Fuel 78 (1999) 1425–1436

www.elsevier.com/locate/fuel

Interaction between emissions of SO2 and HCl in fluidized bed combustors

W. Xie, K. Liu, W.-P. Pan, J.T. Riley*
Department of Chemistry, Materials Characterization Center, Western Kentucky University, Bowling Green, KY 42101, USA
Received 11 September 1998; accepted 22 April 1999

Abstract
The ability to capture SO2 and halogens is one of the most important advantages of fluidized bed combustion (FBC). In order to clarify the
affects of chlorine in the absorption of SO2 emission, experiments involving the addition of PVC to coals were carried out using the 0.3 m ID
bench scale FBC system at Western Kentucky University. During the experiments, PVC was added to three coals in different percentages,
and the mixtures fed into the FBC system. The Ca/S ratio was kept constant at 3:1. The experimental results indicate that chloride addition
dramatically decreased the SO2 concentration in flue gases. The sulfur and chlorine contents in both fly ash and bed ash increased. At the
same time, the utilization efficiency of calcium increased with an increase in the PVC weight percent. This reduction in SO2 is attributed to
more voids on the sorbent particle surface (limestone and/or ash) through transient formation of a mobile halide ion-containing phase
(CaCl2), and to the reactions:

4HCl 1 O2 $ 2Cl2 1 2H2 O …1†

Cl2 1 SO2 1 H2 O $ 2HCl 1 SO3 : …2†

A proposed mechanism involving the interaction between SO2 and HCl is discussed in detail. q 1999 Elsevier Science Ltd. All rights
reserved.
Keywords: Fluidized bed combustion; Municipal solid waste; Dioxins

1. Introduction low combustion temperatures used in an FBC system, which

subsequently can suppress the corrosion of heat exchange
It is well know that the emission of SOx from fossil fuel- tubes sometimes caused by chloride compounds, especially
fired power plants is one of the causes of acid rain. It is also when firing municipal solid waste (MSW) or high chlorine
known that SO2 together with HCl emitted from fossil fuel fuels.
combustion may play roles in the corrosion of boiler compo- Experimental work in the retention of sulfur and halogen
nents and in power plant operational problems [1–6]. In situ species has been studied extensively in this decade. It has
sulfur and halogen capture by limestone is the major advan- been found that sulfur and halogen capture by sorbents is
tage of fluidized bed combustion (FBC). The use of calcium- interrelated [7,8]. A study by Liang and others [9] showed
based sorbents, such as limestone or dolomite, in fluidized that chloride capture shows a large variation with tempera-
bed combustion of coal to reduce sulfur dioxide (SO2) emis- tures, moving from a low of 18% gaseous HCl at 7008C to
sion is a well established technique. Limestone is introduced 99% HCl at 9508C. The resulting product was almost
into the fluidized bed combustor within the temperature entirely in the form of liquid CaCl2. Munzner and Schilling
range 750–9008C, the limestone is rapidly calcined to the [10] reported that a greater recapture of chloride was
porous calcium oxide, which can subsequently react with reached with larger excesses of limestone, or when the
SO2 to form calcium sulfite and calcium sulfate. Also, the Ca/S atomic ratio was greater than 2. However, from pilot
addition of limestone can capture hydrogen chloride to form plant studies and industrial tests [11] it was reported that
liquid or solid phase calcium chloride under the relatively using calcium-based sorbent injection in fluidized bed coal
combustors demonstrated utilization efficiencies of 30%.
* Corresponding author. Tel.: 1 1-502-745-6020; fax: 1 1-502-745- Incineration is an important waste-to-energy technology
6293. used for the disposal of municipal solid waste (MSW) or
0016-2361/99/$ - see front matter q 1999 Elsevier Science Ltd. All rights reserved.
PII: S0016-236 1(99)00070-8
1426 W. Xie et al. / Fuel 78 (1999) 1425–1436

Table 1
Analytical values (moisture is as-received, all other values are reported on a dry basis) for the coals and limestone used in the study

Coal 95010 Coal 95011 Coal 95031 KY Limestone

Proximate analysis (%)

Moisture 2.32 10.07 8.32 0.19
Ash 7.22 9.37 10.78 57.93
Volatile matter 39.97 43.34 37.21 18.90
Fixed carbon 52.82 47.29 52.02 22.98

Ultimate analysis (%)

