AISTech 2019 — Proceedings of the Iron & Steel Technology Conference

6–9 May 2019, Pittsburgh, Pa., USA

DOI 10.1000.377.044

Carbonaceous Materials Transformation in Pulverized Coal Injection

Ka Wing Ng1, Xianai Huang2, Louis Giroux3 and Ted Todoschuk4

1
Natural Resources Canada, CanmetENERGY
1 Haanel Drive, Ottawa, ON, Canada, K1A 1M1
Phone: 613 996 8712
Email: kawing.ng@canada.ca
2
Natural Resources Canada, CanmetENERGY
1 Haanel Drive, Ottawa, ON, Canada, K1A 1M1
Phone: 613 947 6584
Email: xianai.huang@canada.ca
3
Natural Resources Canada, CanmetENERGY
1 Haanel Drive, Ottawa, ON, Canada, K1A 1M1
Phone: 613 996 7638
Email: louis.giroux@canada.ca
4
ArcelorMittal Dofasco Inc.
1390 Burlington Street East, Hamilton, ON, Canada, L8N 3J5
Phone: 905 548 4796
Email: ted.todoschuk@arcelormittal.com

Keywords: Pulverized Coal Injection, Carbon Type, Injection Rig

INTRODUCTION
Pulverized coal injection (PCI) is a common practice in modern-day blast furnace ironmaking. The primary goal of PCI is to
reduce expensive coke consumption on a per tonne basis of hot metal production, commonly referred to as coke rate. The
capability of PCI in reducing the coke rate is quantified by the coke replacement ratio (CRR). CRR is coal specific and is
defined as coke rate reduction by unit mass of PCI. Numerous correlations have been put forward in the literature for
calculating CRR from chemical composition and calorific value of PCI coal.
An important factor that determines the actual CRR is the combustion behavior of PCI coal upon rapid heating. When
pulverized coal is introduced into a hot blast, the coal particles undergo very rapid pyrolysis [1, 2]. Upon heating, the outer
surface of a coal particle first undergoes softening which then propagates toward the particle center, releasing gases as
heating continues, causing the coal particle to swell. Due to the extremely fast heating of coal particles, the outer surface
reaches resolidification temperature prior to its center reaching softening temperature. As a result, coal particles are not able
to maintain a fluid state throughout the whole particle. Once the outer surface resolidifies, internal pressure starts to develop
due to the continuous softening and gas evolution. Eventually, the outer crust of the coal particle erupts from the internal
pressure. The residual solid after eruption forms cenospheres [3]. The eruption of a coal particle also leads to the explosive
release of gas and ejection of tar droplets into the hot blast. The rapid heating of these droplets results in dehydrogenation and
solidification. The solid particle produced is referred to as blast furnace soot [4].
CanmetENERGY has setup an experimental injection rig to simulate the combustion of pulverized coal in conditions similar
to those of industrial blast furnace tuyeres [5]. Since the commissioning of this rig in 2015, more than 20 coal samples have
been tested and valuable information on the combustion behaviour of coals has been collected. The rig and experimental
procedure was recently upgraded to allow systematic evaluation of combustion behavior of coal upon rapid heating.
A carbon type differentiation (CTD) technique was also developed at CanmetENERGY to quantify different types of
carbonaceous materials (char, coke, soot) in blast furnace dust and sludge samples. The technique has been successfully

© 2019 by the Association for Iron & Steel Technology. 427

implemented to monitor pulverized coal injection efficiency of industrial blast furnaces [6]. In this work, the CTD technique
was applied to analyse combustion residues collected in the injection rig to provide insight into the conversion of pulverized
coal into different types of carbonaceous materials.

RESULTS AND DISCUSSION

Pulverized Coal Injection Simulation Rig

Figure 1 shows the schematic of the CanmetENERGY PCI rig setup. The blast, which is heated by two preheating furnaces in
series, can accommodate different levels of oxygen enrichment formed by mixing air, oxygen and nitrogen accordingly. With
this configuration, the blast can be heated to about 700 C. Samples of pulverized coal in the hopper are delivered to the N2
carrier gas by a screw feeder. The sample is fed into a pneumatic transportation system and carried to the pre-heated blast just
before the reactor. The coal sample is then transported by the hot blast into the reactor chamber, which is preheated to over
1200 C. The reaction chamber is an alumina tube of 3.8 cm diameter and 80 cm in length. The estimated residence time of
coal particle in the reacting chamber is about 100 ms. The coal particle rapidly devolatilizes and the volatiles combust inside
the reaction chamber. The uncombusted particles exiting from the reaction chamber are water quenched and collected for
further analysis.

