Effect of coaxial oxygen enrichment on three different coals and their blends

Abstract

Mathematical models appear as very attractive alternatives in blast furnace studies. Therefore, a three-dimensional model is described and validated to simulate the flow and combustion of binary coal blends under simplified blast furnace conditions. The present article aims to analyse the performance of different coals and their blends for conventional and coaxial oxygen enrichment methods. Among the simulated coals, Coal C, which combines one of the smallest particle sizes with one of the highest volatile matter contents, presented the highest burnout and average temperature values. Coaxial oxygen enrichment showed a negative effect on coal combustibility in the cases with 26.9% of O2. This is explained by the large amount of gas being injected by the coaxial lance, resulting in a cooling effect and particle concentration. Also, blending coal A with coal B resulted in an improvement of burnout for both coals, because of the synergetic effect caused by their combustion.

Introduction

Owing to the high productivity and efficiency, blast furnace is the main metallurgical reactor to produce hot metal [1–4]. During its operation, hot air is injected through the tuyeres presenting temperatures around 1100°C [5]. Then, the injected air reacts with coke in the furnace raceway, generating heat and reducing gases, which are vital to obtain hot metal [3].

Blast furnace operation demands high amounts of energy, representing about 70% of the total energy input required by the steel industry [6]. An alternative to reduce costs and to decrease the emissions of carbon dioxide is pulverized coal injection (PCI) [7,8]. It aims in replacing part of coke loaded on the superior part of the reactor [1,2,4,8] by injecting fines of coal, through the tuyeres, in its lower part. Replacing coke is important because it has a higher price than coal, and a large amount of COG is released during its production (coke oven gas) [9].

The injection of pulverized coal was intensified around 1980, during this time, other fossil fuels were also used to be injected as alternative fuels, but coal presented the lowest price and the most abundant reserves among them [10]. Nowadays, PCI technology has been continuously operated with pulverized coal rates around 200 kg/tHM [11,12]. However, further increases in injection rate have been a challenge due to the reduction of permeability caused by the unburned coal, so it needs improvements on char combustion [11,13].

Oxygen enrichment is one of the most important operational parameters to increase coal combustion efficiency in a blast furnace [14]. There are two different methods of increasing the oxygen content in the raceway: one is the enrichment by the blast, and the other by coaxial lance, where room temperature oxygen flows through the annulus of the lance [15,16]. Despite that, the information regarding the application of oxygen enrichments still not enough. As a result, there is a lack of knowledge in how different coals and their blends would behave for different enrichment methods.

The combustion in the blast furnace raceway is a complex process, and detailed measurements are extremely difficult to obtain because of high temperatures and pressures conditions, presence of molten material, lack of accessibility, and inevitable reduction of production [17]. So, in order to maximize production efficiency, there is a necessity of developing new technologies on ironmaking process to analyse and understand heat and mass transport phenomena inside this region of the blast furnace. Mathematical modelling appears as attractive alternatives to study blast furnace raceway.

In the past 50 years, several mathematical models of ironmaking processes have been developed and they have significantly contributed to improve blast furnace operational efficiency [18]. They are mainly represented by continuum-based (CFD) or discrete-based approaches (DEM) [19]. However, continuum modelling requires comparatively less computational power [20], and some models were already developed to predict localized regions, including pulverized coal combustion within the raceway.

Among the CFD models that were developed to study pulverized coal combustion in the raceway, Wijayanta et al. [2,17] have investigated some aspects of coal and biochar injection in a 2D model. While, Shen et al. [4,8,21–28] have carried out many studies, on the flow and combustion behaviours of pulverized coal, using a 3D model validated by a pilot-scale test rig. Also, Zhou et al. [1,14] have developed a 3D model to simulate pulverized coal combustion, but this work was focused on different ways to achieve oxygen enrichment. Nevertheless, none of them have studied the distinct ways of oxygen enrichment for different coals and their blends. Therefore, in order to improve the practice of PCI, it is important to investigate the chemical and physical interactions of different coals injected simultaneously in a blast furnace, so it could be useful to increase the number of coals able to be injected [29].