Ash 7.22 9.37 10.78 57.93
Carbon 79.38 74.08 72.16 11.18
Hydrogen 5.31 5.08 4.82 0.16
Nitrogen 1.63 1.54 1.54 0.00
Sulfur 0.67 3.2 2.38 0.00
Oxygen 5.69 6.72 7.57 30.73
Chlorine (ppm) 1039 118 3070 36
BTU/pound 14 077 13 203 12 842 n/a

refuse-derived-fuels (RDF). However, it is necessary that aromatic compounds, including chlorophenols and poly-
the possibility of the formation of volatile organic chlorophenols, which are precursors of PCDDs and
compounds (VOCs) during combustion be reduced before PCDFs. In contrast, it is also possible that chlorine gas
incineration can reach its full potential. For instance, instead of HCl will more readily attack metal in combustion
polychlorinated organics, including polychlorinated systems to cause severe corrosion.
dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzo- Coal, as a co-firing energy source for municipal solid
furans (PCDF), have been reported as products of MSW wastes, is able to suppress the formation of chlorine-
combustion [12,13] in the parts-per-billion to parts-per-tril- containing organic compounds. Scheidle and co-workers
lion range. These materials, which may be formed under demonstrated that adding lignite coal as an auxiliary fuel
conditions similar to those in which molecular chlorine is to paper recycling residues decreased the levels of dioxins in
formed, have been found in the fly ash and flue gases of fluidized-bed combustion emissions [16]. Similar results
some incinerator facilities in the US and Europe. This has can be inferred from Lindbaurer’s study which showed
slowed or even stopped the construction and operation of that co-firing MSW with 60% coal drastically reduced the
waste-to-energy plants. formation of PCDD/Fs [17]. Banaee and co-workers studied
Municipal solid waste varies considerably in composition the co-combustion of Saran wrap with high-sulfur coals and
[14]. Many organic materials present in the MSW possess a other polymeric materials, and the results showed the ability
chemical structure capable of releasing chlorinated organic of high-sulfur coal to inhibit the formation chlorinated
compounds under pyrolysis and combustion conditions, benzenes [18]. Bonfanti and co-workers investigated the
especially plastics. Most plastics used for household appli- environmental aspects relevant to the co-firing of pulverized
cations are disposable and nonbiodegradable. Plastics coal and RDF in a slag forming combustor, and the results
account for 8% (by weight) of the total amount of municipal show that the micro pollutant emissions were very low [19].
solid wastes and can make up to 20% of the volume of the Ohlsson observed that despite the enhanced level of HCl
wastes [15]. As their usage is anticipated to grow in the due to the addition of RDF, no PCDD/Fs were detected
future, the amount of plastics discarded will also increase. when co-firing high-sulfur coal and RDF pellets [20]. Fran-
PVC is one of the leading plastics in total production. The kenhaeuser and coworkers also addressed the adverse
existence of PVC plastic, and other chloride-containing effects of SO2 in the formation of chlorinated organics
organics such as pesticides, herbicides, flame retardants, during the co-combustion of plastics with coal [21].
etc. in municipal solid wastes, is responsible for the forma- The main objectives of the study reported in this article
tion of corrosive or toxic substances like HCl, other halo- were to study the chloride-sulfur interactions during fossil
genated acids, chlorinated organic materials, and possibly fuel combustion and to investigate new ways for using FBC
PCDD/Fs, in the flue gases and particles from waste incin- systems with high sulfur/high chlorine coals/MSW or RDF
erators. Since MSW in the US generally consists of notable to minimize SOx emission and the emission of chlorinated
quantities of PVC, Saran and other chlorinated polymers, compounds, including PCDDs and PCDFs.
the yield of HCl from combustion of MSW will be significant.
Chlorine gas is a key intermediate in the formation of
chlorinated dioxin compounds. It is generally thought that 2. Experimental
it is molecular chlorine, and not HCl, that reacts with
aromatic compounds such as phenols to produce chlorinated All experimental work was conducted with the 0.1 MWth
 W. Xie et al. / Fuel 78 (1999) 1425–1436 1427

