Bioresource Technology 92 (2004) 8391

Emission performance and combustion eciency of a conical

uidized-bed combustor ring various biomass fuels
W. Permchart a, V.I. Kouprianov
b

b,*

a
Joint Graduate School of Energy and Environment, King Mongkuts University of Technology Thonburi, Bangkok 10140, Thailand
Mechanical Engineering Program, Sirindhorn International Institute of Technology, Thammasat University, Pathumthani 12121, Thailand

Received 17 October 2002; received in revised form 22 May 2003; accepted 5 July 2003

Abstract
This paper summarizes the results of an experimental study on combustion of three distinct biomass fuels (sawdust, rice husk and
pre-dried sugar cane bagasse) in a single uidized-bed combustor (FBC) with a conical bed using silica sand as the inert bed material. Temperature, CO, NO and O2 concentrations along the combustor height as well as in ue (stack) gas were measured in the
experimental tests. The eects of fuel properties and operating conditions (load and excess air) on these variables were investigated.
Both CO and NO axial proles were found to have a maximum whose location divides conventionally the combustor volume into
formation (lower) and reduction (upper) regions for these pollutants. Based on CO emission and unburned carbon content in y ash,
the combustion eciency of the conical FBC was quantied for the selected biomass fuels red under dierent operating conditions.

2003 Elsevier Ltd. All rights reserved.
Keywords: Sawdust; Rice husk; Sugar cane bagasse; NOx ; CO; Formation; Reduction

1. Introduction
Biomass is an important source of energy in tropical
countries, particularly in South and Southeast Asia,
accounting for some 40% of the total regional energy
consumption (FAO-UN, 1997). Combustible residues
and wastes collected on a large scale from agricultural
and forest-related activities such as rice, sugar, palm
oil and wood industries, are used as biomass fuels for
electric power and heat production in these countries.
In Thailand, rice husk, sugar cane bagasse and wood
residues represent the most viable biomass fuels. Recent
assessments indicate a signicant overall power generation potential for these energy sources, of about 1 GW
(NEPO, 2000) to 2 GW (Duval, 2001). Thus, the
growing substitution of combustible biomass for fossil
fuels (particularly, fuel oil/gas) will undoubtedly contribute to energy conservation in the power sector of the
Thai economy.
Biomass fuels with low and moderate moisture contents (W < 50%) are eectively utilized in combustion
*
Corresponding author. Tel.: +66-2-986-9009x2208; fax: +66-2-9869112.
E-mail address: ivlaanov@siit.tu.ac.th (V.I. Kouprianov).

0960-8524/$ - see front matter
2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2003.07.005

systems; otherwise or alternatively, biomass can be used

for thermochemical conversion through gasication and
pyrolysis (Bhattacharya, 1993; Natarajan et al., 1998;
McKendry, 2002). Among the proven combustion
technologies (such as grate-red, suspension-red and
uidized bed systems), the uidized bed technology is
reported to be the most ecient and suitable for converting agricultural and wood residues into energy (van
den Broek et al., 1996; Bhattacharya, 1998; Werther
et al., 2000).
In addition to highly ecient operation, a combustion system should comply with the requirement of
minimizing environmental impact. The emission rate of
various pollutants from the combustion of agricultural
and wood residues depends on fuel analysis, combustor
design and operating conditions. With sucient combustion air, so called unburned pollutants (such as HC,
tar, polycyclic aromatic hydrocarbons, Cx Hy ) are expected at insignicant levels. Meanwhile, due to small S
and Cl contents in these biomass fuels, SO2 and HCl
emissions from the combustion system are considered to
be negligible (Werther et al., 2000).
Apart from the above pollutants, CO and NOx
(generally, as NO) are in eect the major harmful pollutants emitted from biomass combustion in uidized