Then, the major goal of this work is the development of a 3D computational fluid-dynamic model, using ANSYS-CFX® 17.1, to predict the flow and the combustion of blends made of two different coals. Then, an investigation will be carried out to observe the effects of local and conventional oxygen enrichments for the different coals and their blends. The geometry used to develop this work is based on a combustion test rig used between 1998 and 2000 by BHP Newcastle Laboratories [30,31].

Material and methods

Model description

Governing equations for flow

This is a three-dimensional model, where gas phase flow is treated by transport equations of the continuum phase with Eulerian frame. It is described by the steady-state Reynolds averaged Navier-Stokes equations closed by a standard k-ε turbulence model. The equation is written in the coordinate free tensor notation (Equation (1)). (1)

where

is mean velocity,

is density,

is the diffusion coefficient, Φ is the variable, and

is the source term. Values for Φ,

, and

can be replaced for each governing equation presented on Table 1. They include, mass (m), velocity component (u, v, w), turbulence kinetic energy (k), turbulence dissipation rate (

), enthalpy (H), and species mass fractions (Y_i).

Table 1.

Variables Φ, Γeff, and SΦ of the governing equations for gas phase.


Continuity	1	0
Momentum	u, v, w
Energy	H
Species mass fraction	Y_i
Turbulent kinetic energy
Turbulent dissipation rate
; ;

The solid phase is modelled using a Lagrangian approach, and it is treated as a discrete phase without consider interaction between particles. So, the particles are tracked along their discrete trajectories determined by integrating Newton’s second law of motion, considering drag force and dispersion force. The change of particle temperature is governed by three physical processes: convective heat transfer

, latent heat transfer associated with mass transfer

, and the radiative heat transfer

. The governing equations responsible to describe the solid phase are represented on Table 2 [32].

Table 2.

Governing equations for the solid phase [32].

Mass:

Momentum:

Energy:

The drag coefficient, , is given by a modified Schiller and Naumann correlation [32], Equation (2), where it is limited as a maximum value of 0.44. (2) Turbulent dispersion of particles is determined by the turbulent velocity, , lifetime, , and eddy length, , respectively represented by Equations (3)–(5) [33]. (3) (4) (5) where is a normally distributed random number, and is a turbulence constant. Then, there is a full coupling of mass, momentum, and energy between particle and gaseous phase.

Coal devolatilization

Coal combustion is usually considered as a four-stage process: (i) pre-heating; (ii) coal devolatilization into two products, volatile matter (VM) and char; (iii) gaseous combustion of the VM; and (iv) oxidation and gasification of the residual char. The reactions considered in this model is listed in Table 3 based on Shen et al. [21].

Table 3.

Chemical reactions considered in the model.

Chemical Reaction	Description
Raw Coal_(s) → VM_(g) + Char_(s)	Devolatilization
VM_(g) + O_2(g) → CO_2(g) + H₂O_(g)	Gaseous Combustion
Char_(s) + O_2(g) → CO_2(g)	Char Oxidation
Char_(s) + CO_2(g) → 2CO_(g)	Char Gasification (Solution Loss)
Char_(s) + H₂O_(g) → CO_(g) + H_2(g)	Char Gasification (Water Gas)

The devolatilization process is simulated using the two-competing-reactions model [34], Equation (6). Where, a pair of reactions (R₁, R₂) with different rates (

), and volatile yields (

) compete to release the volatiles. (6)

The rate of volatile release is given by Equation (7) [21]: (7)

where

is the mass of the raw coal, and

and

are the rate constants in Arrhenius form (Equation (8)). (8)

On this equation,

represents the Arrhenius rate constant and

is the activation energy in terms of temperature. The parameters used to model the devolatilization process is listed in Table 4. Based on the Ubhayakar model [34], the yield

is defined as the volatile content of coal resulted from the proximate analysis (dry and ash free, DAF). And the enhanced yield

is treated as being related to

Table 4.

Kinetics of devolatilization modelling [21].

Devolatilization Yields	Arrhenius kinetics values
Devolatilization Yields	(s⁻¹)	(K)
	3.7 × 10⁵	18000
	1.46 × 10¹³	30189

Gaseous combustion

The gaseous combustion is described by the reactions between the released volatiles and O₂ to form CO₂ and H₂O. They are modelled using the Eddy Dissipation Model [35], Equation (9), that assumes velocity of reaction is controlled by turbulent diffusion. So, the reaction rate, , is proportional to the time required to mix reactants at the molecular level, and it is defined by the turbulent kinetic energy, , and dissipation, [32]. (9) where is an Eddy Dissipation Model coefficient and is defined as 4.0, is the molar concentration of the reactant components, and is the stoichiometric coefficient for each specie.

