Effect of Biocarbon Addition on Metallurgical Properties of Mill Scale-Based Auger Pressing Briquettes

Abstract

This work focused on the usage of bio-based and secondary iron and steelmaking raw materials. Auger pressing briquettes, cold-bonded agglomerates made from by-products, mainly mill scale (80%), were successfully tested in industrial-scale blast furnace (BF) trials. The briquettes from industrial production were studied in two different laboratory-scale reduction experiments to compare them to laboratory-made briquettes. A blast furnace simulator (BFS) device was utilized in the simulations mimicking the temperature and gas profiles of an actual BF process. A reduction under load (RUL) device enabled the simulation of the physical load under reducing conditions. To determine how the high-temperature properties of the self-reducing briquettes depend on the amount of biocarbon, a bio-based reducing agent (2–10%) was added to the laboratory-scale briquettes that already contained 5.6% total carbon mainly originating from BF dust. For the recipes studied, a weight loss of about 30% under reducing conditions leads to the disintegration of a briquette under load. Based on the BFS experiments, adding biocarbon to the recipe was profitable in terms of a self-reducing effect up to 6% when the total carbon content was 11%. The RUL experiments showed that the structure of the briquette became extremely plastic with the addition of 4% biocarbon which covers 39% of the total carbon contained in the briquette. This was the upper limit for biocarbon addition due to telescoping and disintegration followed by the formation of fines. The industrial briquettes used in the BF corresponded well to the laboratory-made briquettes in terms of metallurgical properties.

1. Introduction

The global need for steel is being met with increasing production. In 2022, 1885.4 million tons of crude steel were produced, which means a 20.6% increase when compared to 2012 rates.¹⁾ At the same time, a lot of work needs to be done in the ironmaking and steelmaking production chain to mitigate the production of CO₂ emissions and improve circular economy actions. Despite the goals to switch to hydrogen-based steelmaking and at the same time increase the share of electric arc furnace processes, blast furnace – basic oxygen furnace (BF-BOF) steelmaking is still the most used process route.²⁾ One of the essential ways to reduce CO₂ emissions in the BF ironmaking process is to use bio-based materials to replace fossil-based carbon which acts as a reducing agent. Biocarbon has been used in several cokemaking and sintering applications, and in BF operation it has the potential to be utilized instead of pulverized carbon injection (PCI).^3,4) Nevertheless, the properties of biocarbon have been studied substantially less from the perspective of briquetting.⁵⁾ In briquetting solutions, the focus is on using secondary raw materials with ferrous and carbonaceous fines. Iron and steelmaking by-products such as dusts, sludges, scales, and slags can be recycled as briquettes which enables utilizing the iron and carbon contents of these products. Various binders are added to the mixture of raw materials to create a structure that withstands transport, storage and charging into the furnace. Furthermore, it is essential to prevent the generation of fines in the high-temperature zones of the process.⁶⁾

The metallurgical properties of biocarbon are essential when it is utilized to substitute a fossil-based material. It has been concluded that biomass can be partly used in coal blends without losing mechanical strength. However, the share of substitutable coal has been 2–10%. In their recent article, Nonaka et al.⁷⁾ used kraft lignin as a binder in coke and found that the tensile strength of coke reached a maximum at 10% lignin content. From a metallurgical perspective, the main concerns are the lower CSR value (coke strength after reduction) and higher CRI value (coke reactivity index).^4,8) Mousa et al.⁵⁾ studied biocarbon briquettes consisting of charcoal fines (62%), molasses (20%), hydrated lime (10%) and cement (8%). They concluded that briquette strength decreases as the biocarbon content is increased. The hot strength under mechanical load was tested and it was found that cracks appeared which led to the disintegration of the briquette. Tang et al.⁹⁾ carried out a numerical simulation of a BF with a maximum charge of 60% biochar composite briquettes which contained bio-based carbon (11.1%), magnetite (72.7%), wüstite (11.25%), metallic iron (0.77%) and gangue (4.67%). They concluded that the optimal charging ratio was 40%.

This work focuses on utilizing secondary raw materials and bio-based reducing agent in BF briquettes. Both industrial and laboratory-scale briquettes were tested in the experiments. Varying amounts of biocarbon were used in the laboratory-scale briquette recipes. In our previous work,¹⁰⁾ the examination of the high-temperature properties reducibility, swelling and cracking, of laboratory-scale cold-bonded auger pressing briquettes was implemented by simulating BF conditions in a laboratory-scale furnace, aiming to achieve different reduction degrees of iron. In this work, the amount of the main raw material of the briquette, mill scale, was gradually decreased in the briquette recipes while the amount of biocarbon was increased. Along with simulating the temperatures and gas atmosphere of an actual BF process, a novel method was used for simulating the physical load caused by coke and other charge materials on the briquettes fed into the BF process. The reducibility was evaluated based on the briquette weight loss in reduction experiments and swelling as a relative change in the external dimensions of the briquette. In addition to swelling, possible cracking and disintegration behavior was studied by real-time monitoring using a video camera. In the experiments in which a physical load was simulated, the compressibility of a briquette was evaluated as the amount of telescoping in relation to the height.

2. Experimental

The experimental part of this study can be divided into two parts. The production of mill scale-based auger pressing briquettes on both industrial and laboratory scale was studied by carrying out tests measuring metallurgical and mechanical properties of the briquettes. The investigation of industrial mill scale briquettes was followed by the experiments on laboratory-scale briquettes in which biocarbon was used as a reducing agent. The main raw materials, mill scales, used in these two batches of briquettes were of different quality. The main production stages and pictures of industrial and laboratory-scale briquettes are seen in Fig. 1.

Fig. 1. Stages of auger pressing briquette production and research, (a) process scheme, (b) finished industrial briquettes and (c) finished laboratory-made briquettes. (Online version in color.)

2.1. Industrial Auger Pressing Briquettes

A total of 2000 tons of mill scale-based industrial briquettes with a diameter of 40 mm were produced using auger pressing technology. The agglomeration method introduced in our previous work¹⁰⁾ is one of the key cold-bonding briquetting techniques, along with vibro pressing and roller pressing. In this type of agglomeration, the briquettes are cylindrical and cut to their final length by gravity. The industrial mill scale auger pressing briquette (IMAPB) tested in laboratory experiments consisted of mill scale (80%), BF dust (18%), bentonite (1.5%) and AMCOM’s binder (0.5%). The binder was specially developed to ensure the functionality of the briquettes in the customer’s BF operation. A higher mill scale content was tested during the briquette production, but it turned out to be an unsuitable solution due to damage to the equipment lining caused by abrasiveness of the mill scale. The chemical composition of the IMAPB sample was analyzed using an X-ray fluorescence (XRF) analysis and the amount of carbon was examined using an LECO elemental analyzer. The chemical composition of the briquette sample is presented in Table 1.

Table 1. Mass fractions (wt.-%) of compounds in industrial mill scale auger pressing briquettes (IMAPB) based on XRF and LECO analyses.

