Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) 2162-2906 (Online) Journal homepage: https://www.tandfonline.com/loi/uawm20

An Alternative to Additional SO3 Injection for Fly

Ash Conditioning

David J. Bayless , Ashikur R. Khan , Srinivas Tanneer & Rajkumar Birru

To cite this article: David J. Bayless , Ashikur R. Khan , Srinivas Tanneer & Rajkumar Birru (2000)
An Alternative to Additional SO3 Injection for Fly Ash Conditioning, Journal of the Air & Waste
Management Association, 50:2, 169-174, DOI: 10.1080/10473289.2000.10464013

To link to this article: https://doi.org/10.1080/10473289.2000.10464013

Published online: 27 Dec 2011.

Submit your article to this journal

Article views: 406

View related articles

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uawm20
TECHNICAL PAPER
An Alternative to Additional SO3 Injection for
Fly Ash Conditioning
David J. Bayless, Ashikur R. Khan, Srinivas Tanneer, and Rajkumar Birru
Department of Mechanical Engineering, Ohio University, Athens, Ohio
ABSTRACT
Small concentrations, approximately 2–10 parts per mil-
lion (ppm), of injected sulfur trioxide (SO3) have improved
particulate collection efficiencies of electrostatic precipita-
tors burning lower-sulfur coal. However, the addition of
extra SO3 not only incurs costs but also presents negative
environmental effects. This work explored a method that
could be applied to existing coal-fired power plants to con-
vert the sulfur dioxide (SO2) already present in the flue gas
to sufficient levels of SO3 for fly ash conditioning as an
alternative to adding SO3 by burning elemental sulfur. Dur-
ing this research, a pre-mixed natural gas flame was used
to promote the conversion of SO2 to SO3 in a drop-tube
furnace with average non-flame, free stream gas tempera-
tures of 450 and 1000 K. SO3 concentrations measured by
wet chemistry and confirmed using elemental balances
of other sulfur species measured by gas chromatography
revealed that as much as 7% of SO2 was homogeneously
transformed to SO3. The results also showed that at low
temperatures, the rate at which SO3 is converted back to
SO2 decreased, thus extending the time period during
IMPLICATIONS
Pulverized-coal-fired power plants that switch to low-
sulfur coal often experience higher ash loading and lower
fly ash resistivity, resulting in higher stack opacity. Many
of these power plants add small concentrations of SO3
from the oxidation of elemental sulfur to lower resistivity
and improve ash collection in the electrostatic precipita-
tor. While this technique may avoid some expense of ret-
rofitting additional precipitator collection areas, the extra
SO3 can lead to the formation of a “blue plume” and re-
lease acid rain precursors. This paper shows that SO2
existing in the flue gas could be used to produce suffi-
cient SO3 through a combustion-based process immedi-
ately upstream of the electrostatic precipitator. This tech-
nique also humidifies the flue gas, which may further en-
hance precipitator performance.
Volume 50 February 2000
Bayless,J. Air
ISSN 1047-3289 Khan, Tanneer,
& Waste and50:169-174
Manage. Assoc. Birru
Copyright 2000 Air & Waste Management Association
which SO3 concentrations would be sufficient for ash con-
ditioning. An additional benefit of this technique is specu-
lated to result from increased flue gas humidity.
INTRODUCTION
The combustion of high-sulfur coal produces correspond-
ing high concentrations of sulfur trioxide (SO3), typically
1–2% of the sulfur dioxide (SO2) concentration. SO3 and
moisture combine to form sulfuric acid, which deposits
on the fly ash surface, creating a conductive film that
lowers the electrical resistance of the ash. Consequently,
the ash surrenders its charge to the grounded collection
plates more easily, making the ash easier to collect than
high-resistivity ash. However, many plants that once
burned high-sulfur coal have switched to low-sulfur coal
to comply with the 1990 Clean Air Act Amendments. Not
only is the ash more resistive because of the lower con-
centrations of SO3, but low-sulfur coals typically have a
higher ash content. Consequently, electrostatic precipita-
tors in these plants experience serious ash collection prob-
lems and often require plant curtailments (or deratings) to
meet local opacity regulations.1,2
One solution is to install more expensive precipita-
tors with larger collection areas. This is usually dismissed
under all but extreme circumstances. Instead, most plants
install flue gas conditioning equipment that injects SO3
into the exhaust ducts before the precipitators. The SO3
is usually produced from combustion of elemental sul-
fur to form SO2, which is further oxidized to SO3 using
catalysts. Small amounts of the generated SO3 are then
