EPA-600/2-75-015
TVA-F75 PRS-5
August 1975
Environmental Protection Technology Series
CONDITIONING OF FLY ASH
WITH SULFUR TRIOXIDE
AND AMMONIA
U.S. Environmental Protection Agency
Office of Research and Development
Washington, 0. C. 20460
-------
EPA-600/2-75-015
TVA-F75 PRS-5
CONDITIONING OF FLY ASH
WITH SULFUR TRIOXIDE
AND AMMONIA
by
Edward B. Dismukes
Southern Research Institute
2000 Ninth Avenue South
Birmingham, Alabama 35205
TVA Research Agreement TV36921A
EPA Contract No. 68-02-1303
ROAP No. 21ADJ-029
Program Element No. 1AB012
TVA Project Officer: James R. Crooks
Power Research Staff
Tennessee Valley Authority
Chattanooga, Tennessee 37401
EPA Project Officer: Leslie E. Sparks
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, North Carolina 27711
Prepared jointly for
POWER RESEARCH STAFF
Tennessee Valley Authority
Chattanooga, Tennessee 37401
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D. C. 20460
August 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:
1. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2. ENVIRONMENTAL PROTECTION TECHNOLOGY
3. ECOLOGICAL RESEARCH
4. ENVIRONMENTAL MONITORING
5. SOCIOECONOMIC ENVIRONMENTAL STUDIES
6. SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to
develop and demonstrate instrumentation, equipment and methodology
to repair or prevent environmental degradation from point and non-
point sources of pollution. This work provides the new or improved
technology required for the control and treatment of pollution sources
to meet environmental quality standards.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No. EPA-600/2-75-015
-------
Abstract
This report summarizes research on the conditioning of fly
ash in coal-burning electric power stations with two flue-gas
additives—sulfur trioxide and ammonia. It presents experi-
mental information regarding the use of these additives to
improve the efficiency of electrostatic precipitation of fly
ash by adjustment in the electrical resistivity of the ash
and by other mechanisms less widely recognized than resis-
tivity modification. The report shows that the primary role
of sulfur trioxide is lowering resistivity from the excessive
values found with ash from low-sulfur coals. It indicates
that the role of ammonia, to the contrary, does not involve
a change in resistivity, despite findings to the contrary
by other investigators. At least for the specific circum-
stances investigated, the research data indicate that condi-
tioning by ammonia involves a space-charge enhancement of the
electric field in the interelectrode space of a precipitator
and, sometimes additionally, an increase in the cohesiveness
of the collected ash. The report not only addresses the
theoretical aspects of conditioning mechanisms but deals with
such practical matters as the effectiveness of each agent as
a function of the concentration added, the facilities used
for adding the agent, the chemical composition of the ash
treated, and the temperature of the ash during conditioning
and precipitation.
This report was prepared in partial fulfillment of research
under Contract 68-02-1303 with the Environmental Protection
Agency. It covers research funded by this agency during
1970-1971 under Contract CPA 70-149 and by the Tennessee
Valley Authority during 1972-1974 under Research Agreement
TV36921A.
ill
-------
CONTENTS
Page
Abstract iii
List of Figures vi
List of Tables viii
Acknowledgments xii
Sections
I Conclusions 1
II Recommendations 4
III Introduction 7
IV General Description of Research on Conditioning by 15
Sulfur Trioxide and Ammonia
V Results of Studies of Conditioning with Sulfur 20
Trioxide
VI Results of Studies of Conditioning with Ammonia 70
VII Discussion of Conditioning with Sulfur Trioxide 116
VIII Discussion of Conditioning with Ammonia 135
IX References 143
X Appendix. Experimental Methods 148
-------
FIGURES
No.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Electrical Resistivity of Fly Ash at Kingston
Unit 5
Concentrations of Sulfur Trioxide at Kingston
Unit 5
Chemical Properties of Fly Ash at Kingston Unit 5
Schematic Diagram of Electrical Sections in Collec-
tor IB at the Bull Run Plant
Current Density Versus Voltage in Collector IB of
the Bull Run Plant (August 1972)
Concentration of Sulfur Trioxide as a Function of
Gas Temperature at the Bull Run Plant (July 1974)
Current Density Versus Voltage in Collector IB of
the Bull Run Plant (July 1974)
Schematic Diagram of Electrical Sections in Collec-
tor 7A at the Widows Creek Plant
Current Density Versus Voltage in Collector 7A of
the Widows Creek Plant (Low-Sulfur Coal, June 1972)
Current Density Versus Voltage in Collector 7B of
the Widows Creek Plant (Low-Sulfur Coal, July 1972)
Current Density Versus Voltage in Collector 1C of
the Bull Run Plant (Ammonia Conditioning, September
1972)
Current Density Versus Voltage in Collector 1C of
the Bull Run Plant (Ammonia Conditioning, October
1972)
Current Density Versus Voltage in Collector 7A of
— f—
22
23
25
51
53
65
68
77
79
81
88
89
98
the Widows Creek Plant (High-Sulfur Coal, November
1972)
VI
-------
No. Page
14. Rapidity of the Effect of Ammonia Injection on the 108
Voltage Supplied to the Inlet Electrical Field of
Gallatin Precipitator 4C
15. Current Density Versus Voltage in Precipitator 4C 109
of the Gallatin Plant
16. Effects of Changes in Ammonia Injection on the 111
Emission of Particulate from Gallatin Precipitator
4C as Indicated by the Bailey Bolometer (20 ppm of
NH3)
17. Effects of Changes in Ammonia Injection and Elec- 113
trode Rapping on the Emission of Particulate from
Gallatin Precipitator 4C as Indicated by the Bailey
Bolometer (20 ppm of NH3)
18. Resistivity as a Function of the Concentration of 120
Injected Sulfur Trioxide
19. Resistivity as a Function of the Sulfate Concentra- 122
tion in Fly Ash
20. Dew Points of Vapor Mixtures of Sulfur Trioxide and 125
Water
21. Resistivity Apparatus Using a Mechanical Cyclone 149
Dust Collector (Cohen and Dickinson*6)
22. Cyclone Probe Inserted in Duct (Nichols27) 151
23. Point-to-Plane Resistivity Probe Equipped for 152
Thickness Measurement (Nichols27)
VI1
-------
TABLES
No. Page
1. Power Plants Investigated During the Research on 19
the Conditioning of Fly Ash
2. Composition of Fly Ash from the Kingston Plant 26
3. Electrical Resistivity of Fly Ash at Cherokee 28
Unit 2
4. Concentrations of Flue Gases at Cherokee Unit 2 29
5. Chemical Properties of Fly Ash at Cherokee Unit 2 29
6. Composition of Fly Ash from the Cherokee Plant 3^
7. Electrical Resistivity and Sulfate Content of Fly 32
Ash at Plant 3
8. Composition of Fly Ash from Plant 3 32
9. Electrical Resistivity of Fly Ash at Cherokee 35
Unit 3
10. Concentrations of Flue Gases at Cherokee Unit 3 35
11. Chemical Properties of Fly Ash at Cherokee Unit 3 36
12. Composition of Fly Ash from Arapahoe Unit 4 39
13. Electrical Resistivity of Fly Ash at Arapahoe 39
Unit 4
14. Concentrations of Flue Gases at Arapahoe Unit 4 39
15. Chemical Properties of Fly Ash at Arapahoe Unit 4 40
16. Composition of Fly Ash from Plant 6 41
17. Electrical Resistivity of Fly Ash at Plant 6 43
18. Concentrations of Flue Gases at Plant 6 44
19. Chemical Properties of Fly Ash at Plant 6 44
Vlll
-------
No. Page
20. Composition of Fly Ash from the Bull Run Plant 46
(August 1972)
21. Electrical Resistivity of Fly Ash at the Bull Run 47
Plant (August 1972)
22. Chemical Properties of Fly Ash at the Bull Run 48
Plant (August 1972)
23. Concentrations of Flue Gases at the Bull Run Plant 49
(August 1972)
24. Precipitator Electrical Data from the Bull Run 52
Plant (August 1972)
25. Composition of Fly Ash from the Bull Run Plant 56
(July 1974)
26. Precipitator Efficiencies at the Bull Run Plant 57
(July 1974)
27. Precipitation Efficiency as a Function of Gas 58
Temperature at the Bull Run Plant (July 1974)
28. Electrical Resistivity of Fly Ash at the Bull Rulr 59
Plant (July 1974)
29. Chemical Properties of Individual Samples of Fly 60
Ash Collected in Thimbles at the Bull Run Plant
(July 1974)
30. Chemical Properties of Composite Samples of Fly 61
Ash Collected in Thimbles at the Bull Run Plant
(July 1974)
31. Chemical Properties of Composite Samples of Fly 62
Ash from Precipitator Hoppers at the Bull Run Plant
(July 1974)
32. Concentrations of Sulfur Trioxide Equivalent to 63
the Concentrations of Sulfate in Fly Ash at the
Bull Run Plant (July 1974)
33. Precipitator Electrical Data from the Bull Run 67
Plant (July 1974)
34. Composition of Fly Ash from a Low-Sulfur Coal at 72
Widows Creek Unit 7 (June-July 1972)
IX
-------
No. Page
35. Electrical Resistivity of Fly Ash from a Low- 74
Sulfur Coal at Widows Creek Unit 7 (June-July 1972)
36. Chemical Properties of Fly Ash from a Low-Sulfur 74
Coal at Widows Creek Unit 7 (July 1972)
37. Concentrations of Flue Gases from a Low-Sulfur Coal 75
at Widows Creek Unit 7 (June-July 1972)
38. Peak Values of Precipitator Secondary Voltages and 80
Secondary Currents at Widows Creek Unit 7 (Low-
Sulfur Coal, July 1972)
39. Primary Voltages and Currents Supplied by the QQ
Transformer-Rectifier Sets at Widows Creek Unit 7
(Low-Sulfur Coal, July 1972)
40. Compositions of Fly Ash from the Bull Run Plant 33
(Ammonia Conditioning, September-October 1972)
41. Electrical Resistivity of Fly Ash at the Bull Run 94
Plant (Ammonia Conditioning, September 1972)
42. Chemical Properties of Fly Ash at the Bull Run 35
Plant (Ammonia Conditioning, October 1972)
