Solid carbonaceous materials transformation during pulverised coal and natural gas co-injection

Abstract

Co-injection of coal and gaseous fuel, such as natural gas (NG), is common practice promoting pulverised coal gasification in blast furnace (BF) ironmaking. The high hydrogen content in NG makes it possible to act as cleaner reductant to reduce iron ore to produce hot metal, which will result in reduced CO₂ emissions during the ironmaking process. In this work, selected parameters affecting the overall performance of the pulverised coal and natural gas co-injection system were studied using the CanmetENERGY injection test rig, including coal injection rate, natural gas rate and blast oxygen enrichment. Increase in combustion intensity promotes conversion of injected coal into solid carbonaceous material with relatively low reactivity with CO₂. Hence, it reduces the competitiveness of combustion residues for oxygen in the raceway to continue the gasification process. NG co-injection reduces the coal combustion intensity but enhances the reactivity of combustion residues by competing the oxygen in the raceway.

Keywords

pulverised coal injection natural gas co-injection blast furnace simulation rig solid carbonaceous transformation

Introduction

Pulverised coal injection (PCI) is a common energy practice in modern blast furnace (BF) operations. The primary goal of PCI is to reduce coke consumption per tonne of hot metal production basis while maintaining sufficient carbon input to the support the ironmaking process. The efficiency of PCI in coke rate reduction is commonly quantified by the replacement ratio of the auxiliary fuel selected for direct injection, which is defined as coke rate reduction per unit mass of injected auxiliary fuel. Significant technology development efforts had been devoted to enhance the replacement ratio of PCI coal.^1–4

One of the important factors that determining the replacement ratio of PCI coal is its combustion behaviour upon rapid heating. When the pulverised coal particle is introduced into the hot blast and rapidly heated up, the coal particle undergoes rapid pyrolysis.^5,6 Devolatilisation, transformation of solid carbonaceous material and gasification by surrounding oxygen occurs simultaneously. Both gasified volatile matter and solid residue are carbonaceous. This study will focus on the properties of the relict solid portion of the injected pulverised coal since the behaviour of the solid carbonaceous materials dictated the gasification kinetics of the process. Therefore, the term ‘carbonaceous material’ in this article refers to the solid carbonaceous combustion residue. Ideally, the coal particle is fully gasified to achieve its maximum coke replacement capacity. However, in practice, injected coal is only partially gasified in the tuyere-raceway because the residence time of coal particles in this region is extremely short (∼20 ms) and may not be sufficient to fully gasify the injected coal.⁷ The devolatilisation and combustion of injected coal particle initiate inside the tuyere. Upon existing the tuyere and entering the raceway, the remaining solid carbonaceous materials leave the oxygen-rich environment and enter the lower oxygen content raceway.¹ Once in the raceway the remaining solid carbonaceous materials competes with surrounding hot coke for oxygen to continue the gasification. The extent of gasification in the raceway depends on the reactivity of the solid carbonaceous materials generated by rapid pyrolysis when travelling through the tuyere. A portion of injected pulverised coal may leave the raceway and be carried upwards by the gas flow without being gasified. Since the temperature drops as the gas moves upwards in the BF, it is more difficult for the combustion residues to be gasified in the upper portion of the furnace. These residues maybe consumed in the BF after the raceway but if not, they will eventually leave the furnace as dust carried in the BF top gas. The incomplete gasification of injected coal represents loss of carbon efficiency⁸ and reduces the coke replacement ratio of the injected coal.

It has been reported that increase the temperature close to the injected PCI coal particles can enhance the ignition and gasification of PCI coal.^9,10 It was reported that Praxair designed of an oxygen–natural gas (NG) diffusion flame burner^9,11 and Paul Wurth developed a co-axial lance for injecting both gaseous injectant and PCI⁹ to improve the combustion of injected solid coal in BF. It is generally believed that NG will combust immediately upon injection thus leading to increase the temperature near to the coal stream and consequently promote the pulverised coal combustion. Moreover, NG is an excellent energy and reductant resource with high heating energy value and high hydrogen content. The high hydrogen content in NG makes it possible to be as cleaner reductant to reduce iron ore to produce metal by injecting NG in BF and reducing consumption of coal and coke, consequently resulting in reduction of CO₂ emissions for ironmaking.¹² Therefore, co-injection of NG and pulverised coal is a common practice in modern ironmaking BF operation.^13,14 By optimising the BF operation parameter, the overall utilisation coefficient of the BF increased can be improved.¹²

The effect of NG injection on the BF performance has been studied extensively.^12,15–17 Injected NG will ignite and rapidly combust at blast temperature (CH₄+ 2O₂= CO₂+ 2H₂O). The H₂O generated subsequently reacts with C in the raceway and decomposes into H₂ and CO. Because of the endothermic H₂O decomposition reaction, NG injection has a cooling effect which reduces the raceway adiabatic flame temperature (RAFT). Therefore, it is important to take the cooling effect into account when considering the co-injection of NG with pulverised coal.