Figure 1. Schematic of CanmetENERGY PCI Rig Setup

As shown in Figure 1, the flow rate of each component gas is controlled by a dedicated mass flow controller. In a typical rig
test, the blast O2 enrichment is increased stepwise (21%, 23%, 25% and 28%) to examine the effect of O2 enrichment on coal
combustion behavior.
The delivery of coal to the hot blast and carried into the combustion chamber is controlled by adjusting the screw feeder
rotation speed. In a typical test, the feeding of coal lasts for 5 minutes. The average coal feeding rate is determined from the
rate of mass change of the hopper as illustrated in Figure 2. Combustion residue is collected for further analysis after 1
minute of injection. When the 4 minutes of sample collection is completed, the feeding of coal is stopped and blast O2
enrichment is changed to prepare for the next test with different blast level of O2 enrichment.

Figure 2. Coal Feeding Rate Measurement

428 © 2019 by the Association for Iron & Steel Technology.

The combustion chamber in the reactor is electrically heated to compensate for the heat loss to sustain the combustion of
coal. In order to eliminate the effect of electrical heat supply on temperature measurement, constant electric power is
delivered to the combustion chamber during the tests. The electric power setting in all tests is to maintain the initial reaction
chamber temperature at 1250 °C prior to coal injection. Since the electrical heat supply is maintained constant, the increase in
temperature recorded corresponds to the energy released by reaction of injectant with oxygen in the blast.
When combustion residue sample collection is completed, the feeding of coal is stopped and the combustion chamber is
allowed to cool down to the initial temperature before coal feeding is resumed at different blast %O2. Temperature at 20 cm
and 38 cm from the injection point is continuously monitored. Figure 3 shows the increase in temperature at these two
locations with different blast %O2 in a typical test.

Figure 3. Temperature Change with Feeding Time

Total Burnout Determination and Ash Tracer Assumption Validation

The combustion residues collected in the injection rig test are evaluated for proximate analysis for total burnout
determination. Total burnout is defined as the weight percentage of combustible materials gasified during the test, Eq. 1.

Total Burnout = (Combustibleinput - Combustibleoutput)/Combustibleinput (Eq.1)

Based on the ash tracer assumption, Eq.1 can be simplified and the total burnout can be calculated from the ash content
before and after combustion, Eq. 2.

Total Burnout = (1 - %Ashinput/ %Ashoutput )/(1 - %Ashinput ) (Eq. 2)

Equation 2 is obtained assuming the composition of ash in the original coal and collected combustion residues is identical. In
order to validate this assumption, an injection rig test was conducted to compare the chemical composition of ash in the
original coal with that of the combustion residues collected at different blast O2 enrichment, Table 1.

© 2019 by the Association for Iron & Steel Technology. 429

 Table 1. Ash Chemistry of Original Coal and Combustion Residue
Original Coal Combustion Residues
Blast O2 -- 21% 23% 25% 28%
Ash % 6.73 17.29 17.86 18.60 20.07
Proximate (db) VM % 35.82 1.81 1.45 1.41 1.44
FC% 57.45 80.90 80.68 79.99 78.49
SiO2 % 52.46 56.82 56.23 57.25 56.45
Al2O3 % 29.53 31.26 31.12 31.73 31.33
Fe2O3 % 8.53 4.84 4.60 4.67 5.18
TiO2 % 1.71 1.85 1.87 1.92 1.92
P2O5 % 0.139 0.080 0.080 0.075 0.071
CaO % 1.98 0.94 0.91 0.89 0.88
MgO % 0.70 0.64 0.65 0.66 0.64
SO3 % 1.74 0.31 0.29 0.22 0.19