injected into the flue gas upstream of the precipitator.
Unfortunately, the injection of SO3 into the flue gas pre-
sents its own set of problems. Often, flue gas condition-
ing produces a bluish plume of sulfate particles that form
outside the stack. In addition, the process inherently adds
acid rain precursors by introducing additional sulfur to the
flue gas. Further, the injection of SO3 alone is rarely suffi-
cient to remediate all particulate collection problems,
Journal of the Air & Waste Management Association 169
Bayless, Khan, Tanneer, and Birru
often necessitating the addition of ammonia or steam to
agglomerate the ash particles.3
An alternative exists that transforms existing SO2 into
SO3 without the use of high-maintenance catalysts. When
SO2 passes through a flame in the presence of O2, a tran-
sient superequilibrium of SO3 forms. This process has been
studied at boiler temperatures and reviewed by Cullis and
Mulcahy.4-10 The motivating factor for many of these ef-
forts was the desire to minimize corrosion in coal-fired
boilers and economizer sections which resulted when the
SO3 reacted with water to form sulfuric acid. The process,
while potentially detrimental to boiler tubes, may be ben-
eficial if applied immediately upstream of the electrostatic
precipitator.
The gas-phase conversion of SO2 to SO3 occurs when
excess O2 forms O radicals through numerous reactions
in a flame, especially the oxidation of CO.10 The oxygen
radicals combine with SO2 to form SO3 through interac-
tion with third bodies (M), given by
SO2 + O + M ↔ SO3 + M (1)
After the SO3 has formed, it decomposes through re-
actions with oxygen and (to a much lesser extent) with
hydrogen radicals, as well as by thermal decomposition,
shown as
SO3 + O ↔ SO2 + O2 (2)
SO3 + H ↔ SO2 + OH (3)
SO3 + M ↔ SO2 + O + M (4)
The duration of the increased SO3 concentration depends
on the concentration of radicals and the temperature of
the gas following the flame zone.4 The lifetime of this
“superequilibrium” is typically very short (less than 200
msec) at high temperatures because the radicals that form
SO3 quickly destroy it.
Applying this concept to low temperatures, such as
those found upstream of an electrostatic precipitator, pre-
sents an interesting possible combination of effects which
can be inferred from a gas co-firing study by Bayless et
al.11 Flames produced in gas flows with mean tempera-
tures as low as 450 K (350 °F) could generate sufficient
oxygen radicals to convert SO2 to SO3. In addition, the
lower temperature could reduce the level of oxygen radi-
cals downstream of the flame, thus retarding the destruc-
tion of SO3. Further, the lower temperature of the gas after
the flame could reduce the rate at which the SO3 ther-
mally decomposes. Combining these effects could
lengthen residence time for the converted SO3, thus in-
creasing the opportunity for fly ash conditioning.
In addition, the flame generated by the combustion
of a hydrocarbon such as natural gas produces relatively
170 Journal of the Air & Waste Management Association
large quantities of water. For example, the combustion of
one molecule of methane produces two molecules of wa-
ter. The addition of water to the exhaust gas of pulverized
coal-fired units can stimulate the agglomeration of fly ash
to form larger particles. Larger particles are more easily
collected in electrostatic precipitators, thus increasing
collection.2,3,12
This work examines the conversion of SO2 to SO3 due
to interaction with a perpendicularly introduced natural
gas flame at average free stream gas temperatures of 1000
and 450 K as a function of O2 and SO2 concentrations.
Results indicate that low temperature conversion occurs
for an extended duration and that the process can be prac-
tically applied.
METHODOLOGY
Experimental Setup
Experiments were performed in a two-stage electrical-
resistance furnace known as the Sulfur Reactivity Experi-
mental Facility, shown in Figure 1. Gas, including SO2,
was preheated in the first stage to limit quenching ef-
fects in the second stage, called the secondary furnace,
where the reactions occurred. A 1.25-in.-diameter reac-
tion tube, made of stainless steel to minimize SO 3
catalyzation,5 was used to contain the reactions in the
secondary furnace.
In the secondary furnace, a premixed natural gas
flame, controlled by varying natural gas and oxygen con-
centrations, was directed perpendicularly to the main
flow. Sampling of gas from the reaction tube was done
with a 0.25-in.-diameter stainless steel tube that could
be positioned at any axial location inside the reaction
tube. The gas stream was transported through a heated
Figure 1. Schematic of the Sulfur Reactivity Experimental Facility.
Volume 50 February 2000

