Professional Documents
Culture Documents
Fupeng Shan, Qizhao Lin, Kun Zhou, Yakun Wu, Wei Fu, Po Zhang, Lanbo Song, Chunyu Shao, Bolun Yi
Fupeng Shan, Qizhao Lin, Kun Zhou, Yakun Wu, Wei Fu, Po Zhang, Lanbo Song, Chunyu Shao, Bolun Yi
Fuel
journal homepage: www.elsevier.com/locate/fuel
h i g h l i g h t s
a r t i c l e i n f o a b s t r a c t
Article history: This study presents an experimental investigation of the ignition and combustion characteristics of single
Received 19 May 2016 biomass pellets in air and O2/CO2 atmospheres containing 21%, 30%, 40%, 50% and 100% oxygen mole frac-
Received in revised form 19 September tions. In this experiment, the temperature of the gas surrounding the biomass pellets was set respectively
2016
at 873 K, 973 K and 1073 K. The rate of co-flow gas was set at 10 L/min, 15 L/min and 20 L/min. Single
Accepted 23 September 2016
Available online 13 October 2016
pellets of Pinus bungeana and rice husk were fixed on a thermocouple and burned in a vertical heating
tube furnace. High-speed photography was used to record images of the combustion process of biomass
pellets. The resulting images exhibit two ignition behaviors: (1) homogeneous ignition of volatile, and (2)
Keywords:
Biomass pellets
heterogeneous ignition on the particles’ surface. After ignition, the combustion was no longer a homoge-
Ignition mechanism neous combustion of volatile. When the oxygen concentration exceeded 50%, the biomass showed a
Combustion characteristic hetero-homogeneous combustion. Similarly, when the concentration of O2 increased, the flame became
Oxy-fuel shorter and more stable. With the same oxygen concentration, and once the N2 was replaced by CO2, the
Single pellet ignition delay, internal ignition temperature and the volatile combustion time increased. Inversely, when
the oxygen concentration exceeded 21% (i.e. under the conditions of oxy-fuel), these parameters were
correspondingly reduced. In other words, the ignition and combustion intensity were increased. The
effects of oxygen concentration and co-flow temperature on ignition and combustion of biomass pellets
were greater than that of co-flow velocity. The study’s results show that the biomass pellets did not break
during combustion.
Ó 2016 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.fuel.2016.09.069
0016-2361/Ó 2016 Elsevier Ltd. All rights reserved.
278 F. Shan et al. / Fuel 188 (2017) 277–284
Oxy-fuel combustion is a promising technology used to pro- and gas temperature on the ignition and combustion
mote carbon dioxide capture and storage (CCS) by burning fuel characteristics.
in a mixture of oxygen and recycled flue gases (rather than in
air, as conventional combustion). The flue gases mainly contain
carbon dioxide. This technique of CCS needs more carbon dioxide 2. Experimental
to work properly, so biomass offers a great advantage. Riaza et al.
[4] have concluded that the combustion of biomass in the oxy- 2.1. Biomass samples
fuel atmosphere could be conducive to capturing carbon dioxide
and enhancing combustion intensity. Current research on biomass Two kinds of biomass pellets were used as the research object,
burning focuses more on co-firing biomass with coal. Some stud- which are Pinus bungeana and rice husk. They were obtained from a
ies suggest that the combustion of biomass-coal blends under pellet factory in Hefei City, Anhui Province, China. The pellets’
oxy-fuel conditions cannot only reduce carbon dioxide emissions diameter was 9 mm. Then the length of biomass pellet was pro-
and atmospheric carbon dioxide, but also control the maximum cessed to 10 mm. A hole with a depth of 5 mm and a diameter of
temperature of coal combustion [5–10]. This technology has been 1 mm was drilled into the biomass pellets. These column of bio-
called ‘‘clean” coal combustion. Other researchers have also stud- mass samples are similar to that in [15–17]. The proximate and
ied the combustion of pulverized biomass particles. Riaza et al. ultimate analyses of the biomasses are given in Table 1 (GB/T
[4] studied the combustion of single biomass particles in air 28731-2012).
and in oxy-fuel conditions in a drop-tube furnace. The tempera-
ture of the furnace wall was set to 1400 K, and the gas atmo-
2.2. Experimental equipment and methods
sphere was air and O2/CO2 containing 21%, 30%, 35% and 50%
oxygen mole fractions. Riaza et al.’s results show that increasing
An electrically-heated tube furnace with an inner diameter of
the oxygen mole fraction in the CO2 background gas enhanced
8.0 cm was used for the experiments, and the schematic diagram
combustion intensity of biomass and decreased the burnout time
is shown in Fig. 1. In this experiment, oxygen (1) and nitrogen
of volatiles and char residues. However, it increased the temper-
(2) were used to form air in the gases mixing chamber (7) before
ature of the burning char particles. Tran et al. [11] have con-
flowing through steady flow chamber (8) into the vertical heating
cluded that the torrefaction process of biomass particles
tube furnace (9). Carbon dioxide (3) and oxygen (mole fractions of
integrated with simulated oxy-fuel combustion flue gases can
21%, 30%, 40%, 50% and 100%) were used to simulate the oxy-fuel
reduce mass and energy yields. However, the biomass would
combustion conditions. The gas flow rate was set respectively
get better grindability. The mass yield (Ymass) and energy yield
to10 L/min, 15 L/min and 20 L/min.