Ash Chemistry Na2O % 0.54 0.43 0.46 0.38 0.33

(adb) K2O % 1.82 2.32 3.26 1.66 2.43
BaO % 0.112 0.097 0.101 0.099 0.095
SrO % 0.153 0.114 0.114 0.114 0.110
V2O5 % 0.064 0.051 0.049 0.049 0.048
NiO % 0.03 0.026 0.024 0.022 0.020
MnO % 0.039 0.024 0.026 0.023 0.026
Cr2O3 % 0.036 0.037 0.044 0.035 0.035
LOF % 0.42 0.16 0.18 0.20 0.24
Basicity 0.166 0.104 0.113 0.093 0.108

As shown, there are some changes in ash chemistry after combustion. In particular, the Fe2O3, P2O5 and CaO contents in the
combustion residues are significantly different from those in the original coal. It is clear that the basicity of the combustion
residues decreased. This implies that some basic components in the ash are lost during combustion. In order to evaluate
impact on burnout calculation, the total burnout is re-calculated based on conservation of SiO2, the most abundant component
in the ash. Since the ash composition is expressed on a dry ash basis, the increase in SiO2 shown in Table 1 is due to the
decrease in some of the other components in the ash. In order words, the amount of SiO2 input in the coal is likely same as
SiO2 output in the residue. Therefore, it is possible to calculate burnout by assuming conservation of SiO2.

Table 2. Effect of Calculation Methods on Total Burnout

Total Burnout
Assumptions
21% O2 23% O2 25% O2 28% O2
Conservation of Total Ash 65.48% 66.81% 68.42% 71.26%
Conservation of SiO2 68.13% 69.04% 71.06% 73.29%

Based on this assumption, the total burnout is calculated and compared with that calculated from ash tracer assumption
(conservation of total ash), Table 2. The calculated total burnout based on ash tracer assumption is about 3% lower than the
one calculated from conservation of SiO2. Since a small portion of ash is gasified during the test, the amount of ash in the
combustion residues is overestimated by the ash tracer assumption. Despite this discrepancy, the difference between both
calculation methods for burnout is small. Taking into account other uncertainties that may occur in the rig test, such as
variation in coal feeding rate, uncertainty in chemical analysis etc., it is legitimate to accept the use of the ash tracer
assumption method in total burnout calculation.

430 © 2019 by the Association for Iron & Steel Technology.

Combustion Behavior of PCI Coal Samples
A series of injection rig tests was conducted to examine the combustion behavior of typical PCI coal samples. Table 3
summaries the properties of coals examined. The coals cover a wide range of VM content (13.76% - 36.04%) and have
significantly different petrographic properties, namely wide Ro range and reactive/inert ratios. Coal D shows no plasticity
and dilatation . Whereas Coal B has highest maximum plasticity, Coal C has highest softening temperature and narrower
melting range.

Table 3. Properties of Coals

A B C D
Proximate (db) %Ash 8.11 7.99 5.19 8.61
%Volatile Matter 36.04 21.56 16.81 13.76
%Fixed Carbon 55.86 70.46 78.00 77.63
Ultimate (db) %C 76.60 81.70 85.90 83.20
%H 5.14 4.51 4.34 3.81
%N 1.70 1.32 1.25 1.17
%S 0.73 0.78 0.94 0.42
%O(by diff) 7.75 3.73 2.41 2.71
Gieseler Plasticity Softening Temp 407 440 467 --
Solidification Temp 458 508 510 --
Melting range 51 68 43 --
Max Fluidity 14 71.3 12.8 --
Ruhr Dilatation %Contraction 23 24 21 --
%Dilatation 30 48 30 --
Petrography V4 – V9 64.4 6.1 0 0
V10 – V15 0 50.2 6.4 0.3
V16 – V21 0 13.6 64.8 32.3
Total Reactive 76.2 72.3 73.1 41.70
Total Inert 23.8 27.7 26.9 58.3
Mean Ro 0.84 1.42 1.68 1.76

The combustion behavior of each coal at different blast O2 enrichment was examined in the injection rig. The coals were
pulverized using the same procedure. Table 4 lists the size distribution of pulverized coals. For all tests, the blast rate was
maintained constant and the coal feeding rates were kept similar to allow direct comparison.
Figure 4 presents the total burnout of the coals at different blast O2 enrichment, and showing the increase in total burnout
with increasing blast %O2. Compared to other coals, the increase in total burnout of Coal D with blast %O2 increase is
smaller, suggesting that combustion of Coal D is less sensitive to oxygen content in the blast than other coals examined. The
total burnout of Coal A is significantly higher than other coals examined in this work, and is likely due to the higher volatile
matter content of Coal A. However, even if the volatile matter content in Coal C is slightly higher than in Coal D, the total
burnout of Coal C was found to be lower than that of Coal D. This observation suggests that the volatile matter content of
coal plays an important role in the combustion behavior but is not necessary the only determining factor.

