Regular Article
Identification of Carbonaceous Materials in Blast Furnace Dust by Micro-Raman Spectroscopy
2024 Volume 64 Issue 3 Pages 538-549
Details
2024 Volume 64 Issue 3 Pages 538-549
Carbonaceous materials (CMs) in blast furnace dust (BFD) contain abundant information about evolution behavior and utilization efficiency of metallurgical coke and injected pulverized coal. A novel method for determination of origin and content of BFD by micro-Raman spectroscopy was proposed in this paper. Ten randomly areas was selected and each one covers an area of 120×120 µm in the surface of briquetting demineralization-BFD sample. Average area was multi-point scanning measured by micro-Raman spectrometer for 140 test points with the same step length. Then, contrasted CM including evolutive coke, coal char and precipitated carbon from deposition of carbon process were prepared in laboratory simulated the process of fuels under gradually heating up condition of blast furnaces (BFs) and tested by micro-Raman spectroscopy. Meanwhile, the raceway coke was also sampled and tested for compared with coke at high temperature. Typical feature of spectrogram of these CMs can be well correspond to the BFD results. Therefore, CMs in BFD could be attributable to different sources, and its content can be counted by statistics of spectrogram. By comparing the identification results of petrography method, this method can measure the small particle size such as soot derived from very fine pulverized coal and volatiles which is hard to measure by the latter. For the investigated BF, unconsumed char escape from raceway and powdery coke in stack contributed the main source of CMs in BFD, while content of overflowing coke powder from high-temperature region of BF and precipitated carbon by deposition carbon reaction is very low.
Due to increasing environmental concerns, optimizing the blast furnace (BF) ironmaking process has become an important task for achieving low-carbon and green ironmaking. Coke plays multiple crucial roles in the ironmaking process, including providing heat energy, performing the role of reducing agent and maintaining permeability of the BF.1) The structure of coke in BF is changed due to the interaction of thermal, physical and chemical conditions. Meanwhile, a lower coke ratio results in a longer detained time for coke,2) which leads to increased breakage effects such as thermal stress, mechanical wear, solution loss reaction. These effects contribute to the deterioration of working conditions in BF.3) PCI (Pulverized coal injection) is used as a substitute for coke to supply heat and act as a reducing agent, resulting in a reduction in coke ratio and pig iron cost. The PCI rate is being increased in recent years owing to the high economics of PCI. However, increasing the PCI rate inevitably increases the amount of char inside the furnace as the residence time available for conversion of the coal particles is extremely short.4) As a result, some pulverized coal enters various parts of BF in the form of unburned char (UBC) and soot.5) When the amount of pulverized coal injection increases beyond a certain extent, a large amount of UBC negatively affects the furnace conditions. This includes an increase in coke load, a deterioration of the permeability of the melting zone and the dripping zone, and a decrease in the replacement ratio of coal and coke. Consequently, there is a significant increase in the carbon content in the furnace dust and a decrease in the BF coal ratio.6) It is worthwhile to find out what amount of coal injected into the BF is most economical under certain operating conditions. However, the fuels undergo a complex physical and chemical evolution process in a high temperature and high pressure state during actual production. Consequently, it is challenging to directly sample and analyze the fuels to evaluate the deterioration behavior of coke and UBC in different parts of the BF. Additionally, determining the appropriate coal injection amount in real time based on the furnace condition is also difficult.
Meanwhile, in the stack zone of BF, CO in the rising gas may also convert to solid carbon due to chemical reaction, especially at temperatures between 400–600°C.7) This reaction occurs when the kinetic and thermodynamic conditions are favorable:
| (1) |
Reduced iron can catalyze the reaction (1) to increase its rate.8,9) When CO and sintering ore come into contact at an appropriate temperature, the carbon precipitates and infiltrates into the micropores of the raw material. Simultaneously, the newly formed reduced iron also diffuses into the carbon. This repetitive process can lead to the breakage of the sinter and the production of powder,10) which hinders the rise of the gas flow and has a negative impact on the BF production. Additionally, the precipitated carbon may adhere to the joints of the lining bricks, causing them to loosen and affecting the stability of the BF. However, there is currently no detailed report on the extent of the carbon evolution reaction in the upper part of the BF and the acquisition of morphological samples.