Fetot	Fe2O3	P	SiO2	CaO	MgO	Al2O3	Zn	C	Basicity B2 (CaO/SiO2)
53.2	76.1	0.02	13	8	1.5	3.2	0.1	7.5	0.62

An X-ray diffraction (XRD) analysis was used to evaluate the main phases to determine the oxidation states of iron. It was determined that magnetite (30.7%), wüstite (33.2%) and hematite (21.0%) were all present in the sample. The chemical composition for the general raw material of IMAPB, Mill Scale #1, is presented in Table 2. As seen from the analysis, the scale had relatively high gangue (SiO₂, CaO, MgO and Al₂O₃) content. The total iron content in the finished briquette was 53.2%, but the proportion of iron oxides was high. In addition, there was some variation in the mill scale particle size. The particle size distribution of the mill scale seen in Fig. 2 shows that the particle size was generally less than 1 mm. However, metallic iron particles up to 8 mm in diameter were present. The presence of metallic iron is an advantage for BF briquettes, because it no longer needs to be reduced.

Table 2. Mass fractions (wt.-%) of compounds in the Mill Scale #1 used in IMAPB production.

Fetot	Femet	P	SiO2	CaO	MgO	Al2O3	Zn
59	2.1	<0.01	4.4	6	0.5	1.1	<0.1

Fig. 2. Particle size distribution of Mill Scale #1 used in IMAPB production.

The briquettes depicted in Fig. 1(b) were tested in industrial blast furnace (1700 m³) trials at Azovstal Iron & Steel Works in Mariupol, in Ukraine. The briquettes were charged into the BF with an iron burden consisting of lump ore (0.19%), sinter (81.84%) and iron ore pellets (17.97%). An FeO content of 55.9% was reached by charging 5% extra mill scale-based briquettes. The approximate coke consumption was 382 kg/tHM and a pulverized coal injection of 110 kg/tHM. The trials were carried out at the beginning of 2022 with a campaign length of 38 days. There were no operational deviations during the trials. The BF exhaust gases contained 23.3% of CO, 21.9% of CO₂ and 5.8% of H₂, with a calorific value of 794 kcal. The company used these briquettes for washing the BF hearth and based on the results of industrial testing, it was decided to continue the production of industrial briquettes. To study the metallurgical properties, IMAPB samples were subjected to laboratory-scale reduction experiments using the same equipment as in the case of the biocarbon briquettes which are described in the next chapter. The results obtained with the IMAPB samples were compared to the briquettes produced in the laboratory.

2.2. Biocarbon Briquette Materials and Cold Strength Testing

The cold-bonded biocarbon briquettes studied in this work were manufactured in smaller batches under laboratory conditions using auger pressing equipment presented in the previous studies.^10,11,12) The briquette production procedure demonstrated in Fig. 1(a) consisted of mixing the dry charge, adding a binder, adding water, aging the charge, preliminary and final auger pressing, and drying the finished briquettes. The mechanical strength tests were performed at AMCOM’s laboratory, and the metallurgical tests were performed at the University of Oulu.

The briquettes consisted of two secondary raw materials originating from iron and steelmaking processes: mill scale formed in the hot rolling process of steel and BF dust which is generated from the top gas of a BF. Additives and binders included biocarbon or charcoal, slaked lime, bentonite, and a binder consisting of both inorganic and organic materials. Six mill scale-based auger pressing briquette recipes called MAPB-X, where X refers to the amount of added charcoal (wt.-%), contained varying amounts of mill scale (69.2–79.2%) and charcoal (0–10%). The amounts of BF dust (15%), slaked lime (3.5%), bentonite (1.7%) and binder (0.6%) were kept constant for comparability. Table 3 shows the share of the raw materials in each recipe.

Table 3. Relative proportions of raw materials in the studied mill scale auger pressing briquettes (MAPB).

Briquette	Share of constituents, wt.-%
	Charcoal	Mill scale	BF dust	Slaked lime	Bentonite	Binder
MAPB-0	0	79.2	15	3.5	1.7	0.6
MAPB-2	2	77.2	15	3.5	1.7	0.6
MAPB-4	4	75.2	15	3.5	1.7	0.6
MAPB-6	6	73.2	15	3.5	1.7	0.6
MAPB-8	8	71.2	15	3.5	1.7	0.6
MAPB-10	10	69.2	15	3.5	1.7	0.6

The mill scale used as the main raw material for the briquettes originated from Eastern Europe and was of higher quality than the one used in the IMAPB samples. The total iron content of the mill scale was high while the gangue content was significantly lower, which can be noticed in the final composition of the briquette. The mill scale, BF dust, slaked lime and bentonite samples were analyzed using XRF analysis and the oxidation states of the mill scale were studied using an XRD analysis. Mill scale #2 was also analyzed using ICP-OES. The chemical composition shown in Table 4 and the XRD result presented in Fig. 3 indicate that the mill scale generally consisted of oxidic iron.

Table 4. Chemical analysis (wt.-%) of Mill Scale #2 used as a raw material.

Fetot	S	P	SiO2	CaO	MgO	Al2O3	Zn	Mn
74.7	0.02	0.05	0.5	0.12	0.1	0.06	0.01	1.1

Fig. 3. The XRD pattern of Mill Scale #2 showing peaks for wüstite (FeO), magnetite (Fe₃O₄) and hematite (Fe₂O₃).

Based on the XRD analysis, wüstite (72.6%), magnetite (22.6%) and hematite (2.7%) were all found to be present in the mill scale sample. Most of the mill scale consisted of wüstite, and only a minor part was hematite.

BF dust is a side stream generated in BF top gas. The fine BF dust fraction (<5 mm) needs processing before recycling¹³⁾ and part of the wet fraction of BF dust, i.e., BF sludge is still landfilled due to its zinc content. The chemical analysis of the BF dust is shown in Table 5.

Table 5. Chemical analysis (wt.-%) of the BF dust used as a raw material.

Fetot	S	P	SiO2	CaO	MgO	Al2O3	Zn	C
26.7	0.69	0.04	7.65	4.95	0.91	2.44	0.33	35

Based on the chemical analyses of the raw materials, the chemical compositions were calculated for the finished briquettes. The chemical compositions of the briquettes are presented in Table 6. As seen from the table, MAPB-0 contained 5.6% total carbon derived from BF dust.

Table 6. Calculated total chemical compositions for briquette qualities MAPB-X, where X denotes the share of added biocarbon (wt.-%), and the calculated CaO/SiO₂ basicity values.

