Energy 203 (2020) 117882

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Cofiring low-rank coal and biomass in a bubbling fluidized bed with

varying excess air ratio and fluidization velocity
Paulo R. Wander a, *, Fla

vio M. Bianchi b, Nattan R. Caetano c, Marcos A. Klunk a,
Maria Luiza S. Indrusiak d
a ~o Leopoldo, 93022-750, Brazil
Mechanical Engineering Graduate Program, Universidade do Vale do Rio dos Sinos, Sa
b
SATC College, Criciúma, Brazil
c
Mechanical Engineering Department, Federal University of Santa Maria, Santa Maria, Brazil
d
Brazilian Section of Combustion Institute, Brazil

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

Article history: Cofiring coal and biomass is one of the alternatives to reduce emissions from thermoelectric power
Received 26 August 2019 plants. The influence of biomass fraction in the mixture is an important research topic to increase coal
Received in revised form substitution and achieve lower GHG emissions. A combustion system based on laboratory-scale bubbling
23 April 2020
fluidized bed technology was used to analyze the burning of different ratios (5, 10 and 15%) of mixed
Accepted 14 May 2020
biomass and low-rank mineral coal. Two operational parameters were varied, fluidization velocity (3umf
Available online 18 May 2020
and 6umf) and excess air ratio (50 and 70%). Previous studies focused mainly on gasification processes,
where excess air is a negative parameter. For combustion processes, high values of excess air are
Keywords:
Bubbling fluidized bed
mandatory. Experimental results showed that fuel mixtures containing up to 15% of biomass gives good
Cofiring results in terms of emissions and temperature stability control. Emissions were slightly increased as well
Fluidization velocity as the operating temperature with the presence of biomass. The operating regimes with higher fluid-
Low-rank mineral coal ization velocity and excess air ratio presented better performance, however, biomass volatiles release at
Biomass the freeboard should be the main concern for this type of application. Therefore, it was concluded that
the combined burning of coal with biomass is a feasible alternative to reduce emissions without
impairing efficiency.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction and mineral processing (calcination of alumina, roasting of ores,

pre-calcination for the manufacture of cement), chemical and
Energy conversion technologies using fluidized bed are envi- pharmaceutical (phthalic anhydride from naphthalene, conversion
ronmentally friendlier than more traditional processes such as of methanol to olefins etc.) [6], physical processing (drying, particle
fixed bed and pulverized fuel combustors [1]. A relevant advantage coating, heat treatment) amongst others [7].
of fluidized bed technology is the possibility of gasification and The mineral coal present in the south of Brazil is mostly of the
combustion of low quality solid fuels [2], such as high ash content sub-bituminous type, with approximately 40% ash and a substantial
coals and biomasses with high moisture content [3], due to the amount of sulfur, around 2%. The raw coal is usually prepared to
good uniformity of bed operating temperature [4]. Among the reach typical requirements suitable to its use in thermoelectric
characteristics of the fluidized beds that provide this advantage, the power plants [6]. Moisture is also present and can hinder com-
high heat and mass transfer rates are the most important ones [2]. bustion by requiring operational adjustments [8], but this is a
Applications of fluidized bed technology in the industry include problem that it has in common with biomass [9]. However, there
energy conversion (steam generation, gasification and incineration) are several other advantages in the combined burning of coal and
[5], petrochemical (catalytic cracking and Fisher-Tropsch synthesis) biomass described in the literature [10].
Currently, the combined burning of coal with renewable fuels
has been a sustainable alternative to reduce the fossil fuels widely
* Corresponding author.
used in industry [11]. Bubbling fluidized bed combustion usually
E-mail addresses: [email protected], [email protected] presents satisfactory operation even with fuels mixed in different
(P.R. Wander).

https://doi.org/10.1016/j.energy.2020.117882
0360-5442/© 2020 Elsevier Ltd. All rights reserved.
2 P.R. Wander et al. / Energy 203 (2020) 117882

their aqueous form produces sulfuric acid, highly corrosive to low

List of Acronyms chromium steels. Sulfur delivered to the atmosphere produces acid
rain and respiratory problems.
ABNT Brazilian Association of Technical Standards The concentration increase of carbon dioxide in atmosphere is
ASTM American Society for Testing and Materials considered as the main mechanism leading to the global warming
NBR Brazilian Standard phenomenon. Renewable fuels such as biomass are considered CO2
GHG Greenhouse Gases neutral and so, combining biomass with a fossil fuel in a combus-
tion process has a positive impact on the overall emission of this
List of Symbols gas.
umf Minimum fluidization velocity (m/s) Nitrogen oxides (mainly NO, NO2 and N2O, referred to as NOx)
l Excess air ratio are formed in the fluid bed combustors mainly in two ways, the
oxidation of the nitrogen present in the fuel (fuel NOx) and re-
actions between the oxygen and the nitrogen present in the air
(thermal NOx). Since thermal NOx is produced at high tempera-
proportions, dealing well with blend characteristics [12]. However, tures (mainly above 1480
C), it represents a small fraction of the
there are still gaps in the understanding of the combined burning NOx produced in fluidized beds, with fuel NOx being responsible for
technology of low-rank coal with renewable fuels such as biomass the largest contribution [26]. It is worth to mention, as described by
[13]. Turns [27], the formation of the prompt NOx from the mechanism
Cofiring low-rank coal and biomass in a bubbling fluidized bed of Fenimore and that is closely linked to the chemistry of the
in terms of temperature and emissions was investigated before combustion of hydrocarbons. From this mechanism, NO is rapidly
[13e15]. However, most of the papers addressed gasification pro- produced in the flame zone, in pre-mixed laminar flames before the
cesses of rice husk biomass [16]. This study considered the coal- formation of thermal NOx. This effect is hardly perceived in
biomass combustion in a bubbling fluidized bed combustor in bubbling fluidized bed combustion due to the dynamics of fuel-air
laboratory scale, varying the percentage of biomass in the fuel and mixture and the small equivalence ratio.
using operating temperature and emissions of CO2, CO, NOx, CxHy The increase of the nitrogen in the fuel generates increase in the
and SO2 as control parameters. Indeed, this is a work about cofiring emissions of NOx fuel. Its formation depends not only on the
employing sawmill biomass residues mixed with low rank coal for amount of nitrogen in the fuel, but how it is combined in fuel
energy generation purposes [17]. structure and on combustor operating conditions. The NO and N2O
Some studies showed interesting results for the cofiring of coal chemistry is complex in fluidized bed combustors [28] where the
with sludge [18] and biomass [19], although the mixture fraction low combustion temperatures led to relatively high levels of N2O.
was limited to 10%. In fluidized bed combustion, the particulate However, with increasing volatile fuel matter, NO increases while
solids of the bed, including in some cases limestone for sulfur N2O tends to decrease [26]. With the increase of biomass per-
compounds control [20], present physical and morphological centage (high volatile content) in the combined firing with mineral
characteristics that affect the molecular phenomena (such as mass coal, this effect may be observed, although in lean combustion and
diffusion) occurring within the particles and or amongst particles. low temperatures, N2O reacts with monoatomic oxygen and is
These characteristics are very useful for the mixture of different quickly converted to NO.
fuels and are important for the design of a fluidized bed reactor Jones et al. [29], observed that the combined burning of bitu-
column [21]. Reactor dimensions, such as its diameter and useful minous coal and wood resulted in reduced emissions of particulate
height, and some operational parameters, like flow and pressure matter and NOx when a fixed bed combustor was used. Jurado et al.
drop, are affected by the particle characteristics [22]. [30] verified the attenuation of SO2 concentrations and increase in
Bubbles formation and bed hydrodynamic behavior depends on HCl concentrations with the increase of the biomass ratio (using
particle density and mean size, as well as fluidization gas, as clas- cereal residues) firing with El Cerrejon South American coal, using a
sified by Geldart [23], defining the fluidization regime and the value pilot unit of pulverized firing under conditions of air burning and
of the minimum fluidization velocity [21,24]. oxy-combustion. Also, under oxy-combustion conditions, Riaza
The mass of fuel present in a fluidized bed combustor consti- et al. [31] observed the reduction of NO emissions when biomass
tutes about 0.5e5% of the total bed solids, the remainder consisting mixtures (olive residues) were burned together with coals (Spanish
of inerts: bed particles, ash from the fuel, adsorbents and other semi-anthracite) in a laboratory-scale flow-fired combustor.
non-combustible materials. The fuel introduced into the equipment Using a bench scale fluidized bed reactor, Aho et al. [32] tested
is quickly surrounded by the bed hot material raising the fuel Chinese fuels, including two coals, two straws and sawdust and
temperature. The heating rate varies from 100
C/s to more than observed the formation of alkaline chlorides (highly corrosive) in
1000
C/s, depending on many factors, among them, the fuel the simple burning of the straws. The study found that by
granulometry and bed agitation [1]. combining the straws with Chinese coals reduced the formation of
The possible pollutants to be formed during the burning of any alkaline chlorides due to its sulphation thanks to the high content
organic fuel are particulate matter (PM), sulfur oxides (SO2 and of aluminum and silicon in the coal.
SO3), carbon monoxide and dioxide (CO and CO2) in addition to The advantages of using fluidized bed combustors include
nitrogen oxides (NO, N2O and NO2). The quantities of these pol- flexibility in the use of fuels, as high ash (60e70%) and low volatile
lutants are naturally dependent on the fuels used, the technologies material (less than 1%) contents, and the lack of moving parts in the
used in the combustion, the conservation state of the equipment system, which gives robustness. Other advantages are the good
and the operating conditions [25]. flexibility regarding fuel granulometry, which reduces preparation
Sulfur is oxidized during combustion resulting in the formation costs, the admissibility of fuels with high moisture content, besides
of sulfur oxides. In bubbling fluidized beds, it is common to use the neutralization of sulfur compounds through the direct appli-
limestone directly in the bed to remove sulfur dioxide. The reten- cation of limestone or dolomite, and low NOx emissions due to low
tion efficiency may be greater than 95% when using proper operating temperatures [1]. However, the main drawbacks in the
amounts of limestone. Sulfur compounds (SOx) are potentially use of bubbling fluidized bed are the erosion occurring mainly in
aggressive to the metallic components of industrial plants, since zones of high density of particles, the need for relatively large areas
 P.R. Wander et al. / Energy 203 (2020) 117882 3