84

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

bed systems. For a selected fuel, CO emission (strongly

aecting the combustion eciency) is a function of operating variables, such as excess of combustion air as
well as combustor load, and can be eectively controlled
by the air supply (Bhattacharya et al., 1984; Kouprianov
and Permchart, 2003). Unlike CO, NOx emissions are
weakly dependent on the combustion conditions, i.e.
bed temperature and excess air, whereas the eect of
fuel-N on the yield of this pollutant is very strong. These
facts conrm the predominant role of the fuel-NO
mechanism of NOx formation in biomass combustion
(Leckner and Karlsson, 1993; Tariq and Purvis, 1996;
Werther et al., 2000).
The inuence of particulate emission on combustion
eciency is rather signicant when ring rice husk, a
high-ash biomass fuel, in uidized bed systems (Bhattacharya et al., 1984; Suwanayuen et al., 1992). The
amount of unburned carbon associated with the particulate matter (mainly y ash) is dependent on the
combustor loading (per cross-sectional area) and excess
air. However, the level of particulate emissions from
combustion systems red with low-ash biomass fuels
(wood residues, bagasse, etc.) is expected to be negligible.
This paper deals with the experimental study of
combustion of distinct biomass fuels (namely, sawdust,
rice husk and sugar cane bagasse) in a conical uidizedbed combustor (FBC) using silica sand as inert material.
Compared with cylindrical and prismatic uidized-bed
combustors, the proposed device is characterized by a
lower pressure drop over the bed, shorter start-up time
and uniform cross-sectional (i.e. radial) proles of
combustion characteristics (Kouprianov et al., 1996;

Permchart and Kouprianov, 2002). The main objectives

of this work were: (1) to study formation and reduction
of the major gaseous pollutants (CO and NO) in the
conical FBC when ring selected biomass fuels, and (2)
to determine the combustion eciency of the device
red with these fuels for dierent operating conditions.

2. Methods
2.1. Experimental set-up
The schematic diagram of the experimental set-up is
shown in Fig. 1. The conical FBC consisted of two parts:
(1) a conical section of 1 m height with a cone angle of
40, and (2) a cylindrical section of 0.9 m inner diameter
and 2 m height. The above elements were made of 4.5mm-thick steel, and the combustor body insulated
with 50 mm ceramic-ber material covered externally by
1-mm-thick galvanized steel.
The design characteristics of the combustor were selected with the aim of providing the bubbling uidized
bed regime of the bed material with a minimum carryover of sand from the device when ring up to 100 kg/h
of biomass fuels in wide ranges of combustor loading
and air ow rate (or excess air). This was achieved by
the proper selection of both the bed cone angle (preventing spouted bed regimes; Jing et al., 2000), and
optimal size (of 0.30.5 mm) of the sand particles
(Bhattacharya et al., 1984).
Under cold operating conditions, the pressure drop
over the uidized bed was about 200 mm H2 O, corresponding to the 20 cm-height static bed of sand used in

Fig. 1. Schematic diagram of the experimental set-up with the conical FBC.

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

85

bon, 5% for hydrogen and 10% for nitrogen. Meanwhile, for the proximate analysis the relative errors are
estimated to be below 5% for moisture in all the tested
fuels as well as for ash in rice husk, and below 10% for
ash in sawdust and bagasse.

these experimental tests. During combustor start-up, the

LPG-ring burner was used for preheating the bed
material.
A 25-hp blower supplied combustion air into the
combustor through the air distributor. This distributor,
comprising nine stand-pipe type nozzles with caps, was
located in the bottom of the conical section and used for
air bubbling. The overall air supply was controlled by
the ow meter (volume ow rate) and/or by the gas
analyzer (for oxygen content monitoring).
The combustor was equipped with a screw type feeder
supplying biomass fuel over the bed, and the fuel feed
rate was controlled by the three-phase inverter. An external cyclone at the top of the conical FBC served for
separation of particulate matter (char, ash and carryover sand particles) from the ue gas leaving the combustor.

2.3. Experimental procedures

The concentrations of major gaseous pollutants (CO
and NO) in ue gas were measured in the experimental
tests along the combustor height above the air distributor when ring the selected biomass fuels. In addition,
O2 concentrations as well as temperatures were detected
along the combustor height and in the ue gas at the
cyclone outlet (see Fig. 1). For each test run, the value of
excess air in the ue gas was determined using the O2
and CO concentrations at the cyclone outlet.
Flue gas was sampled through seven holes arranged
at dierent locations along the ue gas path. For measuring the gas concentrations, the Testo-350 gas analyzer was employed. The relative measurement errors
were of 5% for CO and NO and about 1% for O2 . Seven
chromelalumel thermocouples (of type K) were xed at
dierent levels along the combustor height and at the
cyclone outlet to monitor the temperatures (in relative
error of about 1%) in the ue gas.
The unburned carbon concentrations in y ash were
detected (in relative error of 15% for sawdust and bagasse and 5% for rice husk) in the fuel laboratory in
order to assess associated heat loss.
Two parameters were chosen in this work as independent variables: fuel feed rate (FR) and percent excess
air (EA). For each selected fuel, the conical FBC was
tested at two dierent fuel feed rates, and the required
feed rate was provided by the corresponding rotational
speed of the feeder screw. For maximum combustor
loading, the screw feeder provided 81.5, 82.4 and 70.0
kg/h when ring sawdust, rice husk and bagasse, respectively; in the tests with minimum fuel supply, the
feed rates were 35.0, 37.3 and 31.0 kg/h, respectively.
Lower feed rates for the bagasse were explained by the
higher resistance oered to the long ber particles of this
fuel during their transportation along the screw feeder
(although the fuel was pre-dried).