Briquette	Share of elements, wt.-%									Basicity B2 (CaO/SiO2)
	Fetot	S	P	SiO2	CaO	MgO	Al2O3	Zn	С
MAPB-0	63.2	0.12	0.04	2.5	3.2	0.3	0.7	0.0	5.6	1.30
MAPB-2	61.7	0.12	0.04	2.5	3.2	0.3	0.7	0.05	7.4	1.31
MAPB-4	60.3	0.12	0.04	2.5	3.2	0.3	0.7	0.05	9.2	1.32
MAPB-6	58.8	0.12	0.04	2.4	3.3	0.3	0.7	0.05	11.0	1.33
MAPB-8	57.3	0.12	0.04	2.4	3.3	0.3	0.7	0.05	12.8	1.35
MAPB-10	55.8	0.12	0.04	2.4	3.3	0.3	0.7	0.05	14.6	1.36

The amount of water added was increased along with fine charcoal as seen in Table 7. This is due to the ability of the fine material to absorb more water. The briquettes were cured for 72 h after manufacturing.

Table 7. Amount of water added in the manufacturing process of the MAPB samples and the moisture contents after 72 h of curing.

Sample	Added water (wt.-%)	Moisture (wt.-%) after 72 h
MAPB-0	4.5	2.36
MAPB-2	5.5	2.51
MAPB-4	6	2.88
MAPB-6	7	2.63
MAPB-8	7.5	2.83
MAPB-10	9	2.84

The cold strength properties of the cured briquettes were tested using mechanical crushing strength, drop strength and abrasion strength tests in the same way as in our previous work.¹⁰⁾ The mechanical crushing strength tests were carried out for 7 pieces of briquettes of each quality by crushing along the axis. The drop strength was tested by dropping 3±0.15 kg sets of briquettes three times using a drop height of 2 m. The abrasion strength was tested for 13–16 kg sets of briquettes in a tumble drum as defined in ISO 3271:2015.¹⁴⁾

2.3. Blast Furnace Atmosphere Simulation

The Blast Furnace Simulator (BFS) is a high-temperature tube furnace which has been used in numerous previous works to study changes during reduction for iron ore pellets, coke, sinter and lump ore^{15,16,17,18,19,20,21)} as well as briquettes,^10,22,23) and was utilized in the BF atmosphere simulation phase of this work. The BFS is made of heat-resistant steel and was capable of simulating actual BF process up to 1100°C. Both isothermal and dynamic tests were possible. The weight of the basket containing the sample was continuously measured at desired intervals using the thermal gravimetric analysis (TGA).

In this study, an atmosphere with N₂, CO, and CO₂ gases was used in dynamic BFS reduction experiments. The sample weight was recorded at 10-second intervals. The reduction program was the same as in the 280-minute experiment performed in our previous work to form metallic iron,¹⁰⁾ which included a 40-minute isotherm at 1100°C. The gas compositions throughout the experiments are shown in Fig. 4.

Fig. 4. The BFS reduction program showing the gas atmosphere and temperature profiles as a function of time. (Online version in color.)

2.4. Reduction Under Load Simulation

The Reduction Under Load (RUL) device is a novel laboratory-scale tool that enables to study the agglomerate strength during reduction and gasification reactions in the BF process. It includes a custom-made cylinder, a baseplate and a weight of 25 kg which is placed on top of the sample. The weight is utilized to simulate the pressure of the ferrous burden in BF conditions. Since the weight is a separate part of the device and stands freely on top of the sample, the pressure applied to the sample only depends on the cross-sectional area of the briquettes. Hence, the test does not comply with the reduction under load test introduced in ISO 7992:2015.²⁴⁾ In the standardized test, it is advised to apply a load of 50 kPa ± 2 kPa. For the present non-standardized RUL device, the pressure exerted by the weight on the cylindrical briquettes can be calculated simply as follows:

  

$${\text{P}}_{\text{weight}}\left(\text{kPa}\right)=\frac{\text{F}}{\text{A}}=\frac{{\text{m}}_{\text{weight}}\times \text{g}}{{\sum }_{\text{i}=1}^{\text{n}}\frac{\pi }{4}{\text{d}}_{\text{i}}^2}$$(1)

where P_weight is the total pressure on the briquette, m_weight is the mass of the weight, g is the gravity of Earth, n is the number of samples and d is the diameter of the cylindrical briquette.

To enable heating, the cylinder was placed inside an Entech model ETF 75/17 V gradient furnace manufactured by Entech AB Ängelholm, Sweden. Koskela et al.²⁵⁾ have previously used this furnace to study interaction between coal and lignin briquettes. The furnace has two separately controllable zones, an upper and lower zone. It is possible to heat the furnace up to 1000°C. The three points measured with thermocouples were located in the upper zone, lower zone and inside the cylinder in the immediate vicinity of the briquette sample. The operating principle of the RUL equipment placed inside the furnace is illustrated in Fig. 5.

Fig. 5. General principle of the Reduction Under Load (RUL) device inside a gradient furnace. (Online version in color.)

However, unlike in the case of BFS, the gas compositions used in the gradient furnace were not linearly adjustable with increasing temperature. Thus, the desired gas compositions had to be set manually in segments. A program including segments with different N₂–CO–CO₂-atmospheres was created based on the gas compositions used in BFS experiments. The program is shown in Fig. 6.

Fig. 6. The RUL program with five different segments using an N₂–CO–CO₂ atmosphere.

Due to the segmented nature of the experiment, the gas compositions had to be modified in relation to the program used in the BFS experiments. The Fe–O–CO–CO₂ phase diagram in Fig. 7 is based on data from the HSC Chemistry thermodynamic calculation software. The diagram shows the conditions needed for the stable phase of iron and how they are reached using the current RUL and BFS reduction programs in the temperature range 400–1000°C. The picture shows how reduction takes place first from hematite (Fe₂O₃) to magnetite (Fe₃O₄) above 500°C, then to wüstite (FeO) at around 800°C and finally to metallic iron (Fe) at a temperature above 900°C. As it can be seen, there were differences in reducing conditions between the two furnaces.

Fig. 7. A Bauer-Glaessner diagram indicating the differences between reducing conditions in the Reduction Under Load (RUL) and Blast Furnace Simulator (BFS) equipment.

Due to the structure of the gradient furnace, the lengths of the heating segments varied slightly. Each segment ended after reaching the target temperatures of the segments, i.e., 500, 650, 800, 900 and 1000°C. After terminating the reduction program, the sample was cooled overnight using N₂ gas with a volume flow of 3 l/min. The heating rates and estimated lengths of the segments are listed in Table 8.

Table 8. Estimated segment lengths (min) and heating rates (°C/min) in the RUL experiments.

Segment	Estimated length (min)	Temperature (°C)	Heating rate (°C/min)
1	60	25 → 500	7.9
2	30	500 → 650	5.0
3	30	650 → 800	5.0
4	30	800 → 900	3.3
5	30	900 → 1000	3.3

2.5. Sample Preparation and Test Set Up

The weights of the raw biocarbon briquettes varied between 228–286 g and the lengths between 106–132 mm. The diameter of the briquettes was 30 mm. Three samples of each briquette quality were prepared for each RUL experiment due to the nature of the test, and one sample for each BFS test. To perform RUL experiments stably and reliably and to accurately measure the swelling properties in the BFS experiments, the briquettes were cut into flat-ended pieces with a length of 40 ± 1 mm. A uniform length for the cylindrical briquette samples was important especially for the RUL experiments.