43. Concentrations of Flue Gases at the Bull Run Plant 8e
(Ammonia Conditioning, September 1972)
44. Precipitator Electrical Data from the Bull Run 87
Plant (Ammonia Conditioning, September 1972)
45. Concentrations of Submicron Particles at the Bull 9^
Run Plant (Ammonia Conditioning, October 1972)
46. Composition of Fly Ash from a High-Sulfur Coal at 93
Widows Creek Unit 7 (November 1972)
47. Precipitator Efficiencies with High-Sulfur Coal at 94
Widows Creek Unit 7 (June-July 1970)
48. Electrical Resistivity of Fly Ash from a High-Sulfur 95
Coal at Widows Creek Unit 7 (November 1972)
49. Chemical Properties of Fly Ash from a High-Sulfur 95
Coal at Widows Creek Unit 7 (November 1972)
50. Concentrations of Flue Gases from a High-Sulfur Coal 96
at Widows Creek Unit 7 (November 1972)
-------
No. Page
51. Precipitator Electrical Data from Widows Creek 97
Unit 7 (High-Sulfur Coal, November 1972)
52. Concentrations of Submicron Particles at the Widows 99
Creek Plant (High-Sulfur Coal, November 1972)
53. Composition of Fly Ash from Gallatin Unit 4 101
54. Fly-Ash Emission from Gallatin Unit 4 (Collector 102
4C)
55. Emission of Fly Ash in Various Size Ranges at 103
Gallatin Unit 4 (Collector 4C)
56. Electrical Resistivity of Fly Ash at Gallatin 105
Unit 4
57. Chemical Properties of Fly Ash at Gallatin Unit 4 106
58. Concentrations of Flue Gases at Gallatin Unit 4 106
59. Precipitator Electrical Data from Gallagin Unit 4 107
60. Concentrations of Submicron Particles at Gallatin 107
Unit 4
61. Emission of Fly Ash in Various Size Ranges as a 114
Function of Electrode Rapping and Ammonia Injection
at Gallatin Unit 4
XI
-------
Acknowledgements
Preparation of this report was made possible by the financial
support of the Environmental Protection Agency and the
Tennessee Valley Authority. Much of the technical information
presented was obtained as the result of experimentation by
TVA personnel. Accomplishment of all of the experimental
work described in this report was possible only through the
active cooperation of various utility organizations who made
full-scale conditioning facilities available for study.
Cooperation of TVA and the Public Service Company of Colorado
is explicitly acknowledged. Cooperation of still other com-
panies who have chosen to be anonymous is also noted with
appreciation.
The research described in this report was conducted under the
guidance of Sabert Oglesby, Jr., Vice President of Southern
Research Institute and Head of the Institute's Engineering
and Applied Sciences Department. Technical contributions of
major dimensions were made by Grady B. Nichols, John P. Gooch,
Walter R. Dickson, and Joseph D. McCain of the Institute's
research staff.
xn
-------
SECTION I
CONCLUSIONS
The use of gaseous chemical compounds as "conditioning" agents
for fly ash is an effective way to improve the electrostatic
precipitation of the particulate produced in a coal-burning
power plant. The value of this technology has been long
recognized. Certain aspects of conditioning with sulfur tri-
oxide and ammonia, however, are clarified by the research
described in this report: (1) the circumstances under which
conditioning is effective and (2) the mechanisms by which con-
ditioning can occur (which include processes other than the
most widely known process of lowering the electrical resistiv-
ity of the ash).
CONDITIONING WITH SULFUR TRIOXIDE
Practical Considerations
The most significant function of sulfur trioxide conditioning
is to lower the resistivity of fly ash from low-sulfur coals.
In regard to this process, the principal conclusions from
this research are as follows:
• The various types of injection systems investigated
produce equally effective results when properly
engineered and maintained. (These systems are based
on anhydrous sulfur trioxide, concentrated sulfuric
acid, and catalytically oxidized sulfur dioxide as
source materials.)
• Satisfactory sites of injection include the flue-
gas ducts in locations upstream from a precipita-
tor, upstream from the combination of a precipi-
tator and a mechanical collector, and between the
two types of collectors.
-------
• The minimum required concentration of sulfur tri-
oxide is in the range from 5 to 20 ppm, with the
appropriate value depending upon the flue-gas
temperature and the fly-ash composition.
• Fly ash of widely varying compositions can be
successfully conditioned with sulfur trioxide.
The main effect of compositional variations is
the requirement of more conditioning agent with
high-alkalinity ash.
• Fly ash can be successfully conditioned at temper-
atures ranging from 110°C to at least 160°C and
perhaps near 200°C.
One of the secondary functions of sulfuric trioxide condition'
ing is to increase the cohesiveness of fly-ash particles and
minimize rapping reentrainment. This role of the condition-
ing agent is to be expected if the resistivity of the ash is
not high enough to maintain an electric force as an adequate
restraint against reentrainment.
Theoretical Factors
Lowering the resistivity of fly ash obviously occurs with sul'
fur trioxide conditioning. The process involves the codeposi'
tion of the injected sulfur trioxide and the naturally occur-
ring water vapor from the flue gas onto the surfaces of ash
particles. Either adsorption or condensation of the two
gases leads to increased surface conduction. The mechanism
of conduction on conditioned ash may involve hydrogen ion
migration in a surface film of sulfuric acid or acid-induced
mobility of alkali metal ions normally present in the ash.
Other mechanisms of conditioning apparently include a change
in the cohesiveness of fly ash, as mentioned above, and a
space-charge effect. The first of these mechanisms involves
a modification of the ash surface properties, as does the
lowering of resistivity. The second mechanism involves a
change in the electrical properties of the flue gas rather
than the fly ash.
CONDITIONING WITH AMMONIA
Practical Considerations
The utility of ammonia as a conditioning agent may be more
restricted than that of sulfur trioxide. The use of ammonia
conditioning to overcome the prevalent problem of high resis-
tivity of fly ash from low-sulfur coals, especially those of
-------
Western origin, cannot be recommended on the basis of the
direct results of this research. However, the utility of
ammonia conditioning under these circumstances was not inves-
tigated during this program and needs clarification by
further research.
The value of ammonia as a conditioning agent was established
clearly with fly ash from selected coals of Eastern origin
having both low- and high-sulfur contents. With these fuels,
the resistivity of the fly ash was not excessive, but ammonia
injection was able to improve collection efficiency through
mechanisms involving a space-charge effect in the flue gas
and an increase in the cohesiveness of the ash.
Theoretical Factors
In connection with the mechanisms of ammonia conditioning,
this research led to the following conclusions:
• Ammonia has little if any effect on the resistivity
of fly ash under the specific power-plant condi-
tions investigated. However, this conclusion does
not necessarily mean that the effect of ammonia on
resistivity will always be negligible.
• Ammonia conditioning occurs through a space-charge
mechanism, through which the electric field in the
flue-gas stream between precipitator corona wires
and collection electrodes is enhanced. Reaction
of ammonia with naturally occurring sulfur trioxide
and water vapor to produce fine particles of ammo-
nium sulfate or bisulfate is evidently the first
step in this mechanism. Electrical charging of the
particulate and reduction of the average mobility
of charge carriers evidently occurs subsequently
and serves as the direct cause of the field
enhancement.
• Ammonia conditioning of fly ash of abnormally low
resistivity appears to include an increase in the
cohesiveness of the ash with an attendant reduction
in rapping reentrainment.
-------
SECTION II
RECOMMENDATIONS
The preceding discussion of the conclusions from this research
includes at least some indirect recommendations on condition-
ing with sulfur trioxide and ammonia that are based on present
knowledge. The following discussion offers explicit recommen-
dations on additional research that is needed.
USE OF SULFUR TRIOXIDE
One of the major needs in the power industry is the ability
to predict the electrical resistivity of fly ash to be pro-
duced by fuel from a new coal field and, thus, to predict the
need for conditioning or an alternative method of overcoming
high resistivity. This need can only be satisfied if rela-
tionships are established involving the composition of repre-
sentative core samples of coal, the composition of the fly
ash and the flue gas to be produced by the coal, and the sus-
ceptibility of the ash to conditioning by normal constituents
of the flue gas and by added sulfur trioxide.
Fundamental laboratory work is recommended to provide the
information required for predicting fly-ash resistivity with
and without added sulfur trioxide as a conditioning agent.
Preliminary research of this nature has already been initiated
at Southern Research Institute with the support of the
Environmental Protection Agency. This research should be con-
tinued with attention directed to these specific questions:
(1) What is the relationship of the composition of
fly ash to the composition of the mineral con-
stituents of coal?
