Improvement of Coking Property of Low-Grade Coals by Mild Reduction Treatment

Abstract

With the recent concern over the depletion of high-quality coking coal used for coke production, it is desirable to develop a simple method to convert low-grade coals with poor coking properties into high-quality coking coals. In this study we have focused on the possibility that polymerization reactions caused by the recombination of coal-inherent radicals may occur when coal is heated, which can deteriorate the softening and melting properties of coal. Aiming at removing such radicals, we have proposed a pretreatment method for improving coking property of low-grade coals, in which low-grade coals are treated with a reductant such as formic acid at around 60°C. It was shown that the treatment by either aqueous formic acid or formic acid vapor significantly improved the coals’ thermoplastic performance and enhanced the strength of the resulting coke even though consumption of the reductant formic acid was small enough to hardly change the elemental composition of the coals. We have also succeeded in restoring the coking property of weathered coking coal.

1. Introduction

The amount of high-quality coking coal with coking properties used in the production of coke for steelmaking is only about 10% of the total available coal reserves and is feared to be depleted. Therefore, in the future, low-grade coal with poor coking properties must be used mainly as a coke feedstock.

Coke has a role in the blast furnace as a reductant to reduce iron ore, a fuel to maintain high temperatures, and a spacer to secure distribution paths for gas and molten iron, but coke produced from low-grade coal has poor mechanical strength and cannot serve as a spacer. In general, to produce coke with mechanical strength, coal must be softened and melted during the heating process. However, low-grade coal has poor softening and melting properties.

Various methods have been developed to increase the percentage of low-grade coals used in the coke feedstock. Increasing the packing density of coking coals in coke production is effective, and the Stamp Charging Technology¹⁾ and the Formed Coke Manufacturing Process²⁾ are processes that take advantage of this effect. SCOPE21 (Super Coke Oven for Productivity and Environmental enhancement toward the 21st century) is a process for modifying low-grade coals through pretreatment, which has been commercialized in Japan.³⁾ In the SCOPE21, the softening and melting property of low-grade coals can be improved by rapidly heating raw materials containing low-grade coals to about 350°C. While such physical and thermal treatments are in practical use, treatment of coal using chemicals has mainly been used for demineralization,⁴⁾ and few processes for the purpose of upgrading low-grade coals have been developed.

It is known that radicals are present in coal not only when heated but also originally,^5,6) and it has been reported that radicals can be removed by reduction treatment using chemical substances.⁷⁾ Larsen, et al. reported that the pyrolysis behavior of immature kerogen was significantly altered by removing the inherent radicals.^8,9) Radicals originally present in low-grade coals may recombine upon heating and become cross-linking points for molecules in the coal. In this study, we focused on the possibility that polymerization reaction caused by the recombination of radicals originally existing in coal can deteriorate softening and melting properties, and proposed a mild reduction pretreatment method to remove the radicals, and examined the validity of the proposed method.

Formic acid was used as the reducing agent, which becomes CO₂ gas after reduction and is easily separated from the coal. Formic acid has almost the same boiling point as water and can be supplied as a vapor even at temperatures below 100°C. When gaseous formic acid is used, both the reactant (formic acid) and product (CO₂) of the reductant are gaseous, and considering separation, the process should be suitable for treating solid coal.

2. Experimental

2.1. Samples

Two types of low-grade coals A, B and a coking coal C were used. Elemental compositions of the coals used are shown in Table 1. The coals were ground to less than 1 mm before use.

Table 1. Elemental compositions of the samples used in this study.

Coal	Ultimate analysis [mass%, d.a.f]				Atomic Ratio [−]		Ash [mass%, d.b.]
	C	H	N	O (diff.)	H/C	O/C
A	79.8	5.6	1.3	13.3	0.83	0.13	14.8
B	83.2	5.0	1.0	10.9	0.71	0.10	8.8
C	86.9	4.5	0.8	7.8	0.62	0.07	4.7

2.2. Formic Acid Aqueous Solution Treatment

In an Erlenmeyer flask, 5 g of raw coal and 250 cm³ of 0.5 mol/L formic acid aqueous solution were mixed and the flask was immersed in a water bath kept at 60°C. After 6 hours, the flask was taken out of the water bath and the solution and coal were separated by filtration using a filter paper.

2.3. Formic Acid Vapor Treatment

In a closed 250 cm³ glass container, 2.5 g of raw coal and 7 g of formic acid in a sample tube bottle were separately placed, and the volatilized formic acid and raw coal were brought into contact with each other at 60°C for 2 or 6 hours using the water bath.

2.4. Analyses of the Products

The gases produced during the treatment were collected in a gas bag and analyzed by a gas chromatograph (GC-12A, Shimadzu). The treated coals were subjected to elemental analysis (CHN Corder MT-6M, Yanaco) and thermomechanical analysis (TMA-60, Shimadzu).