2.2. The fuels

The biomass fuels (sawdust, rice husk and sugar cane
bagasse) used in the experimental tests were delivered
from dierent Thai mill companies.
The supplied sawdust was in eect the mixture of
residuals (sawdust) of various local woods (Teng, Maka,
Rang, Pradu and some others). The average particle size
of the mixed sawdust was determined as 0.8 0.8 mm.
The individual particles of rice husk were much
greater in size (of 2.4 8 mm on average). However, this
fact did not involve any problem in feeding the fuel into
the combustor because of relatively low fuel moisture
(W 10.3%).
Attempts to burn as-delivered bagasse failed in preliminary tests because of the high moisture content in
this raw biomass fuel (W 48.8%). Moreover, moisture
negatively aected the feeding of long individual fuel
particles (of about 0.4 21 mm in size) of this ber
residue. The fuel was therefore preliminarily dried to
W 14.4% under room conditions (i.e. by natural
ventilation) for a few days.
Table 1 shows ultimate and proximate analyses as
well as the lower heating values (LHV) of the biomass
fuels used in this work. All the data in Table 1 represent
averages. For the ultimate analysis the relative errors are
believed to be below the following values: 1% for car-

Table 1
Properties of biomass fuels used in the experimental tests: W moisture, A ash, daf dry and ash-free basis, LHV lower heating value (proximate analysis)
Fuel

Ultimate analysis (wt.%, daf)

C

Sawdust
Rice husk
Bagassea (sugar cane)

45.43
48.45
42.00

6.71
6.16
6.58

47.65
44.62
51.00

0.19
0.55
0.26

0.02
0.22
0.16

15.9
10.3
14.4

0.61
19.54
1.84

Proximate analysis (wt.%)

Pre-dried under room conditions from W 48.8% (as-delivered) to W 14.4%.

LHV (MJ/kg)
13.9
12.3
12.6

86

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

combustor top (2.75 m height) increased by 6080 C for

ring rice husk and bagasse, whereas it increased by 160
C for ring sawdust. This tendency, but of a lesser
degree, remained for the reduced combustor loads.
As seen in Fig. 2, at xed EA the load reduction of
more than 2 times led to some decrease in the temperatures. In particular, for EA of about 60%, the bed
temperature dropped by some 90 C for rice husk, 70 C
for sawdust and 50 C for bagasse, remaining, nevertheless, at a level sucient for stable fuel ignition and
combustion. This conrmed the high sustainability of
the combustion process of the selected biomass fuels in
the conical FBC operating in wide ranges of load and
excess air.
The axial O2 concentration proles are shown in Fig.
3 for the same (as in Fig. 2) operating conditions. For
particular excess air, the rate of oxygen consumption
along the combustor height was almost independent of
either nature of biomass fuel or combustor load; at the
same sampling point, the dierences in oxygen concentrations for the fuels of interest did not exceed 1.5% O2 .
The maximum rates of oxygen consumption were observed in the bed region for all the fuels.
It appears that, with variation in EA, the oxygen
concentration gradient (along the height above the air
distributor) is changed. Meanwhile, in a combustion test
at the particular value of EA, the axial oxygen proles
for distinct fuels were found to hold the above tendencies.

For the maximum fuel feed rates, the combustor was

run at ve EA (of about 20%, 40%, 60%, 80% and
100%), whereas in operation with the minimum loads
the fuels were red at three EA (of about 20%, 60% and
100%).