It is expected to analyze the conversion efficiency of fuel in the BF by studying the blast furnace dust (BFD) in literature.11) The BFD is a fine dust that is collected by the BF gas during the iron-making process. BFD contains carbonaceous materials (CMs), such as coke powder and evolution products of pulverized coal, which provide valuable information about the evolution behavior and utilization efficiency of metallurgical coke and injected pulverized coal. The presence of a significant amount of UBC in BFD suggests that the coal injection efficiency is low and the replacement ratio decreases. By determining the source of CMs in BFD, more knowledge about fuel evolution behavior in BF will be acquired, which is beneficial for reducing carbon consumption and maintaining high yield, stability and economic operation.11) Currently, the most studied methods for analyzing the dust include petrographic analysis,12) XRD13) and Raman spectrogram analysis.14) Petrographical analysis can identify different mineral microstructures based on morphology and optical properties. The results commonly show that the ratio of UBC is 10%–20% in the CMs of the BFD.12) This method is commonly used due to its cost-effectiveness and quick sampling.15,16) However, it requires high professional analysis skills, involves subjective factors, and has a heavy workload. The XRD method determines the coke powder content by comparing the aromatic layer stacking height Lc of the (002) peak of the carbon with the standard curve.1) However, the diffraction peaks of various CMs overlap during the diffraction process, causing in peak deformation and potentially leading to misclassification of CMs.
Raman spectroscopy is a useful technique for characterizing the structure of CMs. It can provide information about the proportions of crystalline carbon and amorphous carbon present in these materials. This is reflected in the intensity, area, and width of the Raman characteristic peaks.17) A universal evaluation method can be established to assess the degree of disorder in CMs by analyzing the changes in Raman parameters. Coke is produced through high temperature carbonization of coal, resulting in a more orderly carbon structure. The inconsistent process of pulverized coal and coke passing through the high-temperature zone in the BF leads to differences in carbon structure order, resulting in varying Raman spectral intensity. This characteristic makes it suitable for studying the graphitization degree and microstructure changes during heat treatment evolution of CMs under different temperature.18) However, it is important to note that the dust not only contains CMs but also consists of hematite, lime, periclase, silex etc.19) Due to this heterogeneity, multi-field detection is required. Additionally, the current method primarily relies on single point scanning,1) which results in isolated sample points and incomplete coverage, leading to potential misjudgment due to partial generalization. Compared to traditional Raman spectroscopy, micro-Raman spectroscopy offers higher spatial resolution, operating frequency, operating speed, and sensitivity. It enables analysis of trace samples as small as 0.5–1.0 μm and allows for qualitative, quantitative, and positional analysis by combining spectroscopy and images.20) If micro-Raman spectroscopy is utilized for accurately locating and visualizing the two-dimensional area of the BFD, it will enhance the correlation between the micro-Raman results and the mapped image.20) Nonetheless, there is currently no universally accepted criterion for evaluating CMs in BFD by micro-Raman spectroscopy.
In this study, the demineralization of BFD from the investigated BF is performed as the first step. The micro-Raman technique is used to analyze the BFD. Unlike single-point detection, the test plane is defined through multi-point scanning. To compare and contrast, CM samples are prepared in the lab. The Raman spectrogram of BFD is then classified based on the difference of Raman characteristic parameters of different carbons. This classification is compared with the Raman peak type and characteristic parameters of contrasted samples to determine its composition. Statistical analysis is conducted to determine the content of each CM in the test plane. This establishes an evaluation method for micro-Raman analysis of BFD, as shown in Fig. 1. The proposed method aims to enhance the accuracy in determining the content and source of various CMs. Additionally, the petrographic analysis is compared to this evaluation system.
