Experimental Thermal and Fluid Science 54 (2014) 212–218

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science

journal homepage: www.elsevier.com/locate/etfs

Experimental investigation of ignition behavior for coal rank using

a flat flame burner at a high heating rate
Ryang-Gyoon Kim 1, Dongfang Li 1, Chung-Hwan Jeon ⇑
School of Mechanical Engineering, Pusan Clean Coal Center, Pusan National University, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: The ignition behavior of pulverized coal particles was investigated as a function of different ranks and
Received 5 September 2013 sizes using a flat flame burner at high heating rate conditions (>105 K/s). A high-speed camera was used
Received in revised form 23 November 2013 to image the ignition process. Five coal types (anthracite, medium-volatile bituminous, high-volatile bitu-
Accepted 21 December 2013
minous, subbituminous, and lignite coals) with particle sizes 150–200, 75–90, and <45 lm were tested.
Available online 8 January 2014
The released volatile matter of medium- and high-volatile bituminous coal in the size ranges 150–200
and 75–90 lm underwent homogeneous ignition. When the particle size is <45 lm, high-volatile
Keywords:
bituminous coal underwent homogeneous ignition, while medium-volatile bituminous coal underwent
Pulverized coal particle
High heating rate
heterogeneous ignition. For particle sizes in the range 150–200 lm, anthracite coal exhibited homoge-
Ignition neous ignition after primary fragmentation, whereas lignite coal underwent direct fragmentation and
Fragmentation homogeneous ignition prior to ignition without primary fragmentation.
Coal rank Crown Copyright Ó 2014 Published by Elsevier Inc. All rights reserved.
Particle size

1. Introduction However, combustion problems occur when non-designed coal is

Worldwide, coal continues to be one of the most widely used
primary fuel sources owing to its abundance, low cost, and easy
availability with respect to safety and stability. Applications utiliz-
ing coal have been under investigation for centuries; numerous
types of coal technologies have been developed to produce elec-
tricity energy using a pulverized coal boiler and iron using a blast
furnace. The pulverized coal boiler technology remains the most
widely used to date and is expected to be the dominant method
for coal combustion over the next two or three decades, especially
in the field of power production. The blast furnace process is still
the most widely used technology to produce iron. Its major advan-
tages are a very high production rate and a high degree of heat uti-
lization [1]. However, the pulverized coal boiler and blast furnace
use different coal ranks. In a pulverized coal boiler, middle-rank
coal is used, with a heating value of approximately 6080 kcal/kg.
Pulverized coal injection (PCI) in the case of a blast furnace
requires high-rank coal, whose heating value is above 7500 kcal/kg.
Most power plants and iron making companies in South Korea pro-
duce power and iron using poor-quality imported non-designed
coal, because of the recent increase in the price of high-rank coal.
Abbreviations: PC, pulverized coal; VM, volatile matter.
⇑ Corresponding author. Tel.: +82 51 510 3051; fax: +82 51 582 9818.
E-mail address: chjeon@pusan.ac.kr (C.-H. Jeon).
1
These authors contributed equally to this work.
used in a pulverized coal boiler or a blast furnace, owing to its fuel
characteristics that make it difficult to obtain the optimal operat-
ing conditions. Moreover, the problem of burner damage must be
considered, which occurs because of the ignition mechanisms in-
volved in high-volatile low-rank coal. Therefore, to avoid burner
damage, the ignition behavior as a function of coal rank should
be investigated as fundamental data in order to take advantage
of the new operating procedure of the burner.
Most practical systems deal with combustion under turbulent
conditions, in which the heat release is much faster than that
under laminar conditions. At very high Reynolds numbers, the flow
becomes turbulent. In a turbulent flow, eddies move randomly
back and forth and across the adjacent fluid layers. The flow no
longer remains smooth and orderly. Even chemically non-reacting
turbulent flows are highly challenging because of the above men-
tioned characteristics. The problems become even more complex
when chemical reactions occur, since the turbulent fluid flow is
further coupled with chemical kinetics, and quite often, with phase
changes. For laminar flows, however, the adjacent layers of fluid
slide past one another in a smooth, orderly manner. The velocity,
temperature, and concentration profiles measured in a laminar
flow will be quite smooth [2]. Therefore, many investigators have
preferentially studied the coal combustion and ignition behaviors
under laminar flow conditions before studying these under turbu-
lent conditions, in order to solve the chemically reacting turbulent