Oxidation and gasification of the residual char

Residual char is considered as pure carbon, and its heterogeneous reactions with O₂, CO₂, and H₂O are modelled applying the Gibb Model [32]. This approach considers the oxygen external diffusion on the char particles, also the oxygen diffusion inside the pores of the char particles. So, the change in the char mass is given by Equation (10). (10) The far field oxygen concentration is taken to be the time-averaged value obtained from the gas phase calculation, and is the density of the char. Physically, is the rate of external diffusion, is the surface reaction rate, and represents the rate of internal diffusion and surface reaction, Equation (11). (11)

where

is the carbon oxidation rate defined by the modified Arrhenius equation (Equation (12)) and

is described by Equation (13). The values for the model constants

and

, used for each one of the heterogeneous reactions, are listed on Table 5. (12)

(13)

In the previous equation,

is the void fraction,

is the volume/internal surface area ratio, and

is the resistance to flow in porous media, and

is the pore diffusivity.

Table 5.

Kinetics of the heterogeneous reactions [21].

Reaction	(m s⁻¹ K⁻¹)	(K)
Char_(s) + O_2(g) → CO_2(g)	14	21 580
Char_(s) + CO_2(g) → 2CO_(g)	20 230	39 743
Char_(s) + H₂O_(g) → CO_(g) + H_2(g)	606.9	32 406

Model validity

The proposed model is validated against data from a pulverized coal combustion test rig used between 1998 and 2000 in BHP Newcastle Laboratories [30,31]. The modelled domain, Figure 1, presents an inclined coaxial lance, 6 degrees, positioned on the centre of the reactor. Three different gas streams are injected into the domain, carrier gas for the coal, coaxial gas (to reduce the temperature of the lance), and a hot air blast. As this is a symmetric geometry, a symmetrical plane was created, and the simulated domain consists of one symmetrical half. The properties of the simulated coals are shown in Table 6, where Q factor is used to represent the volatile yield enhancement of devolatilization reaction at high temperature. And their granulometric distribution, using the Rosin Rammler method, is represented by Figure 2.

Figure 1.

Domain dimensions (mm).

Figure 2.

Rosin Rammler distribution for the Coals A, B, and C.

Table 6.

Physical and chemical properties of the simulated coals.

	Coal A	Coal B	Coal C
Moisture (%)	1.2	3.2	2.7
Volatile matter (%)	19.95	32.5	26.85
Ash (%)	9.7	9.8	8.9
Fixed carbon (%)	69.1	54.5	61.6
Sulphur (%)	0.34	0.58	0.33
Gross specific energy (MJ kg⁻¹)	31.94	30.08	30.22
C (%)	89.1	83.5	85.2
H (%)	4.7	5.3	5.1
N (%)	1.7	1.95	1.7
S (%)	0.37	0.60	0.34
O (%)	4.1	8.6	7.77
d₅₀ (μm)	41.0	63.0	48.0
Q Factor	1.1	1.45	1.4

The model validation was done in two steps. First, under the same boundary conditions (Table 7), the results were compared to the ones from the model developed by Shen et al. [21]. The main comparison factor is burnout that can be calculated by Equation (14). (14) where is the ash content of the raw coal and is the ash content of the residual. The burnout evolution, for particles between 10 and 200 μm, obtained by the model is compared to the results from Shen et al. [21], Figure 3. Both curves are very similar, presenting the same tendencies. In the middle part of the reactor, the developed model overpredicts burnout due the recirculation of small particles, which react rapidly, and they are highly affected by the turbulent dispersion.

Figure 3.

Model validation with Shen et al.’s model [21].

Table 7.

Boundary conditions for validation with Shen et al.’s model [21].