Table 2 of the combustor. The hot gases from the combustor are
The operating parameters allowed to enter the wet cyclone scrubber where they are
Fuel compositions: (1) 100% coal met with a wall of water (which keeps the cyclone cool),
(2) 89% coal, 1% PVC and 10% wood which subsequently takes solids from the cyclone into a
pellets holding tank. The operating parameters (air/water flow,
(3) 86.7% coal, 3.3% PVC and 10% wood coal/lime feed, bunker weight, temperatures, and pressure)
pellets
are controlled and logged to file with a Zenith 150 MHz
95010 95011 95031
Fuel feed rate (kg/h) 7.95 8.41 8.72 computer utilizing the labtech software version 3.0.
Limestone feed rate (kg/h) 0.52 2.53 1.95 During the combustion runs any needed changes in the para-
Ca/S molar rate , 3.0 , 3.0 , 3.0 meters could easily be entered into the computer, by acces-
Air flow rate (kg/h) 89.9 89.9 89.9 sing the correct control screen and making the necessary
Fluidizing velocity (m/s) 1.15 1.15 1.15
corrections on line.
Fly ash rate (kg/h) 0.93 2.95 2.59
Bed ash rate (kg/h) 0.47 1.57 1.45 Three coals were used in this study, an Illinois # 6 coal
Oxygen concentration in gas 4–5% (95031; 0.31% Cl and 2.4% S), an Eastern KY coal (95010;
0.10% Cl and 0.67% S) and a western KY coal (95011;
0.01% Cl and 3.2% S). Analytical data for the three coals
bench scale FBC system at Western Kentucky University. A and limestone used in this study are presented in Table 1.
full description of the FBC system has been presented by The limestone came from Kentucky Stone in Princeton, KY.
Orndorff and others [22], so only a brief description is given The coal and limestone both were air dried before being
here. In this project, an underbed continuous fuel/limestone crushed to 24 mesh (4.75 mm). The limestone was also
feeding system was installed at the 0.1 m point above the air used as the bed material in the FBC system. The PVC was
distributor in the FBC system. Six moveable heat exchange mixed with coal in three ratios: 0.5% by weight, 1% by
tubes were added in the bed area of the FBC system. Typical weight, and 3.3% by weight. During combustion runs, lime-
operation of the combustor involves setting the correct fuel/ stone was fed into the system at a constant rate, depending
limestone feed and air flows and then use the moveable heat upon the fuel used.
exchange tubes to adjust the bed temperature to the desired The gases eluting from the furnace were swept into three
setting. Another sixty-six gas heat exchange tubes are in a ice-cooled traps with impingers. The first two traps and
fixed position located approximately one meter from the top impingers contained 0.1 M H2SO4 to trap HCl and the last

Fig. 1. The effect of the PVC/fuel ratio on SO2 emission in the flue gas.
1428 W. Xie et al. / Fuel 78 (1999) 1425–1436

Fig. 2. The effect of the PVC/fuel ratio on the sulfur content in fly gas.

one contained 0.1 M NaOH to trap Cl2 on the basis of the gases was calculated by the average of two sample values.
following reaction: The effects of adding limestone and coal ash during the test
period were considered in the content change of each
Cl2 1 2NaOH $ NaCl 1 NaClO 1 H2 O …3†
compound in the bed material absorption calculation.
A Dionex Model 120 Ion Chromatograph was used to During combustion runs the flue gases at the gas heat
determine the concentration of chloride in both the acidic exchange region were analyzed continuously using on-line
and basic solutions. The mobile phase was Na2CO3/ FTIR spectroscopy and gas chromatography. The major
NaHCO3 and the flow rate was 1.2 ml/min. In all tests, no operating parameters for the experiments were as follows:
chloride was detected in the second impinger, meaning that excess air level—around 1.3; Ca/S ratio—approximately 3;
all HCl was absorbed in the first bottle. Therefore, all chlor- bed temperature was controlled between 1140 and 1160 K.
ide species in the third bottle resulted from the molecular Other details of the combustion conditions are listed in
chlorine produced in the combustion process. Table 2 [25].
The fly ash and bed ash sampling procedures used have
been discussed previously [22–24]. A brief description for
the mass balance test process is as follows: the first samples 3. Results and discussion
of solids (fly ash and bed ash) were collected after an 8 h
stable combustion period for each FBC test. Within the 3.1. SOx Emission
following 2–4 h, the second (10 h after the start) and the
third samples (12 h after the start) of solids (fly ash and bed One of the objectives of this study was to demonstrate
ash) and the first (9 h after start) were collected. The first that SOx emissions decrease as the amount of chloride in the
and second flue gas samples were collected 9 and 11 h after flue gas in the FBC system is increased. PVC is the principal
the start, respectively. The mass flow rates of fly ash and bed source of chloride in the fuels used in the study. Fig. 1 shows
ash were measured in the tests. The concentrations of the that the SO2 emissions decrease as the amount of PVC in the
compounds in the fly ashes and bed ashes were re-calculated fuel blend increases. It can be clearly seen that the SO2
on the basis of the mass flow rates of fly ashes and bed ashes, emissions decrease dramatically when a mixture of 1% by
respectively. The difference in the concentration of each weight of PVC is fed into the combustor, and then leveled
compound in the bed ashes between two adjacent samples off when the PVC content in the fuel mixture increased from
was due to the addition of each compound as captured in the 1 to 3.3 wt%. The sulfur content in the fly ash and bed ash
combustor. The concentration of each compound in the flue increased with an increase in the amount of PVC used in the
 W. Xie et al. / Fuel 78 (1999) 1425–1436 1429