(Yenergy) were determined by the following equation, respectively:
Ignition and combustion experiments of biomass pellets were
Y mass ð%Þ ¼ mmbiochar 100%, Y energy ð%Þ ¼ Y mass HHV
HHV biochar
100%. In
feedstock feedstock conducted under a flow gas condition. The gas temperature sur-
addition to the above coal-based studies, many researchers have rounding biomass pellets was respectively set to 873 K, 973 K
only studied biomass combustion. Most of the prior studies have and 1073 K. The ignition temperature of solid biomass was approx-
been applied to boilers and small heating devices. Grotkjær et al. imately 473 K [12,18]. Therefore, the heating rate of biomass was
[12] studied the ignition mechanism of wheat straw. In this sufficient. The gas temperature was determined by heating control
study, the straw sample was connected to the thermocouple cabinet (13) and a thermocouple (14) which measures feedback
and the supportive rod. Air flow temperature was set to 260– temperature. The furnace was heated until the temperature of
350 °C. The ignition temperature was determined according to the thermocouple (14) reached the set value.
the carbon conversion ratio. The flue gas was collected to mea- The biomass pellet (18) was fixed on a thermocouple (15) with
sure the amount of carbon monoxide and carbon dioxide. Results high-temperature resistant inorganic glue. The thermocouple (15)
of this study indicated that when the superficial gas velocity was supported by a bracket (16) and used to measure the internal
decreased, the ignition temperature increased. This was consis- ignition temperature of biomass pellets. When the gases flow and
tent with the homogeneous ignition mechanism. Boman et al. the temperatures of the gases stream were stable, the biomass was
[13] studied the combustion characteristics and environmental fed into the furnace. The biomass was ignited when the visible
performance of pelletized hardwood of aspen. In this study, bio- flame was observed, and the temperature demonstrator (12) indi-
mass pellets were placed in a sample holder. Their test method cated the internal ignition temperature of biomass. High-speed
had a great influence on the combustion flame, so their research photography (10) was used to record the ignition and combustion
only concentrated on the emissions and slagging. Kuo and Hsi
[14] studied the pyrolysis and ignition of single wooden spheres
in high-temperature airflow. They also analyzed the impact of Table 1
the wood spheres’ diameter, air velocity, and temperature on Proximate and ultimate analyses of the biomass.
their ignition rate. Their conclusion was that biomass anisotropy Component Pinus bungeana (wt%) Ricehusk (wt%)
had the largest effect on ignition time. The above studies are con-
(Ash)ara 4.84 11.67
cerned with the combustion of biomass powder particles (such as (Moisture)ar 6.54 7.45
wood spheres) in oxy-fuel and air. However, few studies have (Volatiles)ar 85.89 73.08
been carried out to examine the ignition and combustion of bio- (Fixed carbon)arb 2.73 7.80
(LHV)cdb (MJ/kg) 19.32 17.53
mass pellets under oxy-fuel conditions. Further research on this
topic is necessary. The present study is therefore useful for future (Carbon)db 47.17 39.88
(Hydrogen)db 6.43 5.54
applications of biomass pellets in power plants, the metallurgical
(Oxygen)db 42.5 36.71
industry, and other areas. (Nitrogen)db 0.07 0.46
In this paper, two kinds of biomass pellets (Pinus bungeana and (Sulfur)db 0.46 0.48
rice husk) were used to investigate their ignition and combustion
LHV = Low heating value.
characteristics under the conditions of oxy-fuel and air in a tube a
ar indicates the abbreviation of ‘as received basis’.
furnace. The thermocouple and high-speed photography were b
Calculated by difference.
c
used to record the effects of gas velocity, oxygen concentration db indicates the abbreviation of ‘as dry basis’.
F. Shan et al. / Fuel 188 (2017) 277–284 279
process of biomass pellets through the observation window (17), at small particles, it is possible to be blown away by the gas flow
a speed of 50 frames per second. and this will increase the emission of fly ash and heat loss.