	Flow rate	Temperature
Hot air blast (20.9 wt-% O₂)	301 Nm³ h⁻¹	1474 K
Coaxial gas (20.9 wt-% O₂)	3.2 Nm³ h⁻¹	600 K
Carrier gas (100% N₂)	2.0 Nm³ h⁻¹	323 K
Coal	25.5 kg h⁻¹	320 K

The second part of the validations is the comparison, in terms of burnout values, to the experimental combustion tests, presented by Shiozawa, for different coals and their binary blends (in a 50% proportion of each coal) [36], Table 8. For the experimental measurements, data is collected at the transit time of 20 milliseconds, 925 mm downstream from the injection point and by averaging the burnout collected at +50 mm, centre line, and −50 mm [36]. For the model predictions, the averaged burnout is collected at the same 925 mm downstream, considering the whole reactor circumference. The whole circumference was chosen to better represent the average burnout, collecting burnout by points can be very unprecise, due all the uncertainties involved in the turbulent dispersion of particles. These uncertainties could justify the differences between the experimental and predicted results. So, qualitatively, the model presented a good prediction of the experimental values.

Table 8.

Second part of the burnout validation.

	Operational conditions				Burnout
	Blast flow (Nm³ h⁻¹)	Blast Temp (K)	Blast O₂ (%)	Coal Rate (kg h⁻¹)	Burnout
					Experimental [36]	Model
Coal A	298	1473	20.8	35.9	55.5	63.27
	298	1475	25.9	52.3	60.6	64.36
	296	1474	21.9	41.8	52.5	60.95
Coal B	303	1474	20.9	40	75.5	65.74
	299	1475	21.8	44	70.7	66.45
	299	1474	25.9	43.9	76.2	71.72
Coal C	304	1474	20.9	39.4	68.9	65.31
	295	1475	21.8	46.8	62.7	63.41
	298	1474	25.6	48.7	73.9	71.35
Blends A-B	300	1473	21.1	38.2	71.5	62.66
	300	1473	21.9	43	66.8	68.52
Blends B-C	300	1473	21.00	24.8	72.9	66.33
	300	1473	25.9	47.2	72.7	67.63

Model application

By using the same coals and domain from the validation cases, the developed model was applied to investigate the different oxygen enrichment methods in the blast furnace raceway, and how coals with different compositions and their blends react under those methods. The conventional and coaxial O₂ enrichment methods are shown in Figure 4. In the conventional, hot air enriched with oxygen is injected by the blast inlet. Whereas, in the coaxial enrichment method, pure oxygen is injected by the coaxial lance. Table 9 lists the fixed operational parameters, in the inlets, which are kept constant in all tests. While, Table 10 shows the boundary conditions for the different oxygen enrichments in each case. In case 1, the reference cases, there is no oxygen enrichment. For cases 2 and 3 oxygen enrichment is applied only in the blast, increasing the content of oxygen for this gas. For cases 4 and 5, pure oxygen is injected by the coaxial lance in a flow rate of 12 and 25 Nm³ h⁻¹, respectively. Then, this set of cases is repeated for Coals A, B, C, and their blends (AB, AC, and BC – in a proportion 50/50 in mass). So, a total of 30 different tests were simulated. All the cases were simulated using the same geometry, with the coaxial lance. However, for the reference and conventional enrichment cases, a typical atmospheric gas (21% of O₂) was injected by the coaxial lance in a low flow rate (3.2 Nm³ h⁻¹). This gas is responsible to cool down the lance in the simulated test rig.

Figure 4.

Schematic diagram showing the oxygen enrichment methods.

Table 9.

Fixed operational parameters in all cases.

	Flow rate	Temperature
Blast Inlet	301 Nm³ h⁻¹	1474 K
Coaxial Inlet	–	320 K
Carrier Gas (100% N₂)	2.0 Nm³ h⁻¹	323 K
Coal	30 kg h⁻¹	320 K

Table 10.

Boundary conditions for the set of cases.

Description	Cases Id	O₂ (Vol.-%)	Coaxial flow rate (Nm³ h⁻¹)
Reference	1	20.9	3.2
Blast enrichment	2	23.9	3.2
Blast enrichment	3	26.9	3.2
Coaxial enrichment	4	23.9	12.0
Coaxial enrichment	5	26.9	25.0

Results and discussion

Figure 5 shows, for all the O₂ enrichment tests, the averaged burnout collected at 925 mm downstream from the injection point. The conventional enrichment method (blast enrichment), in general, presented a better result than the coaxial enrichment method. Therefore, the highest burnout value achieved is 74.85% for Coal C under blast enrichment of 26.9% (Case 3 conditions). This coal has a very attractive combination of small size of particles (d₅₀ equal to 48μm) with high amount of volatile matter (26.85%). Both characteristic that aids to increase the burnout value.