sample line with the aid of a vacuum pump. The gas
sample was then split between a gas chromatograph and
a wet chemistry setup to quantify sulfur species.
Gas chromatography was performed by passing the
sampled gas through a capillary column at a nominal tem-
perature of 573 K before analysis using a flame photo-
metric detector. The resulting chromatogram was analyzed
using HP ChemStation software and compared with chro-
matograms of known samples of sulfur species, including
hydrogen sulfide (H2S) and SO2.
SO3 detection was done using wet chemistry and the
apparatus shown in Figure 2. Basically, a drawn sample
was collected in a hydrogen peroxide solution and tested
for SO3 content using barium perchlorate trihydrate
(Ba(ClO4)2·3H2O) as specified by U.S. Environmental Pro-
tection Agency (EPA) Method 8.13,14 The sample was ti-
trated and analyzed via spectrum for pH, also as specified
by EPA Method 8, which was then used to determine the
concentration of SO3.13,14
Experiments were performed by establishing a flow
of N2, O2, and SO2 through the primary furnace. In the
secondary furnace, the gas flow then perpendicularly
passed through a pre-mixed methane air flame. Experi-
ments were performed to examine the effects of gas tem-
perature, O2 concentration, CH4 concentration, and
residence time on gas phase SO2-to-SO3 conversion. Ex-
cess O2 levels (the level of O2 in the combustion prod-
ucts compared with stoichiometric levels) ranged from 0
to 10%. Natural gas concentration was either 0 or 4%.
SO2 concentration varied from 500 to 3500 parts per
million (ppm) to reflect low-, medium-, and high-sulfur
coals. Average non-flame free stream temperatures of
1000 and 450 K and an average gas velocity of 1.016 m/
sec were used.
Figure 2. EPA Method 6/8 setup.
Volume 50 February 2000
Bayless, Khan, Tanneer, and Birru
Modeling
A model of the reactions was developed to help analyze the
experimental results and to calculate concentrations of radi-
cals needed to quantify the reaction’s kinetics. The model
contained several facets, including gas-phase heating and
combustion, as well as the evolution of gaseous sulfur spe-
cies. Gas temperatures in the furnace were determined from
measured furnace wall temperatures and inlet conditions
by combining heat transfer from the wall with gas phase
combustion. Variations in gas properties with temperature
and species were calculated using Wilke’s mixing model.15
CHEMKIN was used to solve the chemical kinetics for a
reaction mechanism taken from GRI-Mech for natural gas
combustion.16,17 Experimentally measured inputs to the
model included initial gas temperature, gas composition,
gas velocity, and axial wall temperatures. The outputs were
species concentrations and temperatures.
RESULTS
First, experiments were performed to examine conversion
at low free stream temperatures. The reacting stream in
the secondary furnace was sampled at five locations: at
the position of the flame-holder, and at 100, 200, 300,
and 400 msec downstream, with and without a natural
gas flame. Excess oxygen was set at 10% and the initial
SO2 concentration was set at 2000 ppm. SO2 and SO3 lev-
els in the samples were quantified using a gas chromato-
graph with a flame photometric detector and EPA Method
8, as discussed.
The results are shown in Figures 3 and 4, where SO3
concentrations were plotted versus residence time after
the flame. The data indicated that the presence of a natu-
ral gas flame promoted the homogeneous conversion of
SO2 to SO3, even at 450 K. Modeling results indicated that
the majority of oxygen radicals formed from reactions in
the flame; the free stream temperature was too low to
Time after Flame (milliseconds)
Figure 3. Measured SO3 concentrations at 1000 K with 10% excess O2.
Journal of the Air & Waste Management Association 171
Bayless, Khan, Tanneer, and Birru