3. Results and discussion

Based on the measured temperatures and gas concentrations (averaged over time), the corresponding
axial proles along the height above the air distributor
(see Fig. 1) were plotted for the fuels of interest red in
the conical FBC under dierent operating conditions. In
the following sections, the proles of the above variables, as well as other relevant characteristics obtained
for these fuels, are compared for some operating conditions.
3.1. Variables aecting combustion chemistry
Fig. 2 shows the axial temperature proles for maximum and minimum loads, i.e. for dierent FRs, of the
conical FBC ring the fuels at similar EA (of about
60%). These proles seem to be rather uniform and
characterized by small temperature gradients along the
height above the air distribution: positive in the bed
region (for heights up 0.81.5 m), and negative in the
freeboard region.
For quasi-identical operating conditions, the highest
temperatures in the combustor were observed for sawdust. Despite the LHV for bagasse being greater than
that for rice husk, all of the temperature proles for
bagasse were lower because of the higher moisture
content and smaller feed rate for this fuel.
With variation in EA for the xed load, the bed
temperatures remained to be almost unchanged. However, in the freeboard region the temperatures were
found to have a tendency to increase for higher EA. For
example, when EA varied from about 20% to 100% in
the tests with maximum FRs, the temperature at the

3.2. Pollutant formation and reduction

Fig. 4 shows the axial CO concentration proles for
the same operating conditions as provided in Figs. 2 and
3. In all the tests, the axial CO concentration proles
were found to have an extreme (maximum), COmax , at
approximately the same locations (or heights above the
air distributor) which were detected for the maximum
temperatures for corresponding operating variables.
Hence, the combustor volume can be conventionally
divided into formation (lower) and reduction (upper)
1000

Temperature (C)

Temperature (C)

1000

800

600

400

Rice husk: FR=82.4 kg/h, EA=59.6%

Sawdust: FR=81.5 kg/h, EA=61.1%
Bagasse: FR=70.0 kg/h, EA=60.7%

200

800

600

400

Rice husk: FR=37.3 kg/h, EA=61.8%

Sawdust: FR=35.0 kg/h, EA=58.9%
Bagasse: FR=31.0 kg/h, EA=62.3%

200
0

Height above air distributor (m)

Height above air distributor (m)

Fig. 2. Axial temperature proles in the conical FBC for three distinct biomass fuels red under dierent operating conditions.

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

20

O2 concentration (vol.%)

O2 concentration (vol.%)

20

87

15

10

Rice husk: FR=82.4 kg/h, EA=59.6%

Sawdust: FR=81.5 kg/h, EA=61.1%
Bagasse: FR=70.0 kg/h, EA=60.7%

15

10

Rice husk: FR=37.3 kg/h, EA=61.8%

Sawdust: FR=35.0 kg/h, EA=58.9%
Bagasse: FR=31.0 kg/h, EA=62.3%

0
0

Height above air distributor (m)

Height above air distributor (m)

Fig. 3. Comparison of axial O2 concentration proles in the conical FBC for three distinct biomass fuels red under dierent operating conditions.

2.5
Rice husk: FR=82.4 kg/h, EA=59.6%
Sawdust: FR=81.5 kg/h, EA=61.1%
Bagasse: FR=70.0 kg/h, EA=60.7%

2.0

CO concentration (vol.%)

CO concentration (vol.%)

2.5

1.5
1.0
0.5

Rice husk: FR=37.3 kg/h, EA=61.8%

Sawdust: FR=35.0 kg/h, EA=58.9%
Bagasse: FR=31.0 kg/h, EA=62.3%

2.0
1.5
1.0
0.5
0.0

0.0
0

Height above air distributor (m)

Height above air distributor (m)

Fig. 4. Comparison of axial CO concentration proles in the conical FBC for three distinct biomass fuels red under dierent operating conditions.