Figure 5.

Burnout results for the O₂ enrichment study.

Increasing the amount of oxygen is beneficial to increase burnout, because there is more oxygen available to react with the particles and with the volatile released by them, which also tend to generate higher temperature values. Figure 6 shows the average temperature in the domain in each test.

Figure 6.

Average temperatures for the O₂ enrichment study.

Particle size

Size of particles is a major factor for the combustibility of a coal. Small particles have a higher surface area than larger ones, which help them to be heated up and to react faster. In this study, Coals A and C have the smallest sizes of particles with d₅₀ equal to, respectively, 41 and 48 μm, while Coal B presents 63 μm. However, Coal A is a low volatile coal (19.95%) with lower combustibility. The comparison of the burnout evolutions for different group of particles for all coals in their Reference Cases is shown in Figure 7, where the importance of particle size can be seen. Smaller particles (between 20 and 30 μm) show a much higher burnout than the large ones (between 100 and 200 μm). And, due the higher volatile matter (VM) content, for all group of particle sizes, the final Coal B burnouts are higher than the others. So, when particles of the same size are considered, the volatile content is the most important factor to achieve the highest burnouts. Nevertheless, the general results, presented by Figure 5, showed a better performance of Coal C, because of its lower granulometry, which means that, this coal has is a higher content of particles smaller than 60μm in comparison to Coal B.

Figure 7.

Particle sizes comparison for Coals A, B, and C in their Reference Cases.

Volatile matter release

Burnout is highly affected by the volatile matter (VM) release from the coal. This is one of the main stages during coal combustion. The released VM reacts with oxygen, producing heat, and helping on the char combustion. Figure 8 exhibits the amount of VM inside the particles during their trajectory inside the reactor. Due its granulometry and composition, the curve for Coal C decreases faster than the ones for Coal A and Coal B. And, the particle temperatures for Coal C, Figure 9, are considerably higher than the one representing Coal B. Therefore, the faster volatile release (occurred because of the lower granulometry) aids to increase particle temperatures. Also, higher temperatures help the volatile release. So, this combination of cause and effect is another major factor that contributes to achieve higher burnouts. So, in the simulated cases, particle size is the most predominant factor to improve coal combustion, comparing medium and high VM coals.

Figure 8.

Volatile matter evolution for the Reference Cases of all coals and blends.

Figure 9.

Particle temperature evolution for the Reference Case of all coals and blends.

Differences between blast and coaxial O₂ enrichments

Coaxial O₂ enrichment is an alternative to increase the oxygen content being injected. Although, analysing the results from Figure 5, Cases 5 (coaxial O₂ enrichment of 26.9%) usually exhibits the minimum burnouts among the enrichment cases. In this case, to achieve high amount of oxygen, it’s necessary a high flow rate of gas being injected by the coaxial lance, 25 Nm³ h⁻¹ against 3.2 Nm³ h⁻¹ for the conventional cases. This increase of the amount of gas being injected results in a much higher gas phase velocity in the coaxial lance exit region, Figure 10, comparing cases with the same O₂ content for a certain coal. In the regular cases (flow rate of 3.2 Nm³ h⁻¹), the gas phase velocity on the proximity of coaxial lance reaches 14.7 m s⁻¹, while in the coaxial enrichment Case 5, it could reach 117 m s⁻¹.

Figure 10.

Velocity profiles inside the tuyere of Case 3 (a) and Case 5 (b) for Coal B.

This much higher velocity close to the lance exit has some negative impacts on coal combustibility. First, it concentrates the particles in a central region of the reactor. And it’s already known that one way to improve coal combustibility is injecting particles in a more dispersed phase. A more dispersed particle phase increases O₂ availability, consequently, increasing O/C ratio, which benefits the oxidation reactions, increasing burnout. Figure 11 shows the O₂ molar fraction for one enrichment case by the blast (Case 3) and other by coaxial lance (Case 5), both with 26.9% of O₂. The conventional enrichment method provides a more dispersed amount of O₂, which is better to increase O₂ availability.

Figure 11.