when compared to pulverized fuel and circulating fluidized bed 2.1.1. Minimum fluidization velocity
systems, the need for good uniformity of the fuel dispersion in the Based on the methodology suggested by Kunii and Levenspiel
bed, which requires multiple feeders for larger units, besides hav- [21], experimental tests were carried out in order to establish the
ing limitations to operate with partial load [1]. minimum fluidization velocities of the pre-selected bed particles.
Cofiring low-rank coal and biomass in a bubbling fluidized bed The tests to acquire the minimum fluidization velocities used in the
in terms of temperature and emissions was investigated before present work were performed in a cold fluidized bed at room
[10,11,13]. However, those studies considered only varying the temperature and with visualization of the phenomena, once the
percentage of biomass in the fuel as operational control parame- test section was an acrylic tube. The test section comprises also an
ters. Indeed, many studies about cofiring addressed gasification Artek™ radial compressor, model ACRRB 7.5 and a frequency
processes of biomass [16,35,40e42]. Also, the combined burning of inverter WEG, model CFW08. The constructive details of the
coal with renewable fuels has been used to generate power in in- equipment are described by Bianchi et al. [34].
dustry [7e9,11,17e19]. Most of the biomass types considered in
those studies (rice husk, straw, cereal and olive residues, lignite and 2.2. Fuels
sludge) are different from that used in the present work. Some
studies about cofiring has limited the mixture fraction to 10% [19], The fuels used were coal and wood sawdust. Mixtures were
and with some differences, including limestone for sulfur com- made with wood quantities from 5 to 15% in mass basis. From
pounds control [20]. Other studies have verified the emissions of simulation in a drop tube furnace using combined burning of low
cofiring with emphasis in oxy-combustion [30,31,43]. There are rank coal with palm cluster, Darmawan et al. [35] observed better
works using fixed bed combustor to analyze emissions of particu- results in terms of temperature and gas composition for mixtures
late matter [29], circulator fluidized bed [44], pressurized bed [45] between 10 and 25% of biomass mixed with coal.
and, also, in order to observe the formation of alkaline chlorides for
corrosivity analysis [32] or the effect of the biomass particle size 2.2.1. Mineral coal
[47]. Thus, the main novelties brought by this work are the detailed The coal used in this study is the EC 4500 type and was supplied
analysis on the emissions and temperature behaviors applied to by Copelmi Mineraça ~o LTDA, in January of 2016. The company also
cofiring of both, low-rank coal (CE 4500) and biomass (Pinus Elliotti provided some information that will be presented in the results.
sawdust), in a bubbling fluidized bed considering different coal/
biomass ratio, excess air ratio and fluidization velocity. Since these 2.2.2. Biomass
are key parameters for systems design several new studies and The biomass used in this study, furnished by Capivaras Industry
industrial applications that depend on this information may benefit and Trade LTDA, was a wood residue from the processing of pine
from the tendencies presented here to optimize the operational logs for the manufacture of wood boards in its plant located in the
parameters and control the emissions. city of Criciúma - SC. The wood residues were collected directly
from the conveyor belt of the circular saw. This equipment uses
water to cool the cutting blade, so the residue comes out with high
2. Materials and methods
humidity.
This study aims at analyzing the behavior of the combined
2.3. Thermogravimetric analysis
burning of low-rank coal (CE 4500) with biomass (Pinus Elliotti
sawdust) in a bench scale bubbling fluidized bed combustor,
A thermogravimetric analysis of the fuels was performed using a
focusing on the effect on the emissions of the coal/biomass ratio
TGA brand LECO™ Corporation, model 701/604100700 and
used.
applying the method known as "Coal constant".
In this way, it is possible to determine the humidity, ash content,
2.1. Inert particles volatile matter content and the fixed carbon content (by differ-
ence). For this purpose, the following methods were used: ASTM
Bubbling fluidized bed combustors contain a mass of granular D7582-15 for hygroscopic moisture, ash, volatile matter and fixed
solids generally having sizes ranging from 0.1 mm to 1.0 mm, carbon by macro thermogravimetric analysis, and ASTM D 3172-13
depending on the application type. These solids are called bedding for proximate analysis of coals and coke.
material and are characteristically sand, limestone (when burning
coals with high sulfur content), and ashes resulting from coal 2.4. Bubbling fluidized bed combustor
burning when in operation. Inert bed solids account for 97e99% of
all particulate matter present in this equipment [1]. For the experimental tests, a laboratory scale bubbling fluidized
For the current work, two types of candidate particles were bed reactor was used, composed of a cylindrical combustion
studied, a sand widely used in the manufacture of casting molds chamber of 120 mm internal diameter, 4 mm thick walls, and
(casting sand) and a sand used in pool water filters (filter sand). height of 1620 mm, made of stainless steel AISI 310.
Tests were carried out to determine the densities, using NBR 8630 The combustor has at its base a plenum chamber, responsible for
(ABNT, 1984) and NBR 9745 (ABNT, 1987) standards for bulk density stabilizing the flow and absorbing pressure fluctuations of the inlet
and actual density, respectively. The sphericity of the particles was gas, before the perforated air distribution plate in the lower base of
estimated from a correlation indicated by Foust et al. [33] from the the combustion chamber. The plenum has the same diameter of the
measurement of the porosity of the fixed beds. The particle size combustion chamber and useful height of 385 mm.
range and mean diameter were obtained from image analysis using To avoid a large temperature gradient on the combustor wall
scanning electron microscopy, through a SEM microscope model and to ensure the stabilization of the system temperature, an
ZEISS™ EVO MA10 with EDS Bruker QUANTAX™. electric furnace divided in 3 heating zones with 3800 W each, was
Casting sand was selected because of the good particle size erected around the cylindrical body of the equipment, allowing the
distribution and to be in group B of the Geldart’s classification. The furnace to be heated to temperatures close to 1000
C, also
filter sand has a grain size slightly larger than the casting sand and providing the initial heating of the system in the starting process.
was positioned in group ‘D’ of the Geldart’s classification. The upper outlet of the combustion chamber was connected to a
4 P.R. Wander et al. / Energy 203 (2020) 117882