(2) How is the susceptibility of fly ash to condi-
tioning by water vapor and sulfur trioxide
related to the chemical composition of fly ash
and the nature of the gaseous environment?
-------
(3) What are the mechanisms of surface conduction
on conditioned fly ash of varying compositions?
A practical benefit from the fundamental studies described
above will be guidance in determining the range of circum-
stances under which sulfur trioxide will be beneficial.
Information on this matter is already available from the
power-plant studies described in this report. However, an
extension of this information will permit greater confidence
in the results to be achieved with sulfur trioxide condition
ing in full-scale practice.
practical questions deserve attention. One is how to
optimize the design of an injection manifold for sulfur triox-
ide and gain maximum utilization of the agent. A fundamental
Problem in injecting sulfur trioxide is to avoid acid conden-
sation at the interface between the concentrated vapor being
injected and the flue gas serving as a diluent. Another ques-
tion is to determine the merits of injecting sulfur trioxide
upstream from the air preheater. A potential advantage of
injecting the agent at this location is to eliminate the
Possibility of condensation at the interface of the injected
agent and the flue gas.
Another practical matter is to assess the impact of sulfur
trioxide conditioning on the emission of sulfate-containing
Particulates from a power plant. If all of the injected sul-
fur trioxide is collected by the ash, no increase in sulfate
emission occurs. It is evident, however, that sometimes part
°f the sulfur trioxide enters the stack as a vapor and that
it is able to condense as sulfuric acid mist in the plume
leaving the stack. The magnitude of sulfate loss and the cir-
cumstances that control this effect need to be ascertained.
USE OF AMMONIA
major potential application of ammonia conditioning — treat-
of high-resistivity fly ash from low-sulfur Western
coals — has not been realized. It does not appear that suffi-
cient research to determine the value of ammonia conditioning
in this respect has yet been attempted. It is certainly
Desirable to resolve the question of whether ammonia can
control the problem of high resistivity, however, for the use
°f this agent would be less costly, less hazardous, and more
convenient than the use of sulfur trioxide.
approaches are recommended to assess the value of ammonia
as a substitute for sulfur trioxide in treating high-
resistivity fly ash:
-------
(1) Analyze in detail the results of investigations
of ammonia conditioning in Australia and the
Soviet Union and attempt to deduce the probable
effect of this process on ash of interest in the
United States.
(2) Include studies of the fundamental aspects of
ammonia conditioning in the previously described
laboratory studies of sulfur trioxide condition-
ing.
If the results of these two approaches warrant further studies
of ammonia conditioning, a program to conduct such studies in
a pilot-plant or full-scale precipitator should be initiated.
A possible effective mechanism of ammonia conditioning in a
precipitator collecting high-resistivity ash is the suppre-
sion of the electrical manifestations of back corona without
necessarily lowering resistivity. Perhaps some type of space-
charge effect is able to suppress the massive flow of positive
ions that occurs from regions of back corona to the discharge
electrodes in a precipitator/ thus increasing the effective
operating voltage. For practical as well as fundamental
reasons, research to clarify the effect of ammonia on back
corona needs to be undertaken.
A final practical question that needs attention is whether
the space-charge effect encountered with ammonia conditioning
leads to an unacceptable increase in the emission of fine
particulate from a power station. This question was consid-
ered during the research described in this report, but it was
not fully resolved and it needs further consideration.
-------
SECTION III
INTRODUCTION
ply ash emitted from coal-burning electric power stations is
°ne of the most important causes of air pollution in an indus-
trialized community. This particulate matter may be removed
from the effluent gas of a power boiler by several approaches
based on high-efficiency electrostatic precipitators, liquid
scrubbers, and fabric filters. For many years, collection of
fly ash in electrostatic precipitators has been the most com-
monly used method for cleaning gas from power boilers. Even
today, the supremacy of electrostatic precipitators for gas
Cleaning in the power industry continues. However, there are
increasingly stringent requirements for particulate control,
Specially the material of small size. Thus, achieving the
needed efficiency of electrostatic precipitators requires elim-
xnation of unfavorable dust properties wherever possible, such
as the reduction of excessively high electrical resistivity.
EFFECT OF THE ELECTRICAL RESISTIVITY OF FLY ASH ON THE PER-
FORMANCE OF AN ELECTROSTATIC PRECIPITATOR
the efficient removal of fly ash from flue gases in an
electrostatic precipitator, several conditons must be satis-
fied. One of the most important conditions is that the elec-
trical resistivity of the ash deposited on the collector
electrodes shall lie within an appropriate range of values ,
somewhat difficult to define but described by some authors as
having a lower limit of about 1 x 10 7 to 1 x 10 9 ohm cm and
an upper limit of about 1 x 1010 ohm cm.1"3 Other conditions,
however, are of comparable importance and must be satisfied
simultaneously. For example, the gas flow distribution in
the precipitator must be reasonably uniform, the velocity of
the gas passing between the corona wires and the collector
electrodes must not be excessive, and the power supplies must
capable of maintaining an adequate current density at the
electrodes .
-------
If the resistivity of the collected ash is too low, only a
small voltage drop can be maintained across the collected
layer. Hence, the electrical force holding the ash to the
collector electrodes may be so low that reentrainment of ash
particles in the adjacent gas stream becomes severe. On the
other hand, if the resistivity of the collected ash is too
high, either of the following phenomena may occur:
(1) The resistance through the collected layer of
ash will lower the corona current that can be
produced with the normal operating voltage and,
as a consequence, the electric field in the gas
stream and the resulting migration velocities of
negatively-charged fly-ash particl.es toward the
collecting electrodes will be markedly reduced.
Usually, an attempt to overcome the effect of
excessive resistance through the collected ash
with increased voltage and correspondingly
increased electric field will meet with failure
as the result of rapid sparking.
(2) The resistance through the collected layer of ash
may be sufficient to cause electrical breakdown
in the layer with attendant formation of positive
gaseous ions ("back corona" or "reverse ioniza-
tion"), neutralization of negative charges on the
ash particles, and reentrainment of the ash parti-
cles in the gas stream.
Judging from the published literature, problems stemming from
low resistivity are rare, but those stemming from high resis-
tivity are fairly common, especially in power plants burning
low-sulfur coals. High resistivity of ash from low-sulfur
coal is attributed to the low concentration of sulfur trioxide
that is produced in the combustion process and the resulting
failure of the ash to collect sufficient amounts of sulfur
trioxide and water vapor from the gas stream to produce a con-
ductive film on the particle surfaces. Owing to the increas-
ing emphasis on the use of low-sulfur coals to minimize, emis-
sion of sulfur oxides and the simultaneous demands for
improvements in fly-ash collection, increasing efforts are
being made in the power industry to find methods to overcome
the problem of high resistivity.
CONTROL OF ELECTRICAL RESISTIVITY PROBLEMS IN ELECTROSTATIC
PRECIPITATORS
There are, at present, essentially four methods of overcoming
the problem of high resistivity of fly ash:
-------
(1) Operate the electrostatic precipitator at temper-
atures well below the normal range (e.g.r around
120°C), at which enhanced collection of water
vapor can produce a sufficiently conductive sur-
face film to make up for the shortage in sulfur
trioxide.
(2) Operate a so-called "hot precipitator" ahead of
the air preheater with temperature in excess of
300°C, where the volume conductivity of fly ash
(in contrast to the surface conductivity referred
to above) is high enough to permit efficient pre-
cipitation.
(3) Operate a precipitator at customary temperatures
(around 150°C) and inject a chemical condition-
ing agent in the flue gas to make up for the
shortage of naturally produced sulfur trioxide.
(4) Increase substantially the specific collecting
area of the precipitator electrodes (the ratio of
the total electrode area to the volume flow rate
of flue gas).
The best known chemical conditioning agents are sulfur triox-
ide (SOa) and ammonia (NHs). Both of these compounds have
been in use for many years; even so, the circumstances under
which these compounds are most effective and the mechanisms
by which they act (ammonia, especially) are not fully under-
stood.2'3 Other conditioning agents that are in use at pres-
ent or that are potentially useful include sulfamic acid,
ammonium sulfate, and ammonium bisulfate. The mechanisms of
action of these compounds is even more uncertain than those
of sulfur trioxide and ammonia.1*
Although the problem of low resistivity is less common than
the problem of high resistivity, power plants burning high-
sulfur coals and producing high concentrations of sulfur tri-
oxide are subject to this difficulty. There have been at
least two technical papers recognizing low resistivity as a
source of difficulty and reporting success in the use of
ammonia as a flue-gas additive to overcome the problem.5'6
CONDITIONING OF FLY ASH WITH SULFUR TRIOXIDE
In this report as in publications by other investigators,
conditioning of fly ash with sulfur trioxide refers not only
to the use of this compound (SO3) but also to the alternative
use of the chemically equivalent compound sulfuric acid
(H2SOt»). Injection of the vapor of either compound into a
-------
stream of flue gas at customary precipitator temperatures of
about 150°C leads to the conditioning of fly ash only by sul-
furic acid, for sulfur trioxide is quantitatively converted
to the acid by reaction with water vapor. Even though sulfur
trioxide is frequently spoken of as one of the, components of
flue gas, it only occurs as this substance at temperatures in
the gas train above 300°C. As the temperature falls to values
below 300°C downstream from the air preheater, the normally
occurring sulfur trioxide is converted to sulfuric acid.