As mentioned above, reactivity of solid carbonaceous material originating from the rapid pyrolysis of the injected coal is an important factor that determines its competitiveness for oxygen in the raceway and upper BF. The property of the solid carbonaceous material is determined by the carbon transformation of pulverised coal during the rapid heating process. This is dictated by the fundamental properties of the coal as well as the heating environment. The effect of NG co-injection on pulverised coal transformation is another factor that influences the effectiveness of NG co-injection on promoting PCI coal gasification.

To conduct a detailed study on the effectiveness of NG co-injection on pulverised coal gasification, it is necessary to characterise pulverised coal residues leaving the tuyere-raceway region. Due to the extremely hot environment, it is unrealistic to carry out experimental studies directly in an industrial BF. Therefore, small-scale PCI physical simulation is necessary. To address this need, CanmetENERGY-Ottawa has designed and constructed an experimental injection rig to simulate the combustion of pulverised coal in conditions similar to the industrial BF tuyeres-raceway.⁷ Since the commissioning of the facility in 2015, more than 40 coals and numerous blends have been successfully tested. Significant understanding on the transformation of solid carbonaceous materials during PCI has been developed.^7,18,19 The injection rig is further upgraded to enable co-injection of NG with pulverised coal. In this work, a series of tests were conducted to examine the effect of NG co-injection on pulverised coal gasification behaviour and on the solid carbonaceous materials transformation.

Transformation of coal particles upon rapid heating

Prior to discussing the experimental results in detail, it is necessary to understand the transformation of a coal particle during injection rig test. Gray²⁰ described the transformation of a coal particle upon rapid heating; when the coal particle is introduced in the hot blast, a steep temperature gradient is developed within the particle due to the significant difference in temperature between the coal particle and the surrounding atmosphere.

The surface of the coal particle heats up first. When the softening temperature of the injected coal is reached at the surface, the surface of the reactive particle starts to plastify and devolatilise. The devolatilisation results in the reactive particle expansion due to the volatile matter being released. Because of the steep temperature gradient, the interior of the particle is still below the coal's softening temperature and remains in solid form. As heating continues, the interior of the particle also eventually reaches its softening temperature to initiate devolatilisation. However, the devolatilisation at particle surface may have already be complete and have reached the coal's resolidification temperature. In this case, a rigid shell is formed enclosing the interior materials. With the presence of the rigid shell, the volatile matter released in the interior of the shell cannot be evacuated, leading to a rapid pressure build up. This eventually leads to breakage of the shell to form a cenosphere structure. Figure 1(a) shows the appearance of original coal particle for comparison. Figure 1(b)–(e) shows images of combustion residues sample collected in this work to illustrate the formation of the cenosphere structure and the effect of petrographic properties on solid carbonaceous material transformation upon rapid heating. Figure 1(f) shows the petrographic inert particles, which did not occur undergo physical transformation upon rapid heating. Because of the lack of softening and resolidification, the heating on petrographic inert particles do not lead to formation of cenosphere structure.

The entire PCI process can be categorised into three stages, namely devolatilisation, carbonisation stage and solid combustion. The coal injection process and stages can be illustrated in schematic diagram of the blowpipe-lance-tuyere-raceway system shown in the Figure 2(a) and (b). The devolatilisation stage refers to the release of volatile matter due to temperature increase and combustion of released volatile matter by the surrounding oxygen. The extent of coal devolatilisation is enhanced by the increase of high heating rates. Because of the rapid heating of the in PCI, the yield of volatile matters released is anticipated to be higher than the volatile matters content measured by standard laboratory analysis. The devolatilisation of the coal particle leads to plastification and eventually re-solidification as the temperature further increases. The transformation of the solid carbonaceous materials is referred to as carbonisation. The devolatilisation and carbonisation stages occur simultaneously and are analogous to the conventional coking process. In conventional cokemaking, the coal particle is heated slowly. Hence, it provides sufficient time for the plastified carbonaceous materials to re-solidify into carbon forms with different textural structures. Carbon forms in coke depend on petrographic properties of coal as well as thermal history experienced by the coal macerals. In PCI, the heating of coal particles is extremely rapid. It results in the textural structure of the re-solidified carbonaceous material not being as well defined as in cokemaking. In this work, the extend of carbon textual structure formation is referred as maturity of the carbonaceous materials. Solid carbonaceous materials with well-defined textural structure as in conventional cokemaking is referred to as high carbon maturity.