The cut biocarbon briquettes were weighed, measured, and kept in a temperature cabinet overnight at 105°C to remove moisture. The moisture content of the samples varied between 0.4–0.9%. The briquettes were placed differently in the experiments: in the RUL experiments, the briquettes were placed in a vertical position. In the BFS experiments, the briquettes were placed on their side in the sample basket in order to detect possible cracking during the test. The swelling was also more noticeable from this angle.

In the RUL experiments, the biocarbon briquettes were placed in a triangular arrangement on the baseplate. Figure 8 shows the principal view of the three briquette samples A, B and C and the thermocouple placed between them, as well as the samples in the RUL capsule pictured before and after placing the weight. It is notable that testing the industrial mill scale briquettes differed from this: only one piece of industrial briquette was used in the RUL experiment due to the small number of samples.

Fig. 8. Illustration and photos of placing the biocarbon briquettes and weight on the baseplate inside the RUL cylinder. (Online version in color.)

Using this RUL test arrangement and by calculating using Eq. (1), the result for the load on the biocarbon briquette was approximately 57.8 kPa. In the case of industrial briquettes, using one sample with a diameter of 40 mm gave a higher result for the load due to the smaller surface area. The load for an industrial briquette in the RUL experiment was approximately 97.5 kPa. These results are therefore not directly comparable.

2.6. Weight Loss Calculation and External Changes Observation

This work focused on the examination of reducibility, swelling and cracking which are the typical metallurgical properties of iron ore pellets. The samples studied, however, were briquettes. When compared to pellets, briquettes are of different sizes, shapes and contain carbon. It should be noted that the reduction degree cannot be determined due to carbon gasification which contributes to weight loss.

The reducibility was examined based on the weight loss curves obtained from the BFS experiments as well as using mineralogical characterization, which we will return to in the next chapter. However, the actual weight losses were obtained as the difference between the weighing results before and after the experiments in both simulations. Sample disintegration during some of the RUL experiments made it impossible to weigh all of the samples after the experiment.

Swelling was determined by measuring the briquette length and diameters at three points before and after the reduction experiments and thus calculating the change in volume. In this way, the swelling index (%) expressed in the previous study was obtained. Swelling was also monitored in real time using a camera during the BFS experiments. In addition to the swelling, the temperature at which cracking occurred was determined.

It was also necessary to study the plasticity of the briquettes during the RUL experiments. In this study, telescoping (%) indicates the relative decrease in the length of the briquette, and the briquette length after the experiment is compared to the original length of the briquette. Unlike in the case of the swelling index which considers the change in volume, only the change in length was examined here. The telescoping rate was not defined for the IMAPB since the results were not comparable to the MAPB samples.

2.7. Mineralogical Characterization

In order to detect reduction and other possible reactions, the samples were examined using a Digital Olympus DSX1000 light optical microscope (LOM) and Zeiss ULTRA Plus field emission scanning electron microscope (FESEM) with an energy-dispersive X-ray spectroscopy (EDS) elemental analyzer using backscattered images. Mill Scale #2 and BF dust were analyzed using X-ray fluorescence (XRF), X-ray diffraction (XRD) and inductively coupled plasma - optical emission spectrometry (ICP-OES) methods. The carbon content of the BF dust was analyzed using combustion analysis (LECO).

3. Results and Discussion

The experimental results mainly consider the MAPB samples that could be analyzed in more detail. The corresponding results for the IMAPB samples from each test are presented for the results that were comparable.

3.1. Cold Strength Tests

The cold strength test results for the crushing, drop and abrasion strength of all the MAPB and IMAPB samples were compared. The crushing strength test results are shown in Table 9.

Table 9. Crushing strength test results (kg/cm²) for the briquette samples. The results presented are sample averages. SD stands for standard deviation.

Briq.	Diameter (mm)	Mass (g)	Length (mm)	Density (g/cm3)	Compression force (kg)	Strength (kg/cm2)	Strength SD
MAPB-0	30	245.1	102.3	3.44	100.7	0.915	0.2
MAPB-2	30	255.1	108.7	3.32	106.6	0.917	0.1
MAPB-4	30	242.3	106.1	3.23	154.6	1.355	0.1
MAPB-6	30	244.3	113.6	3.04	190.6	1.575	0.1
MAPB-8	30	203.6	96.4	2.99	168.3	1.613	0.2
MAPB-10	30	207.5	107.9	2.72	235.0	2.045	0.2
IMAPB	40	431.0	97.4	3.52	574.7	3.960	1.3

The results show that the crushing strength of the briquettes increased with the amount of added charcoal. For MAPB-0, the compressive strength was only 0.915 kg/cm² on average, while with a 10% charcoal increase (MAPB-10) it was 2.045 kg/cm² on average. The result for MAPB-0 can be considered weak, while the result for MAPB-10 is promising and almost at the same level as in our previous work (2.172 kg/cm² on average).¹⁰⁾ Based on the results, the larger share of finer particles (charcoal and BF dust) in a MAPB sample, the stronger the briquette is, in terms of the cold crushing strength. It is notable that the crushing strength results for IMAPB were excellent. A comparatively better result is probably due to the absence of slaked lime and possibly higher pressure used during production. The result is significantly higher than in the case of MAPB, although the standard deviation was large.

The drop strength test results for each MAPB quality are shown in Fig. 9. The results are presented as shares in which the particle size exceeds 5 mm after the tests. Drop tests were not performed for IMAPB.

Fig. 9. The drop strength test results showing the portions (%) with a particle size exceeding 5 mm for each briquette quality after the tests.

Based on the drop strength tests, the ability of a briquette to withstand shocks caused by falling decreased slightly with charcoal addition. Among the tested briquettes, the briquette quality with the highest proportion of charcoal, MAPB-10, formed fewer fines than the three briquette qualities MAPB-4, MAPB-6, and MAPB-8 with lower proportions of charcoal. Despite this, the differences were small, and the trend is distinguishable. In the case of each briquette, over 95% of the material had a particle size of more than 5 mm.

The results from the abrasion strength tests are shown in Fig. 10. The results show the share of MAPB samples exceeding the particle sizes of 5 and 0.5 mm when the drum was rotated 25, 50, 100 and 200 revolutions. The results indicate the tendency of the briquette to generate fines. For IMAPB samples, there are results available after 25 and 200 revolutions. Sieves with a diameter of 5 and 0.5 mm were used.

Fig. 10. The abrasion strength test results showing the portions (%) with a particle size exceeding a) 5 mm and b) 0.5 mm for the MAPB and IMAPB briquette qualities after the tests.