BFD was collected from the bag filter of a BF with an effective volume of 2400 m3. The production parameters are as follow: coal ratio is 165 kg/t, fuel ratio is 523 kg/t, oxygen enrichment rate is 2.11% , and the structure of coal injection is anthracite: bituminous coal in a ratio of approximately 75:25. Hematite has a maximum Raman characteristic peak at 1320 cm–1, which impacts the judgment of the peak shape of carbon at 1350 cm–1, as shown in Fig. 2. To ensure the accuracy of this experiment and eliminate the influence of non-carbon minerals, it is necessary to remove inorganic minerals from the BFD. The demineralization process is as follows: first, the sample is pickled using a mixture of 500 ml of 38% hydrochloric acid and 500 ml of 40% hydrofluoric acid; then, all the sample should be rinsed with deionized water until it becomes neutral, and dried at a temperature of 105°C for 24 hours or until it reaches a constant weight;21) after that, the powder sample is pressed into a flaky shape with a diameter of 25 mm and a height of 7 mm using a tablet press. To confirm the removal non-carbon minerals, a scanning electron microscope test was conducted. The element content was determined using the area scanning method, as depicted in Fig. 3. The data indicates that demineralization process effectively eliminated the interference from the majority of mineral elements, enabling the subsequent micro-Raman detection.
The types of CMs commonly found in BFD include the evolution products of coke and pulverized coal.22) The temperature experienced by the coke varies in different zones of the BF during the downward process. Upon entering the BF, some breakage and abrasion occur as the coke falls onto the stockline, during charging. As the temperature increases to over 1000°C, the coke begins to oxidize with CO2. In this zone, coke degradation occurs due to mechanical load and mild gasification. Upon entering the cohesive zone, coke gasification with CO2 becomes significant due to the higher reaction rates at higher temperatures. As the coke descends into the dripping zone, the temperature rises to about 1500°C. With the increase of temperature, carbon atoms are dissociated from the molecular structure of coke, transferred to molten iron, and results in an increased breakage for coke. In addition, the deterioration of coke also comes from the erosion of slag and iron. Finally, the coke enters the raceway zone. The temperature in this area is significantly higher than 1500°C, and may even exceed 2000°C. Under the impact of high temperature airflow, the graphitization degree of coke is greatly improved, and the surface loose structure falls off.3) Due to the influence of these different temperatures, it is necessary to prepare the corresponding heat-treated coke at different temperatures in the laboratory for investigation. At the same time, the interval is distinguished by the key inflection point temperature data, as shown in Table 1. As the raceway coke can be directly taken out as an experimental sample,17) and in order to highly simulate the process of coke deterioration in the BF, the coke is extracted from the raceway of this BF tuyere for direct micro-Raman detection, which is referred to as a part of high temperature coke.
| Raw material | Contrasted CMs | Preparation or collected condition | Simulative location of CMs in BF |
|---|---|---|---|
| Charged coke | Coke-low/medium temperature | Heat treated at 1100–1400°C/ Kept for 30 min/N2. | Stack zone, cohesive zone and BF bosh |
| Coke-high temperature | Heat treated at 1500°C/ Kept for 30 min/N2. | Raceway | |
| Blast-stop period/About 1 m front of tuyere/Air-free | |||
| Pulverized coal for PCI | UBC-low temperature | Heat treated at 300–700°C/ Kept for 30 min/N223) | UBC escape from raceway rapidly and then upward movement |
| UBC-high temperature & soot | Heat treated at 1100–1500°C/ Kept for 30 min/N2 | Char reaction and circle round in raceway and then upward movement&Volatiles transfer in BF | |
| Iron ore powder/CO | Precipitated carbon | 500°C/100% CO atmosphere kept for 30 min.9) | Stack zone |
The injected pulverized coal undergoes partial combustion in front of the tuyere due to the short reaction time. Some of the UBC escapes from the raceway and participates in the gasification reaction at around 1100°C. Additionally, the UBC is involved in the direct reduction reaction of iron, as well as the reduction reaction of non-iron elements and the carburizing reaction of pig iron at 1500°C. Another portion of the UBC may be deposited in the slag or dust.6) Hence, this study simulate the evolution of BF pulverized coal at temperatures ranging from 300°C to 1100°C and from 1100°C to 1500°C through bituminous coal pyrolysis. Additionally, the difference between UBC and coke at 1100°C will also be discussed.
To investigate the proportion of carbon in the dust of the BF originating from this process, precipitated carbon was simulated at 500°C in 100% CO atmosphere.
In summary, reparation or collected method of three types of CMs similar to BFD were prepared: coke, UBC and carbon generated by reaction (1) (precipitated carbon). Coke can be further classified into low/medium temperature and high temperature. Pulverized coal can also be divided into low temperature and high temperature, along with soot, as shown in Table 1.