In PCI, combustion of volatile matters occurs as soon as devolatilisation is initiated. The combustion of the transformed solid carbonaceous materials occurs by reacting with the surrounding oxygen. The kinetics of solid combustion depends on oxygen availability in the atmosphere as well as carbon maturity of the transformed solid carbonaceous materials from the carbonisation stage. Generally, solid carbonaceous materials with low carbon maturity react with oxygen at lower temperatures than materials with high carbon maturity. Hence, solid carbonaceous material with a low carbon maturity is generally easier to combust than those with high carbon maturity. Therefore, the transformation of solid carbonaceous materials during the PCI influences the overall combustion behaviour of the process, which will be discussed below.

Experimentation

Injection simulation rig setup

The CanmetENERGY-Ottawa injection rig is designed to physically simulate the combustion behaviour of coal in an oxygen-rich environment, which corresponds to the behaviour of the injected coal in the tuyere-raceway region prior entering the upper region of BF. Figure 3 shows a schematic of the injection rig setup. The blast is prepared by mixing air, oxygen and nitrogen according to different levels of oxygen enrichment by controlling the flow rate of each component gas using individual dedicated mass flow controllers. The blast gas is heated to about 760°C by two preheating furnaces positioned in series.

A picture of the rig is shown in Figure 4(a). The pulverised coal sample in the hopper is delivered to the N₂ carrier gas by a screw feeder, and then carried to the entrance of the reactor furnace and introduced into the preheated blast. The coal sample is then transported by the hot blast into the reaction chamber. Figure 4(b) shows the positions and dimensions of the injection lance and the reactor tube, where combustion of the injected coal occurs. The locations of thermocouples and heating elements are also shown. The reaction chamber is an alumina tube 44 mm in diameter and 600 mm in length. The reaction chamber is electrically heated with constant power input and is stabilised at 1250°C prior to coal injection. Since the electrical heat supply is maintained constant during tests, the change in temperature recorded within the reaction chamber during a test corresponds to the energy released by the combustion of injectants. Temperatures at 200 and 380 mm from the injection point and at the exit of the reaction chamber are continuously monitored. The combustion residues exiting the reaction chamber are immediately water quenched and are collected for further analysis.

Figure 5 shows the design of the injection lance for facilitating coal and NG co-injection. NG and coal are delivered via separate openings in the same lance. Table 1 summarises the related parameters and conditions used during injection experiments. The experimental conditions are designed based on the ratio of solid injectant to blast gas similar to typical industrial operation, which is about 1.2 g/NL. It should be noted that the injection rig is not designed to simulate a particular industrial blast operation. Instead, it is designed for comparing the combustion behaviour of different injectants under the same combustion environment to assist understanding combustion characteristics of injectants. The experimental conditions employed correspond to coal particle residence time of ∼118 ms and heating rate ∼ 10⁴°C/.⁷ The residence time was calculated based on the blast gas rate, reactor dimension, as well as the gas pressure and temperature in the reactor. It is independent of coal rate and NG rate. The residence time may be longer, and the heating rate may be lower than industrial BF operation. However, it enables the comparison of behaviour of the injectants on relative basis. The combustion behaviour of the injected coal is quantified by the total burnout, which is determined based on the ash tracer assumption. A detailed account of total burnout determination method can be found elsewhere.^18,19

Table 1.

Experimental conditions.