Compared to the auger pressing briquettes (APB) studied in our previous work, the abrasion strengths were lower for all the briquettes studied in this work. The briquette which contained the most charcoal, MAPB-10, performed best in the test. The briquette without charcoal, MAPB-0, was the weakest briquette quality in terms of abrasion strength. For an unknown reason, MAPB-6 also differed from the others in terms of the proportion of fines exceeding 5 mm.

There is a clear increase in the compression strength with increasing carbon, but also a tendency to increase fragility and the amount of fines when dropped. The abrasion test did not give a clear correlation with biocarbon content but showed a trend and a significant difference between laboratory and industrial samples.

IMAPB briquettes showed significantly better resistance to abrasion, which may be due to the removal of the slaked lime in the mixture, the larger size of the briquette and the higher pressure of the industrial equipment. With the help of the obtained data, the tendency of the mill scale-based auger pressing briquettes to form fine matter can be evaluated as indicative numerical values.

3.2. Reduction Experiments

Figure 11 presents the final weight losses (%) occurring in the briquette samples during the BFS experiments. These calculations are based on weighing before and after the experiments.

Fig. 11. Relative weight losses (%) of the laboratory-scale auger pressing briquettes (MAPB-X) and industrial-scale briquette (IMAPB) during the BFS experiments.

The weight loss results presented in Fig. 11 show that the weight loss of the briquettes was proportional to the amount of mill scale replaced with charcoal. The weight loss for industrial briquette, IMAPB, was 21.6%. The laboratory-scale briquette without charcoal, MAPB-0, lost 22.3% of its weight during the experiment, while the weight loss for the briquette containing 10% charcoal, MAPB-10, was 34.6%. The increase in the weight loss was quite steady, since the weight loss was 26.0% for MAPB-2, 29.6% for MAPB-4, 31.8% for MAPB-6 and 33.9% for MAPB-8. The TGA data presented as a function of time together with a temperature curve in Fig. 12 gives further information about how the weight losses occurred for each briquette.

Fig. 12. Relative weight changes (wt.-%) of the laboratory-scale auger pressing briquette (MAPB) and industrial-scale briquette (IMAPB) samples in the BFS reduction experiments presented as a function of time using TGA.

The TGA curves show the weight losses which are mainly caused by the reduction of iron and carbon gasification. The weight loss curves are mostly the same shape, and the losses remained low until the temperature of 500°C where the formation of magnetite (Fe₃O₄) by the reduction of hematite (Fe₂O₃) can be expected to occur. At temperatures above 570°C, greater weight losses began to occur. Here, it can be expected that the reduction from magnetite to wüstite (FeO) starts to occur and accelerates as the temperature approaches 800°C. The curve of the non-biocarbon briquette, MAPB-0, clearly differs in shape from those of the other laboratory-scale briquettes above 800°C. The weight loss curves of the other briquettes started to steepen at 850–900°C, while for MAPB-0 this only happened at around 950°C. The differences in the weight losses started to increase at temperatures above 950°C. At these temperatures, wüstite can be expected to start to reduce to metallic iron (Fe). Similar phase transformations from hematite first to magnetite, then to wüstite and further to metallic iron were observed in our previous work, in which they were investigated in interrupted BFS experiments using target temperatures of 500, 800 and 1100°C.¹⁰⁾ During the 40-minute isotherm at 1100°C, further weight losses occurred and about 2.5–4.0% of the total weight was still lost.

The weight loss curve for MAPB-0 is similar in shape to the weight loss curve for IMAPB, but the decrease in weight is clearly slower for MAPB-0, even though the final weight loss result is almost the same. Differences in the curves mainly indicate quality differences and, more precisely, differences in the oxidation levels in the mill scales used in briquette manufacture. The carbon in these two briquettes is practically entirely derived from BF dust.

The formation of metallic iron is often accompanied by sample swelling and cracking. Figure 13 shows the relative volume changes, i.e., swelling indices of the briquettes.

Fig. 13. Swelling indices (%) of the briquette samples (MAPB) during the BFS reduction experiments. No swelling occurred in the case of IMAPB.

Based on the results, the briquette without charcoal swelled slightly (12.7%). The briquette underwent a strong swelling (24.6%) during reduction when only 2 wt.-% of the mill scale was replaced with charcoal. The swelling was greatest (36.7%) in the case of the briquette sample MAPB-4. Thus, the amount of swelling increased when 2–4 wt.-% charcoal was added but already started to decrease (35.0%) when it was increased to 6 wt.-%. When the proportion of charcoal was 8 wt.-%, swelling was almost at the same level (26.6%) as when it was 2 wt.-% (24.6%). With 10 wt.-% charcoal content, the swelling was insignificant (6.4%). It is notable that at this point swelling occurred less than with the briquette without charcoal. The timing of the swelling events can be examined by viewing the video camera image from the BFS. No clear transformations were detected at temperatures below 700°C. Images of each briquette sample at temperatures of 700, 900, 950, 1000 and after isotherm at 1100°C are presented in Fig. 14.

Fig. 14. Briquette samples MAPB-0, MAPB-2, MAPB-4, MAPB-6, MAPB-8 and MAPB-10 pictured during the BFS experiments at the temperatures of 700, 900, 950, 1000, and 1100°C after a 40-minute isotherm. Swelling is illustrated by highlighting the edges of the briquettes and adding the final swelling indices to the images taken slightly before finishing the reduction experiment. The formed cracks are also indicated in the picture. (Online version in color.)

The swelling behavior of the IMAPB samples was different from that of the MAPB samples. The IMAPB sample showed no swelling, but instead shrank up to approximately 10%.

No eye-perceptible swelling occurred below 700°C and it was minor below 900°C. Based on the images taken during the BFS experiments, the strongest swelling took place during the transformation of wüstite to metallic iron above 950°C. In addition, the swelling was more prominent in the width direction. For MAPB-4, the most significant swelling seemed to occur above 1000°C. No cracking occurred in MAPB-0 and MAPB-2. Significant swelling of MAPB-4 and MAPB-6 appeared to lead to the formation of cracks. These cracks can be seen in the BFS camera pictures. MAPB-8 cracked from the bottom and thus the cracks are not visible in the images. However, cracking also occurred in MAPB-10, whose swelling remained remarkably slight. Despite the cracking, each briquette remained intact throughout the experiment. As seen in Fig. 15, the visible changes were swelling, cracking, increased porosity, and the lighting of the sample color.

Fig. 15. A raw briquette and the samples of different MAPB quality pictured after the BFS experiments. (Online version in color.)

The variation in swelling was largely due to the different extent of reduction that took place in the briquettes. It should be noted that carbon gasification also occurred in the experiments, causing an increase in sample weight losses. The differences between the reduction behavior of the briquettes were already manifested by small changes in the amount of charcoal, because the amount of reducible iron decreased continuously as the amount of reducing agent was increased. When the amount of charcoal was increased from 8 to 10 wt.-%, the weight loss no longer increased in the same proportion as when increasing it from 0 to 8 wt.-%. When the total carbon content was low, the reduction rate was lower, swelling was minor, and no cracking occurred in the briquette. From the perspective of reduction, there was excess carbon at the higher carbon concentrations tested. When there was less total iron, significant swelling did not occur, but the structure of the briquette was nevertheless softer, leading to cracking. Higher carbon contents would probably lead to briquette disintegration which did not occur during the BFS experiments carried out in this study.