© 2019 by the Association for Iron & Steel Technology. 431

 Table 4: Pulverized Coal Size Distribution (wt%) for Injection Rig Tests
Size Range (µm)
Sample 250-212 212-150 150-106 106-75 75-45 45-38 38-0
A 4.24% 16.96% 18.55% 20.08% 20.90% 7.09% 12.18%
B 3.90% 12.42% 12.17% 15.73% 26.99% 13.58% 15.21%
C 1.60% 6.51% 11.24% 16.51% 25.64% 10.55% 27.95%
D 1.90% 11.04% 13.60% 20.21% 24.14% 8.98% 20.13%

Figure 4. Total Burnout of Coals

Combustion Residues Carbon Type Analysis

To further develop understanding between coal properties and combustion behavior of coal, carbon type analysis of the
combustion residues collected in injection rig tests was performed. A TGA technique was developed at CanmetENERGY to
quantify the amount of different carbon type (Char, Coke and Soot) in industrial blast furnace dust [6]. Upon rapid heating,
carbon in coal is converted into carbonaceous materials with different maturation level. The conversion depends on
fundamental properties of the original coal as well as thermal history of the combustion residues. Figure 5 shows a SEM
picture of combustion residue collected in an injection rig test. The cenosphere structure due to melting of coal particles and
rapid release of gas can be clearly seen. As illustrated, the morphology of particles with different carbon maturation levels are
significantly different.

Figure 5. Morphology of Combustion Residues

Carbonaceous materials with different maturation levels start to combust in air at different temperatures. By making use of
this characteristic, a specific heating profile is developed to combust coal stepwise for determination of different carbon type.
Figure 6 shows the TGA combustion rate profile of the residues collected in injection rig test at different blast O2 enrichment.
The sample weight loss rate peak below 200 C corresponds to drying of the sample. The first and second major peaks
correspond to combustion of “Char”, 400 C, and “Coke”, 850 C, respectively. By examining the area underneath each peak,

432 © 2019 by the Association for Iron & Steel Technology.

the relative amount of “Char” and “Coke” can be determined. As can be clearly seen, the amount of “Char” decreases while
“Coke” increases as O2 enrichment in the blast increases.
It should be emphasized that the term “Char” and “Coke” are simply used to distinguish between carbonaceous materials
starting to combust in air at different temperatures. “Char” refers to carbon material that starts to combust at relatively low
temperature, whereas “Coke” refers to carbon material that requires higher temperature to initiate combustion. The “Char”
and “Coke” identified by this TGA technique are not two distinct materials. In fact, it is likely that carbonaceous materials in
coal are transformed into a wide range of materials with different chemical and physical properties. The terminology
employed in this work is simply for categorizing carbonaceous materials in combustion residue for easy comparison.

Figure 6. TGA Carbon Type Differentiation

The results obtained from the carbon type analysis is on a combustion residue mass basis. Since the total burnout in each test
is different, the carbon type differentiation results cannot be directly compared. Using the information obtain in TGA analysis
together with burnout determination results, it becomes possible to determine the amount of original coal converted into gas
and different type of carbonaceous materials to facilitate direct comparison between coals.
Figure 7 shows the relationship between amount of feed converted into Char and Coke. As shown, the amount of
material/feed converted into char gradually decreases as the blast %O2 increases. This is due to the fact that char is relatively
easy to combust. As the availability of O2 increases, a larger amount of char is gasified resulting in the increase in total
burnout and reduction in feed conversion into char.
The amount of feed converted into char for Coal A is significantly higher than that of the other coals examined in this work.
This is due to fact that the volatile matter content in Coal A is high. The rapid release of large amount of gas upon rapid
heating results in vigorous eruption of coal particle leading to a larger percentage of feed converted into char. The rate of
decrease in char conversion with increasing blast %O2 among all coal samples examined in this work is very similar despite
the significantly different coal properties. This observation suggests that combustion behavior of char formed by rapid
heating is independent of coal properties.