Stairmand-type solid-gas separation cyclone [21] with an internal

cylindrical body with 80 mm reference diameter.
A rotary type flow meter with a range of 0e400 L/min at
ambient temperature and pressure (21
C and 101.3 kPa) and a
needle valve provided a known flow of air to the test bench. The air
was heated to 500
C by a 3 kW electric air heater before being
inserted into the plenum chamber.
The fuel supply system was provided with a helical feeder type
valve with steel body and screw in polyamide adapted for low flow
and with reasonable precision of adjustment and control. The
feeder was driven by a DC motor with 0e24 V variable power
supply and an analog variable voltage regulator.
The fuel is fed into the reactor 400 mm above the distributor
plate. To prevent premature heating and possible volatilization of
the lighter fractions, the tilted feeder tube was provided with a
water jacket, thus maintaining the temperature appropriate to the
best performance of the equipment.
The air distributor located at the base of the combustor was a
perforated plate type constructed according to the methodology
proposed by Basu [1]. The distributor plate has 384 holes of 1 mm in
diameter and triangular arrangement with pitch to diameter ratio
of 5.5. To prevent the sand from falling through the holes into the
plenum chamber, a 200 mm stainless steel screen was placed on the
top of the distributor. The Silagra 1108 silica and alumina fiber,
commercially named Silplate™ 1108 from Unifrax was chosen due
to the difficulties imposed by the plate holes manufacturing and the
possibility of buckling effect on steel due to temperature variation.
For the ash removal system and bed level maintenance, the plate
was modified to have a central hole through which a 34 mm outer
diameter drain was positioned to allow variation in bed height and
particle removal of the combustor at the end of the experiments.
The equipment also has several type K thermocouples posi-
Fig. 1. Schematic of the bubbling fluidized bed combustor. 1 - inlet air pre-heater, 2 e
tioned near the wall of the combustor, as can be seen in Fig. 1, one in
plenum chamber, 3 e distributor plate, 4 - combustion chamber, 5 - insulation, 6 -
the plenum chamber, to monitor the primary air temperature, 6 feeder, 7 - silo, 8 - thermocouple, 9 - pressure tap, 10 - electric oven, 11 - cyclone.
thermocouples distributed in the combustion chamber, for process
temperatures acquisition, and one thermocouple at the air heater
output to control the heating temperature thereof (not shown in combustion tests: the excess air ratio, the biomass percentage
Fig. 1). mixed with coal and the fluidization velocity.
Thermocouples L1, L2 and L3 were positioned in or close to the For comparison purposes, combustion tests were carried out
bed, respectively 75 mm, 200 mm and 320 mm above the distrib- using only coal, varying the excess air ratio and the fluidization
utor, which means that L3 is on the top of the expanded bed. velocity, and the flue gas components concentration from these
Thermocouples F1, F2 and F3 were at the freeboard, respectively tests were used as reference for the evaluation of the combined
720 mm, 1170 mm and 1600 mm above the distributor, at the top of fuels combustion results. Casting sand was selected as bed material
the freeboard. between two candidate particulates, considering Geldart’s classi-
Pressure taps were also distributed along the structure of the fication and minimum fluidization characteristics.
equipment to monitor the pressure drop as well as check bed The fluidization velocities indicated in Table 1 are relative to the
fluidization. Fig. 1 presents the experimental apparatus. An induced minimum fluidization velocity of the sand. The air flow rates used
draft fan (EAC 1/2 MAR model by Artek™ with a frequency inverter in this work were determined using the selected fluidization ve-
WEG, model CFW100) was also used after the cyclone so the reactor locities and the air excesses were obtained for each fluidization
operated under a slight depression (50e200 Pa) condition, avoiding velocity by changing fuel feed rate. In this way, it was possible to set
the backflow of combustion gas through the feeding tube. the operating speeds in terms of the fluidization velocity. Table 1
A portable gas analyzer of ECIL brand Chemist 500 was used to provides the operational parameters that were tested comprising
evaluate the concentration of combustion gases: O2, CO, NO (cor- the 16 different conditions of the test matrix.
rected internally to NOx), SO2, CO2 and CxHy. The measurements The condition to initiate the fuel feeding was attained when the
were carried out downflow the cyclone, to avoid saturation of the temperature in the bed reached a value around 650
C and the
analyzer filters by particulates, and preserve the equipment. internal pressure of the freeboard became stable. This temperature
Concentration measurements of CO2, CO, NOx in parts per
million (ppm) were converted to mg/Nm3 (273 K and 1 atm), as
suggested by Clarke [36], and corrected to constant oxygen (O2)
Table 1
fraction of 7%, as demanded in environmental regulations and to System operational parameters.
allow comparisons at different excess air ratios.
Parameters Values

2.5. Experimental approach Excess air ratio (%) 50 and 70

Biomass (by mass) (%) 0; 5; 10 and 15
Fluidization velocity (m/s) 3 and 6 umf
Three main parameters were evaluated for the execution of the
 P.R. Wander et al. / Energy 203 (2020) 117882 5