Likewise, in the lower range of temperatures, injected sulfur
trioxide is converted to the acid. (Thermodynamic data sup-
porting the foregoing statements about the conversion of sul-
fur trioxide to sulfuric acid have been published in the
JANAF Tables.7)
Sulfur trioxide has been recognized for many years as a use-
ful conditioning agent for various dusts with high electrical
resistivity. Some of the earliest and most important experi-
mental work on conditioning with sulfur trioxide was conducted
by Chittum;8 later significant work was carried out by White.1
This experimentation was done largely on a laboratory scale,
but it has nevertheless proved valuable in understanding pro-
cesses of conditioning in full-scale precipitators.
Sulfur trioxide was one of a long series of conditioning
agents investigated by Chittum between 1942 and 1945 at the
Western Precipitation Corporation (now a division of Joy
Manufacturing Company).8 Chittum demonstrated the relative
efficiencies of different conditioning agents for various
types of dusts in a "racetrack" apparatus, in which the dust
and conditioning agent were recirculated in an airstream
and precipitated in a point-plane device that permitted deter-
mination of resistivity. Chittum found that in conditioning
basic dusts (such as magnesium oxide), sulfur trioxide and
other acidic compounds were the most effective conditioning
agents. He found on the other hand that in conditioning
acidic dusts (such as boric acid), ammonia and other basic
compounds were most effective. He concluded that the condi-
tioning agent activated the adsorption of water vapor to pro-
vide a conductive surface layer on the suspended particles
and that the activation process was favored by the use of
conditioning agents that opposed the dust in acid-base char-
acter.
Chittum apparently was not directly concerned about the condi-
tioning of fly ash; his main concern in a practical sense was
the conditioning of catalyst dust in the petroleum refining
industry. White, on the other hand, was concerned directly
with the conditioning of fly ash.1 He demonstrated, for
example, the interrelationship between the vapors of water
and sulfur trioxide in conditioning fly ash and the importance
10
-------
of flue-gas temperature in determining the concentration of
sulfur trioxide needed for efficient conditioning. Like
Chittum, White-was concerned with the mechanism of condition-
ing, not just the practical results in terms of precipitation
efficiency. Thus/ White's experimental work included deter-
minations of 'the electrical resistivity of fly ash before and
after treatment with the conditioning agent.
During the decade between 1960 and 1970, practical interest
in sulfur trioxide as a conditioning agent for fly ash accel-
erated. With increased use of low-sulfur coals in the utility
industry, some means of overcoming the attendant loss of
precipitator performance became necessary. As indicated
previously in this report, chemical conditioning of the fly
ash was one of the practical solutions investigated. Some of
the earliest practical steps toward developing the use of sul-
fur trioxide as a flue-gas additive were taken by Lodge-
Cottrell, Ltd., in Great Britain. Various papers published
by representatives of this firm described the practical value
of sulfur trioxide injection and showed the expected effect
of lowered fly-ash resistivity on the basis of field or
in situ measurements of this parameter.9'10 Other papers pub-
lished within the same span of time described the practical
value of sulfur trioxide injection without clarifying the
mode of action.*1 »l3
All efforts to overcome the problem of high-resistivity fly
ash with the injection of sulfur trioxide have not been suc-
cessful. Baxter6 and Watson and Blecher11* described failures
of sulfur trioxide to produce the expected improvement in
precipitation efficiency, even when the level of injection
was far higher than the normal level (producing about 10 to
20 ppm by volume of sulfur trioxide after dilution with flue
gas). Baxter's paper states that sulfur trioxide failed to
act as desired at one power plant, even when the level of
injection was high enough to intensify the plume of particu-
late emitted to the atmosphere (presumably the result of the
formation of a mist of condensed sulfuric acid). The unsuc-
cessful trials of sulfur trioxide conditioning have no simple
explanation. One factor that must be kept in mind, however,
is that a conditioning agent may eliminate the problem of
high resistivity without having any effect on other problems
such as inadequate precipitator size or poor gas distribution
in the precipitator.
Anomalous results from sulfur trioxide conditioning have been
obtained in some instances where high resistivity may not
have been the basic cause of poor precipitator performance.
The experimental work of Dalmon and Tidy15'16 offers a possi-
ble explanation for some of these results. These investiga-
tors found that the efficiency of precipitation of low-
11
-------
resistivity ash containing a high percentage of unburned
carbon could be improved by using sulfur trioxide as a binder
for the individual particles of ash and carbon. They con-
cluded that increased cohesiveness of the fly ash and carbon
deposited on precipitator electrodes could overcome excessive
reentrainment as the factor limiting precipitator performance,
CONDITIONING OF FLY ASH WITH AMMONIA
The value of ammonia as a conditioning agent was first
reported in 1942, in an electrostatic precipitator for recov-
ering catalyst dust in petroleum refining.3 The use of ammo-
nia and related compounds (organic amines) for improving the
precipitation of catalyst dust then became part- of the inten-
sive study undertaken by Chittum at the Western Precipitation
Corporation between 1942 and 1945. The catalyst dust was a
mixture of silicates of various metals and was acidic in its
behavior, producing a pH of less than 7 in a slurry with
water. Chittum found that ammonia and amines were much more
effective than sulfur trioxide and other acids in lowering
the resistivity of the catalyst dust. Chittum's finding, of
course, was consistent with his conclusion that the best con-
ditioning agent is one that opposes the dust in acid-base
character.
Usually, the concentrations of dusts and conditioning agents
in Chittum's laboratory experiments were quite high. Calcula-
tions of the concentration based on the total volume of his
racetrack apparatus and the total amounts of catalyst dust
and ammonia added lead to dust concentrations as high as
2,300 g/m3 (1,000 gr/ft3) and ammonia concentrations as high
as 10,000 ppm by volume. Undoubtedly, the additions of dust
and conditioning agent were made gradually; thus, instanta-
neous concentrations probably were never as high as the values
cited, but they must have greatly exceeded the concentrations
of fly ash and conditioning agents that are found in power-
plant effluents. It is true that Chittum's relative concen-
trations of dust and conditioning agent corresponded roughly
to the relative concentrations of fly ash and conditioning
agents that are common in power-plant effluents; however, it
is likely that the high absolute concentration of ammonia in
Chittum's experiments may have been a critical factor in the
adsorption of ammonia on the surface of the catalyst particles
and the resulting lowering of electrical resistivity.
Generally, the volatility of a gas can be roughly correlated
with the ease of adsorption of the gas on a solid. Ammonia
is highly volatile, having a boiling point of -33°C, and its
adsorption on alumina-silica catalysts at elevated tempera-
tures requires high partial pressures of the gas.17 Thus,
12
-------
adsorption of ammonia on fly ash, which is somewhat similar
in composition to Chittum's alumina-silica catalysts, appears
unlikely at concentrations of the order of 15 to 20 ppm in
flue gas at temperatures around 150°C. Adsorption of ammonia
on fly ash may be greatly aided if the surface of the ash is
acidic; however, the occurrence of an acidic surface material
is probably the result of the adsorption of sulfur trioxide,
and the resistivity of the ash should not in that event be
unacceptably high.
Although it is difficult to visualize the adsorption of ammo-
nia as effective in conditioning a high-resistivity fly ash,
there have been occasions when it may have occurred. Baxter
reported success in conditioning high-resistivity fly ash in
several power plants.6 In each plant, the injection of
15 ppm of ammonia lowered the emission of fly ash from the
precipitator, increased the precipitator power consumption,
lowered the resistivity of the fly ash (measured under labo-
ratory conditions), and lowered the acidity of the ash.
Watson and Blecher reported that the injection of 15 to
20 ppm of ammonia improved the efficiency of collection of an
acidic ash;14 these investigators also reported voltage-
current data for their precipitator indicating that ammonia
suppressed the occurrence of back corona, a phenomenon some-
times caused by high resistivity. Other investigators whose
work has not been published have reported results of precipi-
tator tests indicating that ammonia lowered fly-ash resistiv-
ity.18"20
In contrast to the reports indicating that ammonia condition-
ing lowered fly-ash resistivity from excessive values, other
reports have indicated that ammonia conditioning increased
the resistivity of fly ash when the value was too low. One
instance of this phenomenon was reported by Baxter,6 who
found that 15 ppm of ammonia produced an increase in resis-
tivity (from 5 x 108 to 1 x 1010 ohm cm by laboratory measure-
ment), a decrease in the acidity of the ash, and accompanying
increases in precipitator power consumption and efficiency.
Another apparent instance of increased resistivity was .
described by Reese and Greco.5 These authors discussed a
power plant burning high-sulfur coal and collecting fly ash
at a precipitator temperature of about 130°C. (This plant,
Widows Creek Station in the TVA system, is discussed later in
the section of this report covering experimental work.) The
authors were able to reach acceptable precipitator efficien-
cies by either of two methods: (1) raising the gas tempera-
ture to 155°C or (2) injecting ammonia at a concentration of
about 15 ppm with no temperature change. The authors' assump-
tion was that either method increased the resistivity of the
13
-------
fly ash (by lowering the amount of sulfuric acid vapor con-
densed on the surface of the ash) and thus decreased
reentrainment of fly ash in the precipitator.
