose the coal blends into four groups, which correspond to gas coals (R
om = 0.50–0.89%) fat coals (R
om = 0.80–1.19%), coking coals (R
om = 1.0–1.29%) and lean coals (R
om = 1.30–1.80%).
Figure 2a shows the relationship between the parameters of R
om and the VM of coal blends. The VM yield is found to have declined with the increase in the R
om of coal blends. This is a strong relationship that reflects the rank of the coals [
29,
30]. According to Sahoo et al. [
30] the R
2 for correlation between mean vitrinite reflectance and the volatile matter is 0.89. However, it should be noted that the yield of VM is of technological value, as it determines the amount of gaseous products released during coals heating.
Figure 2b shows the relationship between R
om and FC of coal blends. With an increase in the R
om, the values of FC also increase due to a decrease in the yield of VM and the influence of ash yield, which was within a narrow range of 8.0–8.5% for the studied blends. A similar negative relationship between the R
om and VM and a positive relationship between the R
om and FC were observed in a previous study [
58], however, the correlation coefficients are higher in our study.
It should be noted that the relationship between Rom and VM for blends is more pronounced, which can be explained by the fact that the composition of the coal blends was purposely chosen. In contrast, the dependence of coals can only characterize the actual properties without mixing to produce blends.
Figure 3a shows the relationship between vitrinite content and VM and the R
om. As the content of vitrinite increases, the yield of VM tends to decrease, as was also shown in previous studies [
31,
59,
60], while the R
om increases. For the studied blends, the maximum thickness of the plastic layer of the coal blends tends to decrease with increasing vitrinite content, as shown in
Figure 3b. The sum of caking microcomponents determines the plastic properties of coals. Usually, the maximum formation of the plastic layer corresponds to coals with VM of 30–35 wt.%, as follows from
Table 2,
Table 3,
Table 4,
Table 5,
Table 6,
Table 7,
Table 8,
Table 9 and
Table 10. According to the requirements for industrial coal blends, the maximum thickness of the plastic layer should be within the range of 14–16 mm, ensuring good blend caking in the plastic state [
55]. It is worth noting that the vitrinite values for the blends were in the range limited to 80.2–86.5 wt.%, and the main values for five blends were in a narrow range of 82.9–83.4 wt.% as shown in
Figure 3b, which determined the behavior of the obtained curve.
Figure 4 shows a negative correlation between parameters of inertinite content and the R
om for coal blends. With an increase in the amount of inertinite content in coals, the vitrinite content decreases, leading to a decrease in the R
om for blends. Additionally, no relationship was observed between the content of inertinite and the VM.
3.2. Metallurgical Coke Properties
Table 11 presents the characteristics of proximate analysis, physicochemical and physicomechanical properties of cokes, as well as the final temperatures of its production. According to the proximate analysis results, the ash yields of cokes are 10.6–11.1 wt.%. Changes in the yield of VM indicate the level of carbonization of the obtained coke and decrease with an increase in the final coking temperature. The low sulfur content in the coke is due to the low sulfur content in the blends. The CRI values of cokes decrease from coke 1 to coke 9 and are in the range of 32.9–30.3 wt.%, while the CRS values increase and are between 53.6–58.4 wt.%. Additionally,
Table 11 shows the results of the tumble and abrasion strengths, M
25 and M
10, respectively. It can be seen that the M
25 and M
10 values of cokes are quite stable within the ranges from 87.3 wt.% to 88.8 wt.% and from 7.2 wt.% to 7.8 wt.%, respectively.
It is well known that the CSR parameter negatively correlates with the CRI parameter, and this relationship can be vice versa [
61,
62,
63].
Figure 5 shows that with a decrease in the CRI of coke, the CSR increases. Coke 9 presents the lowest CRI and highest CSR, representing a good metallurgical property. According to [
64] the R
2 is 0.894 for the relationship between CRI and CSR. The resulting coefficient is slightly lower than indicated by [
64] and below the R
2 = 0.977 value indicated in [
46,
61].
In addition, using equations that describe the relationship between CRI and CSR, for example, CRI = 52.547–0.397·CSR in paper [
65], and CSR = −1.6884·CRI + 102.38 [
66], to assess the quality of the obtained coke leads to an underestimation of the values of the parameters in comparison with those obtained in this study. Therefore, the original equation is characterized by better accuracy.
Figure 6a,b show the dependencies of the R
om on CRI and CSR and on Micum 25 and Micum 10 indices, respectively. It has been established that the coke quality indicators improve with an increase in the R
om. It should be noted that the reflectance of vitrinite for coal blend 9 corresponds to a value of 1.11 wt.%. According to Zhang et al. [
67], a high CSR value of coke can be obtained from coal with a reflectance of vitrinite of 1.1–1.2 wt.% and/or from coals with VM of 22–26 wt.%. The relationship between the parameters of R
om and the parameters of CRI (R
2 = 0.99) or CSR (R
2 = 0.98) is stronger than that for M
25 (R
2 = 0.78) and M
10 (R
2 = 0.87). The higher the coefficient of determination (R
2), the better the relationship. This is because the R
om depicts the microstructure and level of coalification, which ultimately affects the CRI and CSR parameters [
67,
68].