Time after Flame (milliseconds)

Time after Flame (milliseconds)

Figure 4. Measured SO3 concentrations at 450 K with 10% excess O2. Figure 5. Measured SO3 concentrations at 1000 K with 2% excess O2.

significantly dissociate O2. The results also indicated that

the super-equilibrium state of the SO3 occurs for an ex-
tended period of time, compared with the results of
Hedley, because of the lower gas stream temperatures.4
The data also indicated that the peak concentration of
SO3 occurs less than 100 msec downstream from the flame,
which is consistent with the results of Hedley.4
The percentage of SO2-to-SO3 conversion, in excess
of the base case (without natural gas), was 3.5% for a free
stream temperature of 1000 K, and decreased to 3.0% at
450 K. The conversion percentage decreased with tem-
Time after Flame (milliseconds)
perature because the lower gas temperature cooled the
flame, which slowed the reactions producing oxygen radi-
cals. Further, the rate at which SO3 decomposed to SO2
apparently decreased at lower temperatures because (1) Figure 6. Measured SO3 concentrations at 450 K with 2% excess O2.
there were fewer oxygen radicals to attack the SO3 out-
side the flame, and (2) modeling results indicated that
the lower temperature reduced thermal decomposition of
SO3 by third body attack (eq 1), further decreasing the
rate at which SO3 was reduced back to SO2. In fact, at 450 K,
SO3 levels remain significantly higher (greater than 10
ppm) than those found without a flame for up to 400
msec. This is crucial to potential practical application.
The next experiments were conducted to examine the
effects of excess oxygen. Using similar conditions to the
data shown in Figures 3 and 4, the excess oxygen in the
secondary furnace gas stream was lowered to 2 and 4%.
The results are shown in Figures 5–8.
The data showed that, as with 10% excess oxygen, the Time after Flame (milliseconds)
natural gas flame promotes the homogeneous conversion
of SO2 to SO3. The results also indicated that the rate at
which the concentration of SO3 decreased was slower at Figure 7. Measured SO3 concentrations at 1000 K with 4% excess O2.
lower temperatures, as was previously discussed. In addi-
tion, these results indicated that excess oxygen levels af- at 450 K. With 4% excess O2, the conversion levels were
fected the conversion percentage at the peak SO 3 3.25% at 1000 K and 2.75% at 450 K.
concentrations. With 2% excess oxygen, the maximum The lower conversion percentage at very low excess
conversion peaked at 2.3% at 1000 K and decreased to 2.0% oxygen levels can possibly be explained by the lack of