The thermomechanical analysis (TMA) was employed to evaluate softening and melting properties of the raw coals and treated coals. In the TMA, the raw or treated coals were stacked about 1 mm thick in a platinum cell with an inner diameter of 5.3 mm, and heated to 900°C in nitrogen at a temperature increase rate of 10°C/min while applying a 0.098 N load with a 4.3 mm diameter rod, and the change in position from the initial rod position was measured. The position was normalized by the initial thickness of the sample bed to obtain ‘normalized displacement’ data.

2.5. Fractionation by Solvent Extraction

Solvent extraction of raw and treated coals was performed to determine if low molecular weight components increase with treatment. The details of the extraction process are as previously reported.^10,11) In a single extraction operation, coal is separated into three components: soluble (extracted at high temperature and still soluble in solvent at room temperature), deposit (extracted at high temperature but deposited as a solid at room temperature), and residue (not extracted at the extraction temperature). Molecular weight of the soluble should be the lowest and that of the residue should be the highest among the three fractions. Here, the extraction was performed using 1-methylnaphthalene as a solvent at 350°C.

2.6. Preparation and Evaluation of Coke

Approximately 2 g of coking coal or treated coal was charged into a stainless steel tube with an inner diameter of 16.8 mm, heated to 900°C at 10°C/min in nitrogen, and held for 30 minutes to prepare the coke.¹²⁾ Indirect tensile strength tests (AGS-10kNJ, Shimadzu) were performed on the resulting cylindrical coke to evaluate the strength of the coke.

3. Results and Discussion

3.1. Formic Acid Aqueous Solution Treatment

Table 2 shows the yields of treated coals when A coal or B coal was treated with formic acid aqueous solution at 60°C. The yields of treated coals were almost 100%. Figure 1 shows the thermomechanical analysis curves of the raw coal and formic acid aqueous solution treated coal. The rod position changed more significantly for the formic acid-treated coals than for the raw coals at around 400°C, where softening and melting occurs, indicating that the treatment clearly improved the softening and melting properties.

Table 2. Yields of coals treated in formic acid aqueous solution (HCOOH aq.) or hydrochloric acid (HCl aq.).

Coal	Yield [kg/kg, d.b.]
	HCOOH aq.	HCl aq.
A	0.955	0.976
B	0.961	0.984

Fig. 1.

Thermomechanical analysis curves of the raw coals and coals treated in formic acid aqueous solution (HCOOH aq.). (Online version in color.)

Figure 2 shows the indirect tensile strength of coke prepared from raw coal and formic acid aqueous solution treated coal. The strength of coke increased after the treatment with formic acid aqueous solution regardless of the coal type, indicating the effectiveness of formic acid aqueous solution treatment. The strength-enhancing effect of the coke in the proposed method could be attributed to the improved softening and melting properties mentioned earlier. Since formic acid is both a reducing agent and an acid, it is possible that the strength was increased by acid-induced demineralization. Therefore, we performed a similar treatment with hydrochloric acid with the same acidity as the formic acid solution used and compared the strength of the coke. Figure 2 shows that the strength increased more when treated with formic acid solution than when treated with hydrochloric acid, confirming the effectiveness of the reducing effect of formic acid.

Fig. 2.

Strength of coke prepared from coals treated in formic acid aqueous solution (HCOOH aq.).

3.2. Formic Acid Vapor Treatment

Table 3 shows the yields of treated coals when A or B coal was treated with formic acid vapor at 60°C for 2 or 6 hours. As in the case of the formic acid aqueous solution treatment, the yield of the treated coals was almost 100%.

Table 3. Yields of coals treated by formic acid vapor.

Coal	Yield [kg/kg, d.b.]
	2 h	6 h
A	0.992	1.083
B	0.978	0.995

Figure 3 shows the thermomechanical analysis curves of the raw coal and formic acid vapor treated coal. For the 2 h treated A coal and the 2 h and 6 h treated B coals, the rod position changed more significantly for the formic acid vapor treated coals than for the raw coals at around 400°C, where softening and melting occurs, indicating that the treatment clearly improved the softening and melting properties. For the 6 h treated A coal, unlike the other treated coals, a rod drop was observed by 200°C, and the drop around 400°C was not much different from that of the raw coal. Table 3 shows that only in the case of the 6 h treatment of A coal, the yield of the treated coal was slightly higher than 1. This suggests that there may have been physical or chemical adsorption of formic acid on the coal under this treatment condition, and that the rod drop to 200°C in TMA may have occurred as the adsorbed formic acid was desorbed. This adsorption and desorption of formic acid may have negatively affected the softening and melting properties of A coal. This behavior was not observed in the 2 h treatment of A coal and the 2 h and 6 h treatments of B coal, which have improved softening and melting properties of the both coals.