Fig. 3. The effect of the PVC/fuel ratio on the sulfur content in bed ash.

fuel. Fig. 2 illustrates that the sulfur content in the fly ash used as the bed material, and the change of run conditions
increases as the PVC/fuel ratio increases. Likewise, Fig. 3 was such that no new limestone was added to the combustor
shows that the sulfur content of the bed ash increases with with the fuel. The samples of fly ash and bed ash were col-
an increase in the PVC/Fuel ratio. One explanation of the lected 8 h after the limestone feed to the combustor had
increased capture of SOx by limestone is that the transient stopped. In the presence of limestone, there is a significant
formation of a mobile halide ion-containing phases (i.e. improvement in the sulfur capture in the flue gas. Accordingly,
CaCl2) can modify the surface of the partially sulfated the sulfur content in both fly ash and bed ash increases. The
sorbent particles to form more voids on the surface, which material balance data for sulfur are presented in Table 4.
provide diffusion paths for HCl and SOx toward the interior
of a limestone particle, leading to increased SO2 capture 3.2. Chloride emission
[26]. However, a more complex but possible mechanism
involves the interaction of Cl2 with the SO2 from the For all experimental runs presented in this article, no
combustion of sulfur in the fuel producing SO3. Molecular molecular chlorine was detected in the flue gases, as
chlorine produced in the combustion process through the measured by the ion chromatographic analysis of chloride
Deacon reaction in the presence of excess oxygen is a key in the NaOH trapping solutions. In the case of the HCl
organic chlorinating agent is replaced by HCl, which is less emissions study, a difference in the behavior of chloride
likely to cause any chlorination of organic species. was noted for different amounts of sulfur in the fuels used
The results of the effects of limestone on the capture SO2 in the FBC system. Fig. 4 illustrates that a higher concen-
are shown in Table 3. In this experiment, limestone was tration of hydrogen chloride occurred in the flue gas from
Table 3
The effect of limestone on the distribution of sulfur

Coal 95010 with 3.3 wt% PVC Coal 95031 with 3.3 wt% PVC

With limestone No limestone a With limestone No limestone a

SO2 emissions in the flue gas (ppm) 9.91 136.86 17.62 1240.29
Sulfur content in fly ash (%) 0.95 0.32 1.30 1.37
Sulfur content in bed ash (%) 0.98 1.53 5.16 4.96
a
No new limestone was fed with fuel, bed material, and limestone.
1430 W. Xie et al. / Fuel 78 (1999) 1425–1436