Parameters	Experiment conditions
Coal particle size, μm	100% < 150
Carrier N₂ injection rate, m³/s	2.5 × 10⁻⁴
Blast gas flowrate, m³/s	2.78·×10⁻³
Sampling distance, mm	∼600 from injection point
Blast gas temperature, °C	∼760
Reactor furnace temperature, °C	1250
Residence time, ms	118.5

Coal sample property

Table 2 shows the properties of PCI coal used in this work. The same high volatile coal was used in all tests. The coal sample was pulverised to 100%< 150 µm. Co-injection of the pulverised coal with NG was conducted to study the effect of NG injection on coal combustion behaviour. To study the effect of various parameters, a central composite design experimental plan was developed (Table 3). Three parameters including coal feeding rate, NG rate and blast oxygen enrichment were selected for examination. The effect of each parameter at three different levels was studied. In the experiment, the total hot blast flowrate (air + O₂) was maintained identical between tests. Only the ratio between air and O₂ was varied between experiments to examine the effect of combustion behaviour in different level of O2 availability. By maintaining the total hot blast rate identical, the influence of residence time was eliminated to allow direct comparison of oxygen availability, coal rate and NG rate on combustion behaviour. A total of seven injection rig tests were conducted following the standard injection operating procedures. All tests were conducted on the same day to minimise variation between tests.

Table 2.

PCI coal property.

Chemical property			Thermal rheology property
Proximate (db)	%Ash	6.73	Gieseler Fluidity	Initial softening temperature (°C)	393
	%VM	35.82		Max Fluid Temperature(°C)	435
	%FC	57.45		Solidification Temperature (°C)	463
Ultimate (db)	%C	76.90		Melting Range (°C)	70
	%H	5.1		Max Fluidity (ddpm)	621
	%N	1.69	Ruhr Dilatation	Softening Temperature (°C)	356
	%S	0.89		Max Contraction Temperature (°C)	405
	%O (by diff)	8.73		Max Dilatation Temperature (°C)	432
Ash Chemistry (adb)	%SiO₂	52.46		%Contraction	22
	%Al₂O₃	29.53		%Dilatation	30
	%Fe₂O₃	8.53	FSI	FSI	6.5
	%TiO₂	1.71	Petrography Property
	%P₂O₅	0.139	Maceral Analysis	Vitrinite	65.70
	%CaO	1.98		Semifusinite	4.4
	%MgO	0.7		Exinite	4.8
	%SO₃	1.74		Total Reactive	74.9
	%Na₂O	0.54		Semifusinite	8.9
	%K₂O	1.82		Micrinite	8.6
	%LOF	0.42		Fusinite	3.4
	Basicity Index	0.166		Mineral Matter	4.2
Calorific value	HHV (J/kg)	32.2 × 10⁶		Others	0
				Total Inert	25.1
			Reflectance	Mean Ro	0.86

Table 3.

Central composite design experimental plan.

Test	Coal rate	NG rate		Blast %O₂
Unit	(g/min)	(m³/s)	(SLPM)	(%)
1	22.5	0	0	25
2	22.5	5.83 × 10⁻⁵	3.5	25
3	22.5	1.17 × 10⁻⁴	7	25
4	22.5	5.83 × 10⁻⁵	3.5	23
5	22.5	5.83 × 10⁻⁵	3.5	27
6	25	5.83 × 10⁻⁵	3.5	25
7	20	5.83 × 10⁻⁵	3.5	25

Combustion residue characterisation

Figure 6 illustrates the analysis procedures of combustion residue sample collected in each injection rig test for characterisation of the combustion behaviour of the injected coal.

In each injection rig test, proximate analysis of the collected combustion residue sample was conducted to determine total burnout. The total burnout is defined as the weight percentage of combustible materials in the injected coal that has been gasified during the test (Equation (1)).

T o t a l B u r n o u t = \frac{C o m b u s t i b l e_{i n p u t} - C o m b u s t i b l e_{o u t p u t}}{C o m b u s t i b l e_{i n p u t}}

(1)

Based on the ash tracer assumption, the total burnout is calculated by the simplified equation from the ash content before and after combustion (Equation (2)).

T o t a l B u r n o u t = \frac{1 - \frac{% A s h_{i n p u t}}{% A s h_{o u t p u t}}}{1 - % A s h_{i n p u t}}

(2)