was defined from the ignition point proposed by Garcia [25] beds). Table 3 shows the velocities measured.
considering the volatile matter of coal and far exceeding the igni- Pang et al. [39] showed that the minimum fluidization velocity
tion point. variation with temperature depends upon the particle size. The
Fuel samples were mixed manually, since the small quantities minimum fluidization velocity decreases with an increase in tem-
used per test day (up to 10 kg) do not require specific equipment for perature, considering small particles (Geldart’s group B). The trend
good homogeneity. Subsequently the mixture was conditioned in is reversed for large particles (Geldart’s group D). Operating tem-
the feed silo from where it is fed into the combustion chamber at perature in a fluidized bed affects the interphase momentum ex-
the feed rate established in the test matrix. After bed temperatures change coefficient, and consequently, the drag force exerted by the
stabilization, data from the flue gas and the temperature conditions gas on the particle. The value of interphase momentum exchange
of the equipment were acquired. The procedure was done for each coefficient increases with an increase in temperature for Geldart’s
condition of the test matrix. group B particles, whereas, it decreases for group D particles.
For each test performed, three flue gas samples were taken at Nevertheless, solid phase axial velocity decreases with an increase
intervals ranging from 3 to 5 min. Temperature information along in temperature for both group B and D particles [39].
the reactor was monitored by supervisory software on the Field- Agu et al. [40] developed correlations to estimate the residence
Chart NOVUSTM platform. Data were stored for subsequent time necessary for complete conversion of biomass using bed
appreciation. temperature and pressure time histories. It was found that the
residence time decreases with increasing air flowrate, as in this
2.6. Measurement uncertainty work. Also, the model of Agu et al. [40] used the predicted amount
of biomass accumulated in the bed for determination of the mini-
The random errors analysis of experimental uncertainty is a mum fluidization velocity, slugging velocity and other bubble-
result of small independent and uncontrollable uncertainty sour- induced bed properties.
ces, estimated through statistical analysis. The total uncertainty
was achieved performing the method of propagation of indepen- 3.2. Characterization of the fuels
dent uncertainties, as proposed by Kline and McClintock [37]. This
method was applied to the results of chemical species concentra- To carry out the stoichiometry calculations of the fuels and
tion and temperature. mixtures used in this study, the ultimate analysis of the mineral
For the uncertainty regarding the measurement of temperature coal and biomass was carried out. Also, sizing bands were defined
both equipment accuracy and statistics were considered. This value based on the grain size test and the decision making considered the
is consistent with the uncertainty in measuring the standard de- characteristics of the equipment such as combustion chamber
viation during the operation. The uncertainty of the thermocouple diameter, associated velocities and feeder geometry.
in the measured range was also considered. Thus, the propagation For the design of the combustor feeder and subsequent tests, it
of the uncertainties for temperature yields the value of ±4 K. was necessary to prepare the coal using a roller mill to pulverize it.
Several measurements were made for each operational condi- The granulometric test using the DNER-ME 080/94 method was
tions tested in this work, considering various chemical species. applied to the coal powder. The analysis of the biomass, Pinus
Thus, the measurements uncertainty related to the emissions were residue, used as received, showed that its calorific value is similar to
propagated with those reported by the equipment manufacturer that of the CE 4500 coal, however, presenting higher moisture
information. Since the measurement uncertainty is different case content when compared to the mineral coal. Table 4 shows the
by case, the insertion of error bars in the graphs would be characteristics of these fuels.
confusing. Therefore, a statement with the maximum and mini- From the ultimate analysis, three main differences are observed:
mum uncertainties is provided for each gas measurement. nitrogen and sulfur content are much greater for coal, what may
have an effect in nitrogen and sulfur oxides emissions, and the high
3. Results and discussion oxygen content in biomass, resulting in smaller air needs and an
easier combustion.
3.1. Analysis of inert material The granulometric analysis of coal was performed using DNER-
ME 080/94 and a significant percentage of fines (particles below
Table 2 shows the main characteristics of the casting sand par- 0.3 mm) was observed, indicating the high friability of the Brazilian
ticles used in the bed. southern coal when subjected to mechanical processing. The par-
Applying the methodology described for the determination of ticle diameters selected were in the range of 0.59e2.00 mm aiming
the minimum fluidization velocity of the bed particles, velocity the operation stability. Particulates smaller than 0.5 mm tend to be
graphs were set by the pressure drop and the velocities were more easily entrained by the fluidization gas and particles larger
defined from the intersection of the characteristic curves of the than 2.00 mm do not perform well in the feeders used in the tests.
fluidized regime and the flow deceleration in the bed [38]. Samples were obtained using a batch type vibrating screen appa-
Duplicate tests were performed using two bed heights, defined ratus. A classified coal sample was analyzed at the scanning elec-
as one and two times the tube diameter of the experimental tron microscopy to verify its geometric characteristics.
apparatus (typical values for application in bubbling fluidized Fig. 2 shows the microscopic image of the coal. A relatively good
homogeneity of the particles of the classified sample was noted.
There were too few elongated grains, characterizing the good
Table 2 sphericity of the fuel sample, ensuring good sliding through the
Physical characterization of casting sand.
tilted feeder tube to the combustor. Also, some fines were observed,
Characteristic Methods Values aggregated to larger particles, which eventually favors the flow.
Granulometry (mm) DNER-ME 080/94 250e350 For biomass, after three days of natural drying of the residue, a
Actual density (kg/m3) NBR 9745 (1987) 2263 wide particle size range was observed, but it was decided to discard
Bulk density (kg/m3) NBR 8630 (1984) 1514 particles below 0.59 mm and above 2.00 mm, to facilitate the
Sauter mean diameter (mm) 290 operation of the combustor avoiding inconsistencies with the
Geldart’s classification B
combustor bench scale, especially regarding the fuel feeder.
6 P.R. Wander et al. / Energy 203 (2020) 117882

Table 3
Minimum experimental fluidization velocity of bed particles.

Particle Mean diameter (mm) Bed Height (m) Minimum fluidization velocity umf (m/s)

Casting sand 295 0.114 0.122

0.114 0.124
0.228 0.161
0.228 0.158

Table 4
Characteristics of the fuels.

Characteristic Methods CE 4500 Biomass Type 1

Higher Heating Value (kJ/kg) ASTM D 5865-13 19,121 19,550

Lower Heating Value (kJ/kg) Calorimetric Pump Isoperibol 18,368 18,190
Total delivery humidity (%) ASTM D2013/D2013M-12 13.45 19.91
Hygroscopic humidity (%) ASTM D7582-15 1.57 3.54
Ultimate analysis
Carbon (%) ASTM D5373-2014 48.32 52.44
Hydrogen (%) ASTM D5373-2014 3.46 6.29
Nitrogen (%) ASTM D5373-2014 0.84 0.04
Oxygen þ Halogens (%) ASTM D3176-15 8.39 41.23
Proximate analysis
Ashes (%) TEG 701 Thermogravimetric 37.68 n.d.
Volatile Matter (%) TEG 701 Thermogravimetric 28.26 87.92
Fixed Carbon (%) TEG 701 Thermogravimetric 34.06 12.08
Total Sulfur (%) ASTM D4239-14E2 1.31 n.d.

Classified biomass was also analyzed by scanning electron mi-

croscopy, showed in Fig. 3, and different from those of coal, it has
more elongated profiles following the fibers that are characteristic
of this type of material. Due to the lower sphericity, a greater dif-
ficulty in sliding along the tilted feeder tube of the combustor was
expected.
The biomass density values lower than that of coal also suggest
complexity in the gravity type fuel feeding, indicating greater flow
difficulty when used alone or added in large proportions to coal.
Therefore, to secure the homogeneity of the mixture, it was decided
to use up to 15% by mass of biomass in the mixture with mineral
coal in this work.

3.2.1. Thermogravimetry
According to the method presented previously, thermogravi-
metric analysis of the fuels was carried out to determine the

Fig. 3. Biomass scanning electron microscopy.

composition of hygroscopic moisture, volatile content, ash content

and fixed carbon. Table 5 presents the results found.
The thermogravimetric curves (TGA) of the fuels are shown in
Fig. 4.
The higher moisture loss of the biomass in the first temperature
interval (0e105
C) as well as a large difference in the mass loss for
biomass in the period of volatiles loss (250e950
C) was observed.

Table 5
Fuel proximate analysis.

Analysis/Testing Coal Copelmi CE 4500 (%) Pinus Biomass (%)

Hygroscopic moisture 5.12 11.45

Ash content 37.68 0.35
Volatile material content 27.45 74.32
Fixed carbon content 34.87 25.33
Fig. 2. CE4500 scanning electron microscopy.
 P.R. Wander et al. / Energy 203 (2020) 117882 7

Fig. 4. Mass loss of fuels subjected to thermogravimetric analysis.