0894-1777/$ - see front matter Crown Copyright Ó 2014 Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.expthermflusci.2013.12.017
 R.-G. Kim et al. / Experimental Thermal and Fluid Science 54 (2014) 212–218
flow. The mechanism of coal combustion on laminar flow condition
has been elucidated by previous investigations. The combustion of
most solid fuels involves two major steps: (i) devolatilization that
occurs during the initial heating; (ii) subsequent combustion of the
released volatile matter and porous solid residue (char) of the first
step. Ignition occurs between these two steps. Many previous stud-
ies [3] have investigated the ignition process of pulverized coal
(PC) particles, which may occur by either homogeneous ignition
of the released volatile matter or heterogeneous ignition on the
char surface depending on the coal rank [4–9], particle size [10],
coal-feeding rate [11], heating rate [9], and ambient gas conditions,
i.e., gas composition [6,7,12,13], oxygen concentration [6], and
temperature [11].
Ignition behaviors based on the particle size and coal type un-
der low heating-rate conditions have been studied By Chen et al.
[9], who examined the ignition of anthracite, bituminous coal,
and lignite by thermogravimetry (TG) and differential thermal
(DT) analyses, with PC sizes ranging from 37 to 4000 lm. They
proposed that with an increase in the coal quality (from lignite
to bituminous to anthracite), the ignition type changed from
homogeneous to hetero-homogeneous to heterogeneous, respec-
tively. Moreover, the ignition behavior of bituminous coal changed
from a hetero-homogeneous to homogeneous process with
increasing coal particle sizes, whereas in anthracite and lignite
coal, the behavior did not change with change in the range of
the particle sizes examined. However, Chen et al. [9] conducted
their study under low heating-rate conditions(<101 K/s). There-
fore, the investigation of reactors with heating rates similar to that
of PC boilers is required. McLean et al. [4] used a flat flame burner
designed for high heating rates (>105 K/s) and a shadowgraph to
observe the condensed matter surrounding the coal particles with
a mean mass size of 65 lm corresponding to the ignition of
bituminous coal—suggesting that homogeneous ignition initially
occurs. The shadowgraph technique offers the advantage of
detecting the soot cloud. Molina and Shaddix [13] used CCD
(Charged-Coupled Device) camera images with LFR to capture ex-
cited CH* emissions from the coal flame to determine the ignition
point. These results showed a significant ignition delay in char
compared to coal particle ignition, using coal from the same seam
and under similar experimental conditions as the McLean study
[4]. Shaddix and Molina [12] also used a CCD camera with a very
short exposure time (20 ls) to investigate the ignition and
devolatilization behavior of single particles of bituminous and
subbituminous coals by changing the ambient gas conditions.
The results showed that highly volatile bituminous coal ignited
with a high-temperature soot cloud; in contrast, this cloud was
not obvious during the ignition of subbituminous coal. All the
aforementioned studies describe how coal rank and particle sizes
affect PC particle ignition at various conditions; however, the
ignition phenomenon of PC particles, in a wide size range, at very
high heating rates (>105 K/s) as a function of coal rank is yet to be
thoroughly investigated.
Herein, we report our investigations on the ignition behavior of
PC particles of different ranks and sizes using a high-speed camera
at high heating-rate conditions (>105 K/s) in a flat flame burner
which is developed for setting up a similar heating rate of large
industrial scale. Five types of coal were examined, namely
anthracite, medium-volatile bituminous, high-volatile bituminous,
subbituminous, and lignite coals.
2. Experimental method
Five coal types of different ranks were investigated in this
study; Table 1 shows their properties. Coal rank was defined
using the ASTM method. Low-rank coals, which are lignite,
213
subbituminous coals, and high-volatile bituminous coals, are
classified according to the heating value and not the volatile con-
tent. Further, the basis of the volatiles content when comparing
high rank coals which are medium volatile, low volatile bitumi-
nous coals and anthracite are compared on a dry ash-free basis
(Table 1). Experiments were carried out in an entrained-type flat
flame burner at Pusan Clean Coal Center, South Korea. A schematic
of this system is shown in Fig. 1. The burner was designed with a
honeycomb structure and a circular cross section in order to ensure
symmetry in temperature and gas-composition conditions along
the horizontal plane. A quartz tube with a rectangular cross section
was set above the burner to isolate the inner reacting zone and de-
crease the heat loss. A non-premixed flat flame was created using
CO and H2 as a fuel and O2 as an oxidizer to provide a high-temper-
ature condition above the burner. N2 was also used to control the
temperature, oxygen concentration, and velocity of post flame
gas flows. The PC particle inlet was designed at the center of the
burner using N2 as a carrier gas. The calculated composition of
the post flame gas is shown in Table 2. The gas-temperature profile
along the burner centerline was measured with a 125 lm type R
thermocouple and corrected for radiation losses. The corrected
gas temperature as a function of the height above the burner is
shown in Fig. 2. In a blast furnace, pulverized coal particles are
injected into the hot blast from a lance with a small diameter.
The temperature and oxidant gas of the hot blast are 1300–1500 K
and air, respectively [14]. In a pulverized coal boiler, the coal particle
that passes through the burner nozzle encounters a 1700 K flame
zone [15]. Therefore, the experimental conditions were chosen as
21.6 mol% O2 and 1600 K.
PC particles were fed into the burner at a feeding rate of <1 mg/
min to ensure a single-particle feeding condition and to guarantee
that the reacting zone conditions would not be affected by the heat
of combustion of the PC particles. The ignition behavior of the PC
particles was captured via a Photron FASTCAM SA4 high-speed
camera. Because the ignition process occurs at the speed of
101 ms (including gas-phase and heterogeneous reactions under
high heating-rate conditions), a frame rate of 10,000 frames/s
was selected. In order to capture the PC particles before ignition
when no light is released, a backlight using a Photron HVC-UL
device was oriented toward the high-speed camera.
3. Results and discussion
3.1. Effect of coal rank: homogeneous and heterogeneous ignition
behavior
The effect of coal rank for particles in the size range
150–200 lm are shown in Fig. 3. Coal particles are injected into
the burner, and they are preheated by the flame in the downstream
direction. Then, volatile matter is evolved, and ignition occurs on
the particle surface or the gas phase. After almost all the volatile
matter is evolved, char combustion becomes dominant. Finally,
ash is produced after the char combustion is complete. The mass
of the coal particle decreases with an increase in the residence time
during the combustion progress. As mentioned in the Introduction
section, PC particle ignition can occur via either gas-phase combus-
tion of the released VM (homogenous ignition) or heterogeneous
combustion of the particle surface (heterogeneous ignition). Bitu-
minous (coals A, D) and subbituminous coals (Adaro coal) showed
a typical combustion behavior as seen in Fig. 3. As the coal particle
heated up after injection into the burner, homogenous ignition
occurred after the gas-phase VM was released from the particle.
As more gas-phase VM was released, the corresponding combus-
tion flame became bigger and surrounded the particle. The solid
coal particles remained black prior to the extinction of the
214 R.-G. Kim et al. / Experimental Thermal and Fluid Science 54 (2014) 212–218