Fig. 3.

Thermomechanical analysis curves of the raw coals and coals treated by formic acid vapor. (Online version in color.)

Figure 4 shows the indirect tensile strength of coke prepared from the raw coal and formic acid vapor treated coals. Even with formic acid vapor treatment, which is a simpler process than the wet treatment, the strength of coke increased regardless of coal type, indicating the effectiveness of formic acid vapor treatment. For B coal, coke strength increased with treatment time, and after 6 h of treatment, the strength was more than six times higher than that of the raw coal coke. For A coal, treatment for 2 h increased the strength to 5 times that of the raw coal coke, but when the treatment time was extended to 6 h, the coke strength decreased slightly. These results are in good agreement with the results of the thermomechanical analysis, which showed that the softening and melting properties were improved for the treated coals other than the 6 h treated A coal.

Fig. 4.

Strength of coke prepared from coals treated by formic acid vapor (HCOOH (g)).

3.3. Discussion on the Structural Change of Coals through the Treatment

Next, to examine the changes that occurred in the coal as a result of the treatment, elemental compositions were analyzed and the results are shown in Table 4. Formic acid solution treatment slightly increased the carbon content in both coals, but there was no significant difference in elemental composition before and after the treatment.

Table 4. Elemental compositions of raw coals and coals treated in formic acid solution.

Coal	Ultimate analysis [mass%, d.a.f]				Atomic Ratio [−]		Ash [mass%, d.b.]
	C	H	N	O (diff.)	H/C	O/C
A raw	79.8	5.6	1.3	13.3	0.83	0.13	14.8
A treated	80.4	5.6	1.3	12.7	0.83	0.12	7.5
B raw	83.2	5.0	1.0	10.9	0.71	0.10	8.8
B treated	83.9	5.1	1.0	10.0	0.73	0.09	7.3

Figure 5 shows the solvent extraction results for the raw coals and 2 h formic acid vapor treated coals. The yield of extractable fractions (soluble and deposit) of the raw A coal was 35%, while the formic acid vapor treatment increased the extractable fractions to 39%. The yield of extractable fractions of the raw B coal was 36%, but the extractable fractions increased significantly to 46% after formic acid vapor treatment. These results are consistent with the TMA results that the 2 h formic acid vapor treatment improved the softening and melting properties of the coals.

Fig. 5.

Yield of each fraction obtained through solvent extraction from the raw coals and 2 h formic acid vapor treated coals.

Figure 6 shows the change over time of formed CO₂ amount in the formic acid vapor treatment. Since the amount of CO₂ produced was negligible when the control experiment using only coal or only formic acid was performed, it can be said that the CO₂ produced during the treatment was due to the reaction between coal and formic acid.

Fig. 6.

CO₂ amount formed in the formic acid vapor treatment. (Online version in color.)

When formic acid reacts as a reducing agent, it produces CO₂ as shown in Eq. (1),   

$$\mathrm{H}\mathrm{C}\mathrm{O}\mathrm{O}\mathrm{H}\to \mathrm{C}{\mathrm{O}}_2+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}$$(1)

Therefore, considering that the CO₂ produced during the treatment was derived from formic acid, the amount of formic acid reacted during the 2 h treatment was as little as 0.23 mmol/g-coal for A coal and 0.18 mmol/g-coal for B coal. Even if formic acid-derived H is added to the coal according to Eq. (1), the hydrogen content increases by only about 0.4 mmol/g-coal, or about 0.04%, and it is difficult to detect this with the accuracy of elemental analysis.

The fact that the treatment increased the extraction yield of coal despite the small amount of reduction by formic acid is verified below by rough estimation. Assuming an average molecular weight of 500 g/mol of coal, the number of molecules per gram of coal is 2 mmol. If the amount of CO₂ formed from formic acid is 0.2 mmol/g-coal, the H radical donation by Eq. (1) is 0.4 mmol/g-coal, which stabilizes a number of radicals equivalent to 20% of the 2 mmol molecules per 1 g of coal. Since one cross-link is formed by two radicals, the number of cross-links suppressed is 10% of the total number of molecules in this case.

Cross-link formation may occur in the following cases:

1. Cross-linking between two low molecular weight compounds, resulting in the disappearance of two low molecular weight compounds.

2. Cross-linking between low and high molecular weight compounds, resulting in the loss of one low molecular weight compound.

3. Cross-linking between two high molecular weight compounds, or intramolecular cross-linking, with no loss of low molecular weight compounds.