2.2. Micro-Raman AnalysisThe samples were experimentally analyzed using micro-Raman spectroscopy (Horiba LabRAM HR Evolution, Japan). The test conditions were as follows: the laser wavelength was 532 nm, each test measured a 120 μm×120 μm area with 100 points in steps of 12 μm. The exposure time for each test point was 10 s, and the wave number range for the test was 800–1800 cm–1. The test involved taking measurements on 10 randomly selected areas of the sample. After obtaining the original spectrogram data, it was processed as shown in Fig. 4. Because of significant jaggedness of the original image, appropriate smoothing was applied and the baseline was subtracted. Subsequently, the data was normalized to facilitate comparison between different spectrograms of samples. Figure 4(b) illustrates the typical band of CM, with the D band positioned around 1350 cm–1, the G band around 1580 cm–1, and the valley known as the V band appearing around 1500 cm–1.
In order to accurately assess the differences between the D, G, and V peaks in certain spectrogram, a curve-fitting procedure was employed. Following the fitting mode described in literature,23) a five-peak sub-peak fitting approach was chosen, as depicted in Fig. 5. The first-order Raman bands and vibration modes of disordered CMs can be found in Table 2.23)
| Band | Wavenumber/cm–1 | Vibration mode |
|---|---|---|
| G | 1580 | Ideal graphitic lattice (E2g-symmetry) |
| D1 | 1350 | Disordered graphitic lattice (A1gsymmetry) |
| D2 | 1620 | Disordered graphitic lattice (E2g-symmetry) |
| D3 | 1500 | Amorphous carbon |
| D4 | 1200 | Disordered graphitic lattice (A1gsymmetry) |
A tube furnace (TL1200, China) is utilized for heat treatment. The CO2 generated by reaction (1) is detected using a mass spectrometer (THERMO Star, Germany). The traditional method of petrographic (PETRO 200AI, China) is used as a comparison. The sample is prepared following the guidelines provided in GB/T 16773-2008 ‘Method of preparing coal samples for the coal petrographic analysis’. The sample should exhibit a smooth surface, clear optical tissue interface, and distinct characteristics after treatment.24) The macerals are observed using a microscope under reflected light, oil immersion, and objective lens magnification. Differences in structure and optical properties of coke and pulverized coal are used to distinguish the CM. The components from pulverized coal consist of slightly deformed coal particles, undeformed particles, and deformed particles. The macerals from coke include fluidal texture, mosaic texture, leaflet texture, and fusinite. The classification is illustrated in Fig. 6.
The microstructure, spatial pattern, and optical anisotropy of CMs were analyzed using a scanning electron microscope (Gemini SEM 300, Germany). The maximum image resolution exceeds 5 nm at an accelerating voltage of 1–30 kV. The flake sample on the microscope’s sample table has a thickness of approximately 7 mm and a diameter of 25 mm. The working distance is around 7.5 mm, the acceleration voltage is 5 kV, and the electron beam spot measures 5 nm. Element analysis is conducted on special shape points observed in the visual field. To facilitate the analysis of its shape and appearance characteristics, the BFD powder is detected by attaching it to a conductive adhesive, ensuring the removal of most non-carbon minerals.25)
In the case of coal/char/coke, which are typical disordered non-graphite CMs, the Raman spectrogram shows distinct peaks at 1350 cm–1 and 1580 cm–1, known as the D peak and G peak. The V peak is the minimum point between these two peaks. In more ordered CMs, the Raman spectrogram exhibits a narrower peak width, with the intensity of the D peak being almost equal to that of the G peak. On the other hand, in disordered CMs, the peak width of Raman spectrogram is broadened26) and the intensity of D peak is much lower than that of the G peak.