172 Journal of the Air & Waste Management Association Volume 50 February 2000

Time after Flame (milliseconds)
Figure 8. Measured SO3 concentrations at 450 K with 4% excess O2.
generated oxygen radicals from the flame, because of less
overall oxygen available to the flame. Modeling results
indicated that in a system with imperfect mixing, such as
a drop-tube furnace or a ducted-exhaust system, a reduc-
tion in excess oxygen would slow some of the CO-to-CO2
oxidation, thus producing fewer oxygen radicals.
Excess oxygen levels did not appear to affect the rate
at which SO3 decomposed to SO2. Results indicated that
the decomposition rate depended on downstream (of the
flame) oxygen radicals and temperature, both of which were
virtually unaffected by O2 content in the gas stream at low
temperatures.
The effects of initial SO2 levels on the conversion process
are displayed in Figures 9 and 10. Previous work at higher tem-
peratures indicated that the conversion percentage increased
at lower initial SO2 levels.4,7 Two tests were performed, one us-
ing 500 ppm initial SO2 concentration, and the other using
3500 ppm. These levels were chosen to reflect the SO2 levels for
low- and high-sulfur coals to examine the effect of coal sulfur
content. In both cases, the average free stream non-flame tem-
perature was held at 450 K with 10% excess oxygen.
Time after Flame (milliseconds)
Figure 9. Measured SO3 concentrations at 450 K using 500 ppm
SO2 with 10% excess O2.
Volume 50 February 2000
Bayless, Khan, Tanneer, and Birru
Time after Flame (milliseconds)
Figure 10. Measured SO3 concentrations at 450 K using 3500 ppm
SO2 with 10% excess O2.
The peak conversion percentage of SO3 increased for
the case of 500-ppm inlet SO2 concentration (Figure 9),
compared with both the 2000- and the 3500-ppm (Figure
10) cases. The maximum SO2-to-SO3 conversion percent-
age using 500 ppm was 7%, significantly higher than the
3–4% at 2000- and 3500-ppm inlet SO2 concentrations.
These results were similar to the conversion trends previ-
ously studied at higher gas temperatures and indicate that
sufficient levels of SO3 can be generated even when burn-
ing low-sulfur coal.7,8
SUMMARY AND CONCLUSIONS
This work investigated a potential method for SO3 gen-
eration to replace external SO3 injection to condition fly
ash prior to entering an electrostatic precipitator. The re-
sults indicated that application of natural gas combus-
tion upstream of the electrostatic precipitator converted
existing SO2 to SO3, thus eliminating the need to inject
additional sulfur into the gas stream. In addition, the com-
bustion of the hydrocarbons may add sufficient moisture
to the flue gas to increase ash agglomeration.
Results indicated that the presence of an energetic
hydrocarbon flame generated enough radicals to convert
sufficient SO2 to SO3 for conditioning fly ash, even at low
temperatures (450 K). Further, SO3 levels remained elevated
longer in low free stream temperatures than in high tem-
peratures, strengthening the potential for practical appli-
cations. Two mechanisms may be responsible. First, lower
temperatures may reduce thermal decomposition and re-
duce free steam oxygen radicals downstream of the flame,
thus slowing the reactions that revert SO3 to SO2.
Second, initial SO2 concentration and free stream
excess oxygen level also influenced conversion. As SO2
concentration decreased, the percentage of conversion to
SO3 increased. As the percentage of excess O2 increased,
the conversion percentage of SO2 to SO3 also increased,
but it did not significantly change above 4% excess O2. At
Journal of the Air & Waste Management Association 173
Bayless, Khan, Tanneer, and Birru
levels of both parameters typically found in the exhaust
of coal-fired power plants, a natural gas flame could pro-
mote an initial 2–7% conversion of SO2 to SO3, if done
just prior to the precipitator. Further, the results indicate
that this technique converted the greatest percentage of