Table 1
Proximate and ultimate analysis of different coals.

Coal FX A D Adaro Kideco

Rank Semi-anthracite Medium-volatile bituminous High-volatile bituminous Subbituminous Lignite
Proximate analysis (as received) (%) Moisture 1.2 1 2.5 5.2 32.2
Volatile matter 12.4 23 34.7 50.7 40.8
Fixed carbon 78.6 66.9 54.4 39.9 24.7
Ash 7.8 9.1 8.4 4.2 2.3
Ultimate analysis (dry ash-free basis) (%) C 92.3 90.5 84 70.7 67.8
H 4.5 5.1 5.7 4.9 6.1
O 0.9 2.8 7.9 23.4 24.3
N 1.8 1.2 1.8 0.9 1.1
S 0.5 0.4 0.6 0.1 0.7
Heating value (dry ash-free basis) (kcal/kg) 8109 7379 6914 6013 4200

Fig. 1. Schematic of flat flame burner and high-speed camera imaging system.

Table 2 (b) images 5, 6, (c) images 5, 6), assumed to be tar combustion

Calculated compositions of post flame gases (mol%).
CO2 H2O O2
20.9% 8.4% 21.6%
gas-phase volatile combustion; however, the particles became
luminous at the end of that the combustion process, indicating
the initiation of heterogeneous ignition. Thus, bituminous and
subbituminous coals could undergo both homogeneous and
heterogeneous ignition processes.
A comparison of bituminous (coal A, D) and subbituminous
(Adaro) coal samples revealed different volatile combustion behav-
iors. In bituminous coal, prior to the extinction of the gas-phase
volatile combustion, a condensed tail with its direction oriented
along the gas flow streamline formed above the particles (Fig. 3
[4,16]. Tar combustion continued even after the occurrence
N2 heterogeneous ignition. This result is in agreement with the sche-
49.1%
matic representation of the formation of soot tails as proposed by
Fletcher [17]. However, in subbituminous (Adaro) coal, a long tar
combustion tail was not observed during volatile combustion.
Previous studies [12,18–21] have shown that most subbituminous
coals have a relatively low content of tar during the devolatiliza-
tion process, while bituminous coals have a larger tar content.
3.2. Effect of coal rank: fragmentation ignition behavior
Anthracite (FX) and lignite (Kideco) coals exhibited an entirely
different phenomenon compared to that seen in bituminous and
subbituminous coal samples (Fig. 3). Prior to ignition, most of the
coal particles underwent the fragmentation. At the start of the
ignition, the particle fragmented into two or more pieces – a process
 R.-G. Kim et al. / Experimental Thermal and Fluid Science 54 (2014) 212–218 215