As discussed above, coal transforms into solid carbonaceous materials with different carbon maturity upon rapid heating. The characteristics of the solid carbonaceous materials in the combustion residue samples collected in this work were quantified by the carbon type differentiation (CTD) analysis technique.^7,18 The CTD technique examines the combustion characteristics of the sample using thermogravimetric analysis (TGA). The unburnt solid carbonaceous materials start to combust in air at different temperatures depends on its carbon maturity. Figure 7 shows a typical collected residue combustion rate under a defined temperature profile employed in CTD analysis. The first peak corresponds to materials in the residues collected with combustion initiation temperature below 475°C and is referred to as ‘Char_CTD.’ The second peak corresponds to material with combustion initiation temperature at 600°C and is referred to as ‘Coke_CTD.’ The term ‘Char_CTD’ and ‘Coke_CTD’ are used to distinguish two type of solid carbonaceous materials based on the temperature required to initiate combustion in air. Char_CTD refers to solid carbonaceous material that starts to combust at relatively low temperature. It corresponds to the injected coal transformed into carbonaceous materials with relatively low carbon maturity. On the other hand, Coke_CTD refers to resultant carbonaceous material with relatively high carbon maturity. It should be noted that Char_CTD and Coke_CTD are not two unique materials. In fact, it is likely that the injected coal transforms into a wide range of carbonaceous materials with different chemical and physical properties. The terminology employed in this work is simply for categorising the combustion residues based on their combustion characteristic to facilitate comparisons. With the total burnout and CTD analysis results, the injection coal can then be quantified into gaseous and solid combustion products. Hence, detailed comparison of the effect of injection conditions on injected coal conversion can be made.

As discussed above, solid combustion residue entering raceway competes with surrounding hot coke for oxygen to continue combustion. The reactivity of the solid carbonaceous material is an important factor that determines its behaviour in the raceway. In this study, the CO₂ reactivity of the collected combustion residue from the injection rig test is quantified by another developed TGA technique.¹⁹ The reactivity is assessed by heating the collected combustion residue sample from room temperature to 1100°C in inert gas (Ar) atmosphere. At 1100°C, the gas atmosphere is changed to 100% CO₂ and the sample is allowed to be completely gasified. To compare the reactivity between samples, a new parameter, α₂₀, is defined as the percentage weight loss of the sample on a dry-ash-free basis after 20 min of reaction time with CO₂.

Results and discussion

Combustion intensity

The increase in temperature during combustion of fuels injected into reaction chamber indicates the combustion intensity. Figure 8 shows the effect of coal feeding rate on combustion intensity while maintaining NG rate and blast O₂ enrichment constant. In the figures, the temperature increase is in reference to the stabilised temperature of the reaction chamber prior to injection, which is at 1250°C. Measurements were made at 200 mm (8a), 380 mm (8b) from the injection point as well as at the exit of the reaction chamber (8c). It should be pointed out that the thermocouple tips at 200 and 380 mm from the injection point are located on the outside of the reactor tube while the thermocouple tip at the reactor exist is located at the middle of the blast flow. The difference in locations result in the measured temperature at the exit of reactor is significantly higher. However, the relative temperature change at the particular locations reveals the difference in combustion intensity due to the change in test conditions.

The increase in coal feeding rate did not significantly influence the combustion intensity at early stage of injection as shown by temperature increase at 200 and 380 mm did not vary significantly. When pulverised coal was introduced into the hot blast, devolatilisation occurs rapidly. Therefore, majority of combustion at the early stage originated from the combustion of volatile matters as well as the co-injected NG. Experimental data suggested that the volatile matters released in the early stage did not significantly vary when coal feeding rate was increased as evidenced by no significant change in combustion intensity at the early stage of combustion. It is likely due to the cooling effect by introducing larger amount of coal and resulted in reducing the heating rate of coal particles to initiate devolatilisation.

As the process progressed, the combustion of solid carbonaceous material was initiated when travelling towards the exit of the reaction chamber. The combustion kinetics was determined by the oxygen transfer to the solid particle surface inside the plume. The increase in coal feeding rate increased the yield of volatile matters generated and solid to gas ratio of the plume. The available oxygen was first consumed by the volatile matters and NG at the outer layers of the coal plume and NG. It leads to reduction in oxygen available for combustion of solid carbonaceous material at interior of the plume. Hence, the combustion rate of solid carbonaceous materials decreased as coal feeding rate increase leading to the observed decrease in temperature at the exit of the reaction chamber, Figure 8(c).

Figure 9 shows the effect of NG rate on combustion intensity in the reaction chamber while maintaining coal feed rate and blast O₂ enrichment constant. The temperature increase at all three locations in the combustion chamber were significantly reduced by increasing NG rate. The experimental results clearly indicate the cooling effect of NG injection, which is consistent with the data reported in the literature.^15–17 The primary cause of the cooling is the high energy demand in cracking of NG. Another influential factor was the reduction in oxygen availability for coal combustion. As NG rate increased, the amount of oxygen in blast consumed by combustion of NG also increased. Hence, the amount of oxygen available for combustion of coal was reduced, which led to the reduction in coal combustion intensity as indicated by the decrease in temperature in the entire reaction chamber.