Figure 7. Percentage Conversion of Coal into Char and Coke

© 2019 by the Association for Iron & Steel Technology. 433

As shown in Figure 7, the conversion of feed into coke gradually increases as the availability of O2 in the blast increases. As
discussed above, coke refers to carbonaceous materials in combustion residues, which are relatively difficult to combust.
Unlike char, coke is difficult to be gasified even by increasing the availability of O2 in the blast. In fact, experimental results
reveal that increase in blast %O2 favors the formation of coke.
Also, the conversion of feed into coke between coals vary significantly. The relatively low conversion into coke of Coal A is
likely due to rigorous eruption of coal particles as discussed above. The rigorous eruption results in the limited time available
for carbonization of coal into carbonaceous materials with higher degree of maturity. Therefore, the amount of coke formed
from Coal A is relatively small. For other coals with relatively low volatile matter content, the coal particles maintain their
integrity for longer time due to the milder release of volatile matter. This provides longer time for carbonization to occur and
results in larger amount of feed converted into coke.
It was also observed that the conversion into coke for Coal C is exceptionally high despite fact that its volatile matter is not
the lowest among coals examined in this work. This observation suggests that rheological and petrographic properties of coal
play an important role in amount of feed converted into coke upon rapid heating. Further study is needed to improve our
understanding of the role of coal properties on conversion behavior.
To summaries the above discussion, char is produced from the eruption of coal particle upon rapid heating. The severity of
eruption is determined by the amount of volatile matter released and affecting the amount of char produced. The amount of
coke that can be produced is likely related to the duration that the coal particle maintains its integrity. Properties of char are
independent from those of the original coal and char is relatively easy to be gasified. By increasing the blast oxygen content,
a larger amount of char can be gasified and the total burnout is enhanced. However, coke is relatively difficult to combust.
Once coke is formed, increasing oxygen content in the blast does not result in significantly larger amount of coke being
gasified. In fact, the amount of feed converted into coke is slightly increased. The total burnout achievable is strongly
affected by the degree of feed conversion into coke, which is dictated by the fundamental properties (rheology and
petrography) of coal selected for direct injection.

CONCLUSIONS
Using the CanmetENERGY injection rig, systematic study of combustion behavior of four PCI coals with very different
fundamental properties was carried out. Besides measuring the total burnout at different blast oxygen enrichment, carbon
type differentiation (CTD) analysis of the combustion residues collected in injection rig tests were also performed to develop
understand in carbonaceous material transformation upon rapid heating. Char is produced by eruption of coal particle. The
amount of feed converting into char is strongly affected by the volatile matter content of the coal. The combustion behavior
of char is similar regardless of its origin. Coke is relatively difficult to gasify. The extend of coal conversion to coke cannot
be reduced by increasing oxygen content in the blast. Hence, it determines the total burnout can be achieved.

REFERENCES
1. J.C. Chen and S. Niksa, “Coal Devolatilization during Rapid Transient Heating. 1. Primary Devolatilization”, Vol. 6,
No. 3, Energy and Fuels, 1992, pp 254-264.
2. J.C. Chen, C. Castagnoli and S. Niksa, “Coal Devolatilization during Rapid Transient Heating. 2. Secondary
Devolatilization”, Vol. 6, No. 3, Energy and Fuels, 1992, pp 264-271.
3. V.R. Gray, “The Role of Explosive Ejection in the Pyrolysis of Coal”, Vol. 67, No. 9, Fuel, 1988, pp 1298-1304
4. S. Pipatmanomai, A.H. Herod, T.J. Morgan, N. Paterson, D.R. Dugwell and R. Kandiyoti, “The Fate of Volatiles from
Injectant Coals and Soot Formation in Blast Furnace”, Vol. 18, No. 1, Energy and Fuels, 2004, pp 68-76.
5. Ray S, Giroux L, MacPhee T, et al. Evaluation of PCI coals in new injection facility at CanmetENERGY-Ottawa. In:
AISTech - Iron and Steel Technology Conference Proceedings, 2015. p. 926-37.
6. Ng KW, Giroux L, MacPhee T, Todoschuk T, Taggart L, Scott G, “Carbon type differentiation technique for
diagnosing pulverised coal injection efficiency”, Ironmaking & Steelmaking, 2016, 43:3, 214-219.

434 © 2019 by the Association for Iron & Steel Technology.