Timing started when biomass pellets were injected into the fur-
Test error is mainly from the following three aspects: the ther-
nace and stopped once a visible flame emerged. The time measured
mocouple (14), the thermocouple (15) and ignition delay time. The
between both events is the ignition delay. Compared with
previous work has carried out a detailed analysis of these errors in
untreated biomass, biomass pellets have a higher bulk density.
this laboratory [19]. The gas temperature was obtained when the
Thus, the rate of devolatilization is slow, and ignition takes a long
gas flow was stable. The thermal radiation of furnace wall would
time. As shown in Figs. 3 and 4, three variables were selected to
produce error in the measurement of thermocouple (14). Accord-
study the changes of the ignition point in different conditions,
ing to the heat balance equation, the relative error of the temper-
which are oxygen concentration, gas temperature and gas flow
ature measured by the thermocouple (14) is about 3–8%. This error
rate. The following phenomena were observed from the charts:
is within the allowable range. The time interval between particles
(i) at the same oxygen concentration, the ignition delay in the
entering the furnace until they reached their designated position
nitrogen atmosphere was shorter than that in the atmosphere of
was approximately 0.4–0.8 s. The response time of the thermocou-
carbon dioxide. It has been confirmed that this is related to the
ple (15) is 0.5 s. So the error caused by the heating process of the
properties of carbon dioxide [5,20–22]. The specific heat capacity
interval can be disregarded; Ignition delay time was obtained by
of carbon dioxide was relatively larger than that of nitrogen and
high speed photography. The time interval between the two pic-
the diffusion rate of oxygen in the carbon dioxide was relatively
tures is 0.02 s. The max error of ignition delay is 0.4%. This error
lower than N2. The decrease of co-flow temperature led to a
is minor and can therefore also be disregarded.
decrease of heating rate on the surface of particles so that the igni-
tion delay increased. (ii) At the same temperature, the ignition
delay decreased with the increase of oxygen concentration and
3. Results and discussion
gas flow rate. When the oxygen concentration was less than 50%,
the decrease of the amplitude became large. This phenomenon
3.1. Phenomena observed
was confirmed by the heterogeneous ignition theory. Due to the
effects of high oxygen concentration and gas turbulence, the
As shown in Fig. 2, the ignition and combustion process of bio-
increase of oxidant and heat transfer enhanced the ignition of bio-
mass pellets were recorded by high-speed photography. Each row
mass pellets. Biomass pellets showed heterogeneous ignition char-
represents the combustion process of biomass pellets, and the ver-
acteristics. (iii) At the same gas flow rate, when the gas
tical column shows the oxygen concentration which increases
temperature was the same, oxygen concentration has less effect
from top to bottom. Each row has five pictures. The first one was
on ignition delay. In contrast, at the same oxygen concentration,
taken at the start of ignition, whereas the fifth one represents the
the influence of the co-flow temperature was quite large. This is
end of volatile combustion and the beginning of char combustion.
because the increase of temperature accelerated the heat transfer
As shown in the pictures, the ignition of biomass pellets at low
and the release of volatile matter.
oxygen concentrations was homogeneous (oxygen (21%) and car-
bon dioxide atmosphere was more obvious). Under all the condi-
tions of temperature and gas atmosphere, the flame of biomass 3.3. Internal ignition temperature of the single biomass pellets
became more and more luminous with the increase of oxygen con-
centration. A phenomenon was observed that biomass pellets did In this experiment, the internal ignition temperature of biomass
not break during combustion process. This was different with the pellets was measured. Although there is a certain temperature gra-
combustion of biomass powder [4]. Once the biomass broke into dient with the particle surface, their temperature change trend is
280 F. Shan et al. / Fuel 188 (2017) 277–284
Air
0.21
Oxygen concentration, xo2
0.30
0.40
0.50
1.00
Air
0.21
Oxygen concentration ,xo2
0.30
0.40
0.50
1.00
1073K 10L/min(e)
Air
0.21
Oxygen concentration, xo2
0.30
0.40
0.50
1.00
Fig. 2. Several groups of typical sawdust pellets ignition and combustion process under different conditions were displayed.
similar. Zhou [19], Ponzio [23], Kijo-Kleczkowska [24] and Lu et al. with the increase of oxygen concentration. As Fig. 2 above shows,
[25] confirmed this phenomenon. The value of internal ignition at the same temperature, when the oxygen concentration
temperature was obtained according to Fig. 5. exceeded 50%, the biomass was experiencing heterogeneous igni-
As shown in Fig. 6, when the oxygen concentration was less tion. This kind of ignition mode was also observed by Ponzio [23]
than 50%, the internal ignition temperature of biomass pellets and Essenhigh [26].