regions with regard to the location of COmax in the

combustor.
As seen in Fig. 4, the rate of CO formation for rice
husk was much greater than that for sawdust and bagasse. This could be explained by higher fuel-ash concentration and coarser char particles for rice husk: up to
200 lm, against 5 lm for sawdust and 10 lm for bagasse, as follows from y ash analyses. These factors
result in higher char holdup in the bed region and,
consequently, lead to higher values of COmax for rice
husk in comparison with sawdust and bagasse. For the
same reasons, and because of the lower temperatures,
the rate of CO formation for bagasse is higher than that
for sawdust.
The apparent dierence in the axial CO concentration
proles for the selected biomass fuels conrms a signicant role of heterogeneous reactions (on the char
surface), of char-carbon with CO2 and water vapor, at
temperatures typical for uidized bed combustion
(Tillman et al., 1981). These reactions are basically followed by fuel in-bed devolatilization and further oxidation of the carbon-based components, released from
the fuel particles with volatiles, to CO (Werther et al.,
2000). In addition to the eect of the char particle size,
the above reactions explain an elevated rate of CO for-

mation in the bed region for the case of ring high-ash

rice husk.
In the reduction region, the CO is basically oxidized
in the chain termination reaction with OH radicals as
well as by oxygen directly (Tillman et al., 1981). For the
selected biomass fuels, the rates of CO reduction were
found to be in an apparent correlation with CO concentrations in the freeboard region; accordingly, the
highest rates were determined for the rice husk.
Reduction in combustor loading at xed EA led to a
signicant diminishing of both formation and reduction
rates of CO when ring rice husk, whereas the eect of
the load on these processes was relatively weak for ring
sawdust and bagasse (see Fig. 4).
A comparison of the axial NO concentration proles
is shown in Fig. 5. Similar to CO, all the axial NO
concentration proles possesses a maximum, NOmax , the
locations of which coincide with those for COmax for
corresponding operating conditions.
For the temperatures provided in Fig. 2, NO would
be expected to originate from the fuel-nitrogen (Tillman
et al., 1981; Winter et al., 1999; Werther et al., 2000).
Basically, fuel-NO can be formed through: (1) combustion of the nitrogenous species released with volatile
matter (such as HCN, NH3 ), and (2) oxidation of the

88

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

600

Rice husk: FR=82.4 kg/h, EA=59.6%

Sawdust: FR=81.5 kg/h, EA=61.1%
Bagasse: FR=70.0 kg/h, EA=60.7%

500

NO concentration (ppm)

NO concentration (ppm)

600

400
300
200
100

Rice husk: FR=37.3 kg/h, EA=61.8%

Sawdust: FR=35.0 kg/h, EA=58.9%
Bagasse: FR=31.0 kg/h, EA=62.3%

500
400
300
200
100
0

0
0

Height above air distributor (m)

Height above air distributor (m)

Fig. 5. Comparison of axial NO concentration proles in the conical FBC for three distinct biomass fuels red under dierent operating conditions.

nitrogen retained in the char. These reactions, resulting

in rapid formation of NO, are most likely to proceed in
the bed region of this FBC with the bottom air injection.
Meanwhile, in zones with volume O2 concentrations
lower than 1012%, the NH3 concentration is probably
elevated due to the rapid formation of NH3 from HCN
(Winter et al., 1999) as well as because of the emission of
NH3 released with volatiles from fuel particles present in
these zones. In the upper combustor region (with lower
O2 concentrations, as seen in Fig. 3) this may lead to NO
reduction through its reaction with NH3 , followed by
formation of nitrogen gas and water vapor, i.e. neutral
products. The alternative mechanisms of NO reduction
in the upper region of the uidized bed combustor involve reactions of NO with carbon and CO on the char
surface (Leckner and Karlsson, 1993; Werther et al.,
2000) which are highly probable when ring high-ash
fuels.
It is the high fuel-N and ash contents as well as the
large size of particulates (char and ash) that explain the
high rates of NO formation (in the bed region) and reduction (in the freeboard region) in the case of ring rice
husk. Apparently, the fuel-N has a strong eect on NO
formation, as follows from comparison of experimental
results depicted in Fig. 5. With reduction in the combustor load, the rate of NO formation (or NOmax , respectively) becomes somewhat lower corresponding to