At the third level (600e750
C) relative to fixed carbon loss, it is means up to 60 mm above the bed.
noted that coal mass loss is higher than biomass and also spent During the experiments bed temperatures were kept between
more time in this step to reach a constant mass. 800 and 860
C and operation was considered in steady state. The
For a better understanding of the phenomena associated to the average temperatures read on the thermocouples inside the bed
test, the thermogravimetric derivative curves from the base fuels and the freeboard were as follows: L1, L2 and L3 ¼ 837
C;
were analyzed, as shown in Fig. 5. Before pyrolysis, biomass initi- F1 ¼ 852
C, F2 ¼ 821
C and F3 ¼ 800
C. In the denser region (L1,
ates the water loss in a faster way than coal. The higher peak for L2 and L3) the thermal inertia of the sand bed is responsible for the
biomass in the curve derived at the second interval (250e950
C) stable temperatures despite the inherent intermittency of the
infers that the volatiles release starts at a lower temperature and screw feeder. Only L3 point, located at the top of the bed, at tran-
occurs much faster than coal, indicating that the diffusivity and sition zone, undergoes a slight downward fluctuation, probably
external convection is greater for biomass. caused by the cold fuel entering the zone.
For the fixed carbon burning region, no significant difference The freeboard thermocouple F1 is the closest to the bed and
was observed in the reaction rate of the fuels from the burning suffers significant variation of temperatures caused by feeding
peak. However, coal required a longer time for all carbonaceous intermittence. It was noted that when a larger amount of fuel is
contents to be converted compared to biomass. In this way it is released by the feeder, the fast combustion of the volatiles heats the
possible to verify a higher reactivity in the combustion of biomass lower zone of the freeboard, causing a pulsating effect on the
fixed carbon despite of the higher fixed carbon content in the coal. temperature at the F1 position.
The F2 and F3 positions located in the middle and top of the
freeboard, follow in a smoother way the temperature fluctuations
3.3. Operating conditions
presented by the thermocouple F1, demonstrating that the burning
of the volatiles is concentrated in the region closer to the feeder
The bed height at rest was 240 mm and the fuel was fed at a
(400 mm above the distributor plate).
height 160 mm above the bed of particles. During fluidization, that

Fig. 5. DTG analysis of the thermogravimetric derivatives curves of fuels.

8 P.R. Wander et al. / Energy 203 (2020) 117882

A significant decrease in temperature at position F3 (top of the 3.5. Emissions analysis

combustor) was observed, which must have been caused by the
heat transfer to the uninsulated metal parts. 3.5.1. CO2 emissions
The concentration of CO2 in the exhaust gases is usually asso-
ciated with combustion efficiency and quality. Thus, an increase in
concentration is expected when the operating conditions are
3.3.1. Analysis of ashes by loss on ignition
improved through better mixing (turbulence), suitable tempera-
To evaluate the combustion efficiency, ash samples collected at
ture and excess air. Fig. 7 shows the results of CO2 concentration
the cyclone were evaluated through the loss on ignition analysis.
found during the tests. Measurement uncertainties for CO2 pre-
The methodology was based on ASTM C25-17. Table 6 presents the
sented the highest and lowest values of 1.4 and 0.7%, at condition of,
results obtained.
l ¼ 1.5, 3 umf, C95B05 and, l ¼ 1.5, 3 umf, C85B15, respectively.
The results show that the fuel conversion was very good for all
Although slight, an increase in CO2 production is observed for
the fuel mixtures tested, thus the conditions employed in the test
larger excess air ratio and also for higher fluidization rates. In
matrix were adequate to combustion using bubbling fluidized bed.
addition, increasing the biomass portion in the coal blend leads to a
Ashes presented a light gray aspect and there was no significant
reduction in CO2 concentration. The increase in CO2 concentration
change in carbon concentration in the ashes with the increase of
with excess air ratio means that probably there is a better com-
biomass in the mixture suggesting that the biomass had sufficient
bustion due to the increased air availability. This trend is clearly
residence time to achieve a good carbon conversion.
more pronounced for higher fluidization velocity, which means
higher turbulence. Considering that biomass presents more oxygen
in its composition, an increase in CO2 concentration would be ex-
3.4. Analysis of the temperature in the combustor pected, what didn’t happen, indicating a worse combustion due
probably to biomass particles devolatilization and volatiles partial
Fig. 6 shows the average bed temperature of thermocouples L1, burning above the fluidized bed.
L2 and L3 (refer to Fig. 1) as a function of the fuels and conditions Biomass has a larger volatile fraction and its flames tend to be
used. The denomination C100B0 is a sample with 100% of coal and longer in cofiring cases, widening the reaction zone [9]. In this
0% of biomass, and so on. widespread combustion zone, the residence time is lower than
The increase of biomass portion in the fuel mixture leads to an required for a complete combustion. Therefore, the results for
increase in the system temperature. A slight increase in tempera- cofiring presents lower CO2 and higher CO (as shown in the next
ture is also observed for higher fluidization rates due to the increase section). Although differences in CO2 concentrations do not exceed
in turbulence, leading to higher heat and mass transfer rates. On the 2e3% when increasing biomass fraction, there is a tendency to
other hand, it is possible to verify a lower bed temperature differ- decrease CO2 concentration, mainly for lower operating speeds (3
ence among the various fuel mixtures when operated at the higher umf). One of the possible reasons for this behavior is the premature
fluidization velocity (6 umf) caused by the greater turbulence drag of volatiles released by the biomass before a complete con-
leading to a better mixing. The temperature is slightly reduced for version. With increased fluidization air velocity (6 umf), a lower
greater excess air ratio, which means in practice less fuel for the tendency to decrease CO2 concentrations relative to the amount of
same air quantity to be heated. biomass was observed, probably due to the greater turbulence both
The rise in the mean temperature of the system is due to the at the bed and at the freeboard, which enhances also the com-
increase in the biomass fraction in the mixture, as can be observed bustion efficiency of the 100% coal fuel.
from the average bed temperature in the steady state operation. The results of other authors corroborate with those achieved in
This is mainly due to the smaller concentrations of inert material in this work. Kook et al. [42] found that as the reaction temperature
the biomass when compared to the high ash content coal used. The and equivalence ratio were increased, the concentrations of CO2 gas
higher operating temperatures of the equipment are decisive fac- decreased, what agree with the observed in the present work
tors in heat and power applications. (Fig. 6). Slight increases in CxHy concentrations were also observed
All operating conditions showed a stable operation of the by Refs. [42] with increasing the bed temperature. Kumar and Singh
equipment as a function of temperature homogeneity, mainly with [43] employed four kinds of biomass with coal in a bubbling flu-
mixtures containing up to 10% biomass. However, in the burning of idized bed in air fired and oxycombustion. The measured per-
the mixture with 15% of biomass in the fluidization regime of 6 umf centages of NOx, CO2 and other gases are highly influenced by air
there was greater difficulty to obtain system stability due to the flowrate. Also, the emissions and combustion efficiency were
high turbulence and the increase of the volatile content. Bharath investigated. The effect of particle size distribution on combustion
et al. [41] applied rice husk and coal in a lab-scale bubbling bed efficiency is additionally studied. The experimental results show
reactor, Indian coal with 36% ash and Indian rice husk with 22% ash, that a maximum combustion efficiency of 97% was accomplished
considering proper particle sizes, air flow rates and steam flow with 75%coal/25%biomass.
rates. The results showed that when rice husk is added, all the
performance metrics presented an upward trend. The authors 3.5.2. CO emissions
concluded that a proper choice of fluidizing regime, particle sizes Carbon monoxide is also widely used for the qualitative analysis
and equivalence ratio contribute to optimum performance. of combustion efficiency. This component has its concentration
limited by environmental legislation and its control is necessary for
legal compliance. Fig. 8 shows the CO data acquired during the
Table 6
experiments. Measurement uncertainties for CO presented the
Loss on ignition analysis for ashes.
highest and lowest values of 51.3 and 7.4 mg/m3, at conditions of,
Biomass fraction (%) Weight loss (%) l ¼ 1.7, 6 umf, C85B15 and, l ¼ 1.5, 6 umf, C95B05, respectively.
0 1.44 Increasing the amount of biomass in the mixture leads to an
5 1.10 increase in CO production in accordance with the already shown
10 1.05 CO2 reduction. As expected, a reduction in CO production is
15 1.29
observed for higher excess air and higher fluidization rate.
 P.R. Wander et al. / Energy 203 (2020) 117882 9

Fig. 6. Bed temperature for 3 umf (left) and 6 umf (right). - - C100B0,  - C95B05, :- C90B10 and, ;- C85B15.