Table 4
The percent distribution of sulfur in the different phases

Coal 95010 1 0% PVC 1% PVC 3.3% PVC

Before burning
Feed rate (kg/h): 7.95 7.95 7.95
Sulfur content in fuel (wt%): 0.67 0.60 0.58
Feed amount of sulfur (kg/h): 5.33 × 10 22 4.77 × 10 22 4.61 × 10 22
After burning
Solid phase
Bed ash
Content (wt%) 0.41 0.60 0.98
Flow rate (kg/h) 0.47 0.47 0.47
Amount (kg/h) 1.93 × 10 23 2.82 × 10 23 4.61 × 10 23
Bed material
Content change (wt%): 0.076 0.059 0.057
Material inventory (kg): 50 60 60
Amount (kg/h): 0.038 0.0354 0.0342
Fly ash:
Content (wt%): 0.36 0.50 0.95
Flow rate (kg/h): 0.93 0.93 0.93
Amount (kg/h): 3.35 × 10 23 4.65 × 10 23 8.84 × 10 23
Subtotal 0.043278 0.04287 0.04765
Wt % in total sulfur: 81.20 89.87 103.36
Gas phase
Concentration in flue gas (ppm): 100.05 27.64 9.91
Temperature of flue gas (8C): 150 150 150
Flow rate of flue gas (kg/h): 89.9 89.9 89.9
Flue gas volume (m 3/h): 113.12 113.12 113.12
SO2 volume in gas (m 3/h): 0.0113 3.13 × 10 23 1.12 × 10 23
Sulfur amount (kg/h): 9.93 × 10 23 2.88 × 10 23 1.03 × 10 23
Wt % in total sulfur: 18.63 6.04 2.23
Total 99.83 95.91 105.59
Coal 95031 1 0% PVC 1% PVC 3.3% PVC
Before burning
Feed rate (kg/h): 8.72 8.72 8.72
Sulfur content in fuel (wt%): 2.38 2.12 2.06
Feed amount of sulfur (kg/h): 0.208 0.185 0.180
After burning
Solid phase
Bed ash
Content (wt%) 2.06 3.24 6.76
Flow rate (kg/h): 1.45 1.45 1.45
Amount (kg/h) 0.0299 0.0470 0.0980
Bed material
Content change (wt%) 0.22 0.16 0.07
Material inventory (kg) 60 60 60
Amount (kg/h): 0.133 0.0954 0.042
Fly ash
Content (wt%) 1.04 1.22 1.31
Flow rate (kg/h) 2.59 2.59 2.59
Amount (kg/h) 0.0269 0.0316 0.0339
Subtotal 0.1898 0.1740 0.1739
Wt.% in total sulfur: 91.25 94.05 96.61
Gas phase
Concentration in flue gas (ppm) 305.65 100.04 17.62
Temperature of flue gas (8C) 150 150 150
Flow rate of flue gas (kg/h) 89.9 89.9 89.9
Flue gas volume (m 3/h) 113.12 113.12 113.12
SO2 volume in gas (m 3/h) 0.0346 0.0113 1.99 × 10 23
Sulfur amount (kg/h) 0.0303 9.93 × 10 23 1.75 × 10 23
Wt % in total sulfur 14.57 5.37 0.97
Total 105.83 99.41 97.58
 W. Xie et al. / Fuel 78 (1999) 1425–1436 1431

Fig. 4. The effect of the PVC/fuel ratio on HCl emission in the flue gas.

Fig. 5. The effect of the PVC/fuel ratio on the content of chloride in the bed ash.
1432 W. Xie et al. / Fuel 78 (1999) 1425–1436

Fig. 6. The effect of the PVC/fuel ratio on the content of chloride in the fly ash.

the combustion of the low sulfur fuel (95010), as compared The effects of the PVC/fuel ratio on the chloride contents
to the HCl in the flue gas from the combustion of the high in the fly ashes is presented in Fig. 6. The chloride contents
sulfur fuel (95031) when PVC was added to the coals. in fly ashes increase dramatically with an increase of the
However, the HCl contents in the flue gases are reversed PVC/fuel ratio from zero to 3.3% by weight. Compared to
for these two coals when no PVC is added, even though the the results shown in Fig. 5, it is obvious that the change of
high sulfur coal (95031) also has about 30 times more chlor- chloride contents in fly ashes is more than that in bed ashes.
ine compared to the low sulfur coal. As expected, HCl emis- It is known that the optimum temperature for SO2 capture by
sion in the flue gases increased with an increase in the limestone in an FBC system is around 8508C in the fluidized
amount of PVC added to the fuel. It is known that absorption bed itself. However, the favored temperature for HCl reten-
of SO2 by calcium oxide causes pore closure of limestone, tion is relatively lower, and normally less then 6508C. HCl
resulting in the reduction of sulfation, because the molar retention by calcium oxide also depends on the concentra-
volume of the final product, CaSO4, is much greater than tion of HCl in the flue gases [31]. Compared with bed ash,
that of CaO. This is one of the reasons that the utilization of fly ash has a greater chance to absorb chloride since the
calcium in limestone is only 30–40%. However, the sulfation temperature of the flue gas decreases along the height of
behavior may help reduce HCl emissions. In contrast to the the combustor, resulting in greater absorption of chloride.
case of sulfation, it is assumed that chloride trapped on The material balance data for chloride are presented in
the surface of limestone is transported from the surface Table 5.
into the bulk of the absorbing particle by diffusion or by The relationship between the chloride and sulfur contents
the migration of CaCl2 to other parts of the particle. With an in the combined ashes is shown in Fig. 7. A high sulfur
increase of sulfur content in the coal and progression of content is always accompanied by a high chloride content
sulfation, the CaSO4 layer on the surface of the limestone in these ashes. The different sulfur contents in the fuels, and
particle becomes thicker, and diffusion into the particle the different retention characteristics of sulfur are also
becomes more difficult, eventually stopping the reaction. expressed in the figure. For instance, fuel 95011 has the
So this layer can suppress the escape of chloride in the highest sulfur content, leading to the sharpest curve
vapor at a combustion temperature of 8508C, which is among the three fuels. In contrast, only a slight change
higher than the melting point of CaCl2. This suggested occurred for fuel 95010, which has the lowest sulfur
behavior is also supported by the chloride contents in the content. Consequently, it can be said that the SOx capturing
bed ashes from the three coals, as shown in Fig. 5. efficiency is sensitive to the chloride content in fuels.
 W. Xie et al. / Fuel 78 (1999) 1425–1436 1433