Therefore, compared with the ‘cold’ strength, the physicochemical indicators of coke quality, namely CRI and CSR, better reveal the coke’s relationship with the coals’ nature. This is because the parameters CRI and CSR better characterize the structural changes in the obtained cokes due to a chemical reaction with CO
2 and subsequent determination of the mechanical strength, in contrast to the method of determining the ‘cold’ strength. Therefore, reflectance parameters describe the average degree of order of the molecular structure of organic matter and are a useful tool for the characterization of such heterogeneous carbon materials as raw or pyrolyzed. According to the Nippon Steel Corporation (NSC) model, it should also be pointed out, showing the dependence of the CSR index and the reflectance of vitrinite and inertinite [
69,
70]. The CSR increases with increasing R
om up to a value of about 1.4 wt.%, and at any level of R
om, coals with an inertinite content of 15–25 wt.% allow obtaining optimal CSR values. The paper [
63] also reported that the maximum CSR value for each rank level of coals is obtained with optimum inertinite content. However, Kosina and Heppner [
71] found that bituminous coals with the reflectance of vitrinite within the range of 0.80–0.90% and high inertinite content had no significant influence on coke mechanical properties expressed by Micum indices. In contrast, a distinct decrease in M40 values with increasing inertinite contents was observed for higher rank coals.
Figure 7a shows the relationship between the inertinite content of coal blends and CRI or CSR. As shown in
Figure 7a, with an increase in inertinite content in the coal blend, the CRI values increase. It is worth noting that in paper [
71] also with an increase in inertinite, the CRI increases, but the correlation coefficient is 0.696. The CSR index has the opposite effect, which is consistent with the results in [
69]. Thus, for the investigated blends, it was found that the coke quality deteriorates with increased inertinite content by more than 14%.
Figure 7b shows the relationship between the parameters of inertinite content in coal blends and indicators M
25 or M
10. An increase in inertinite content negatively correlates with the M
25 indicator and, contrariwise, shows a positive relationship with the M
10 indicator. In addition, relationships (R
2 = 0.91 for M
25 and 0.90 for M
10) indicate that the inertinite content of coal blends significantly influences ‘cold’ mechanical strength than on the CRI and CSR. However, many of the obtained M
25 values are within the standard deviation range of 0.85–0.90. Thus, a low proportion of reactive macerals inhibits the agglomeration of coal particles, creating a less stable coke structure, according to Pearson [
72].
Since vitrinite is the main active component in coal, which contributes to forming a non-volatile liquid mass during coking, the influence of vitrinite content in coal blends on coke quality indicators was investigated. It is well-known that non-volatile liquid masses obtained from coking coals of different grades are firstly fused and then interact with inert components to form a coke [
40].
Figure 8a shows the relationship between the parameters of the vitrinite content in coal blends and CRI or CSR.
With an increase in the vitrinite content within the coal blend, the CRI of coke decreases (R
2 is 0.80). This is consistent with paper [
73], where R
2 is 0.748. In turn, this has the opposite effect on CSR, which is also consistent with [
73] where R
2 is 0.659. However, the obtained dependences of the content of inertinite and vitrinite contents within blends with CRI are characterized by higher correlation coefficients than in paper [
73]. The decrease in CRI and increase in vitrinite content can be explained by the fact that softened ones originate from vitrinites, fuse, and enter into interfacial reactions with each other in the process of carbonization of a plastic mass. The higher fusion of the plastic mass is beneficial for mixing, and the interfacial reactions of the plastic mass are relatively sufficient in the coking process. Thus, cokes obtained from coal blends with a higher vitrinite content may have fewer defects in carbon structures, which show higher resistance to CO
2 [
40]. Noteworthy, a strong dependence exists between the parameters of the vitrinite content in coal blends and the M
25 or M
10 indices, which corresponds to R
2 = 0.97 and R
2 = 0.90, respectively, as shown in
Figure 8b, which confirms the pattern of obtaining coke with good strength with an increase in the proportion of vitrinite in the blend.
It is known that the final coking temperature affects the decline of CRI and the increase in CSR [
2,
74]. The final coking temperature is the maximum temperature of the coal pyrolysis process and can be considered a factor influencing the coke quality. A comparison of the results of determining the quality indicators of coke with the final temperature of coking shows a decrease in reactivity and an improvement in CSR with an increase in the final temperature of coking, as shown in
Figure 9a. Less strict relationships were obtained between final coking temperature and Micum 25 or Micum 10 indices (
Figure 9b). However, this also confirms that, in addition to the petrographic properties of the blend, the effect of the final coking temperature as a technological factor is important for further improving the coke quality.
Considering the strong relationship between CRI with final coking temperature and petrographic characteristics of coals, it was also determined which of these parameters has a greater impact on coke quality. Therefore, a two-component linear regression model was obtained, describing the dependence of CRI on final coking temperature (FCT) and inertinite content (IC), which includes the following Equation (2):
The regression results showed a greater significance of final coking temperature on CRI. This conclusion can be of great practical importance, since the composition of coal blends takes into account, first of all, the properties of the coal macerals, while the coking temperature is assumed to be constant.
Analysis of the relationship between CRI or CSR with indicators M
25 and M
10 displays medium coefficients of determination. As can be seen from
Figure 10a,b M
25 increases and M
10 decreases with decreasing CRI and increasing CSR. A good relationship between coke quality indicators confirms the influence of the petrographic properties of coal blends and the final coking temperature on the properties of industrial coke.
Consequently, the research results show that under conditions of an unstable raw material base for coking, multi-component coal blends, changes in coking regimes, petrographic parameters of the blend, proximate and caking indices of coal blends are closely related to the quality parameters of industrial coke.