Fig. 7. Influence of biomass on CO2 emissions at 3 umf (left) and 6 umf (right). - - C100B0,  - C95B05, :- C90B10 and, ;- C85B15.

Fig. 8. Influence of biomass on CO emissions to 3 umf (left) and 6 umf (right). - - C100B0,  - C95B05, :- C90B10 and, ;- C85B15.

Nevertheless, CO production increases for the 15% biomass con- oxidation occurs. Those processes depend on fuel properties such
centration at higher fluidization rate, probably due to some insta- as particle size, moisture and volatile fraction. Once volatiles are
bility during the operation that might have affected the results. partly released before reaching the bed, it undergoes an incomplete
The moisture inside biomass vaporizes first, then volatile gases combustion at the freeboard, heating the region over the bed. This
are burned depending on the heating rates, and lastly the char is corroborated by the presence of higher levels of CO and the
10 P.R. Wander et al. / Energy 203 (2020) 117882

reduction of O2 consumption. The biomass char is fluidized fuels. Gong et al. [45] observed that total nitrogen oxide emissions
together with the coal and burns very quickly. increase with the bed temperature while decrease with the oper-
Despite the increase of CO concentration with the addition of ating pressure. The results show that most operating parameters
biomass in all the tests, the values obtained are low. A better have an opposite effect on NOx and N2O emissions, and the N2O
combustion is obtained with higher excess air ratio and greater emissions mainly depend on the bed temperature. Increasing the
fluidization velocity. This suggests that complete oxidation of operating pressure can significantly suppress the fuel-N conversion
biomass volatiles in the freeboard region did not occur. to NOx but enhance its conversion to N2O. With the rise of the
In a similar work, Xie et al. [44] investigated experimentally the excess air level and fluidization number, NOx emissions grow
emissions in a bench scale circulating fluidized bed combustor for distinctly while N2O emissions remain almost unchanged.
coal combustion and co-firing coal and biomass. It was observed Wisut et al. [46] observed that at different bed temperatures, the
that an increase in the biomass shares resulted in an increase of the total conversion of fuel-nitrogen into NOx and N2O was almost
CO concentration. According to the authors, biomasses are less constant, but the increase of bed temperature enhances the NOx
dense than coal having a smaller terminal velocity than that of coal production as well as N2O reduction.
or sand, leading to devolatilization and char combustion occurring
well above the bed, where O2 concentration is already lower. The 3.5.4. CxHy emissions
high volatile content of the biomass also favors the reduction of O2 CxHy are hydrocarbons produced during the pyrolysis and
concentration, leading to a worse combustion efficiency, corrobo- burning reactions of fuels. The presence of CxHy as well as CO in
rated by the slightly higher levels of CO and lower levels of CO2. exhaust gases are indicators of incomplete oxidation of fuels. Fig. 10
shows the results of CxHy concentrations found during the tests.
3.5.3. NOx emissions Measurement uncertainties for CxHy presented the highest and
Nitrogen oxides are found in exhaust gases from boilers and lowest values of 0.17 and 0.041%, at condition of, l ¼ 1.7, 3 umf,
combustors in the industry. Considering the formation routes (fuel C100B0 and, l ¼ 1.7, 3 umf, C90B10, respectively.
NOx, thermal NOx and immediate NOx), due to the composition of The increase of the biomass portion in the coal mixture leads to
the fuels used and the technologies applied in this study, NOx an increase in the production of CxHy. Despite the higher volatile
measured concentrations are possibly more related to the fuel content, the presence of biomass produces more unburned hy-
mechanism. Fig. 9 shows the concentrations of NOx found in the drocarbons, mainly because the system was not specifically
tests. Measurement uncertainties for NOx presented the highest developed for the proper burning of volatiles, which causes a
and lowest values of 150.8 and 12.5 mg/m3, at condition of, l ¼ 1.5, dispersion in the flame and incomplete burning at this stage of
3 umf, C90B10 and, l ¼ 1.5, 6 umf, C90B10, respectively. combustion.
An increase in the NOx production is observed for larger excess An increase in CxHy production is also observed for higher
air ratio and, for mixtures with higher biomass fraction, a slight fluidization rates due to the greater difficulty in burning the vola-
increase for higher fluidization rates. Moreover, the increase of the tiles, which were blown out of the combustion region. In this sit-
biomass portion in the fuel mixture leads to a reduction in the NOx uation, fuel residence time is decreased and there is no time for a
production due to the lower amount of nitrogen in the biomass in complete conversion. Turbulence, which had a small positive effect
relation to the coal. on CO emissions, did not help in the case of CxHy as there was a
It is possible to observe that for the higher fluidization (6 umf) a clear increase in emissions at higher air speed. On the other hand,
small tendency to decrease NOx formation when the biomass mass the production of CxHy is reduced with the increase of excess air, as
fraction in the fuel mixture was increased. However, for the fluid- expected.
ization condition of 3 umf, the mixture with 15% of biomass showed
a decrease of almost 30% compared to coal alone for an excess air 3.5.5. SO2 emissions
ratio of 1.7. The sulfur compounds were evaluated considering the biomass
In addition, a small increase in concentrations were observed as fractions in the fuel as can be observed in Fig. 11.
a function of the excess air in the tests. However, emissions of NOx An increase in the concentrations of SO2 is observed for all the
present a more complex dependency on the nitrogen content of fuel mixtures directly linked to the increase of the fluidization

Fig. 9. Influence of biomass on NOx emissions at 3 umf (left) and 6 umf (right). - - C100B0,  - C95B05, :- C90B10 and, ;- C85B15.
 P.R. Wander et al. / Energy 203 (2020) 117882 11

Fig. 10. Influence of biomass on emissions of CxHy at 3 umf (left) and 6 umf (right). - - C100B0,  - C95B05, :- C90B10 and, ;- C85B15.

Fig. 11. Influence of biomass on SO2 emissions to 3 umf (left) and 6 umf (right). - - C100B0,  - C95B05, :- C90B10 and, ;- C85B15.

velocities suggesting that turbulence favors the formation of this satisfactorily. Ash analysis at both the cyclone and the bed after the
component. Measurement uncertainties for SO2 presented the tests indicated good conversion with small percentages of uncon-
highest and lowest values of 273.9 and 105.7 mg/m3, at condition of verted material. Images analyses suggest that biomass may present
l ¼ 1.5, 3 umf, C100B0 and, l ¼ 1.5, 3 umf, C90B10, respectively. feeding problems due to its low sphericity.
According to these uncertainties, differences related to biomass Concerning biomass fraction in the mixture, its increase led to a
fractions are not conclusive. slight increase in both CO and CxHy concentrations. CO2 and NOx
Sulfur content in biomass is lower than in coal and therefore, concentrations decrease when biomass fraction increases, while for
lower SO2 emissions, mainly as a result of dilution, were expected, SO2 there was no significant change in its concentration, despite the
but the interaction with ash and sand components into the bed may reduced sulfur percentage in biomass composition.
have catalyzed sulfur oxidation reactions. The sulfur chemical ki- The average temperature of the system increased with more
netics and interaction with ash were not considered in this work. biomass in the mixture. The results presented here might be
According to Xie et al. [44], in co-firing, biomass share increases explained mostly due to the high volatiles percentage of biomass
SO2 emissions, nevertheless they observed that the relationship and its high reactivity, showed in TGA analysis. When the fuel is
between SO2 emission and fuel sulfur content was non-linear. introduced above the fluidized bed, part of these volatiles is
Helmer et al. [47] observed many advantages of firing wood/coal released and burn before reaching the bed, what is perceived
mixtures over the burning of each fuel individually. According to through the large fluctuations in freeboard temperature, which
the authors, wood in the mixture slightly reduced SO2 emissions, increase when the biomass fraction was increased, corresponding
but combustion temperature and excess air were more crucial pa- to feeder intermittence. The combustion of the volatiles at the
rameters influencing SO2 and NOx emissions. freeboard is not enhanced by turbulence as in the bed and the
lower residence time impair its completeness, what explain higher
CO and CxHy and smaller CO2 concentrations when biomass is
4. Conclusions increased. NOx decrease is explained due to the smaller amount of
nitrogen in biomass composition, confirming that fuel NOx is the
The bed fluidized at most conditions initially proposed. Results main mechanism, since temperatures are too low for thermal NOx.
presented are for conditions where fluidization performed
12 P.R. Wander et al. / Energy 203 (2020) 117882