3.3. Reduction Under Load Experiments

Figure 16 shows the relative weight losses of samples IMAPB, MAPB-0 and MAPB-2 during the RUL experiments. The total weight losses after the RUL experiments have been compared to the weight losses during the BFS experiments at 1000°C. The weighing result of the sample MAPB-4 after the RUL experiment was unreliable and the weights of samples MAPB-6, MAPB-8 and MAPB-10 could not be weighed at all due to disintegration.

Fig. 16. Weight losses of the samples IMAPB, MAPB-0 and MAPB-2 after the RUL experiments presented as columns. The corresponding weight losses during the BFS experiments at 1000°C are presented as lines for comparison.

Based on the total weight losses during the RUL experiments, the reducing conditions of the RUL experiments closely matched the conditions of the BFS experiments. The relative weight loss was 10.7% for the IMAPB, 9.8% for MAPB-0 and 13.11% for MAPB-2 samples. Based on the TGA curves of the BFS experiments, the final weight loss for each briquette was 7–13% greater in the RUL experiment compared to the BFS experiment at a temperature of 1000°C. For the MAPB-4 sample, which disintegrated during the RUL experiment, a weight loss of about 20% was measured. However, the result cannot be considered reliable, and it is higher than expected compared to the corresponding weight loss in the BFS experiment (14%). As seen from the results of the BFS experiments, the final weight loss of MAPB-0 was greater than that of IMAPB, but the weight loss of MAPB-0 exceeded that of IMAPB only during the isotherm. Figure 17 presents the temperature curve of the RUL experiment for MAPB-0 as a function of time.

Fig. 17. The temperature curve and gas compositions of the RUL experiment carried out for MAPB-0 presented as a function of time.

As seen in Fig. 17, the RUL experiment was terminated, and cooling started immediately after reaching 1000°C. The experiment for MAPB-0 lasted for 166.3 min. However, the total experiment times varied between 166–178 min. For this reason, the boundaries of the heating segments would be inaccurate if all the temperature curves were presented in the same figure. The most significant differences in the temperature curves were noticeable in the 5^th heating segment. Figure 18 illustrates how the temperatures measured for MAPB-6 and MAPB-10 suddenly dropped during the heating.

Fig. 18. The temperature drops measured for MAPB-6 and MAPB-10 at 973 and 989°C were the most noticeable differences between the heating curves for the samples during the 5th segment of the RUL experiment.

The heating curves show that the temperature measured for MAPB-6 dropped from 973 to 965°C when the RUL test had been running for 161.8 min. For MAPB-10, a similar phenomenon was observed at 989°C, when the temperature dropped to 977°C after 160.5 min. This observation probably indicates the moment when the briquettes collapsed. As a consequence, the thermocouple placed between the samples moved, causing a delay in reaching the final temperature. At the corresponding temperatures, in the BFS experiments, the weight loss was at its highest. Simultaneously, cracking occurred and in the case of MAPB-6, clearly distinguishable swelling was observed. An image of each briquette after the RUL experiments is shown in Fig. 19.

Fig. 19. The effect of the RUL test on the structure of the briquette samples, (a) raw sample, (b) MAPB-0 after the RUL experiment, (c) MAPB-2 after the RUL experiment, (d) MAPB-4 after the RUL experiment, (e) MAPB-6 got stuck in the equipment after the RUL experiment, (f) MAPB-8 got stuck in the equipment after the RUL experiment and (g) MAPB-10 disintegrated while removing the weight after the RUL experiment. (Online version in color.)

As seen in the pictures, the briquettes sagged under the load and this effect became stronger when increasing the proportion of charcoal in the briquette. The briquettes became sticky when disintegrating and stuck to the baseplate and the weight. Only one out of the three MAPB-4 samples could be removed from the RUL equipment intact. The remaining three briquette qualities (MAPB-6, MAPB-8 and MAPB-10) completely disintegrated when removed. Telescoping which indicates the relative change in height was measured for MAPB-0 and MAPB-2 as the average of the three samples tested, for MAPB-4 based on the intact briquette and for MAPB-6 and MAPB-8 roughly before removing the samples from the RUL equipment. MAPB-6 and MAPB-8 could be lifted to measure the telescoping, but MAPB-10 already disintegrated during lifting. Thus, no result for the telescoping of MAPB-10 was obtained. The results for samples with up to 8% biocarbon content are shown in Fig. 20.

Fig. 20. Telescoping rates for the briquette samples during the RUL experiments.

Based on the results, the briquettes showed plastic behavior even without added biocarbon and collapsed by 7%. MAPB-2 already had clearly visible depressions from the equipment, and the telescoping rate was on average 11.3%. MAPB-4 collapsed by 23%, and the rest of the samples did so by more than 60%. Reliable measurements were not possible for MAPB-4, MAPB-6, MAPB-8 and MAPB-10. Thus, with a charcoal content of more than 2%, the briquettes could not properly withstand the reducing conditions and the load caused by the weight at the same time. Based on the RUL experiments, it can be concluded that the structure of the briquette is very sensitive to changes that increase the carbon content and correspondingly decrease the iron content of the product. IMAPB, which was tested using only one briquette at a time in the experiment, flexed and collapsed only slightly, even when significantly more weight was applied to it than the MAPB samples. Pictures of the IMAPB sample before and after the RUL experiment are presented in Fig. 21.

Fig. 21. The effect of the RUL experiment on the IMAPB sample, (a) the briquette before the experiment, (b) the briquette stuck to the bottom of the weight after the experiment and (c) the bent briquette removed intact from the equipment after the experiment. (Online version in color.)

3.4. Phase Transformations

Differences between the structures of the raw samples of biocarbon briquettes due to the replacement of mill scale with charcoal were observed. The FESEM-EDS images in Fig. 22 show the 100× and 300× magnifications of the microstructure of the raw briquette samples MAPB-0, MAPB-4 and MAPB-10.

Fig. 22. Field emission scanning electron microscope (FESEM) images of MAPB-0, MAPB-4 and MAPB-10 raw samples showing the phases with iron oxide (wüstite), slag components, charcoal, and epoxy used in the sample polishing.

From the microstructure of the three briquette samples with different biocarbon contents, it can be seen that the charcoal fines are evenly mixed between the mill scale grains, binding them together. Bonding improves the cold strength properties of the briquettes, as can be seen from the abrasion test results in Fig. 10. Iron oxides, mainly wüstite, and the phases between them consisting of slag components and especially the charcoal added in sample MAPB-10 stand out clearly from the structure.

The effect of the reduction experiment on the briquettes can be seen in the FESEM images taken from the briquette samples after subjecting them to BFS experiment. Figure 23 shows the microstructure for the reduced MAPB-0, MAPB-4, MAPB-10 samples.