In a significant fundamental study of various conditioning
agents for fly ash, Dalmon and Tidy compared ammonia and
various other compounds for conditioning fly ash that was
highly basic in character16 (in opposition to the acidic fly
ash treated by Baxter6 and by Watson and Blecher ). Dalmon
and Tidy burned a low-sulfur paraffin oil in the laboratory
to produce a gas stream essentially free of sulfur oxides and
added the fly ash and each of the conditioning agents to this
gas stream before it entered a small electrostatic precipita-
tor. They found that ammonia produced a slight increase in
the efficiency of the precipitator without producing a measur-
able change in the resistivity of the ash; they then posed
this question:21 "Could the small quantities used be influ-
encing the voltage-current characteristic to enable a high
working voltage?"
Dalmon and Tidy also found that if sulfur dioxide was added
to the gas stream to simulate the actual concentration pro-
duced by a low-sulfur coal, the effects of ammonia injection
were inconsequential unless hydrogen chloride was added also
(this gaseous compound is present in flue gas at varying con-
centrations depending upon the amounts of chloride salts that
are present in the coal). The authors attributed the rela-
tionship involving ammonia and hydrogen chloride to the
formation of small gas-borne solid particles of ammonium
chloride, the coprecipitation of this material with the fly
ash, and the resultant lowering of the resistivity of the ash,
Ammonium ion was usually not found in ash analyses, but
chloride ion was found. Loss of ammonium ion was attributed
to the reaction of this ion with the basic constituents of
the ash and the release of ammonia to the gas stream. Hence,
the authors concluded that the effect of ammonia in the pres-
ence of hydrogen chloride was to aid the uptake of chloride
ion by the ash.
It is reasonable to expect that the combination of ammonia,
sulfur trioxide, and water vapor could produce particles of
ammonium bisulfate or ammonium sulfate and condition fly ash
by a mechanism analogous to the one described by Dalmon and
Tidy. It is reasonable to propose as an alternative that the
electrical charging of the particles of either ammonium salt
could increase the electric field in the precipitator and
effectively "condition" the ash through a space-charge effect.
Experimental evidence for the latter mechanism of action by
ammonia is discussed at length in the sections of this report
dealing with experimental work.
14
-------
SECTION IV
GENERAL DESCRIPTION OF RESEARCH ON CONDITIONING
BY SULFUR TRIOXIDE AND AMMONIA
Experimental studies of conditioning of fly ash with sulfur
trioxide and ammonia were carried out by Southern Research
Institute during 1970-1971 under Contract CPA 70-149 with the
Environmental Protection Agency22 and during 1972-1974 under
Research Agreement TV36921A with the Tennessee Valley Author-
ity.23 Most of the work for EPA was concerned with sulfur
trioxide as the conditioning agent; but the studies included
a limited investigation of ammonia conditioning. The later
work supported by TVA included an extension of the studies of
sulfur trioxide conditioning; however, it consisted primarily
of efforts to resolve questions about ammonia conditioning
that were not answered during the first investigation.
OBJECTIVES OF THE RESEARCH
During 1970, the existence or planned installation of facili-
ties for injecting sulfur trioxide in various power plants in
the United States created the opportunity for Southern
Research Institute to investigate the effectiveness of this
conditioning agent in overcoming problems caused by high-
resistivity fly ash from domestic coals. One of the most
ambitious programs of installing sulfur trioxide conditioning
facilities was undertaken by the Public Service Company.of
Colorado, which has now installed such facilities for eight
power units in three plants.2**'25 Before the negotiations
for a research contract between the Institute and the Environ-
mental Protection Agency were completed, an agreement between
the Institute and the Public Service Company had been reached
to permit members of the Institute staff to conduct research
at the power plants to be equipped with conditioning facili-
ties. .
The objective of the planned program of research with such
sulfur trioxide facilities that might become available
(including others as well as those in Colorado) was to
15
-------
investigate the role of each of the following in determining
the effectiveness of conditioning: (1) coal composition,
(2) fly-ash composition, (3) flue-gas composition and temper-
ature, (4) concentration of sulfur trioxide injected, (5)
source of the sulfur trioxide injected, and (6) location of
the point of sulfur trioxide injection in the flue-gas train.
The results of the investigation were expected to provide
general, practical guidelines of value to utility companies
which might need to predict the value of chemical condition-
ing in power plants encountering the problem of high-resis-
tivity fly ash. The results of the investigation were also
expected to shed light on some theoretical aspects of the
conditioning process that were not clearly understood.
The key to the experimental approach was to make; in situ
measurements of the electrical resistivity of fly ash with
and without conditioning. Changes in resistivity as a result
of conditioning but not changes in precipitator performance
were to be measured. The rationale for this approach was to
focus attention on the specific process by which improved
efficiency could be expected, assuming that no other problem
was important in controlling the efficiency. No effort was
made to diagnose causes of poor precipitator performances
other than high-resistivity fly ash.
Experience with various devices for measuring resistivity
proved to be necessary before acceptable measurements could
be made. An apparatus of the cyclone type that was developed
by Cohen and Dickinson26 and used in investigations by Lodge-
Cottrell, Ltd., 9>10 was used briefly but then discarded
because of its lack of reliability. Another cyclone device
and a point-plant apparatus were then developed and found to
be more satisfactory in their performance. Descriptions of
these two resistivity probes have been given by Nichols of
Southern Research Institute in another report. 7 A brief
discussion of the use of these probes for resistivity measure-
ments and the methods employed to obtain related experimental
data is given in the Appendix of this report.
Approximately one year after the. research under the contract
with the Environmental Protection Agency was completed, the
second research program with the Tennessee Valley Authority
was started. The objectives of this program were twofold:
(1) to assist TVA in evaluating the planned installation of
a sulfur trioxide injection facility in one power plant and
(2) to investigate the use of ammonia conditioning in various
other power plants. The research was to be concentrated on
conditioning with ammonia for the purpose of elucidating the
mechanisms of conditioning by this agent.
16
-------
The interest of TVA in ammonia as a conditioning agent steins
from the work of Reese and Greco,s who used ammonia success-
fully to improve the collection of fly ash produced from a
high-sulfur coal. As pointed out in Section III of this
report, Reese and Greco concluded that ammonia increased the
resistivity of the fly ash from its normally low value either
by neutralizing the surface layer of excess sulfur trioxide
on the fly ash or by neutralizing sulfur trioxide in the gas
phase and incidentally producing small solid or liquid parti-
cles of ammonium sulfate or ammonium bisulfate, as shown by
the following equations:
2NH3(g) + S03(g) '+ H20(g) —» (NH,J2SCMs)
NH3(g) '+ S03(g) + H20(g) —» NHi,HS(Ml,s)
The ammonium sulfate shown as the product of the first reac-
tion would exist as a solid in a wide range of flue-gas
temperatures, as indicated in the equation. The ammonium
bisulfate shown as the product of the second reaction would
exist as a liquid above 144°C or as a solid below this temper-
ature; hence, the physical state is indicated in the equation
as either liquid or solid. (There is some uncertainty about
the temperature where ammonium bisulfate solidifies. The
indicated temperature is based on the data of Kelley et al.28
Thermodynamic information indicating that either of the two
reactions can occur under flue-gas conditions is also included
in the data published by Kelley.)
The initial investigation by Southern Research Institute
failed to confirm either of the postulates. Another investi-
gation by McLean29 was also unsuccessful in determining the
mechanism of ammonia conditioning under the circumstances
encountered by Reese and Greco, Thus, additional research
was undertaken at the Institute to elucidate the mechanism of
ammonia conditioning.
ACCOMPLISHMENTS OF THE RESEARCH PROGRAMS
Under the program for EPA, conditioning by sulfur trioxide
was investigated in ten precipitators in nine different power
stations.22 In several of these precipitators, only naturally
occurring sulfur trioxide was available to condition the ash.
However, in four precipitators, including three operated by
the Public Service Company of Colorado, some method of inject-
ing sulfur trioxide was available. For still another precip-
itator operated by another utility company during trials of
sulfur trioxide as a conditioning agent for fly ash from a
low-sulfur Western coal, the test data were acquired and com-
pared with the data obtained by direct experimentation.
17
-------
Under the program for TVA, a study was made of conditioning
with sulfur trioxide in one of the TVA precipitators, and
studies were also made of the conditioning with ammonia in
four TVA precipitators in three different plants.23
A list of the more important power plants and their precipita-
tors that were investigated during the two research programs
is given in Table 1. Two plants can only be referred to by
code numbers (Plants 3 and 6) because the utility companies
permitted publication of the data from these two plants under
the stipulation that neither the name nor location of either
plant would be identified. The information in Table 1
includes properties of the coal burned at each plant, the
flue-gas temperature in the precipitator, and very general
descriptions of the properties of the fly ash.
18
-------
Table 1. POWER PLANTS INVESTIGATED DURING THE RESEARCH ON THE CONDITIONING OF FLY ASH
Plant
No.a
1
2
3
4
5
6
7
8
9
10
11
Name
Kingston
Cherokee
(Not identified)
Cherokee
Arapahoe
(Not identified)
Bull Run
Widows Creek
Widows Creek
Bull Run
Gallatin
Unit
5
2
-
3
4
-
1
7
7
1
4
Precipitatorb
-
-
-
'-
-
-
IB
7B
7A
1C
4C
Ownerc
TVA
PSCo
-
PSCo
PSCo
-
TVA
TVA
TVA
TVA
TVA
Conditioning
agent
Noneb
SO 3
SO3
S03
SO 3
SO 3
S03
NH3
NH3
NH3
NH3
Coal
Wt %
S
2.1
0.6
M>.5
0.5
0.5
0.6
0,85
0.9
3.5
1.2
4.0
Wt %
Ash
19.6
7.9
-
8.6
5.9
11.8
15.99
17.6
13.8
17.0
14.2
Gas
temp,
°C
175
143
110
154
135
160
1259
132
132
125
138
Fly ash
log
pd
12.0
11.2
11.7
12.0
12.6
12.3
10.39
11.5
8.3
10.5
8.6
pHe
6.1
7.0
11.0
10.0
11.1
8.1
4.89
5.2
10.3
4.5
8.6
for reference later in this report.