Considering temperature, one possible reason to its increase process limitations. Energy 2018;143:615e23. https://doi:10.1016/j.energy.
2017.11.065.
with more biomass is the higher percentage of inert material in ~ a R, Vielmo HA. Analysis of the effect on the mechanical
[5] Caetano NR, Catalun
coal, since both fuels have equivalent heating values. injection engine using doped diesel fuel by ethanol and bio-oil. International
As expected, increasing excess air decreases system tempera- Review of Mechanical Engineering 2015a;9(2):124e8. https://doi:10.15866/
tures. The emissions of CO, CxHy and SO2 also decreased with ireme.v9i2.4341.
[6] Caetano NR, Soares D, Nunes RP, Pereira FM, Schneider PS, Vielmo HA, van der
greater excess air ratio, mainly due to greater air availability for a Laan FT. A comparison of experimental results of soot production in laminar
better combustion. The increase in CO2 concentration is due to the premixed flames. Open Eng 2015b;5:213e9. https://doi:10.1515/eng-2015-
same reason, while the relative stability of NOx emissions might be 0016.
[7] Akram M, Tan CK, Garwood R, Thai SM. Vinasse e a potential biofuel e cofiring
attributed to bed temperature, according to literature. with coal in a fluidised bed combustor. Fuel 2015;158:1006e15. https://
The best mixing due to turbulence when using the 6 umf fluid- doi.org/10.1016/j.fuel.2015.06.036.
ization regime led to a smaller difference in bed temperatures [8] Hu M-C, Huang A-L, Wen T-H. GIS-based biomass resource utilization for rice
straw cofiring in the Taiwanese power market. Energy 2013;55:354e60.
when the fuel blend was varied. The most important parameters https://doi.org/10.1016/j.energy.2013.03.013.
that change with fluidization regime are turbulence and residence [9] Ndibe C, Grathwohl S, Paneru M, Maier J, Scheffknecht G. Emissions reduction
time for the reactions. Apparently, at the 6 umf fluidization regime, and deposits characteristics during cofiring of high shares of torrefied biomass
in a 500 kW pulverized coal furnace. Fuel 2015a;156:177e89. https://doi.org/
turbulence improved overall combustion indicated by smaller CO 10.1016/j.fuel.2015.04.017.
and greater CO2 concentrations, but the smaller residence time [10] Loeffler D, Anderson N. Emissions tradeoffs associated with cofiring forest
increased CxHy concentrations. biomass with coal: a case study in Colorado, USA. Appl Energy 2014;113:
67e77. https://doi:10.1016/j.apenergy.2013.07.011.
The combined burning of Brazilian mineral coal with pine wood
[11] Narayanan KV, Natarajan E. Experimental studies on cofiring of coal and
biomass is a potential solution for the minimization of greenhouse biomass blends in India. Renew Energy 2007;32(15):2548e58. https://doi:10.
gases, presenting good operation and control behavior for con- 1016/j.renene.2006.12.018.
[12] Collazo J, Pazo
JA, Granada E, Saavedra A,
Eguía P. Determination of the spe-
centrations of up to 10% biomass in the mixture and fluidization
cific heat of biomass materials and the combustion energy of coke by DSC
regimes up to 6 umf, requiring only trivial NOx and SOx control analysis. Energy 2012;45(1):746e52. https://doi:10.1016/j.energy.2012.07.
devices for compliance with legislation. 018.
In general, greater excess air ratio and fluidization velocity [13] Ndibe C, Maier J, Scheffknecht G. Combustion, cofiring and emissions char-
acteristics of torrefied biomass in a drop tube reactor. Biomass Bioenergy
achieved better results and the main difficulties were related to the 2015b;79:105e15. https://doi.org/10.1016/j.biombioe.2015.05.010.
release of volatiles before biomass reaches the bed, leading to [14] Huang C-W, Li Y-H, Xiao K-L, Lasek J. Cofiring characteristics of coal blended
greater emissions due to partial conversion. with torrefied Miscanthus biochar optimized with three Taguchi indexes.
Energy 2019;172:566e79. https://doi:10.1016/j.energy.2019.01.168.
[15] Caetano NR, Venturini MS, Centeno FR, Lemmertz CK, Kyprianidis KG.
Declaration of competing interest Assessment of mathematical models for prediction of thermal radiation heat
loss from laminar and turbulent jet non-premixed flames. Thermal Science
and Engineering Progress 2018;7:241e7. https://doi.org/10.1016/
Dr. Wander reports grants from CNPq - Conselho Nacional de j.tsep.2018.06.008.
Desenvolvimento Científico e Tecnolo
gico, during the conduction of [16] Barmina I, Valdmanis R, Zake M. The effects of biomass co-gasification and co-
the study. firing on the development of combustion dynamics. Energy 2018;146:4e12.
https://doi:10.1016/j.energy.2017.04.140.
CNPq is a traditional Brazilian funding agency that gave funds to
rez-Jeldres R, Cornejo P, Flores M, Gordon A, García X. A modeling approach
[17] Pe
build the reactor and some scholarships for students. to co-firing biomass/coal blends in pulverized coal utility boilers: synergistic
effects and emissions profiles. Energy 2017;120:663e74. https://doi:10.1016/
j.energy.2016.11.116.
CRediT authorship contribution statement [18] Tan P, Ma L, Xia J, Fang Q, Zhang C, Chen G. Co-firing sludge in a pulverized
coal-fired utility boiler: combustion characteristics and economic impacts.
Paulo R. Wander: Methodology, Writing - review & editing, Energy 2017;119:392e9. https://doi:10.1016/j.energy.2016.12.084.