defined as primary fragmentation. Some of the primary fragments

separated from each other at very high velocities (4 m/s), while
others rotated and accelerated owing to the momentum changes
caused by separation during primary fragmentation (Fig. 4). VM
was released and ignited homogeneously after primary fragmenta-
tion. Homogeneous ignition occurred in a manner similar to that of
bituminous and subbituminous coals, although its diffusion flame
was much smaller compared to that of bituminous and subbitumi-
nous coals because of the latter’s lower VM content. Most particles
of lignite coal underwent fragmentation directly prior to ignition
without undergoing primary fragmentation; this coal fragmenta-
tion was shorter than that for anthracite coal.

3.3. Effects of particle size

In order to show the effect of the particle size, medium-volatile

(A) and high-volatile (D) bituminous coals were selected (Fig. 5).
Fig. 2. Temperature distribution in vertical direction above burner along centerline. Microscopic images of the particles showed that both

(a) Anthracite coal (FX)

(1) (2) (3) (4) (5)

(b) Medium-volatile bituminous coal (A)

(1) (2) (3) (4) (5) (6) (7)

(c) High-volatile bituminous coal (D)

(1) (2) (3) (4) (5) (6) (7)

(d) High-volatile subbituminous coal (Adaro)

(1) (2) (3) (4) (5) (6) (7)

(e) Lignite coal (Kideco)

(1) (2) (3) (4)

Fig. 3. Ignition behavior as a function of coal rank with particles sizes 150–200 lm.
216 R.-G. Kim et al. / Experimental Thermal and Fluid Science 54 (2014) 212–218

Fig. 4. Fragmentation of FX coal with particle sizes 150–200 lm: (a) coal particle before fragmentation; (b) primary fragmentation; (c) fragmentation with the ignition of
released volatile matter.