Table 5
The percent distribution of chloride in the different phases

Coal 95010 1 0% PVC 1% PVC 3.3% PVC

Before Burning
Feed rate (kg/h) 7.95 7.95 7.95
Chloride content in fuel (wt%) 0.104 0.671 1.97
Feed amount of chloride (kg/h) 8.27 × 10 23 0.0533 0.157
After Burning
Solid phase
Bed ash
Content (wt%) 0.045 0.106 0.211
Flow rate (kg/h) 0.47 0.47 0.47
Amount (kg/h) 2.12 × 10 24 4.98 × 10 24 9.92 × 10 24
Bed material
Content change (wt%) 0.045 0.0517 0.155
Material inventory (kg) 50 60 60
Amount (kg/h) 2.25 × 10 23 0.0310 0.0929
Fly ash
Content (wt%) 0.021 0.100 0.342
Flow rate (kg/h) 0.93 0.93 0.93
Amount (kg/h) 1.95 × 10 24 9.33 × 10 24 3.18 × 10 23
Subtotal 2.66 × 10 23 0.0324 0.0971
Wt.% in total chloride: 32.16 60.79 61.85
Gas phase
Concentration in flue gas (ppm) 50.83 151.14 249.86
Temperature of flue gas (8C) 150 150 150
Flow rate of flue gas (kg/h) 89.9 89.9 89.9
Flue gas volume (m 3/h) 113.12 113.12 113.12
Chloride volume in gas (m 3/h) 6.66 × 10 23 0.0171 0.0260
Amount (kg/h) 6.21 × 10 23 0.0185 0.0473
Wt % in total chloride 75.09 34.71 30.13
Total 107.25 95.5 91.98
Coal 95031 1 0% PVC 0.5% PVC 1% PVC
Before burning
Feed rate (kg/h) 8.72 8.72 8.72
Chloride content in fuel (wt%) 0.307 0.591 0.874
Feed amount of chloride (kg/h) 0.0268 0.0515 0.0762
After burning
Solid phase
Bed ash
Content (wt%) 0.107 0.120 0.129
Flow rate (kg/h) 1.45 1.45 1.45
Amount (kg/h) 1.55 × 10 23 1.74 × 10 23 1.87 × 10 23
Bed material
Content change (wt%) 0.0134 0.0461 0.0761
Material inventory (kg) 60 60 68.5
Amount (kg/h) 8.04 × 10 23 0.0277 0.0521
Fly ash
Content (wt%) 0.0773 0.122 0.159
Flow rate (kg/h) 2.59 2.59 2.59
Amount (kg/h) 2.00 × 10 23 3.16 × 10 23 4.12 × 10 23
Subtotal 0.0116 0.0326 0.0581
Wt % in total chloride: 43.28 63.30 76.25
Gas phase
Concentration in flue gas (ppm) 124.49 131.14 115.52
Temperature of flue gas (8C) 150 150 150
Flow rate of flue gas (kg/h) 89.9 89.9 89.9
Flue gas volume (m 3/h) 113.12 113.12 113.12
Chloride volume in gas (m 3/h) 0.0141 0.0148 0.0131
Amount (kg/h) 0.0152 0.0160 0.0141
Wt % in total chloride 56.72 31.07 18.50
Total 100 94.37 94.75
1434 W. Xie et al. / Fuel 78 (1999) 1425–1436

Fig. 7. The relationship between the sulfur and chloride contents in ash (combination of fly ash and bed ash).

Fig. 8. The effect of the PVC/fuel ratio on a Ca utilization.