decreased obviously with the increase of oxygen concentration. As shown in Fig. 7, although the internal ignition temperature of
However, when the oxygen concentration surpassed 50%, the inter- biomass particles decreased slightly with the increase of oxygen
nal ignition temperature of biomass pellets decreased gradually concentration, the effect of co-flow temperature was obvious. From
F. Shan et al. / Fuel 188 (2017) 277–284 281
45 25 10
Vf=10L/min Vf=10L/min Vf=10L/min
Vf=15L/min Vf=15L/min Vf=15L/min
40 8
Vf=20L/min 20 Vf=20L/min Vf=20L/min
τd(s)
τd (s)
τd (s)
6
35 15
45 20 10
Vf=10L/min Vf=10L/min Vf=10L/min
40 Vf=15L/min Vf=15L/min Vf=15L/min
8
Vf=20L/min 15 Vf=20L/min Vf=20L/min
35
τd (s)
τd (s)
τd (s)
30 6
10
25 open symbols=air open symbols=air
open symbols=air 4 filled symbols=CO2
filled symbols=CO2 filled symbols=CO2
20
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
60 60 Tg=1073K
60 Tg=1073K
open symbols=air Tg=1073K open symbols=air open symbols=air
Tg=973K filled symbols=CO2 Tg=973K filled symbols=CO2 Tg=973K
50 filled symbols=CO2 50 50
Tg=873K Tg=873K Tg=873K
40 40 40
30 30
τd (s)
τd (s)
τd (s)
30
20 20 20
10 10 10
0 0 0
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
60 Tg=1073K
60 Tg=1073K
60 Tg=1073K
open symbols=air open symbols=air open symbols=air
filled symbols=CO2 Tg=973K filled symbols=CO2 Tg=973K filled symbols=CO2 Tg=973K
50 50 50
Tg=873K Tg=873K Tg=873K
40 40 40
30 30 30
τd (s)
τd (s)
τd (s)
20 20 20
10 10 10
0 0 0
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
the graph, it indicates that the 50% oxygen concentration should be effect of oxygen. The heterogeneous ignition theory also suggested
treated as a critical point. The experiment results showed that the that the internal ignition temperature depended on the oxidation
effect of temperature on biomass activity was greater than that the temperature [23].
282 F. Shan et al. / Fuel 188 (2017) 277–284
at moment of ignition(K)
at moment of ignition(K)
360
320 300
open symbols=air open symbols=air open symbols=air
340 filled symbols=CO 2 filled symbols=CO 2 filled symbols=CO 2
300 280
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
Vf=10L/min
at moment of ignition(K)
at moment of ignition(K)
at moment of ignition(K)
Vf=10L/min Vf=10L/min
Vf=15L/min 340 Vf=15L/min
Vf=15L/min
350 Vf=20L/min
Vf=20L/min Vf=20L/min
380
330
open symbols=air
340 filled symbols=CO2
320
370
330
open symbols=air
open symbols=air
310
filled symbols=CO2
filled symbols=CO2
400
Tg=1073K Tg=1073K Tg=1073K
at moment of ingnition(K)
at moment of ignition(K)
at moment of ignition(K)
Tg=973K Tg=973K 380 Tg=973K
380 380
Tg=873K Tg=873K Tg=873K
360
360 360
340
340 340
Tg=1073K
at moment of ignition(K)
at moment of ignition(K)
400 Tg=973K 400 Tg=973K 400 Tg=973K
340 340
340
320 320
320
open symbols=air open symbols=air 300 open symbols=air
300
300 filled symbols=CO 2 filled symbols=CO 2 filled symbols=CO 2
280
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
(d) (e) (f) 20L/min
Fig. 7. At the same gas velocity condition, biomass internal ignition temperature changes with the oxygen concentration. Biomass samples are: rice husk (a–c) and
Pinus bungeana (d–f).
70 70 65
V =10L/min V =10L/min V =10L/min
f f f
volatile burnout time(s)
65 V =15L/min V =15L/min
volatile burnout time(s)
f f 60 V =15L/min
f
65
V =20L/min V =20L/min V =20L/min
60 f f f
open symbols=air 55
open symbols=air
55 60 filled symbols=CO 2
filled symbols=CO 2
50
50
open symbols=air 55
45
45
filled symbols=CO2
40 50 40
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
V =15L/min
volatile burnout time(s)
f V =15L/min 55 V =15L/min
f f
55 55
V =20L/min V =20L/min V =20L/min
f f f
50
50 50
45
45 45
open symbols=air 40
open symbols=air open symbols=air
filled symbols=CO 2 filled symbols=CO2 filled symbols=CO 2
40 40 35
20 40 60 80 100 20 40 60 80 100 20 40 60 80 100
oxygen concentration(vol%) oxygen concentration(vol%) oxygen concentration(vol%)
data graph also displayed this phenomenon. Although the high volatile ignite on the particles’ surface. In brief, the mode of igni-
flow velocity and the heating rate increased the speed of diffusion tion is oxygen-sensitive.
of oxygen to the surface of the particles and devolatilization rate, Different combustion modes undergo different combustion pro-
the lower oxygen concentration was incapable of making the cesses. In this experiment, the combustion process can be divided
284 F. Shan et al. / Fuel 188 (2017) 277–284