Among the spectrograms of BFD, four types were identified and classified based on the intensity of the Raman characteristic peak, as shown in Fig. 7. The presence of varying carbon sources in BFD and its residence time in different temperature regions of the BF lead to discrepancies in carbon ordering, which is reflected in the disparity of peak intensities. While the peak positions of these types of spectrograms are essentially the same, the relative intensity of the D, G, and V peaks can distinguish them. As mentioned above, the D band represents the graphitic defect fractions while the G band represents the normal graphite structure. The intensity ratio of the D and G peaks can be categorized into three simple categories: ID < IG, ID ≈ IG, and ID > IG. The intensity of the V peak is also considered in the criterion due to the evident low trough observed in some data. The determination of two threshold values, namely 0.15 and 30%, provides a clear boundary for distinguishing between different types of spectrogram near these values. Additionally, statistical analysis reveals that the frequencies of scanned points assigned to these four types differ significantly. The percentage of different types of points to the total points was calculated as the CMs content. Specifically, the content from highest to lowest is 76.75%, 21.16%, 1.78%, and 0.28%. Theoretically, as the temperature changes during the pyrolysis process, the microcrystalline structure of carbon undergoes a transition towards ordering, eventually leading to complete graphitization. This transition, along with the presence of graphite and disordered carbon, allows for the differentiation of various CMs. Yu et al.27) prepared a mixture of different proportions of chars (0, 10, 20, 30, 40 and 50%) and drilled coke to simulate BFD, and analyzed the BFD and simulated mixture by Raman spectroscopy. Raman intensity parameters were utilized to estimate the mass ratio of chars to CMs in BFD obtained from two different BFs. The ranges of the mass ratios were found to be 8–10% and 10.5–12%. In contrast to previous studies, this paper directly detects BFD and minimizes the error caused by heterogeneity through multi-point detection. This allows for more accurate calculations and reveals the difference between single-point scanning and multi-point scanning.
Figure 8 displays the spectrogram of UBC-low temperature, which was prepared by pyrolysis of bituminous coal as Table 1. The spectrogram reveals that the intensity of the D peak is significantly lower than that of the G peak, indicating a high degree of disorder in the char. As the heat treatment temperature reaches 700°C, the intensity of the D peak slightly increases and its position shifts to a lower wavenumber, aligning with findings reported in the literature.26) Additionally, a strong D4 peak band is observed at 1200 cm–1, leading to a slower rise in the left waveform. The presence of this band suggests that the CM possesses more structural defects.
Figure 9 shows the spectrogram of UBC-high temperature. The results indicate that at temperatures between 1100–1500°C, the intensity of the D peak is higher than that of the G peak, and the V peak gradually decreases with increasing heat treatment temperature. In contrast to UBC-low temperature, the D peak and G peak width of UBC-high temperature decrease, suggesting a gradual tendency towards ordering of carbon atoms in the coal-char structure. Chang et al.28) indicated that when the temperature exceeds 923°C, tar and volatiles partially evolve into soot during the pyrolysis stage. Coal with high tar yield typically exhibits higher soot generation yield, which reaches its maximum at 1327°C. Considering the existence of soot, it is combined with UBC, collectively referred to as UBC-high temperature, as shown in Fig. 9.
Figure 10 displays the Raman spectrogram of coke subjected to heat treatment at various temperatures. It is evident that as the temperature rises, the D and G peaks become narrower. When comparing the contrasted CMs, it is apparent that the trend of coke above 1500°C differs from that of other samples at approximately 1250 cm–1 (D4). Subsequent testing confirms that the spectrogram shape of tuyere coke corresponds to the coke at 1500°C. The D4 peak represents the defect structure of carbon, and the smaller the area value is, the more the graphitized structure is. This means that the heat-treated coke above 1500°C has a more perfect graphite structure. Taking this transformation as a boundary, the coke is divided into two groups, which are Figs. 10(a) and 10(b). The width of the D and G peaks is associated with the quantity of amorphous carbon structure and crystalline carbon structure.29) As the heat treatment temperature increases, the D and G peaks become steeper, indicating a more ordered carbon structure and a reduction in the amorphous carbon structure. Furthermore, the V peak is proposed to also reflect the degree of carbon structure order.30) In the case of stable D and G peak intensities, a lower V peak intensity indicates a lower amorphous carbon content, suggesting a more uniform carbon structure. After the temperature exceeds 1200°C, the gasification reaction between the sample and CO2 starts to increase gradually. At this point, the sample is influenced by both the ordering and gasification reactions. During this process, CO2 molecules collide with the sample, leading to defects in the growth process of the six-membered ring network plane and the stacking process of microcrystallines, resulting in the formation of microcrystalline fragments.31) Consequently, the V peak shifts upwards during this stage. This is the reason why the V peak of coke-1200/1300/1400°C is greater than coke-1100°C. This further supports the rationality of the type classification.