Fig. 23. Field emission scanning electron microscope (FESEM) images of the MAPB-0, MAPB-4 and MAPB-10 samples treated in the BFS and elemental maps of the Fe and O contents in the samples.

As seen from Fig. 23, the self-reduction properties of the samples differ from each other. With MAPB-0, the structure has remained intact, but the gasified carbon has not been fully sufficient to reduce the wüstite to metallic iron. A large part of the mill scale grains were reduced at the edges, as the cores remained as oxidic iron after the reduction experiment. With MAPB-4, the structure is clearly more fragmented, and visually iron was detected only in metallic form in the sample. Maps measuring the Fe and O concentrations of the structure show clear FeO phases in the structure of MAPB-0. However, the oxidic iron has been almost completely reduced to metallic iron in sample MAPB-4, i.e., already with a 4% biocarbon addition which can also be seen by comparing the elemental maps. In MAPB-10, most of the strongly bound charcoal in the briquette was gasified, causing a complete reduction of iron oxides, but also embrittlement of the briquette structure. In the images taken from MAPB-6, no clear areas with Fe and O were observed, based on which it can be concluded that it is not necessary to add more than 6% biocarbon in terms of self-reduction. In the reduced MAPB-4, MAPB-6, MAPB-8 and MAPB-10 samples, the residual oxygen was mostly bound to calcium, silicon and manganese phases.

3.5. Effects of Biocarbon Addition

The most illustrative and simplest way to compare tested auger pressing briquettes is to consider the ratio of iron and carbon contained in the samples. When evaluating reducibility, the amount of iron oxides is essential. In this review, as only the total iron content is considered, the metallic iron possibly contained in the raw samples causes some inaccuracy. Figure 24 presents the comparation between sample Fe_tot/C ratios before the BFS experiments together with the weight loss results from the experiments. The results from this work and our previous study on APB are included.

Fig. 24. Fe_tot/C ratios before the BFS experiments and relative weight losses during the experiments presented for the MAPB and IMAPB samples studied in this work and the auger pressing briquette (APB) sample studied in the previous work.

In Fig. 24, the briquette recipes are arranged according to the Fe_tot/C ratios from the largest to the smallest. As the Fe_tot/C ratio decreases, self-reduction increases, leading to greater weight loss in the BFS experiment. For IMAPB, a deviation is observed which can probably be partly explained by the metallic iron particles present in Mill Scale #1 used in the briquetting process. No visible metallic iron was detected in the MAPB samples, in contrast to the IMAPB samples, which had visible iron particles on the surface of the briquettes. Mill Scale #2 was obtained for analysis and no metallic iron was detected. The metallic iron content of Mill Scale #1 may have differed from that indicated in Table 2 due to quality variation. Based on the results for samples MAPB-8 and MAPB-10, it was found that increasing the carbon content no longer significantly increased the weight loss. For the auger pressing briquette examined in our previous study, which is referred to in Fig. 24 as APB, a higher weight loss was obtained with a lower relative amount of carbon compared to this work. In the previously studied briquette, the carbon originated almost entirely from BF dust. Based on these results, a suitable Fe/C ratio from the point of view of reducibility is around 5, but this depends on the amount of oxidized iron in the briquette raw materials. However, the briquette becomes plastic in high-temperature conditions already at about an Fe_tot/C ratio of 6.3.

When comparing the results of this work with the previous one,¹⁰⁾ it can be stated that swelling was more significant with the auger pressing briquettes containing biocarbon. It should be noted that the APB samples did not use a separate reducing agent, but the coal contained in the BF dust acted as a reducing agent. Using charcoal as a reducing agent appeared to result in lower weight loss than by using only carbonaceous by-products. Based on the mineralogical characterization, complete reduction occurred with 6% biocarbon content, but coal gasification continued to decrease the weight.

Table 10 contains an overview of how the replacement of mill scale with charcoal affects the high temperature properties of the briquette. Each recipe has been evaluated from the perspective of BF use.

Table 10. Main properties of biocarbon briquettes (MAPB-X) and industrial briquettes (IMAPB) based on the results from the BFS and RUL experiments and the suitability for BF use.

Recipe	BFS observations			RUL observations		Suitability for BF use
	Reducibility	Swelling	Cracking	Telescoping	Disintegration
MAPB-0	Moderate	Moderate	No	Low	No	Suitable
MAPB-2	Moderate	High	No	Moderate	No	Suitable
MAPB-4	High	High	Minor	High	Yes	Limited
MAPB-6	High	High	Minor	Very high	Yes	Unsuitable
MAPB-8	High	High	Minor	Very high	Yes	Unsuitable
MAPB-10	High	Low	Minor	Very high	Yes	Unsuitable
IMAPB	Moderate	No	No	Low	No	Suitable

Only two out of six biocarbon briquette recipes, MAPB-0 and MAPB-2, can be considered suitable for BF use based on the reduction experiments. These briquettes had weaker cold compression strength properties but were less prone to generate fines as a result of dropping. The high telescoping rate of MAPB-4 limits its usage in BF processes due to the generation of fines. The rest of the biocarbon briquettes, i.e., MAPB-6, MAPB-8 and MAPB-10, are not suitable for industrial use as such, because they cannot withstand a load simultaneously with reducing conditions. IMAPB, which has already been successfully tested in industry, also proved to be usable based on the laboratory tests. The results show the connection between cracking, telescoping and reducibility. The briquettes that showed cracking during the BFS experiment also showed significant telescoping and disintegration behavior during the RUL experiment. The disintegration of the briquette structure seems to be due to higher reducibility. The related phenomena are noticeable in the briquettes whose weight loss results in the BFS experiment were 29.8% or more.

Of the examined briquettes that can be assessed as suitable for BF use, MAPB-4 contained the largest proportion of biocarbon. In this recipe, 39% of the fossil carbon contained in the briquette was replaced with bio-based carbon. For MAPB-2, the corresponding share is 30%.

4. Conclusions

(1) Based on the mechanical strength tests, increasing the share of charcoal fines in the briquette recipe improves the cold compression strength of the briquette, but increases the generation of fines caused by dropping and abrasion. This data is important for technology, for determining the strength of briquettes and fines generation during transportation.

(2) The industrially and laboratory-produced briquettes have very analogous high-temperature properties. Weight losses in the reduction experiments were fairly similar and the industrial briquette showed no swelling or cracking.

(3) Based on the cracking behavior during the BFS experiments, even a minor replacement of mill scale with biocarbon significantly decreased the strength of the briquette in reducing conditions.

(4) The briquettes showed significant swelling when a small amount of biocarbon was added to the recipe. This is due to the higher reduction rate caused by the self-reducing effect. The briquettes with a separate reducing agent seemed to be more prone to swelling compared to the briquettes in which carbon was derived from by-products. Briquette swelling during reduction was strongest when 4% of the total composition was biocarbon.