Not listed if only one precipitator is used for the unit specified.
jTVA = Tennessee Valley Authority; PSCo = Public Service Company of Colorado.
p = electrical resistivity, ohm cm (unconditioned ash).
epH at equilibrium for a slurry consisting of 1 part of unconditioned ash and 300 parts of
fdistilled water (weight proportions).
Water vapor was investigated briefly as a conditioning agent.
-------
SECTION V
RESULTS OF STUDIES OF CONDITIONING WITH SULFUR TRIOXIDE
The various studies of conditioning with sulfuf--trioxide can
be logically classified by the source of the sulfur trioxide
or the circumstances under which conditioning occurred (e.g_.,
gas temperature and fly-ash composition). The first type of
classification has been selected, more or less arbitrarily,
as the basis for presenting the experimental results. Hence,
the primary headings throughout this section designate the
sources of sulfur trioxide, which are of four types: (1) the
gas produced naturally by the combustion of a coal of moderate
sulfur content and (2) the gases injected from (a) stabilized
liquid sulfur trioxide, (b) concentrated liquid sulfuric acid,
and (c) sulfur trioxide from the catalytic oxidation of sulfur
dioxide.
CONDITIONING BY NATURALLY OCCURRING SULFUR TRIOXIDE
The Kingston plant in the TVA system was one of several
plants investigated under Contract CPA 70-149 in which condi-
tioning of fly ash occurred as the result of the presence of
naturally occurring sulfur trioxide rather than the injection
of this compound and, in limited experimentation, as the
result of the injection of water vapor as a conditioning
agent.22 These data are of special interest because they show
very well the interdependence of sulfur trioxide and water
vapor in the normal process of conditioning and the impor-
tance of temperature in regard to the effectiveness of this
process.
During the experimental work at the Kingston plant, the coal
being burned in the unit under investigation, Unit 5, was of
moderate sulfur content and moderate-to-high ash content.
Analyses of various samples of the coal led to the following
results as averages:
Sulfur, 2.1%
Ash, 19.6%
20
-------
Determinations of the electrical resistivity of the fly ash
entering the precipitator of Unit 5 were made with both a
cyclone apparatus and the point-plane apparatus as the flue-
gas temperature was varied through a wide range, roughly 150
to 190°C. The data were obtained with an electric field of
2.5 kV/cm in either resistivity device and are plotted in
Figure 1. Several features of the data in this figure are of
interest:
• Water vapor not injected. The resistivity values
obtained with either apparatus reached a maximum
at a temperature of about 175°C. The order of
magnitude of the maximum was 1 x 1013 ohm cm with
fly-ash samples in the cyclone apparatus and
1 x 1012 ohm cm with samples in the point-plane
device. The appearance of maxima in both
resistivity-temperature curves reflects the pre-
dominance of volume conduction at higher tempera-
tures and surface conduction at lower temperatures.
• Water vapor injected. Resistivity values found at
temperatures around 165°C were lowered substan-
tially by the injection of water vapor. The
reduction in resistivity was found to be one to
two orders of magnitude, depending upon the method
of measurement. To secure the resistivity data,
water vapor was added by the crude method of pump-
ing a stream of water into the boiler. It was
intended that this process should essentially
double the normal concentration of water vapor,
about 7% by volume, that was produced from combus-
tion of the coal, and it was found that this
objective was satisfied as discussed in connection
with flue-gas analyses in a following paragraph.
Samples of flue gas were collected at sampling ports ahead of
the air preheater at a temperature of about 340°C and ahead
of the precipitator at temperatures of about 150 to 190°C.
The apparent concentrations of sulfur trioxide are plotted
in Figure 2 and discussed in the next paragraph. The con-
centrations of sulfur dioxide averaged around 1500 ppm by
volume, about as expected with a coal containing about 2%
sulfur. The concentrations of water vapor averaged about 7%
by volume without injection or 14% by volume with injection
and thus showed the desired increase as the result of the
injection process.
As indicated by Figure 2, the comparative values of sulfur
trioxide ahead of the air preheater and ahead of the precipi-
tator depended on the temperature at the latter location.
At temperatures around 180°C, the concentration of sulfur
21
-------
10
11*
10
1 3
H
6
E-i
CO
H
CO
10
12
10
i i
10
i o
140
A, B;
C>
WITHOUT H20
INJECTION
WITH H20
INJECTION
O POINT-PLANE SAMPLES
D CYCLONE SAMPLES
I J.
150 160 170 180 190
TEMPERATURE, °C
200 210
Figure 1. Electrical resistivity of fly ash
at Kingston Unit 5
22
-------
20
£
o
ui
PM
O
o
H
1
tt
IS
w
O
u
15
10
UPSTREAM FROM
PRECIPITATOR
UPSTREAM ~
FROM AIR
PREHEATER
/
O WITHOUT H20 INJECTION
® WITH H20 INJECTION
140
150
160 170 180
TEMPERATURE, °C
190
330
340
Figure 2
Concentrations of sulfur trioxide
at Kingston Unit 5
23
-------
trioxide at the inlet to the precipitator exceeded the con-
centration at the higher temperature upstream from the^pre-
heater. This difference was probably due to the reaction of
sulfur dioxide and oxygen to produce sulfur trioxide within
the preheater; this reaction is favored by the lowering of
temperature and the availability of catalytically-active sur-
faces. As the temperature ahead of the precipitator became
lower, however, the concentration of sulfur trioxide
decreased. Moreover, when water vapor was injected, the con-
centration of sulfur trioxide decreased even more. Compari-
son of Figures 1 and 2 shows a marked correspondence between
the values of fly-ash resistivity and sulfur trioxide con-
centration. It is apparent that decreases in resistivity
values and sulfur trioxide concentrations occurred simultan-
eously. Decreases in both parameters were favored by
decreasing temperature or increasing water-vapor concentra-
tion: they signify that increasing adsorption of sulfur tri-
oxide and water vapor occurred on the surface ofthe fly-ash
particles as either change occurred in the flue gas. The
process of collection of sulfur trioxide by the ash was
clearly adsorption rather than condensation, because the flue-
gas temperature was consistently above the dew point of the
sulfur trioxide-water mixture.3"'31
Analyses of fly-ash samples collected in the two resistivity
probes consisted of determinations of (1) the equilibrium pH
of a slurry of 0.1 g of each sample with 30 ml of distilled
water and (2) the weight percentage of the sample dissolved
as sulfate ion. The results of the pH and sulfate determina-
tions are shown in Figure 3. Although the data in this figure
are scattered, they still give evidence of decreasing pH
values and increasing sulfate concentrations as the collection
temperature was lowered or as water vapor was injected. The
indicated changes in pH and sulfate are consistent with the
changes in the electrical resistivity of the ash and the con-
centration of sulfur trioxide in the flue gas as shown in
Figures 1 and 2.
Analyses of fly ash from Kingston Unit 5 which show the,over-
all composition are given in Table 2. The most noteworthy
aspect of this composition—fairly typical of fly ash from
Eastern coals—is the low percentage of calcium oxide compared
with the percentages of lime in fly ash from Western coals,
which are discussed later in this report.
24
-------
10
en
Ou
5.0
5.5
6.0
6.5
7.0
1^
I \ • 1 I
\O v
0 \ WITH H20
\INJECTION
\ °° \
— o >v —
WITHOUT \O O °
O
~ 00 ~~
1 1 II
10 150 160 170 180 19
0.7
0.6
<*>
tat
§ 0.5
SULFATE CONS
o
•
£>>
0.3
0.2
0 1^
III!
^•S* VJITH H20
® "^-^ INJECTION
'^'^s.
0 ^
^^^Sv^ 9
0 0 ^ ^^^^
WITHOUT H2O
INJECTION
1 II 1
10 150 160 170 180 19
TEMPERATURE, °C
TEMPERATURE, °C
Figure 3. Chemical properties of fly ash
at Kingston Unit 5
-------
Table 2. COMPOSITION OF FLY ASH
FROM THE KINGSTON PLANT
Component
Li2O
Na20
K20
MgO
CaO
A1203
Fe20a
Si02
Ti02
P20s
SO 3
Weight percentage
0.06
0.42
3.7
1.3
1.1
30.3
9.9
52.4
1.7
0.51
0.45
CONDITIONING BY THE INJECTION OF VAPOR FROM STABILIZED LIQUID
SULFUR TRIOXIDE
The most frequently used source of sulfur trioxide for condi-
tioning fly ash, particularly in foreign power plants, has
been the stabilized liquid form of this compound. The sta-
bilized compound is available in the United States as the
product of Allied Chemical Corporation known as "Sulfan."
Despite the fact that injection systems based on liquid sulfur
trioxide have been widely used/ only one system of this type
was available for investigation under Contract CPA 70-149. 2
This system was in use at Unit 2 of the Cherokee plant of the
Public Service Company of Colorado. However, limited data
from another plant with a Sulfan system (referred to as
Plant 3) were made available by another Western utility
company and are included in a later section of this report.