(a) 150–200 µm

Coal A

Coal D

(b) 75–120 µm

Coal A

Coal D

(c) <45 µm

Coal A

Coal D

Fig. 5. Ignition behavior of bituminous coal as a function of particle size using medium-volatile bituminous coal (Coal A) and high-volatile bituminous coal (Coal D).
 R.-G. Kim et al. / Experimental Thermal and Fluid Science 54 (2014) 212–218
Fig. 6. Ignition behavior as a function of coal rank and size.
medium- and high-volatile bituminous coal particles with size
150–200 lm exhibited similar phenomena, with homogeneous
and heterogeneous ignition occurring in sequence (Fig. 5a). For
the medium- and high-volatile bituminous coal particles
75–90 lm in size, delay times for both homogeneous and hetero-
geneous ignition were shorter than that exhibited by 150–200 lm
sized particles (Fig. 5b). However, for particles <45 lm, no volatile
combustion flame was detected for medium-volatile bituminous
coal prior to heterogeneous ignition. Initially, the solid particle
was partially ignited, exhibiting a dark red color, which subse-
quently spread over the entire surface and made it luminous
(Fig. 5c). These observations indicate that medium-volatile bitumi-
nous coal particles <45 lm are directly ignited in a heterogeneous
manner without homogeneous ignition; however, in comparison,
high-volatile bituminous samples exhibited distinctly different
phenomena. For particles <45 lm, significant volatile combustion
was observed prior to heterogeneous ignition, with homogeneous
ignition occurring as the initial ignition event. These results
can be attributed to the fact that the high-volatile bituminous
coal (D) has a more volatile content comparing with the
medium-volatile bituminous coal (A).
Finally, through careful observation of the ignition behavior, a
summary of the ignition behavior as a function of coal rank and
size can be proposed, as illustrated in Fig. 6. As mentioned before,
the ignition behavior is described through homogeneous/heteroge-
neous ignition and primary fragmentation/fragmented homoge-
neous ignition based on the effect of coal rank and size.
4. Conclusion
This study investigates the ignition behavior of five coals
varying in ranks and sizes under high heating-rate conditions.
Bituminous (A, D) and subbituminous (Adaro) coals with particle
sizes 150–200 lm underwent homogeneous ignition of the
217
released VM and heterogeneous ignition of char particles.
Anthracite coal (FX) underwent homogeneous ignition after pri-
mary fragmentation, whereas lignite coal (Kideco) simultaneously
underwent direct fragmentation and homogeneous ignition prior
to ignition, but not primary fragmentation. Both medium- and
high-volatile bituminous coal samples (coals A, D, respectively)
with particle sizes 150–200 and 75–90 lm exhibited similar
phenomena, with homogeneous and heterogeneous ignition occur-
ring in sequence. For particles <45 lm, high-volatile bituminous
samples underwent homogeneous ignition, while medium-volatile
bituminous coal exhibited heterogeneous ignition owing to a
reduced volatile content.
References
[1] W.P. Hutny, G.K. Lee, J.T. Price, Fundamentals of coal combustion during
injection into a blast furnace, Prog. Energy Combust. Sci. 17 (1991) 373–395.
[2] K.K. Kuo, Principles of Combustion, John Wiley & Sons, 1986 (chapter 7).
[3] R.H. Essenhigh, M.K. Misra, D.W. Shaw, Ignition of coal particles: a review,
Combust. Flame 77 (1989) 3–30.
[4] W.J. McLean, D.R. Hardesty, J.H. Pohl, Direct observations of devolatilizing
pulverized coal particles in a combustion environment, Symp. Int. Combust. 18
(1981) 1239–1248.
[5] M. Gieras, R. Klemens, P. Wolanski, Experimental and theoretical study of
ignition of single coal particles at zero gravity, Acta Astronaut. 13 (1986) 231–
239.
[6] R. Khatami, C. Stivers, Y.A. Levendis, Ignition characteristics of single coal
particles from three different ranks in O2/N2 and O2/CO2 atmospheres,
Combust. Flame 159 (2012) 3554–3568.
[7] R. Khatami, C. Stivers, K. Joshi, Y.A. Levendis, A.F. Sarofim, Combustion behavior
of single particles from three different coal ranks and from sugar cane bagasse
in O2/N2 and O2/CO2 atmospheres, Combust. Flame 159 (2012) 1253–1271.
[8] Y.A. Levendis, K. Joshi, R. Khatami, A.F. Sarofim, Combustion behavior in air of
single particles from three different coal ranks and from sugarcane bagasse,
Combust. Flame 158 (2011) 452–465.
[9] Y. Chen, S. Mori, W.-P. Pan, Studying the mechanisms of ignition of coal
particles by TG-DTA, Thermochim. Acta 275 (1996) 149–158.
[10] J.B. Howard, R.H. Essenhigh, Mechanism of solid-partical combustion with
simultaneous gas-phase volatiles combustion, Symp. Int. Combust. 11 (1967)
399–408.
218 R.-G. Kim et al. / Experimental Thermal and Fluid Science 54 (2014) 212–218
[11] Y. Liu, M. Geier, A. Molina, C.R. Shaddix, Int. J., Pulverized coal stream ignition
delay under conventional and oxy-fuel combustion conditions, Int. J.
Greenhouse Gas Control 5 (2011) S36–S46.
[12] C.R. Shaddix, A. Molina, Particle imaging of ignition and devolatilization of
pulverized coal during oxy-fuel combustion, Symp. Int. Combust. 32 (2009)
2091–2098.
[13] A. Molina, C.R. Shaddix, Ignition and devolatilization of pulverized bituminous
coal particles during oxygen/carbon dioxide coal combustion, Symp. Int.
Combust. 31 (2007) 1905–1912.
[14] K. Ishii, Advanced pulverized coal injection technology and blast furnace
operation, Elsevier Science Ltd, 2000, (chapter 3).
[15] G.L Borman, K.W Ragland, Combustion engineering, McGraw-Hill, 1998.
Chapter 16.
[16] W.R. Seeker, G.S. Samuelsen, M.P. Heap, J.D. Trolinger, The thermal
decomposition of pulverized coal particles, Symp. Int. Combust. 18 (1981)
1213–1226.
[17] T.H. Fletcher, J. Ma, J.R. Rigby, A.L. Brown, B.W. Webb, Soot in coal combustion
systems, Progr. Energy Combust. Sci. 23 (1997) 283–301.
[18] J.D. Freihaut, M.F. Zabielski, D.J. Seery, A parametric investigation of tar release
in coal devolatilization, Symp. Int. Combust. 19 (1982) 1159–1167.
[19] K.L. Smith, L.D. Smoot, T.H. Fletcher, R.J. Pugmire, The Structure and Reaction
Processes of Coal, Plenum Press, 1994.
[20] Y.A. Levendis, R. Sahu, R.C. Flagan, G.R. Gavalas, Post-ignition transients in the
combustion of single char particles, Fuel 68 (1989) 849–855.
[21] I.W. Smith, The combustion rates of coal chars: a review, Symp. Int. Combust.
19 (1982) 1045–1065.