If the average number of low molecular weight compounds lost through crosslink formation is hypothetically one per crosslink, the treatment can increase the number of low molecular weight compounds by 10%. Based on the above estimates, it can be said that even a reduction with formic acid of about 0.2 mmol/g-coal is sufficient to increase the extraction yield of coal.

3.4. Application of the Proposed Method to Weathered Coking Coal

The applicability of the proposed method to weathered coking coal was examined. Simulated weathered coal was prepared by oxidizing coking coal C at 130°C for 90 minutes under an air stream, and the simulated weathered coal was treated with formic acid vapor at 60°C for 2 hours. As shown in Fig. 6, the amount of CO₂ produced during formic acid vapor treatment was 0.26 mmol/g-coal, similar to that produced during the treatment of A and B coals. Figure 7 shows the strength of coke made from the raw C coal, simulated weathered coal, and simulated weathered coal treated with formic acid vapor, respectively. Although weathering reduced the strength of the coke by about half, formic acid vapor treatment restored the strength to the same level as that of the raw coal coke. The proposed method was also shown to be effective in reviving weathered coking coal.

Fig. 7.

Strength of coke made from the raw C coal, simulated weathered coal, or simulated weathered coal treated with formic acid vapor.

Figure 8 shows the thermomechanical analysis curves of the raw C coal, simulated weathered coal, and simulated weathered coal treated with formic acid vapor. Weathering reduced the rod drop and softening and melting properties, but formic acid vapor treatment restored the rod drop to the same level as that of the original coal. These results correspond well with the coke strength results shown in Fig. 7.

Fig. 8.

Thermomechanical analysis curves of the raw C coal, simulated weathered coal, and simulated weathered coal treated with formic acid vapor. (Online version in color.)

4. Conclusions

As a method for reforming low-grade coal into a coke feedstock comparable to coking coal, we have proposed a pretreatment method for reduction using formic acid under mild conditions of about 60°C, and demonstrated its effectiveness. The amount of reduction in the treatment was so slight as to have little effect on the elemental composition of the coal, yet it was found to improve softening and melting properties and increase coke strength. The proposed method was also effective in reviving weathered coking coal.

This study has been carried out as the national contract research project named “CO₂ Ultimate Reduction System for Cool Earth 50 ‘COURSE50” by the New Energy and Industrial Technology Development Organization (NEDO). The authors are grateful to NEDO for their assistance.

References

• 1)   S. H.  Krishnan,  P. S.  Dash,  M.  Guha,  D.  Kumar and  D. P.  Deshpande: ISIJ Int., 44 (2004), 1150. https://doi.org/10.2355/isijinternational.44.1150
• 2)   Y.  Miura and  T.  Okuhara: Trans. Iron Steel Inst. Jpn., 23 (1983), 700. https://doi.org/10.2355/isijinternational1966.23.700
• 3)   S.  Nomura: Fuel Process. Technol., 159 (2017), 1. https://doi.org/10.1016/j.fuproc.2017.01.016
• 4)   M.  Rahman,  D.  Pudasainee and  R.  Gupta: Fuel Process. Technol., 158 (2017), 35. https://doi.org/10.1016/j.fuproc.2016.12.010
• 5)   L.  Petrakis and  D. W.  Grandy: Free Radicals in Coals and Synthetic Fuels, Elsevier, New York, (1983), 19.
• 6)   T.  Ikoma,  O.  Ito,  S.  Tero-Kubota and  K.  Akiyama: Energy Fuels, 12 (1998), 1363. https://doi.org/10.1021/ef980130h
• 7)   J. W.  Larsen,  H.  Parikh and  D. C.  Doetschman: Energy Fuels, 15 (2001), 1225. https://doi.org/10.1021/ef010104a
• 8)   J. W.  Larsen,  R.  Ashida and  D. C.  Doetschman: Energy Fuels, 19 (2005), 2216. https://doi.org/10.1021/ef0501288
• 9)   J. W.  Larsen,  R.  Ashida,  P.  Painter and  D. C.  Doetschman: Energy Fuels, 20 (2006), 1220. https://doi.org/10.1021/ef050368n
• 10)   R.  Ashida,  K.  Nakagawa,  M.  Oga,  H.  Nakagawa and  K.  Miura: Fuel, 87 (2008), 576. https://doi.org/10.1016/j.fuel.2007.02.035
• 11)   R.  Ashida,  A.  Hagimoto,  K.  Miura,  N.  Okuyama,  M.  Hamaguchi,  T.  Shishido and  N.  Komatsu: J. Jpn. Inst. Energy, 91 (2012), 48 (in Japanese). https://doi.org/10.3775/jie.91.48
• 12)   R.  Ashida,  K.  Iwase and  K.  Miura: ISIJ Int., 57 (2017), 37. https://doi.org/10.2355/isijinternational.ISIJINT-2016-276