O₂ molar fraction profiles of Case 3 (a) and Case 5 (b) for Coal B.

Other negative effect is generated by the high velocity combined with the cooling effect caused by a large amount of gas under low temperature (320 K). Then, the elevation of temperature is further from the lance, Figure 12. In a blast furnace, as in this simulated test rig, there is a limited raceway length. So, a maximum amount of coal must be burned until a certain distance from the lance. Therefore, an increase of particle temperature closer to lance is beneficial to achieve higher burnout values. Figure 13, shows a comparison of particle temperature evolution for all cases injecting Coal B. Case 5 has the slowest elevation of particle temperature, reaching the lowest value, which explains the lowest burnout compare to the other cases.

Figure 12.

Temperature profiles of Case 3 (a) and Case 5 (b) for Coal B.

Figure 13.

Particles temperature for Coal B cases.

The high gas phase velocity also has impacts on the particle residence time inside the raceway. As can be seen in Figure 14, for Coal B, particles present the lowest residence time in Case 5, which means that they had less time to react. There is an increase of residence time comparing the reference (Case 1) with the blast enrichment cases (Case 2 and Case 3), this is justified by the higher average temperature, which decreases the density of the gas phase, reducing the intensity of the drag force acting over the particles. These effects observed for Coal B, shown in Figures 10–14 are also observed for the other coals and blends.

Figure 14.

Particle residence time for Coal B cases.

Effect of blending coals

The only case where the blend showed an improvement in burnout compare to both single coal cases was for Blend AB, Figure 5. This blend is the combination of two very different coals. Coal A presents the smallest granulometric distribution, and the lowest volatile matter content. While Coal B has the largest size of particles and the highest volatiles content. Figure 15 exhibits the burnout evolution for the single coal cases (Coal A and Coal B) and for Coal A and B in the blend, but being isolated from each other to analyse burnout. Burnout from Coal B in the blend had the highest difference from the single coal test in the middle part of the test rig (around 0.5 m distant from the lance). While, Coal A in the blend had the highest difference in the final region (after 0.8 m). The improvement in Coal B is explained by the higher thermal contribution provided by Coal A, due its composition and particle size. And the increasing in burnout for Coal A could be justified by a lesser competition for O₂ to the heterogenous reactions in the final part of the reactor, because there is lower carbon content in Coal B composition compared to Coal A. This synergetic effect is an option to increase the performance of both coals.

Figure 15.

Burnout comparison for single Coals A and B and in the blend.

Conclusions

In this study, a three-dimensional computational model of coal combustion was developed and validated. Then, it was applied to study the effects of two different O₂ enrichment methods, conventional and by the coaxial lance, for different coals and their blends. The main conclusions of this study are summarized as follows:

Coal C presented the best burnout values among the other coals and blends. This good performance is due a smaller particle size combined with a high volatile matter content. Both features have major influences in coal combustion. Smaller particles have higher surface area and also are more easily dispersed, improving its combustibility. Coal C presents a faster VM release compare to the Coal B and it achieves a higher temperature due its particle size. So, for the studied cases, particle size was the most predominant factor, when moderate VM coal is compared to a high VM coal.

After a certain point, coaxial O₂ enrichment starts to prejudice coal combustibility due the cooling effected caused by the larger amount of gas being injected by the coaxial lance, which delays the increase of temperature. The higher velocity on the coaxial exit zone also concentrates and accelerates the particles, decreasing its residence time. Those harmful effects for the coaxial O₂ enrichment method are observed in Cases 5, decreasing burnout.

Blending Coal A with Coal B resulted in a higher burnout for both coals. This is due a synergetic effect caused by the combined qualities of each coal. Coal A is smaller and has a higher colorific power. While, Coal B has more volatiles and it is more easily consumed.

Footnotes

Acknowledgements

The authors are grateful to the Metallurgical, Materials and Mining Engineering Postgraduate Program and the Laboratory of Simulations (LaSim), both located in UFMG. The authors would also like to acknowledge the financial support provided by CAPES (‘Coordenação de Aperfeiçoamento de Pessoal de Nível Superior’).

Disclosure statement

No potential conflict of interest was reported by the author(s).