Cherokee Plant, Unit 2
The Cherokee plant of the Public Service Company of Colorado
consists of four units with power-production ratings varying
from 100 to 350 MW. During 1971, the installation of a
Sulfan injection system to treat flue gas from three units—
Units 1, 2, and 4—was completed. During July of 1971,
shortly after service to Unit 2 was started, the investigation
described in this report was carried out at that unit. Unit
2 is rated at 110 MW, and it is equipped with both a mechani-
cal collector and an electrostatic precipitator. Sulfur tri-
oxide is injected into the duct between these two gas-cleaning
devices.
26
-------
The coal burned at the Cherokee station is normally a low-,
sulfur coal from Routt County, Colorado. It produces a high-
resistivity fly ash that requires conditioning for effective
collection in an electrostatic precipitator. Measured values
of resistivity of the unconditioned ash have ranged from
1.6 x 10ll ohm cm at 143°C in Unit 2 to 1.0 x 10*2 ohm cm at
154°C in Unit 3 of the Cherokee station.
At the time the conditioning study was carried out at Cherokee
Unit 2, a sample of the coal was analyzed and found to have
the following properties:
Sulfur, 0.6%
Ash, 7.9%
These values appear to be fairly representative, for they
agree with values obtained on another occasion when a study
was made of conditioning at Cherokee Unit 3.
The point-plane probe was used for all of the resistivity
determinations at Cherokee Unit 2. Considerable difficulty
was encountered in making these determinations because of the
presence of an abnormally high concentration of unburned
carbon in the ash and the tendency of the fly-ash samples to
undergo electrical breakdown even at very low applied fields.
The only available means of overcoming this difficulty was to
make measurements at very low electric fields, of the order of
0.1 kV/cm or less. (In other plants, resistivity data were
usually taken at substantially higher fields, which have the
effect of lowering the apparent values of resistivity as dis-
cussed by Nichols.27)
The resistivity data obtained without the injection of sulfur
trioxide and with the injection of this agent at concentra-
tions of 13 and 27 ppm* are presented in Table 3. Later in
this report (Figure 18, page 120), the data are shown graphi-
cally in comparison with data from other plants. Both the
table and the figure indicate that 13 ppm of sulfur trioxide
depressed the resistivity to a greater degree than was neces-
sary and suggest that a concentration of only 5 ppm might have
been adequate. Unfortunately, determinations of the
*These concentrations were reported by the plant operators,
as were the concentrations of conditioning agents in most of
the other plants investigated. The only experimental deter-
minations of concentrations of conditioning agents that were
made by personnel of Southern Research Institute are discussed
explicity later in this report.
27
-------
Table 3. ELECTRICAL RESISTIVITY OF
FLY ASH AT CHEROKEE UNIT 2
Injected S03
concn, ppm
0
•13
13
27
27
27
Resistivity, a
ohm cm
1.6 x 1011
1.4 x 107
5.6 x 107
2.5 x 107
3.8 x 107
5.0 x 107
Temperature , 143°C
resistivity at concentrations below 13 ppm could not be made
because a failure in the injection system caused an extended
interruption in its use.
The effect of sulfur trioxide on the resistivity of the ash
had the expected corollary effect on the precipitator volt-
ages, currents, and spark rates. For the inlet section of
the precipitator, the time-average voltage and current were
37 kv and 80 mA at a spark rate of 25/min without injection
of sulfur trioxide. For the same section, the values were
41 kV and 720 mA at a spark rate of only 5/min with injection
of 27 ppm of the conditioning agent.
Concentrations of flue gases with and without injection are
given in Table 4. These data show that the concentrations of
sulfur trioxide ahead of the air preheater and ahead of the
precipitator were quite low without injection, only 2 to
3 ppm. The concentration at the entrance of the precipitator
(downstream from the plane of injection) , however, increased
sharply with injection. Indeed, the concentration reached
values almost as high as the reported injection values /sug-
gesting that only a small fraction of the conditioning agent
was collected on the surface of the fly-ash particles, at
least at the precipitator inlet. Thus, the flue-gas analyses
as well as the resistivity data give evidence of excessive
rates of injection of sulfur trioxide.
Information about the properties of the fly ash is given in
Table 5 . The pH and sulf ate values indicate that a substan-
tial pickup of the normally occurring sulfur trioxide by the
fly ash occurred during passage of the ash through the air
preheater. The decrease in pH and the increase in sulf ate
28
-------
Table 4. CONCENTRATIONS OP FLUE GASES
AT CHEROKEE UNIT 2
Injected SO3
concn, ppm
0
13
27
Temp , a
°C
368
368
143
143
143
143
143
Concentrations
SO 3 , ppm
2
2
3
12
11
26
23
SO 2 / ppm
507
512
473
452
442
460
420
^HaO, %
-
7.4
7.5
The higher temperature indicates sample collection up-
stream from the air preheater. The lower temperature
is for the flue gas entering the precipitator.
Table 5. CHEMICAL PROPERTIES OF FLY ASH
AT CHEROKEE UNIT 2
Injected SO3
concn, ppm
0
13
27
Temp , a
368
143
143
143
143
143
143
143
143
Asn properties
pH
7^3
6.6
6.8
6.9
7.1
6.7
7.4
Sulfate, wt %
0.29b
0.97
1.28
1.42
1.51
1.53b
1.85
1.76
1.89
The higher temperature indicates sample collection
upstream from the air preheater. The lower temper-
. ature is for the flue gas entering the precipitator,
The samples having these properties were collected
in a cyclone sampler; all other samples were col-
lected with the point-plane resistivity probe.
29
-------
concentration may have been influenced to some degree, how-
ever, by the removal in the mechanical collector of coarse
fly-ash particles with a lower-than-average surface-to-mass
ratio. The pH and sulfate data obtained for conditioned ash
show no significant change in pH but show the significant
increase in sulfate concentration that was expected. It is
to be noted that pH data for slurries of fly ash in water are
not necessarily valid indicators of the surface properties of
the ash. Frequently, the pH values of slurries are lower
when contact is first made between ash and water than when
equilibrium is reached after prolonged stirring. The explana-
tion for this phenomenon is assumed to be the existence of an
acid film on the surface and the presence of soluble excess
base in the interior of the ash particles. It is to be noted
further that although sharp increases in sulfate concentra-
tions were produced by the injection of sulfur trioxide, they
corresponded to small fractions of the concentrations injected,
A reasonable estimate of the fly-ash concentration treated is
2.3 g/m3 (1 gr/ft3); the increased sulfate for this amount of
fly ash would require concentrations of only 1 to 2 ppm of
sulfur trioxide in the flue gas.
As a result of the unexpected brevity of the investigation at,
Cherokee Unit 2, no sample of fly ash large enough for com-
plete analysis was collected. Hence, the only data for the
overall composition that can be cited was obtained on another
occasion in an investigation of Unit 3. These data are given
in Table 6. There is at least a possibility that the composi-
tion shown in Table 6 is not representative of the ash that
was conditioned at Unit 2. The basis for emphasizing this
possibility is that the ash from Unit 3 was substantially more
basic than the ash from Unit 2. A possible explanation for
this difference other than an actual difference in composition
is that the ash was collected at a somewhat higher temperature
at Unit 3 (154°C rather than 143°C) and may thus have gained
less acid by the adsorption of normally occurring sulfur tri-
oxide.
It is of interest to cite the results of conditioning at
Cherokee Unit 2 as recently reported by the representatives
of the utility company.25 The reported efficiencies of the
precipitator operating at full load (110 MW) are 94.0% without
conditioning and 95.2% with conditioning at a concentration
of about 15 to 20 ppm. These efficiencies are to be compared
with a design value of 94.2% at a gas velocity of 1.51 m/sec
(4.95 ft/sec) and a specific collecting area of
31.3 m2/(m3/sec) (159 ft2/[1000 ft2/min]). The effect of con-
ditioning in the. instance of Cherokee Unit 2 is thus not very
dramatic. (It is to be noted that the efficiencies cited are
for the precipitator alone, not the combination of mechanical
collector and precipitator.)
30
-------
Table 6. COMPOSITION OF FLY ASH
FROM THE CHEROKEE PLANT3
Component
Li2O
Na20
K2O
MgO
CaO
A12O3
Fe20a
SiO2
TiO2
P;>05
SO 3
Weight percentage
0.02
1.7
1.2
1.9
6.8
25.7
4.8
54.6
1.1
1.0
1.0
Collected at Unit 3 of the Cherokee
plant and assumed to be representa-
tive of fly ash from Unit 2, although
no information for fly ash obtained
directly from Unit 2 is available.
Plant 3
The data for this plant were provided for this study with the
understanding that the name of the plant and the operating
utility company would not be cited in the Institute's reports.
All of the data subsequently presented, except that dealing
with the fly-ash composition, were obtained by the utility
company.
Plant 3 is located in one of the Western states and burns a
low-sulfur coal that is typical of that region. It does not
employ a mechanical collector; it uses only an electrostatic
precipitator for gas cleaning, and the precipitator operates
at 110°C. Despite the low temperature, the resistivity of
the fly ash is high enough to warrant conditioning of the ash.
Trials were thus made with sulfur trioxide as the conditioning
agent and with Sulfan as the source of the agent.
The results of some of the trials of sulfur trioxide condi-
tioning are shown in Table 7. Clearly, the addition of the
conditioning agent lowered the resistivity of the ash and
increased the sulfate content. No information relative to
the flue-gas composition, however, is available for presenta-
tion .