Figure 10 shows the effect of blast O₂ enrichment on combustion intensity in the reaction chamber while maintaining coal feed rate and NG rate constant. Clearly, the increase in blast O₂ enrichment increased the oxygen availability for combustion with both NG and coal in the reaction chamber. This resulted in more severe combustion of coal and higher temperature was reached within the entire combustion chamber.

Coal burnout

The total burnout quantifies the amount of combustible material in the coal that is converted into gaseous products. Figure 11 shows the effect of the injection conditions on the total burnout of the injected coal. The coal burnout decreased with increasing coal feed rate and NG injection rate and increased with increasing blast oxygen enrichment. The observed change in burnout aligned with change in combustion intensity discussed above. As discussed above, the increase in coal feeding and NG rates led to a decrease in the combustion intensity of coal due to the reduction of oxygen availability. Hence, total burnout decreased as these two parameters increased. Increasing the blast oxygen enrichment improved the oxygen availability for coal combustion. It enhanced the coal burnout when maintaining fuel input constant.

Solid carbonaceous materials conversion

Properties of the solid carbonaceous materials produced by devolatilisation and carbonisation of the injected coal play an important role in overall gasification behaviour of PCI. To develop further understanding of the effect of NG co-injection on the combustion of coal in tuyeres as well as in the raceway, detailed analysis of combustion residues collected in the injection rig tests was performed.

Using the CTD technique mentioned in Combustion residue characterisation section and described in detail in Reference,¹⁸ the type of solid carbonaceous material in the combustion residue samples collected in the experiments was quantified. Figure 12 shows the portion of Char_CTD in combustion residues under different injection conditions. The amount Char_CTD detected in the sample is a balance between the formation of Char_CTD during the transformation of the injected coal and the gasification of the produced Char_CTD by the surrounding oxygen.

As discussed above, the increase in coal injection rates reduced the solid combustion intensity. This resulted in reduction in burnout and hence reduction in gasification of Char_CTD in the combustion residues collected, Figure 12(a). Moreover, the reduction in burnout by increase in NG rate also resulted in the increase of Char_CTD remains in the combustion residues, Figure 12(b). The oxygen availability increased as the blast O₂ enrichment also increased. This resulted in a larger amount of Char_CTD gasified and reduction in Char_CTD remained in the collected combustion residue, Figure 12(c).

Char_CTD is defined as solid carbonaceous materials requiring a relatively low temperature to initiate the reaction with oxygen. Hence, it is relatively easy to be gasified. Therefore, the amount of Char_CTD remaining in the combustion residue strongly depends on the oxygen availability in the environment. Figure 13 shows the dependence of the conversion of injected coal to Char_CTD on oxygen availability. The oxygen availability is represented by the oxygen to carbon (O/C) molar ratio in the system. This takes into account all oxygen and carbon sources in the system including all fuels (carbon and NG) and oxygen in the blast. The conversion of the injected coal into Char_CTD in the combustion residue decreases with increases in available oxygen. As discussed above, the amount of remaining Char_CTD is a balance between the Char_CTD formation during the transformation of the injected coal and extend of combustion during the test. The decrease in conversion to Char_CTD with increasing oxygen availability is mainly due to the increase in Char_CTD combustion instead of reduction in Char_CTD formation.

As discussed above, Coke_CTD refers to solid carbonaceous materials require a relatively high temperature to initiate its combustion with relatively high carbon maturity in comparison of Char_CTD. The carbon maturity of the solid carbonaceous material relates to thermal history experienced by the coal particle. Therefore, the amount of Coke_CTD detected in the combustion residue is closely related to the combustion intensity inside the reaction chamber during a test. The measured temperature increases when the combustion intensity enhances. Figure 14 shows the highly linear relationship of Coke_CTD with temperature, indicating increase in reactor temperature raise the formation of Coke_CTD.

Figure 15 shows the remained amount of Coke_CTD in the collected combustion residues. As discussed above, the combustion intensity decreases with increasing coal feeding rate and NG rate and increases with increasing blast O₂ enrichment. In other words, the solid carbonaceous materials experienced relatively lower temperatures as the coal feeding rate and NG rate increases. On the other hand, solid carbonaceous materials experiences relatively higher temperatures as the blast O₂ enrichment increases. As a result, the remained amount of injected coal converted to Coke_CTD decreased with increasing coal feed rate and NG rate, whereas the coal conversion to Coke_CTD increased with increasing temperature due to blast O₂ enrichment, as observed.