 W. Xie et al. / Fuel 78 (1999) 1425–1436 1435

3.3. Calcium utilization Compared to

To reduce SOx emission, the Ca/S mole ratio is normally CaO 1 SO2 1 1=2O2 ! CaSO4 …7†
chosen to be 2.5–3, both in lab studies and in conventional The use of limestone as the bed material and continuous
power or industrial FBC systems. That means that the feeding of limestone with the fuel in fluidized bed combus-
conversion rate from porous calcium oxide to calcium tion will keep excess CaO in the combustor. Subsequently,
sulfate (CaSO4) is only about 30 t 40%. However, in the the concentration of molecular chlorine in the flue gas is
presence of HCl in the combustor, the utilization of calcium minimized on the basis of reaction (6) at temperatures
is improved to higher levels, as much as 100% more than below 10008C.
when only SOx exists in system. This improvement is illu- It has been well established that halogenated species are
strated in Fig. 8. In the reaction with chlorine a number of good flame inhibitors [29]. According to Bulewicz [30]
spherical aggregates and large voids are formed on the there is a phenomenon of halogen inhibition of oxidation
absorbing particles, and consequently more SOx and HCl of CO and other species in an FBC system. Thus it might be
can easily penetrate into the limestone particle, to react expected that the concentration of SO2 may increase due to
with new calcium oxide. This increases the absorption rate its incomplete combustion to SO3 in the presence of chlor-
of HCl and SOx sulfation, and increases the level of conver- ide. Reaction (7), however, can promote the oxidation of
sion of CaO to CaSO4 and CaCl2. SO2 –SO3 in different ways through the Deacon reaction in
Interaction between SO2 and HCl. A possible mechanism the presence of chloride and oxygen-rich conditions. These
for the interaction between SO2 and HCl can be proposed two effects probably compete with each other during coal
from the combination of the results from previous reports combustion.
[27] and those given in this paper. Thermodynamic data
shows the Deacon reaction
4. Conclusions
4HCl 1 O2 $ 2H2 O 1 2Cl2 …4†
From experimental investigations in a fluidized bed
is favored over the range of temperatures from 300 to combustor on the influence of chlorine on sulfur capture,
1500 K and is exothermic at 258C (H ˆ 2 114 KJ/mol, it was shown that the presence of HCl will promote SO2
DG ˆ 2 76 KJ/mol). An increase in temperature will capture by the bed material and fly ash particles. However,
cause the equilibrium to move toward the reactants, which the presence of sulfur leads to an apparent reduction in the
will lower the conversion of HCl to Cl2. Before the equili- formation of molecule chlorine. As a result, minimization of
brium is reached, however, the reaction is predominantly the formation of the PCDDs and PCDFs during the combus-
kinetically-controlled. A rise in temperature will lead to tion of high sulfur and high chlorine fuels occurs.
more products. The equilibrium constant for reaction (4)
at 1120 K is about 1097 [8]. Typically, the oxygen and
water concentrations in the flue gas in an FBC are around Acknowledgements
5%. Therefore, when 1000 ppm HCl is present in the flue
gas, the equilibrium value for Cl2 would be 1.48%, which is The authors wish to thank the United States Department
very much higher than the actual concentration of Cl2 of Energy for the financial support through grant number
present in the flue gas. Yang [28] reported that in the case DE-FG-94PC 94211.
where the reaction takes place in a steady moving flow and
no catalysts are present, the reaction is far from equilibrium. References
Consequently, a higher temperature will lead to a higher
reaction rate, meaning more Cl2 will be produced. Also it [1] Minchener AJ, Lloyd DM, Stringer J. In: Meadowcroft DB, Manning
should be noted that Le Chatelier’s Principle indicates that MI, editors. Corrosion resistant materials for coal conversion systems,
the addition of oxygen to the system to enhance the combus- New York: Applied Science Publishers, 1996. pp. 299.
tion process would tend to form more Cl2. [2] Hancock P, Hurst RC, Sollars AR. Influence of chloride contamina-
tion of the atmosphere on the mechanism of the oxidation of iron.
When SO2 is present from the combustion of sulfur in Metallurgical chemistry-applications in ferrous metallurgy, London:
coal, a most interesting and important reaction is that SO2 Iron and Steel Institute, 1972:415–418.
may be attacked by Cl2 to form SO3 and HCl: [3] Hurst RC, Johnson JB, Davies M, Hancock P. Sulphate and chloride
attack of nickel-based alloys and mild steels. Deposition and corro-
Cl2 1 SO2 1 H2 O $ 2HCl 1 SO3 …5† sion in gas turbines, New York: Wiley, 1973:143–157.
[4] Wright IG, Sethi VK, Krause HH, Stringer J. Effect of chlorine on
materials performance in fireside circuit pulverized coal-fired boilers.
During fluidized combustion, SO3 will be absorbed more
Heat-resistant materials. Proceedings of the First International
easily by limestone than SO2, according to the reactions Conference, Fontana, WI, 1991:415–427.
represented by the following equations: [5] Stringer J, Wright IG. Erosion/corrosion in FBC boilers. EPRI Work-
shop on Wastage of In-Bed Surface in Fluidized Bed Combustor,
CaO 1 SO3 ! CaSO4 …6† Argonne National Laboratory, IL, November 1987.
1436 W. Xie et al. / Fuel 78 (1999) 1425–1436