ORCID

Diego Silva de Deus Vieira

Leandro Rocha Lemos

References

Zhou

Liu

Wang

et al. Effect of local oxygen-enrichment ways of oxygen-coal double lance on coal combustion. ISIJ Int. 2017;57(2):279–285. doi:10.2355/isijinternational.ISIJINT-2016-535.

Wijayanta

Alam

Nakaso

et al. Combustibility of biochar injected into the raceway of a blast furnace. Fuel Process Technol. 2014;117:53–59. doi:10.1016/j.fuproc.2013.01.012.

Kuang

Recent developments in the modeling and simulation of blast furnace ironmaking. Steel Res Int. 2017;87(9999):1–25. doi:10.1002/srin.201700071.

Shen

Guo

et al. Three-dimensional modelling of in-furnace coal/coke combustion in a blast furnace. Fuel. 2011;90(2):728–738. doi:10.1016/j.fuel.2010.08.030.

Majeski

Runstedtler

D’Alessio

et al. Injection of pulverized coal and natural gas into blast Furnaces for Iron-making : lance positioning and design. ISIJ Int. 2015;55(7):1377–1383. doi:10.2355/isijinternational.55.1377.

Castro

Nogami

Yagi

Three-dimensional multiphase mathematical modeling of the blast furnace based on the multifluid model. ISIJ Int. 2002;42(1):44–52. doi: 10.2355/isijinternational.42.44

Liao

Shen

Modelling the injection of upgraded brown coals in an ironmaking blast furnace. Powder Technol. 2017;314:550–556. doi:10.1016/j.powtec.2016.11.005.

Shen

Maldonado

Guo

et al. Computational fluid dynamics study of pulverized coal combustion in blast furnace raceway. Ind Eng Chem Res. 2009;48:10314–10323. doi:10.1021/ie900853d.

Ballal

NB.

Some challenges and opportunities in blast furnace operations. Trans Indian Inst Met. 2013;66(5–6):483–489. doi:10.1007/s12666-013-0308-0.

10.

Chen

Lucas

JA.

Pulverized coal burnout in blast furnace simulated by a drop tube furnace. Energy. 2010;35:576–581. doi:10.1016/j.energy.2009.10.028.

11.

Castro

Araújo

GDM

Mota

IDO

et al. Analysis of the combined injection of pulverized coal and charcoal into large blast furnaces. J Mater Res Technol. 2013;2(4):308–314. doi:10.1016/j.jmrt.2013.06.003.

12.

Kalkreuth

Borrego

Alvarez

et al. Exploring the possibilities of using Brazilian subbituminous coals for blast furnace pulverized fuel injection. Fuel. 2005;84:763–772. doi:10.1016/j.fuel.2004.11.007.

13.

Sahajwalla

Kong

et al. Chemical structure of chars prepared under conditions prevailing in the blast furnace PCI operation. ISIJ Int. 2002;42(8):816–825. doi:10.2355/isijinternational.42.816.

14.

Zhou

Huo

Wang

et al. Effect of oxygen-coal lance configurations on coal combustion Behavior. Steel Res Int. 2017;88(1):1–9. doi:10.1002/srin.201600197.

15.

Chung

Han

Lee

JH.

Coke properties at tuyere level in blast furnace with pulverized coal injection. Met Mater. 1996;2(1):1–7. doi:10.1007/BF03025940.

16.

Chung

J-K

Hur

N-S.

Tuyere level coke coal injection characteristics In blast furnace with pulverized. ISIJ Int. 1997;37(11):9–1. doi:10.2355/isijinternational.37.119.

17.

Wijayanta

Alam

Nakaso

et al. Numerical study on pulverized biochar injection in blast furnace. ISIJ Int. 2014;54(7):1521–1529. doi:10.2355/isijinternational.54.1521.

18.

Ueda

Natsui

Nogami

et al. Recent progress and future perspective on mathematical modeling of blast furnace. ISIJ Int. 2010;50(7):914–923. doi:10.2355/isijinternational.50.914.

19.

Miao

Zhou

et al. CFD-DEM simulation of raceway formation in an ironmaking blast furnace. Powder Technol. 2017;314:542–549. doi:10.1016/j.powtec.2016.11.039.

20.

Okosun

Silaen

Zhou

CQ.

Review on computational modeling and visualization of the ironmaking blast furnace at Purdue University Northwest. Steel Res Int. 2018;1900046:1–16. doi:10.1002/srin.201900046.