diminishing bed temperature. The above facts conrm

the fuel-NO mechanism of the NO formation in combustion of the selected fuels.
The nitric oxide represents in eect the overall NOx
emissions from a biomass combustion system. In this
study, the NO2 concentrations were found to be 01
ppm for all of the tests. The N2 O formation in biomass
combustion is reported to be relatively small or negligible; moreover, even though N2 O has been formed, it
may be rapidly destroyed in the ame by active radicals
and molecules (Kn
obig et al., 1998; Winter et al., 1999;
Werther et al., 2000).
Fig. 6 shows the eects of excess air on the COmax and
NOmax , i.e. on the formation rate of CO and NO in the
bed region, when ring the fuels at maximum combustor
loading. Note that in the region with EA < 50%, CO
formation drastically increases with lowering excess air;
in the region with EA > 50%, the COmax are gradually
reduced (but with quite low rates) whilst EA is increased. Thus, emission of this pollutant can be controlled by proper air supply. However, the NOmax are
slightly greater for higher EA, which is typical for the
fuel-NO formation route, in spite of the dilution eects
caused by excess air. Hence, for the fuels red at maximum combustor loads, the values of excess air of
5060% seem to be optimal for minimizing the environmental impact of the conical FBC red with the
500

Rice husk: FR=82.4 kg/h

Sawdust: FR=81.5 kg/h
Bagasse: FR=70.0 kg/h

400

NOmax (ppm)

COmax (vol.%)

300
200
Rice husk: FR=82.4 kg/h
Sawdust: FR=81.5 kg/h
Bagasse: FR=70.0 kg/h

100
0

0
0

20

40

60

80

Excess air (%)

100

120

20

40

60

80

100

120

Excess air (%)

Fig. 6. Eects of operating conditions on the maximum CO and NO concentrations in the conical FBC for three distinct biomass fuels.

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

selected fuels, unless the ue gas is not utilized in other

combustion systems. For the minimum loads, the optimum values of EA are somewhat lower (because of
lowering in CO formation) and can be selected in the
range of 4050%.
Along a gas path combustor topcyclone outlet
(see Fig. 1) the gas ow is subject to signicant transformation. Moreover, a substantial reduction in the ue
gas temperature over this path was observed in the tests
(e.g. by 180230 C for the maximum loads, and the
greater reductions corresponded to higher EA) mainly
because of the non-insulating gas duct connecting the
combustor body with the cyclone. Despite the relatively
low temperatures, both CO and NO reductions were
found to occur in the cyclone (due to longer gas residence time) in all tests.
Table 2 provides the CO reduction ratios over the
path combustor topcyclone outlet as well as CO
emissions from the conical FBC (i.e. CO concentration
at the cyclone outlet) for the fuels red at maximum
loads for dierent EA. Substantial reductions, especially
for EA > 50%, were detected for rice husk indicating the
strong eect of the particulate concentration in ue gas
on CO reduction. Similar to COmax , the CO emission
from the combustor was lowered with an increase in EA.

89

However, despite greater values of COmax for rice husk,

the corresponding CO emissions for this pollutant were
found to be lower than those for ring bagasse at EA of
40100% which is explained by dierent reduction rates.
Table 3 shows the NO reduction ratios over the path
combustor topcyclone outlet and NO emissions
from the conical FBC for ring the fuels at the same
operating conditions, as provided in Table 2. The
highest NO reductions (of about 3040 ppm) in the
cyclone were observed for rice husk red at the maximum combustor load and EA of 40100%. However,
among selected fuels, rice husk remained to be the
strongest source of NO emission because of the higher
fuel-N content. The NO emissions were correlated with
NOmax ; meanwhile, the eects of EA on NO emissions
were found to be a bit stronger than of those on the
NOmax values.
3.3. Combustion eciency
For the estimation of combustion eciency (as percentage of the LHV), the heat losses owing to incomplete combustion (accounting the CO emission) and
unburned carbon contained in particulate matter were
determined (Basu et al., 2000). Table 4 shows these heat

Table 2
CO reduction over the path combustor topcyclone outlet when ring distinct biomass fuels in the conical FBC under dierent operating conditions
Biomass fuel red

Fuel feed rate (kg/h)

Sawdust

81.5
81.5
81.5

Rice husk

Bagasse

Excess air (vol%)

CO (vol%) at the
2.75 m FBC height

CO (vol%) at the
cyclone outlet

Reduction ratio

16.9
61.1
99.8

1.401
0.116
0.080

0.940
0.076
0.035

1.49
1.53
2.29

82.4
82.4
82.4

16.3
59.6
100.7

1.586
0.596
0.374

1.319
0.116
0.065

1.20
5.14
5.75

70.0
70.0
70.0

17.3
60.7
101.1

1.529
0.312
0.101

1.001
0.198
0.083

1.53
1.58
1.22

Table 3
NO reduction over the path combustor topcyclone outlet when ring distinct biomass fuels in the conical FBC under dierent operating conditions
Biomass fuel red