CO₂ reactivity of combustion residues

In industrial operation, solid combustion residue competes for oxygen sources with surrounding hot coke to continue the gasification when entering the raceway. The competitiveness of the combustion residue plays an important role in its gasification behaviour in the raceway. The competitiveness of the combustion residue samples collected in the experimental works were quantified by α₂₀, which is defined as percentage weight loss of the sample on a dry ash free basis after reacting with CO₂ for 20 min at 1100 °C in TGA. It should be noted that the procedure is not intended to reproduce the raceway conditions in the α₂₀ test. Instead, a standard test condition is applied to allow direct comparison on reactivity with CO₂ between samples on a relative basis.

Figure 16 shows the α₂₀ of combustion residue collected. The CO₂ reactivity of the combustion residues was closely related to the Coke_CTD content, Figure 17. This was because the Coke_CTD corresponded to the solid carbonaceous material in the combustion residue with relatively high maturity and low reactivity compared to Char_CTD. Hence, combustion residue with a high Coke_CTD content exhibited lower reactivity with CO₂.

Regression analysis

The discussion above reveals that the parameters affecting the performance of pulverised coal and NG co-injection are inter-related. To facilitate visualisation of the influence of each parameter examined in this work, namely coal feeding rate, NG rate and blast O₂ enrichment, a regression analysis was performed using the experimental data obtained, Table 4. It should be noted that the result of the analysis is specific to the injection equipment used in this work and serves to only illustrate the relationship between the parameters examined in this work.

Table 4.

Correlation between combustion behaviour and experimental factors.

Factors	Exit temperature (°C)	Burnout (%)	Reactivity (%)
Coal feed rate (g/min)	−4.95	−0.66	0.05
NG injection rate (m³/s)	−7.50	−0.68	0.43
%O₂ (%)	11.04	1.36	−0.92
Intercept	214.97	36.76	110.49
R₂	0.84	0.86	0.80

The coal feeding rate and NG rate negatively impacted the combustion intensity as represented by temperature at the exit of the reaction chamber, whereas the blast O₂ enrichment has a positive impact. The high combustion intensity arises from high burnout of the injected coal. Therefore, the effect on the total burnout of the injection coal follows the same trend as the combustion intensity. The enhancement of mixing via lance design to improve oxygen transfer is essential for enable effective high coal injection rates. Further study is still needed to prove the concept. When NG co-injection is considered, the availability of oxygen to sustain coal combustion becomes an important factor.

The combustion intensity influences the transformation of coal into different types of solid carbonaceous materials, namely Char_CTD and Coke_CTD. This is because the transformation process is strongly influenced by temperature to which the coal particles are subjected. As combustion intensity increases, conversion to Coke_CTD increases and conversion to Char_CTD decreases. Due to reactivity of Char_CTD is higher than that Coke_CTD, the influence of the examined parameters on combustion reactivity is in an opposite trend to total burnout or combustion intensity. The observation highlighted the importance in balancing combustion intensity and the competitiveness of combustion residue for oxygen sources in the raceway, in order to achieve maximum gasification of the injected coal.

Conclusions

The gasification behaviour of PCI coal is an important factor in determining its performance as coke replacement. A systematic evaluation of the effect of coal injection rate, NG co-injection rate and blast oxygen enrichment on combustion behaviour of the injected coal and on the characteristics of solid combustion residue products was conducted in this work. Experimental results reveal that:

Oxygen availability and mixing of coal particle and oxygen are two most important factors determining the combustion intensity in the raceway.

Combustion intensity affects the thermal history experienced by the coal particles, and thus affects the transformation of coal particles into different solid carbonaceous materials, Char_CTD and Coke_CTD. Increased combustion intensity promotes the conversion of the injected coal into relatively inert solid carbonaceous material, Coke_CTD. It decreases the competitiveness of the combustion residues carried into the raceway for oxygen to continue the gasification. Therefore, it is essential to achieve a balance between the combustion intensity in raceway and the reactivity of combustion residues in order to achieve maximum gasification of the injected coal for maximum coke replacement potential.

NG co-injection reduces the coal combustion intensity and enhances the reactivity of combustion residues by competing with the oxygen in the raceway, implying it can be used as adjustor to supreme gasification of the injected coal and coke replacement.

Under the same injection and combustion condition, coal properties determine the carbonaceous transformation of injected coal to different types of carbon residue. This will need further research to determine the linkage between the solid combustion residue and coal macerals.