When considering only the criterion of strength, it becomes challenging to differentiate between coke-low temperature and UBC-high temperature Therefore, curve fitting was performed on both spectrogram to identify the differences. It was observed that the AD4/Aall of coal differs from that of coke. This value represents the ratio of the area of the D4 peak to the total area, and usually represents the defect structure in the CMs. The comparison of Figs. 10 and 11 shows a weakening of the D4 peak, indicating a gradual decrease in defects in the carbon structure with increasing temperature, as depicted in Fig. 12. Hence, CMs can be clearly classified.
During the SEM detection process, distinct and unique morphologies and structures were observed. Figures 13(a) and 13(c) reveal the presence of a smooth and angular block material wrapped on the surface of numerous rough surface materials. The scanning confirmed that this block material contains few iron element. This indicates that despite the demineralization process, traces of iron still remain. Considering the SEM results, it can be inferred that under appropriate conditions, the reaction (1) occurs on the surface of the sinter, leading to the formation of precipitated carbon that covers and protects the iron compounds from being dissolved by the acid solution.32)
The mass spectrum analysis was conducted on the CO2 before and after the introduction of CO. The results, depicted in Fig. 14, reveal that the iron ore powder sample undergoes a vigorous reaction at temperatures around 400°C–500°C upon the introduction of CO, resulting in a significant production of CO2. This reaction is responsible for the occurrence of precipitated carbon. This temperature range offers advantages for reaction (1) under thermodynamic conditions.33) During this phase, a significant amount of carbon precipitated. As the temperature rises beyond the advantageous zone, there is a gradually decreases of CO2 content and a reduction in the precipitated carbon.
Further examination of the Raman spectrogram of the precipitated carbon indicates that the intensity of its D peak is considerably higher than that of the G peak, as shown in Fig. 15. At the same time, the trough intensity of its V peak is significantly low. These characteristics make it different from other evolutions of CMs that conventionally enter the BF, but in the analysis of BFD, it is found that type (4) in Fig. 7 can correspond to it, thus confirming the existence of precipitated carbon in BFD.
Through the analysis and generalization of the above data, five types of Raman spectrograms were identified, as shown in Fig. 16.
Firstly, the intensity of D, G, and V peaks is used as the criterion for BFD classification which corresponds CM types of (1), (3), (4) in BFD, as shown in Fig. 7. Then, when it is challenging to differentiate between ID, IG, and IV materials, curve fitting is necessary. To distinguish between them, the difference in D4 peak area is chosen as the standard. As a result, type (2) in the BFD classification is further refined to (2-1) and (2-2), as depicted in Figs. 17(c)–17(f). Finally, the data from the contrasted samples are plotted and compared with the BFD classification curve. It is observed that the contrasted samples of BF exhibit good consistency with the revised classification standard, as shown in Fig. 17.
Following the classification standard, the acquired BFD data is evaluated and classified, and the data points from the test are categorized and counted. As depicted in Fig. 18, the BFD contains a significant amount of the evolution products of coke and pulverized coal. The former constitutes 35.5% of the CM in BFD, while the latter comprises 64.22%.
The content of UBC-low temperature is 21.16%. The conversion temperature of these UBC does not exceed 700°C. The generation of UBC-low temperature may be attributed to the insufficient burning time in the raceway and the combustion of coal particles occurs individually. This phenomenon can be better understood when the coal has a high vitrinite content, as the volatile matter released instantaneously is readily combustible and utilizes the surrounding oxygen to sustain the burning process. Consequently, this buffers the flammable conditions of the residual carbon particles, leading to a significant deposition of UBC-low temperature. The content of UBC-high temperature and soot is 43.06%, and the existence of this part of CMs increases the total carbon content in BFD. Due to the large amount of bituminous coal injected into this BF, the release of volatiles may produce soot. Their presence can lead to negative effects on furnace conditions, such as an increase in coke load and a deterioration of the permeability in the melting zone and the dripping zone.