Fuel feed rate (kg/h)

Sawdust

81.5
81.5
81.5

Rice husk

Bagasse

Excess air (vol%)

NO (ppm) at the
2.75 m FBC height

NO (ppm) at the
cyclone outlet

Reduction ratio

16.9
61.1
99.8

109
123
144

80
112
125

1.36
1.10
1.15

82.4
82.4
82.4

16.3
59.6
100.7

136
191
219

127
157
172

1.07
1.22
1.27

70.0
70.0
70.0

17.3
60.7
101.1

91
116
131

77
96
121

1.18
1.21
1.09

90

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

Table 4
Heat losses and combustion eciency (LHV%) for the conical FBC ring distinct biomass fuels at the maximum fuel feed rate and dierent values of
excess air
Biomass fuel red

Fuel feed rate (kg/h)

Excess air (vol%)

Sawdust

81.5
81.5
81.5

16.9
61.1
99.8

3.51
0.39
0.16

0.01
0.02
0.02

96.48
99.59
99.82

Rice husk

82.4
82.4
82.4

16.3
59.6
100.7

4.29
0.53
0.34

12.47
13.15
18.35

83.24
86.32
81.31

Bagasse

70.0
70.0
70.0

17.3
60.7
101.1

3.72
0.81
0.53

0.02
0.03
0.06

96.26
99.16
99.41

losses along with combustion eciencies for the selected

fuels red at maximum combustor loads for dierent
EA.
The highly ecient combustion, over 99%, of sawdust
and bagasse in the conical FBC is achieved for
EA > 50%. Taking into consideration the above discussion, EA of 5060% seems to be sucient to provide
highly ecient combustion of these fuels with minimized
environmental impact.
The combustion eciency for rice husk is much lower
because of higher heat loss associated with unburned
carbon which is aected by both the high ash content and elevated concentration of unburned carbon in
particulate matter collected from the cyclone. Owing to
the opposite eect of EA on the above heat losses, the
total heat loss for rice husk possesses a minimum at EA
of 5060% corresponding to the maximum combustion
eciency, as seen in Table 4. Thus, the above recommendation regarding the optimal EA proposed for
sawdust and bagasse is also applicable to the ring of
rice husk in the conical FBC.

4. Conclusions
An ecient and sustainable operation of the conical
FBC was performed when ring sawdust, rice husk and
pre-dried sugar cane bagasse in wide ranges of the
combustor load and excess air.
The axial temperature proles in the conical FBC
were fairly uniform for all the fuels and weakly aected
by fuel analysis and operating conditions. For particular
excess air, the rate of oxygen consumption along the
combustor height was almost independent of either the
nature of the biomass fuel or the combustor load.
Both CO and NO axial proles possessed a maximum
whose location divided conventionally the combustor
volume into formation (lower) and reduction (upper)
regions for these pollutants. CO formation rates, and
consequently COmax , for rice husk were much greater

Heat loss
owing to incomplete
combustion

Heat loss
owing to unburned
carbon

Combustion eciency

than those for sawdust and bagasse for similar operating

conditions because of coarser char particles and higher
ash concentration leading to higher char hold up in the
bed region and signicant contribution of char-carbon
to CO formation. The COmax for distinct fuels were
rapidly diminished with an increase in excess air of up to
5060% demonstrating however a weak dependence on
excess air in the region of 60100%. The rates of CO
reduction for distinct biomass fuels were found to be in
an apparent correlation with CO concentrations in the
freeboard region; accordingly, the highest rates were
determined for the rice husk.
The NOmax was strongly aected by fuel-nitrogen and
weakly dependent on the operating conditions conrming a fuel-NO formation mechanism for all the fuels.
The rate of NO reduction for rice husk was much
greater than those for sawdust and bagasse due to the
signicant role of heterogeneous reactions (on the char
surface) in the freeboard region and ash collecting device
(cyclone).
Of the studied fuels, sawdust is the most environmentally friendly biomass fuel whereas the ring of rice
husk is accompanied by a noticeable environmental
impact.
For the maximum combustor load and excess air of
50100%, a combustion eciency of over 99% could be
achieved when ring sawdust and bagasse. For the case of
ring rice husk at the maximum load, the combustion
eciency was lower because of higher losses owing to
unburned carbon. The maximum combustion eciency
of 86% for ring rice husk was obtained for excess air of
about 60%; an increase in excess air of up to 100% resulted
in a deterioration of the combustion eciency (to 81%).