Footnotes

Acknowledgements

Experiments and material evaluations were conducted at CanmetENERGY-Ottawa under the skilful contributions of B. Barrass and V. Omatsu-Baas. The authors also gratefully acknowledge expertise of CCRA (Canadian Carbonization Research Association) for providing their industrial expertise and materials.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by CCRA and Office of Energy Research and Development of Natural Resources Canada.

ORCID iD

Xianai Huang

References

Hutny

Lee

Price

. Fundamentals of coal combustion during injection into a blast furnace. Prog Energy Combust Sci 1991; 17: 373–395.

Mathieson

Truelove

Rogers

. Toward an understanding of coal combustion in blast furnace tuyere injection. Fuel 2005; 84: 1229–1237.

Guo

Zulli

Rogers

, et al. Three-dimensional simulation of flow and combustion for pulverised coal injection. ISIJ Int 2005; 45: 1272–1281.

Sahajwalla

Harris

. Coal char reactivity and structural evolution during combustion-factors influencing blast furnace pulverized coal injection operation. Metall Mater Trans B Process Metall Mater Process Sci 2001; 32: 811–820.

Chen

Niksa

. Coal devolatilization during rapid transient heating. 1. Primary devolatilization. Energy Fuels 1992; 6: 254–264.

Chen

Castagnoli

Niksa

. Coal devolatilization during rapid transient heating. 2. Secondary pyrolysis. Energy Fuels 1992; 6: 264–271.

Ray

Giroux

MacPhee

, et al. Evaluation of PCI coals in new injection facility at CanmetENERGY-Ottawa. AISTech-Iron Steel Technol Conf Proc 2015; 1: 926–937.

Busser

. Blast furnace energy balance and recovery. 2022.

Babich

Hippe

Lin

, et al. Intensifying the PC conversion by means of coke oven gas, Germany. 2019.

10.

Babich

Senk

Born

. Interaction between co-injected substances with pulverized coal into the blast furnace. ISIJ Int 2014; 54: 2704–2712.

11.

Halder

. An experimental perspective on Praxair’s hot oxygen technology to enhance pulverized solid fuel combustion for ironmaking blast furnaces. n.d.

12.

Wang

Zhou

, et al. Numerical analysis of natural gas injection in Shougang Jingtang blast furnace. Metals 2022: 12: 2107.

13.

Babich

Yaroshevskii

Formoso

, et al. Co-injection of noncoking coal and natural gas in blast furnace. ISIJ Int 1999; 39: 229–238.

14.

Jampani

Gibson

Pistorius

. Increased use of natural gas in blast furnace ironmaking: mass and energy balance calculations. Metall Mater Trans B Process Metall Mater Process Sci 2019; 50: 1290–1299.

15.

Nielson

Okosun

Damstedt

, et al. Tuyere-level syngas injection in the blast furnace: a computational fluid dynamics investigation. Processes 2021; 9: 1447.

16.

Albul

Kobelev

Radyuk

, et al. Effect of natural gas flow rate and temperature on the processes occurring in a blast furnace air tuyere with a heat-insulating insert in the blast channel. Steel Transl 2022; 52: 1020–1026.

17.

Mahanta

Sarkar

Sen

, et al. Consequence of natural gas injection in blast furnace: a critical appraisal using a thermodynamic and evolutionary computation approach. Can Metall Q 2022; 61: 1–13.

18.

Huang

Giroux

, et al. Carbonaceous materials transformation in pulverized coal injection. 2019; AISTech2019 Proc Iron Steel Technol Conf 2019: 427–434. https://doi.org/10.33313/377/044

19.

Huang

Giroux

, et al. Effect of coal properties on combustion behavior during pulverized coal injection. 2019.

20.

Gray

. The role of explosive ejection in the pyrolysis of coal. Fuel 1988; 67: 1298–1304.

Solid carbonaceous materials transformation during pulverised coal and natural gas co-injection

Abstract

Keywords

Introduction

Transformation of coal particles upon rapid heating

Experimentation

Injection simulation rig setup

Coal sample property

Combustion residue characterisation

Results and discussion

Combustion intensity

Coal burnout

Solid carbonaceous materials conversion

CO2 reactivity of combustion residues

Regression analysis

Conclusions

Footnotes

Acknowledgements

Declaration of conflicting interests

Funding

ORCID iD

References

CO₂ reactivity of combustion residues