21.

Shen

Guo

et al. Three-dimensional modelling of coal combustion in blast furnace. ISIJ Int. 2008;48(6):777–786. doi:10.2355/isijinternational.48.777.

22.

Shen

Guo

et al. Model study of the effects of coal properties and blast conditions on pulverized coal combustion. ISIJ Int. 2009;49(6):819–826. doi:10.2355/isijinternational.49.819.

23.

Shen

Guo

et al. A three-dimensional numerical study of the combustion of coal blends in blast furnace. Fuel. 2009;88(2):255–263. doi:10.1016/j.fuel.2008.08.013.

24.

Shen

AB.

Modelling of injecting a ternary coal blend into a model ironmaking blast furnace. Miner Eng. 2015;90:89–95. doi:10.1016/j.mineng.2015.12.009.

25.

Shen

Austin

et al. Modelling in-furnace phenomena of pulverized coal injection in ironmaking blast furnace: effect of coke bed porosities. Miner Eng. 2012;33:54–65. doi:10.1016/j.mineng.2011.10.014.

26.

Shen

Austin

et al. CFD study of in-furnace phenomena of pulverised coal injection in blast furnace: effects of operating conditions. Powder Technol. 2012;223:27–38. doi:10.1016/j.powtec.2011.07.020.

27.

Shen

Shiozawa

Austin

et al. Model study of the effect of bird’s nest on transport phenomena in the raceway of an ironmaking blast furnace. Miner Eng. 2014;63:91–99. doi:10.1016/j.mineng.2014.01.008.

28.

Shen

Characterization of coal burnout in the raceway of an ironmaking blast furnace. Steel Res Int. 2015;86(6):604–611. doi:10.1002/srin.201400333.

29.

Castro

Baltazar

AWS

Silva

AJ.

A three dimensional three phase reactive flow for simulating the pulverized coal injection into the channel of the blast furnace raceway. WIT Trans Eng Sci. 2005;50:207–216.

30.

Guo

Zulli

Rogers

et al. Three-dimensional simulation of flow and combustion for pulverised coal injection. ISIJ Int. 2005;45(9):1272–1281. doi:10.2355/isijinternational.45.1272.

31.

Mathieson

Truelove

Rogers

Toward an understanding of coal combustion in blast furnace tuyere injection. Fuel. 2005;84(10):1229–1237. doi:10.1016/j.fuel.2004.06.036.

32.

ANSYS . ANSYS CFX-Solver theory guide. Canonsburg (PA): ANSYS Inc; 2016.

33.

Gosman

Ioannides

Aspects of computer simulation of liquid-fueled combustors. J Energy. 1983;7(6):482–490. doi: 10.2514/3.62687

34.

Ubhayakar

Stickler

Von Rosenberg

et al. Rapid devolatilization of pulverized coal in hot combustion gases. 16th Symp Combust. 1977:427–436. doi: 10.1016/S0082-0784(77)80342-1

35.

Magnussen

Hjertager

BH.

On mathematical modeling of turbulent combustion with special emphasis on soot formation and combustion. Symp Combust. 1976;16(1):719–729. doi:10.1016/S0082-0784(77)80366-4.

36.

Shiozawa

Numerical modelling of multiphase flow in raceway of ironmaking blast furnace. The University of New South Wales (Thesis), 2013.

Effect of coaxial oxygen enrichment on three different coals and their blends

Abstract

Introduction

Material and methods

Model description

Governing equations for flow

Variables Φ, Γeff, and SΦ of the governing equations for gas phase.

Governing equations for the solid phase [32].

Coal devolatilization

Chemical reactions considered in the model.

Kinetics of devolatilization modelling [21].

Gaseous combustion

Oxidation and gasification of the residual char

Kinetics of the heterogeneous reactions [21].

Model validity

Physical and chemical properties of the simulated coals.

Boundary conditions for validation with Shen et al.’s model [21].

Second part of the burnout validation.

Model application

Fixed operational parameters in all cases.

Boundary conditions for the set of cases.

Results and discussion

Particle size

Volatile matter release

Differences between blast and coaxial O2 enrichments

Effect of blending coals

Conclusions

Footnotes

Acknowledgements

Disclosure statement

ORCID

References

Differences between blast and coaxial O₂ enrichments