SO2 to SO3 for systems using low-sulfur coals that cause
ash resistivity problems.
The overall results indicated that this technique of-
fers the possibility of application in practical systems that
require lower fly ash resistivity without adding external
sulfur. The conversion percentage found in this work was
adequate for ash conditioning, even at residence times
up to 400 msec from the flame, with a greater percentage
of SO3 available for delivery at lower temperatures. Fur-
ther, fly ash found in practical applications might further
enhance the SO3 concentration because of catalytic con-
version by metals in the ash.
ACKNOWLEDGMENTS
This work was funded by the Ohio Coal Development
Office and the Link Foundation. The authors would like
to thank Professor Michael Prudich and engineers at
American Electric Power for their guidance.
REFERENCES
1. Bayless, D.J.; Brumfield, A. American Electric Power’s Project Probe—
Power Engineering Education through Internship. In Proceedings of
the 60th American Power Conference, Illinois Institute of Technology,
Chicago, IL, April, 1998; pp 146-150.
2. Wright, R.; Woracek, D.; Kapuscinski, J. Advances in SO3 Gas Plant
Design and Control. In Proceedings of the American Power Conference,
Illinois Institute of Technology, Chicago, IL, April 1992; pp 1187-
1192.
3. Ashworth, R.; Beshai, R.; Hill, H.; Chong, C.; Opatrny, J.; Keen, R.;
Sanyal, A.; Sommer, T.; May, T.; Krueger, M. Cost-Effective ESP Per-
formance Enhancement by Flue Gas Humidification. Presented at
International Joint Power Generation Conference, American Society
of Mechanical Engineers, 1993; pp 1-11.
4. Hedley, A.B. Fuel 1967, 40, 142-147.
5. Barrett, R.E.; Hummell, J.D.; Reid, W.T. ASME J. Eng. Power 1966, 88A,
165-172.
6. Levy, A.; Merryman, E.L. Comb. Flame 1965, 9, 229-240.
174 Journal of the Air & Waste Management Association
7. Nettleton, M.A.; Stirling, R. Formation and Decomposition of Sulfur
Trioxide in Flames and Burned Gases. Presented at Twelfth Interna-
tional Symposium on Combustion, The Combustion Institute, Pitts-
burgh, PA, 1969; pp 635-641.
8. Fenimore, C.P.; Jones, G.W. J. Chem. Phys. 1965, 69, 3593-3597.
9. Merryman, E.L.; Levy, A. Sulfur Trioxide Flame Chemistry—H2S and
COS Flames. Presented at Thirteenth International Symposium on
Combustion, The Combustion Institute, Pittsburgh, PA, 1971; pp 427-
436.
10. Cullis, C.F.; Mulcahy, M.F.R. Comb. Flame 1972, 18, 224-288.
11. Bayless, D.J.; Schroeder, A.R.; Olsen, M.G.; Johnson, D.J.; Krier, H.;
Peters, J.E.; Buckius, R.O. Comb. Flame 1996, 106, 231-240.
12. White, H. Industrial Electrostatic Precipitation; Addison-Wesley: Read-
ing, PA, 1963.
13. Emissions Measurement Technical Information Center, New Source
Performance Standards, Test Method 6. Determination of Sulfur Di-
oxide Emissions from Stationary Sources; http://134.67.104.12/html/
emtic/cfrprom.htm, 1998.
14. Emissions Measurement Technical Information Center, New Source
Performance Standards, Test Method 8. Determination of Sulfuric Acid
Mist and Sulfur Dioxide Emissions from Stationary Sources; http://
134.67.104.12/html/emtic/cfrprom.htm, 1998.
15. Wilke, C.E. J. Chem. Phys. 1950, 18, 517-520.
16. GRI-Mech 2.1; http://euler.Berkeley.edu/gri_mech, 1998.
17. Kee, R.J.; Miller, J.A.; Jefferson, T.H. CHEMKIN: A General Purpose Prob-
lem Independent, FORTRAN Chemical Kinetics Code; Sandia Labs, 1980;
Sandia Report +SAND80-8003.
About the Authors
David Bayless (corresponding author) is an assistant pro-
fessor in the Department of Mechanical Engineering
(webme.ent.ohiou.edu) and a researcher in the Ohio Coal
Research Center (www.ent.ohiou.edu/~ohiocoal) at Ohio
University. His address is 248 Stocker Center, Ohio Univer-
sity, Athens, Ohio 45701-2979. He can be reached by phone
at (740) 593-0264, by fax at (740) 593-0476, and by e-mail
at bayless@ohio.edu. Ashikur Khan, Srinivas Tanneer, and
Rajkumar Birru are graduate students in the department.
Volume 50 February 2000