Acknowledgements
The authors wish to acknowledge sincerely the nancial support provided by the Energy Policy and
Planning Oce, Ministry of Energy (Thailand).

W. Permchart, V.I. Kouprianov / Bioresource Technology 92 (2004) 8391

References
Basu, P., Cen, K.F., Jestin, L., 2000. Boilers and Burners. Springer,
New York.
Bhattacharya, S.C., 1993. State-of-the-art of utilizing residues and
other types of biomass as an energy source. RERIC International
Energy Journal 15 (1), 121.
Bhattacharya, S.C., 1998. State of the art of biomass combustion.
Energy Sources 20, 113135.
Bhattacharya, S.C., Shah, N., Alikhani, Z., 1984. Some aspects of uidized bed combustion of paddy husk. Applied Energy 16, 307316.
Duval, Y., 2001. Environmental impact of modern biomass cogeneration in Southeast Asia. Biomass and Bioenergy 20 (4), 287295.
FAO-UN, 1997. Regional study on wood energy today and tomorrow
in Asia. Food and Agricultural Organization of the United
Nations, Bangkok, Thailand.
Jing, S., Hu, Q., Wang, J., Jin, Y., 2000. Fluidization of coarse
particles in gas-solid conical beds. Chemical Engineering and
Processing 39, 379387.

mand, L.-E., Leckner, B., 1998. Comparison
Kn
obig, T., Werther, J., A
of large- and small-scale circulating uidixed bed combustors with
respect to pollutant formation and reduction for dierent fuels.
Fuel 77 (14), 16351642.
Kouprianov, V.I., Permchart, W., 2003. Emissions from a conical FBC
red with a biomass fuel. Applied Energy 74, 383392.
Kouprianov, V.I., Utistham, T., Boonmun, S., Jatanakul, W., Suttisonk, B., 1996. Experimental investigation of sawdust burning in
the conical bed combustor. In: Proceedings of the 6th National
Chemical Engineering Conference, Bangkok, pp. 18.
Leckner, B., Karlsson, M., 1993. Gaseous emissions from circulating
uidized bed combustion of wood. Biomass and Bioenergy 4 (5),
379389.

91

McKendry, P., 2002. Energy production from biomass (Part II):

conversion technologies. Bioresource Technology 83, 4754.
Natarajan, E., Nordin, A., Rao, A.N., 1998. Overview of combustion
and gasication of rice husk in uidized bed reactors. Biomass and
Bioenergy 14 (5/6), 533546.
NEPO, 2000. Thailand biomass-based power generation and cogeneration within small rural industries. In: Final Report Conducted
by Black & Veatch, National Energy Policy Oce, Bangkok,
Thailand.
Permchart, W., Kouprianov, V., 2002. Design and experimental study
of a conical uidized bed combustor ring sawdust. In: Proceedings
of the 1st International Conference on Sustainable Energy Technologies, Porto, Portugal, Section LC, pp. 16.
Suwanayuen, S., Chayawattana, T., Surachpakorn, Y., Tia, S.,
Tangsathitkulchai, C., 1992. Combustion of rice hull in a uidized
bed furnace. ASEAN Journal on Science and Technology for
Development 9 (2), 107115.
Tariq, A.S., Purvis, M.R.I., 1996. NOx emissions and thermal
eciencies of small scale biomass-fuelled combustion plant with
reference to process industries in a developing country. International Journal of Energy Research 20 (1), 4155.
Tillman, D.A., Rossi, A.J., Kitto, W.D., 1981. Wood combustion.
Academic Press, New York.
van den Broek, R., Faaij, A., van Wijk, A., 1996. Biomass combustion for power generation. Biomass and Bioenergy 11 (4), 271
281.
Werther, J., Saenger, M., Hartge, E.U., Ogada, T., Siagi, Z., 2000.
Combustion of agricultural residues. Progress in Energy and
Combustion Science 26 (1), 127.
Winter, F., Wartha, C., Hofbeuer, H., 1999. NO and N2 O formation
during the combustion of wood, straw, malt waste and peat.
Bioresource Technology 70, 3949.