vio M. Bianchi: Methodology, Investigation, Re- [19] Mun T-Y, Tumsa TZ, Lee U, Yang W. Performance evaluation of co-firing
Supervision. Fla various kinds of biomass with low rank coals in a 500 MWe coal-fired po-
sources. Nattan R. Caetano: Writing - original draft, Formal anal- wer plant. Energy 2016;115:954e62. https://doi:10.1016/j.energy.2016.09.
ysis. Marcos A. Klunk: Writing - original draft, Formal analysis, 060.
[20] Catalun ~ a R, Shah Z, Pelisson L, Caetano NR, Da Silva R, Azevedo C. Biodiesel
Visualization. Maria Luiza S. Indrusiak: Methodology, Writing -
glycerides from the soybean ethylic route incomplete conversion on the diesel
review & editing, Funding acquisition. engines combustion process. J Braz Chem Soc 2017;28:2447e54. https://
doi.org/10.21577/0103-5053.20170100.
[21] Kunii D, Levenspiel O. Fluidization engineering. second ed. Washington:
Acknowledgments Butterworth-Heinemann; 1991.
[22] Cremasco MA. Operaço ~ es Unit
arias em Sistemas Particulados e Fluid-
The authors acknowledge the Brazilian agency CNPq (National odina^micos. second ed. Sa ~o Paulo: Blucher; 2014, ISBN 978-85-212-0855-6.
[23] Geldart D. Types of gas fluidization. Powder Technol 1973;7(5):285e92.
Council of Technological and Scientific Development e Brasília, DF, https://doi:10.1016/0032-5910(73)80037-3.
Brazil) for the research funding (Proc. 40.6902/2013-5), and the [24] Grace JR, Avidan AA, Knowlton TM. Circulating fluidized beds. Heidelberg:
generous assistance of all the people from the company who Springer Netherlands; 1997, ISBN 978-94-010-6530-6.
[25] Garcia R. Combustíveis e Combust~ ao Industrial. second ed. Rio de Janeiro:
granted us access to their database and perception information. ^ncia; 2013. -13: 978-85-719-3303-3.
Intercie
Author Flavio M. Bianchi also acknowledges CNPq for the grants [26] Miller BG, Tillman DA. Combustion engineering issues for solid fuel systems.
(Proc. 375968/2015-6 and 374288/2016-0) Burlington, MA: Academic Press; 2008, ISBN 978-01-237-3611-6.
[27] Turns SR. An introduction to combustion: concepts and applications. third ed.
New York: McGraw-Hill; 2012. ISBN13: 9780073380193.
References [28] Obras-Loscertales M, Mendiara T, Rufas A, Diego LF, García-Labiano F, Gaya
n P,
Abad A, Ada
nez J. NO and N2O emissions in oxy-fuel combustion of coal in a
[1] Basu P. Combustion and gasification in fluidized beds. first ed. Boca Raton: bubbling fluidized bed combustor. Fuel 2015;150:146e53. https://doi:10.
Taylor & Francis; 2006, ISBN 978-0849333965. 1016/j.fuel.2015.02.023. j.fuel.2015.02.023.
[2] Collot AG. Matching gasification technologies to coal properties. Int J Coal Geol [29] Jones JM, Ross AB, Mitchell EJS, Lea-Langton AR, Williams A, Bartle KD.
2006;65:191e212. https://doi.org/10.1016/j.coal.2005.05.003. Organic carbon emissions from the co-firing of coal and wood in a fixed bed
[3] Catalun~ a R, Shah Z, Venturi V, Caetano NR, Da Silva BP, Azevedo CMN, Da combustor. Fuel 2017;195:226e31. https://doi:10.1016/j.fuel.2017.01.061.
Silva R, Suarez PAZ, Oliveira LP. Production process of di-amyl ether and its [30] Jurado N, Simms NJ, Anthony EJ, Oakey JE. Effect of co-firing coal and biomass
use as an additive in the formulation of aviation fuels. Fuel 2018;228:226e33. blends on the gaseous environments and ash deposition during pilot-scale
https://doi:10.1016/j.fuel.2018.04.167. oxy-combustion trials. Fuel 2017;197:145e58. https://doi.org/10.1016/
[4] Flegkas S, Birkelbach F, Winter F, Freiberger N, Werner A. Fluidized bed re- j.fuel.2017.01.111.
[31]

Riaza J, Alvarez L, Gil MV, Pevida C, Pis JJ, Rubiera F. Ignition and NO emissions
actors for solid-gas thermochemical energy storage concepts - modelling and
 P.R. Wander et al. / Energy 203 (2020) 117882 13

of coal and biomass blends under different oxy-fuel atmospheres. Energy bubbling fluidized bed gasification reactor. Fuel 2019;253:1414e23. https://
Procedia 2013;37:1405e12. https://doi:10.1016/j.egypro.2013.06.016. doi.org/10.1016/j.fuel.2019.05.103.
[32] Aho M, Envall T, Kauppinen J. Corrosivity of flue gases during co-firing Chinese [41] Bharath M, Raghavan V, Prasad BVSSS, Chakravarthy SR. Co-gasification of
biomass with coal at fluidised bed conditions. Fuel Process Technol 2013;105: Indian rice husk and Indian coal with high-ash in bubbling fluidized bed
82e8. https://doi:10.1016/j.fuproc.2011.05.020. gasification reactor. Appl Therm Eng 2018;137:608e15. https://doi:10.1016/j.
[33] Foust AS, Wenzel LA, Clump CW, Maus L, Andersen BL. Princípio das operaço ~ es applthermaleng.2018.04.035.
unit

arias. second ed. Rio de Janeiro: LTC; 1982, ISBN 978-8521610380. [42] Kook JW, Choi HM, Kim BH, Ra HW, Yoon SJ, Mun TY, et al. Gasification and tar
[34] Bianchi FM, Aquino TF, Kestering DA, Dalpont G, MArcello RR, Neto JM, removal characteristics of rice husk in a bubbling fluidized bed reactor. Fuel
Indrusiak MLS, Zinani F, Wander PR. Bancada Experimental para Estudos de 2016;181:942e50. https://doi:10.1016/j.fuel.2016.05.027.
~o de Combustíveis So
Fluidizaça
lidos em Sistemas de Leito Fluidizado Circu- [43] Kumar R, Singh RI. An investigation in 20 kW th oxygen-enriched bubbling
lante. Gramado: Anais do 4
CBCM; 2013. fluidized bed combustor using coal and biomass. Fuel Process Technol
[35] Darmawan A, Budianto D, Aziz M, Tokimatsu K. Cofiring assessment of 2016;148:256e68. https://doi:10.1016/j.fuproc.2016.02.037.
hydrothermally-treated empty fruit bunch and low rank coal in a drop tube [44] Xie J, Yang X, Zhang L, Ding T, Song W, Lin W. Emissions of SO2, NO and N2O
furnace. Energy Procedia 2017;105:1545e50. https://doi:10.1016/j.egypro. in a circulating fluidized bed combustor during co-firing coal and biomass.
2017.03.473. J Environ Sci 2007;19(1):109e16. https://doi:10.1016/s1001-0742(07)60018-
[36] Clarke AG. Industrial air pollution monitoring. Heidelberg: Springer 7.
Netherlands; 1998, ISBN 978-94-010-7143-7. [45] Gong Z, Shao Y, Pang L, Zhong W, Chen C. Study on the emission character-
[37] Kline S, McClintock F. Describing uncertainties in single-sample experiments. istics of nitrogen oxides with coal combustion in pressurized fluidized bed.
Mech Eng 1953;75:3e8. Chin J Chem Eng 2019;27(5):1177e83. https://doi:10.1016/j.cjche.2018.07.
[38] Zhang D, Fan P, Wu D, Li Y. Pressure drop across a fixed bed reactor with 020.
mechanical failure of catalyst pellets described by simplified ergun’s equation. [46] Wisut W, et al. Combustion and emission characteristics of a Thai lignite
China Particuol 2005;3(1e2):23e5. https://doi:10.1016/s1672-2515(07) with a bubbling fluidized bed. Hokkaido Kogvo Gijutsu Kenkyusho Hokoku,
60157-2. 70, 14-20, in 98/04003 Combustion and emission characteristics of a Thai
[39] Pang L, Shao Y, Zhong W, Liu H. Experimental investigation on the coal lignite with a bubbling fluidized bed. Fuel Energy Abstr 1998;39(5):374e5.
combustion in a pressurized fluidized bed. Energy 2018;165:1119e28. https://doi:10.1016/s0140-6701(98)93998-4.
https://doi:10.1016/j.energy.2018.09.198. [47] Helmer WA, Stokke DD, Sun J. A case study of fluidized-bed combustion of
[40] Agu CE, Pfeifer C, Eikeland M, Tokheim L-A, Moldestad BME. Measurement wood/coal mixtures. Part A. The effect of wood particle size. For Prod J
and characterization of biomass mean residence time in an air-blown 1999;48(3):46e9.