During the temperature range of 1100°C–1400°C, coke undergoes degradation as a result of mechanical load and gasification. The content of coke in this zone is 33.72%, indicating that a significant amount of coke is not able to fulfill its thermal effects, reduction, and provide structure role due to the gasification and the erosion caused by slag and iron. It is indicated that this batch has poor mechanical properties and is prone to breakage. On the other hand, the coke-high temperature content is 1.78%. It shows that the temperature of these cokes is between 1500°C and 2000°C. After experiencing thermal stress and alkali metal damage, they are converted into CMs which with higher graphitization degree in the BF. Based on these findings, BF operators can adjust the amount of pulverized coal injection and coke addition to optimize the correlation between BF fuel ratio and BF operation process.
The amount of precipitated carbon is only 0.28%. This is because at temperatures below 300°C, the rate of reaction (1) is nearly zero due to kinetic reasons. When the temperature exceeds the thermodynamic advantage range, the reaction also weakens. This reaction necessitates high conditions and is limited to the upper part of BF, resulting in a lower content.
3.4. Comparison between Micro-Raman Method and Petrographic AnalysisThe same samples were tested using both the petrographic analysis method and the micro-Raman method. The results of the micro-Raman method and petrographic analysis are presented in Table 3. The petrographic analysis method only divides the CM in the BFD into two categories, whereas the micro-Raman analysis method divides it into five categories, indicating a more detailed classification. In the traditional petrographic analysis, the content of samples obtained from pulverized coal accounted for only 26%. In contrast, the micro-Raman method yielded a higher content of samples obtained from pulverized coal, reaching 64.22%. This difference can be attributed to the particle size of carbon generated at high temperature, which falls between 15–615 nm,4) and the sampling point distance of petrographic analysis is between 3×105 ~ 5×105 nm. Small carbon particles are uniformly mixed with bakelite powder and can be artificially disregarded, as illustrated in Fig. 19. The micro-Raman method has a scanning point distance of approximately 1×104 nm and a spot diameter of 1×103 nm, enabling precise measurement of the small particle size of UBC-high temperature and soot. A comparison between the two methods reveals differences in the classification of CMs.
| Petrographic analysis | Micro-Raman analysis | |||
|---|---|---|---|---|
| Origin Source | Field number | Proportion | Point number | Proportion |
| PCI coal | 130 | 26% | 296 (UBC-low temperature) | 21.16% |
| 603 (UBC-high temperature and soot) | 43.06% | |||
| Coke | 370 | 74% | 472 (Coke-Low/Medium temperature) | 33.72% |
| 25 (Coke-high temperature ) | 1.78% | |||
| Rising gas | / | / | 4 (Precipitated carbon) | 0.28% |
| Total | 500 | 100% | 1400 | 100% |
In order to enhance the combustion efficiency of BF fuel, this study expected to focus on optimizing the amount of coal blending and fuel ratio through micro-Raman analysis of CMs in BFD. However, it should be noted that the results may vary across different BFs and batches of raw materials, and further analysis and research are required considering the specific raw materials and furnace conditions. In order to better understand the coke deterioration in the BF, an increase in the number of samples and additional Raman parameters can be added to refine the study. As the number of samples increases, it is recommended to develop a measurement system that includes a corresponding detection instrument for online detection of CMs in the sample. This will aid in guiding the determination of coal injection amount, fuel ratio, and coke quality.
(1) To detailedly analyze the CMs in BFD, a micro-Raman method was developed and its feasibility and accuracy were confirmed through experimental exploration.
(2) The system classifies the BFD based on the differences between ID, IG, IV. Spectrogram with minimal strength differences are further classified based on the value of AD4/Aall, resulting in the division of BFD into five categories: UBC-low temperature, UBC-high temperature, coke-low/medium temperature, coke-high temperature, and the precipitated carbon.
(3) The analysis revealed the presence of precipitated carbon in BFD, with a content of only 0.28%.
(4) Petrographic analysis only allows for the classification of CM in the BFD into two types, whereas the micro-Raman analysis method divides it into five types. This indicates that the micro-Raman analysis provides a more detailed classification.
The authors gratefully acknowledged the National Natural Science Foundation of China (No. 52374347), Shaanxi Provincial Department of Education service local special project (22JC042) and Key Research and Development Program of Shannxi (No. 2021GY-128).