(5) Based on the RUL experiments carried out for the various briquette recipes, the structure of the briquette becomes very plastic below 1000°C when only 4% biocarbon is used. Amounts greater than this were considered critical. Collapsing and disintegration behaviors were exhibited under a load, indicating that the tested briquette recipes with 4% charcoal or more are not suitable for use in the BF process due to generation of fines. In this case, the Fe_tot/C ratio in the briquette recipe was over 6.

(6) Based on the studied briquette recipes, 30–39% of the fossil carbon contained in self-reducing briquettes can be replaced with bio-based carbon without losing usability.

Acknowledgements

The authors would like to appreciate the fund support from Business Finland as a part of the Towards Carbon Neutral Metals (TOCANEM) research program, grant number 40693/31/2020. The author, Olli Vitikka, deeply appreciates the personal grant from the Association of Finnish Steel and Metal Producers of the Technology Industries of Finland Centennial Foundation. The Center for Material Analysis (CMA) is acknowledged for providing mineralogical characterization services. Tommi Kokkonen is acknowledged for his technical support in the laboratory work at the University of Oulu.

References

• 1)  World Steel Association: World Steel in Figures 2023, (June 7, 2023), https://worldsteel.org/steel-topics/statistics/world-steel-in-figures-2023/, (accessed 2023-10-27).
• 2)   K.  Ma,  J.  Deng,  G.  Wang,  Q.  Zhou and  J.  Xu: Int. J. Hydrog. Energy, 46 (2021), 26646. https://doi.org/10.1016/j.ijhydene.2021.05.095
• 3)   S.  Safarian: Fuel, 339 (2023), 127401. https://doi.org/10.1016/j.fuel.2023.127401
• 4)   E.  Mousa,  C.  Wang,  J.  Riesbeck and  M.  Larsson: Renew. Sust. Energ. Rev., 65 (2016), 1247. https://doi.org/10.1016/j.rser.2016.07.061
• 5)   E.  Mousa,  M.  Kazemi,  M.  Larsson,  G.  Karlsson and  E.  Persson: Appl. Sci., 9 (2019), 5288. https://doi.org/10.3390/app9245288
• 6)   M.  Geerdes,  R.  Chaigneau,  I.  Kurunov,  O.  Lingiardi and  J. A.  Ricketts: Modern Blast Furnace Ironmaking: An Introduction, 3rd ed., IOS Press, Amsterdam, The Netherlands, (2015), 28. ISBN 978-1-61499-498-5
• 7)   H.  Nonaka,  T.  Tomono,  Y.  Nagashima and  M.  Kitao: ISIJ Int., 63 (2023), 1539. https://doi.org/10.2355/isijinternational.ISIJINT-2023-046
• 8)  ISO 3271: 2015, Determination of coke reactivity index (CRI) and coke strength after reaction (CSR).
• 9)   H.  Tang,  Z.  Liu and  Z.  Guo: ISIJ Int., 62 (2022), 642. https://doi.org/10.2355/isijinternational.ISIJINT-2021-450
• 10)   O.  Vitikka,  M.  Iljana,  A.  Heikkilä,  I.  Tkalenko,  N.  Koriuchev,  D.  Shehovsov,  A.  Malkki and  T.  Fabritius: Minerals, 12 (2022), 868. https://doi.org/10.3390/min12070868
• 11)   N.  Koriuchev,  O.  Kovtun,  D.  Shehovsov and  I.  Tkalenko: EJENG, 8 (2023), 4. https://doi.org/10.24018/ejeng.2023.8.4.3072
• 12)   I.  Tkalenko,  O.  Kovtun,  N.  Koriuchev,  L.  Platonov and  D.  Shehovsov: EJENG, 8 (2023), 6. https://doi.org/10.24018/ejeng.2023.8.6.3106
• 13)   J.  Rieger and  J.  Schenk: J. Sustain. Metall., 5 (2019), 295. https://doi.org/10.1007/s40831-019-00220-2
• 14)  ISO 3271: 2015, Iron ores for blast furnace and direct reduction feedstocks — Determination of the tumble and abrasion indices.
• 15)   M.  Iljana,  O.  Mattila,  T.  Alatarvas,  V.-V.  Visuri,  J.  Kurikkala,  T.  Paananen and  T.  Fabritius: ISIJ Int., 52 (2012), 1257. https://doi.org/10.2355/isijinternational.52.1257
• 16)   M.  Iljana,  O.  Mattila,  T.  Alatarvas,  V.-V.  Visuri,  J.  Kurikkala,  T.  Paananen and  T.  Fabritius: ISIJ Int., 53 (2013), 419. https://doi.org/10.2355/isijinternational.53.419
• 17)   M.  Iljana,  A.  Kemppainen,  T.  Paananen,  O.  Mattila,  E.  Pisilä,  M.  Kondrakov and  T.  Fabritius: Int. J. Miner. Process., 141 (2015), 34. https://doi.org/10.1016/j.minpro.2015.06.004
• 18)   A.  Kemppainen,  K.  Ohno,  M.  Iljana,  O.  Mattila,  T.  Paananen,  E.-P.  Heikkinen,  T.  Maeda,  K.  Kunitomo and  T.  Fabritius: ISIJ Int., 55 (2015), 2039. https://doi.org/10.2355/isijinternational.ISIJINT-2015-023
• 19)   A.  Heikkilä,  M.  Iljana,  H.  Bartusch and  T.  Fabritius: Steel Res. Int., 91 (2020), 2000047. https://doi.org/10.1002/srin.202000047
• 20)   A.  Abdelrahim,  M.  Iljana,  M.  Omran,  T.  Vuolio,  H.  Bartusch and  T.  Fabritius: ISIJ Int., 60 (2020), 2206. https://doi.org/10.2355/isijinternational.ISIJINT-2019-734
• 21)   M.  Iljana,  A.  Abdelrahim,  H.  Bartusch and  T.  Fabritius: Minerals, 12 (2022), 741. https://doi.org/10.3390/min12060741
• 22)   A.  Kemppainen,  M.  Iljana,  E.-P.  Heikkinen,  T.  Paananen,  O.  Mattila and  T.  Fabritius: ISIJ Int., 54 (2014), 1539. https://doi.org/10.2355/isijinternational.54.1539
• 23)   A.  Abdelrahim,  M.  Aula,  M.  Iljana,  T.  Willms,  T.  Echterhof,  S.  Steinlechner,  D.  Mombelli,  C.  Mapelli,  M.  Omran,  S.  Preiss and  T.  Fabritius: Steel Res. Int., 93 (2022), 2100472. https://doi.org/10.1002/srin.202100472
• 24)  ISO 7992: 2015, Iron ores for blast furnace feedstocks — Determination of reduction under load.
• 25)   A.  Koskela,  H.  Suopajärvi and  T.  Fabritius: Fuel, 324 (2022), 124823. https://doi.org/10.1016/j.fuel.2022.124823