[6] Wright IG, Stringer J. Fly ash erosion of boiler convection banks. [20] Ohlsson OO, Shepherd P. Results of performance and emission test-
Presented at EPRI/Florida Power and Light Seminar on Use of Coal ing when co-firing blends of RDF/Coal in a 440 MW cyclone fired
in Oil-design Utility Boilers, FL, 2–4 December 1980:1–18, section combustor in combustion modeling. Co-firing and NOx Control,
5. FACT-American Society of Mechanical Engineering, 1993;17:173.
[7] Johnson I, Lence JF, Shearer JA, Smith GW, Swift WM, Treats FG, [21] Frankenhaeuser M, Manninen H, Virkki J, Kojo I. The effect of the
Turner CB, Jonke AA. Support studies in fluidized-bed combustion. chlorine/sulfur ratio on organic emissions from the combustion of
Argonne National Laboratory Quarterly Report ANL/CEN/FE-79-8, mixed fuels. NESTE Final report, Porvoo, Finland, 1992.
1979. [22] Orndorff WW, Su S, Napier J, Bowles J, Li H, Li D, Smith J, Pan
[8] Griffin RD. A new theory of dioxin formation in municipal solid W-P, Riley JT. Proceedings, 12th International Coal Testing Confer-
waste combustion. Chemosphere 1986;15:9–12. ence, Cincinnati, OH, 1996:67–75.
[9] Liang DT, Anthony EJ, Leowen BK, Yates DJ. Proceedings of the [23] Xie W, Su S, Li H, Pan WP, Riley JT. Prepr Pap-Amer Chem Soc, Div
11th International Conference on FBC, Vol. 2, Montreal, Canada, 21– Fuel Chem 1997;42(1):345.
24 April 1991:917. [24] Xie W, Pan W-L, Shen D, Pan W-P, Riley JT. Prepr Pap-Amer Chem
[10] Munzner H, Schilling DH. Proceedings of the 8th International Soc, Div Fuel Chem 1997;42(1):364.
Conference on FBC, Volume III, Houston, TX, 18–21 March [25] Pan WP, Riley JT. Behavior of chlorine during coal combustion in an
1985;1219. AFBC System. Final Report to the Electric Power Research Institute,
[11] Fink CE. Presented at the Joint Symposium on Dry SO2 and Simulta- W09002-13, 1997:78.
neous SO2/NOx Control Technology, San Diego, CA, 1984. [26] Matsukata M, Tacked K, Miyatani T, Ueyama K. Chem Engng Sci
[12] Liberti A, Goretti G, Russo MV. Chemosphere 1992;12:661. 1996;11:2529.
[13] Olie K, Vermeulen PL, Hutzinger O. Chemosphere 1997;6:455. [27] Atkins PW. Physical chemistry, 4th edn.. New York: Freeman, 1990.
[14] Agrawal RK. Compositional analysis of solid waste and refuse [28] Yang XD. A chemical pathway to the formation of chlorinated
derived fuels by thermogravimetry. In: Earnest C M, editor. Compo- compounds during combustion. MS Thesis, Western Kentucky
sitional analysis by thermogravimetry, ASTM, STP 997Philadelphia, University, 1996.
PA: American Society for Testing and Materials, 1988. pp. 259. [29] Dixon-Lewis G, Simpson RJ. Sixteenth Symposium (International)
[15] Kreith F. Solid Waste Management: 1989–1990, 47-91-NCSL-1, on Combustion. The Combustion Institute, Pittsburgh, PA,
1991. 1976:1111–1120.
[16] Scheidle K, Wurst F, Kuna RP. Chemosphere 1986;17:2089. [30] Bulewicz EM, Janicka E, Kanderfer S. The Tenth International
[17] Lindbauer RL. Chemosphere 1992;25(7–10):1409. Conference on Fluidized Bed Combustion, San Francisco, CA,
[18] Banaee J, Larson RA. Waste Mgmt 1993;13:77. 1989:163.
[19] Bonfanti L, Cioni M, Riccardi J, Rossi C. Fresenius J Analyt Chem [31] Bramer EA. Flue gas emission from FBC. Coal Sci Technol
1994;348:136. 1995;22:51.