31
-------
Table 7. ELECTRICAL RESISTIVITY AND SULFATE
CONTENT OF FLY ASH AT PLANT 3
Injected SO3
concn , ppm
0
10
18-20
Resistivity,6*
ohm cm
4.5 x 10 M
2.3 x 10 10
7.0 x 109
Sulfate,
wt %
0.17
0.31
0.38
Determined with a point-plane apparatus at an
unspecified electric field with the tempera-
ture near 110°C.
A sample of fly ash from Plant 3 was obtained and analyzed at
Southern Research Institute. It is not known whether this
ash had the same composition as the ash that was conditioned,
but the results of the analysis showing the weight percentages
of the major constituents are presented in Table 8. The ash
that was analyzed was highly basic; it produced a pH value of
about 11 in a slurry with distilled water.
Table 8. COMPOSITION OF FLY ASH
FROM PLANT 3
Component3
CaO
A1203
Fea03
SiOz
Weight percentage
6.3
29.6
3.8
53.0
Only a limited number of the compo-
nents were determined, as in the
analyses of some of the other samples
of ash discussed later in this report,
CONDITIONING BY THE INJECTION OF VAPOR FROM CONCENTRATED SUL-
FURIC ACID
Sulfuric acid is an alternative to stabilized liquid sulfur
trioxide as a source of sulfur trioxide for conditioning fly
ash. The acid is a more commonly used chemical commodity
than stabilized liquid sulfur trioxide. It is customarily
32
-------
available for industrial use as a mixture of sulfuric acid ,
with water, typically 93 wt-% acid and 7 wt-% water (known as
66°Bd acid).
One of the disadvantages of sulfuric acid is that this com-
pound requires higher temperatures for vaporization than sul-
fur trioxide. The boiling points of mixtures of the acid
with water vary with the proportions of the components that
are present. The maximum boiling point is approximately 325°C
for an azeotropic mixture consisting of about 98.5% of the
acid and 1.5% of water.31 Of course, in a vaporizer swept
with a carrier gas such as dry air, vaporization will produce
partial pressures of sulfuric acid and water totaling less
than 1 atm at temperatures below the normal boiling point of
the mixture being used.
Sometimes the vaporization of sulfuric acid is accompanied by
the decomposition of the vapor of the acid to the vapors of
sulfur trioxide and water, as shown by the following equation:
HzSO^fg) —» S03(g) + H20(g)
The extent of decomposition depends upon the partial pressure
of sulfuric acid produced and the concentration of water
vapor present in the vaporizer, as well as the temperature of
the carrier gas used to sweep the vaporizer. The extent of
decomposition of the vapor produced from the azeotrope at the
normal boiling point (about 325°C, the temperature of vapori-
zation in the absence of carrier gas) is about 40%.31 The
extent of decomposition decreases as the temperature is
lowered or as the partial pressure of sulfuric acid or water
vapor is increased. The extent of decomposition under any
specified conditions can be calculated from the thermodynamic
data in the JANAF Tables.7 In the vaporization of sulfuric
acid, extremely high temperatures must be avoided to prevent
the decomposition of the acid vapor to sulfur dioxide/ oxygen,
and water vapor, as shown by this equation:
H2S(Mg)'--*• -S02(g) + *502(g) + H2O(g)
If dry air is the carrier gas sweeping the vaporizer, the
decomposition as shown above becomes extensive at temperatures
above 650°C.32
Cherokee Plant, Unit 3
Whereas Units 1, 2, and 4 of the Cherokee plant were equipped
with a Sulfan injection system during 1971, as stated previ-
ously on page 26, Unit 3 of this plant was equipped with a
sulfuric-acid injection system the previous year. A study of
conditioning at Cherokee Unit 3, a 150-MW plant, was conducted
33
-------
during October of 1970 when the injection manifold was located
in the duct upstream from the mechanical collector. Later
during the year, the injection manifold was temporarily relo-
cated in the duct between the mechanical collector and the
electrostatic precipitator in an effort to improve the effec-
tiveness of the conditioning system. No measurements of the
electrical resistivity were made while the injection manifold
was relocated; however, analyses were made of fly-ash samples
taken from the hoppers of the ash collection system at that
time.
In the injection system at Cherokee Unit 3, the sulfuric acid
was commercial 66°Be" acid. The acid was vaporized with hot
combustion gases from a natural-gas burner; the inlet and out-
let temperatures of the gas stream passing through the vapor-
izer were about 540 and 370°C, respectively. The gas stream
passing from the vaporizer to the injection nozzles was heated
to avoid condensation of the acid vapor. The estimated ratio
of the partial pressure of sulfur trioxide to the partial
pressure of sulfuric acid was 12.4:1.0, based on the assump-
tion that the gas stream leaving the vaporizer had a tempera-
ture of 370°C and a water-vapor partial pressure of 0.05 atm
(produced mostly by the combustion of natural gas and
increased only slightly by the partial decomposition of the
acid vapor).
Analyses of the coal being burned at Cherokee Unit 3 during
the conditioning studies yielded these results:
Sulfur, 0.5%
Ash, 8.6%
These concentrations of sulfur and ash are close to the values
for coal from the same source—-Routt County, Colorado—that
was sampled at Unit 2 on another occasion (page 27).
Two series of resistivity determinations were made, one with
only coal as the fuel and another with a mixture of coal and
natural gas as fuels. In both series, resistivity devices of
the cyclone type were employed, and resistivity values were
determined at an applied electric field of about 2.5 kV/cm.
The results of the determinations are given in Table 9. They
indicate that very little change in resistivity occurred,
despite the fact that the concentrations of injected sulfuric
acid ranged up to 33 ppm on one occasion and 44 ppm on another
occasion.
The results of flue-gas analyses are given in Table 10. The
data for sulfur dioxide show the expected increase in concen-
tration as the fuel was changed from coal and natural gas to
34
-------
Table 9. ELECTRICAL RESISTIVITY OF FLY ASH
AT CHEROKEE UNIT 3
(Validity questionable as discussed in text)
Temp , a
. °C
149
154
Injected HaSOi,
con en/ ppm
0
6
17
26
33
0
13
26
33
44
Resistivity,
ohm cm
2.0 x 10 12
9.3 x 1011
7.8 x 10 ll
5.6 x 10 :1
5.0 x 10 ll
1.0 x 10 12
1.0 x 10 12
6.0 x 10 :1
6.3 x 10 lz
6.6 x 10 J1
The lower temperature was recorded with a
mixture of coal and natural gas as the
fuel. The higher temperature was recorded
with only coal as the fuel.
coal alone. The data for sulfur trioxide indicate that little
of this substance was present under any of the sampling condi-
tions, not even when the maximum concentrations of sulfuric
acid were injected.
The results of determinations of pH and dissolved sulfate in
aqueous slurries of the samples of fly ash that had been
collected for resistivity determinations are given in
Table 11. These data indicate that slight decreases in pH
and increases in sulfate concentration occurred as the result
of the injection of sulfuric acid. The increases in sulfate
are quite small in comparison with the increases that would
have been found if all of the injected sulfuric acid had been
deposited on the ash. If it is assumed that the concentra-
tion of fly ash treated with sulfuric acid ahead of the
mechanical collector was about 0.9 g/m3 (0.4 gr/ft3), the
deposition of 4,4 ppm of sulfuric acid on the ash would have
increased the sulfate concentration by about 12%, whereas the
reported injection of 44 ppm of sulfuric acid caused an actual
increase of only 0.35%.
In view of the fact that the sulfuric acid injected could not
be accounted for by analysis of either flue gas or fly ash in
the duct entering the electrostatic precipitator, analyses
were made of fly ash from the hoppers of the mechanical
35
-------
Table 10. CONCENTRATIONS OP FLUE GASES
AT CHEROKEE UNIT 3
Temp , a
°C
395
149
395
154
Injected HaSOH
concn, ppm
0
0
6
17
26
33
0
0
13
26
33
44
Concentrations
S03, ppm
2
<1
<1
<1
<1
2
<1
1
1
<1
2
3
SOa , ppni
226
222
222
210
217
216
314
358
357
349
349
357
H2O, %
-
9.5
-
-
-
—
-
_
7.7
-
-
8.4
The temperature of 395°C indicates sampling upstream from
the air preheater, and the temperatures of 149 and 154°C
indicate sampling at the precipitator inlet. The tempera-
ture at the latter location was 149°C with coal and
natural gas as a mixed fuel or 154°C with coal only as
the fuel.
Table 11.
CHEMICAL PROPERTIES OP PLY ASH
AT CHEROKEE UNIT 3
Injected ^SO^
concn, pjpm
0
13
26
33
44
Ash properties3
PH
10.0
9.9
9.9
9.8
9.5
Sulfate, wt %
0.77
0.90
1.09
1.00
1.12
For samples collected at the precipitator
inlet at a temperature of 154°C.
36
-------
collector and the electrostatic precipitator. The first sam-
ples were collected with the injection manifold in its orig.-
inal location upstream from the mechanical collector. In
these samples/ the material from the mechanical collector was
expected to have a disproportionately high concentration of
sulfate, accounting for the absence of the injected sulfuric
acid in the duct entering the precipitator; however, the
analysis of this material failed to show the expected con-
trast with the analysis of the material from the precipitator.
Later hopper samples were collected after the injection mani-
fold had been moved temporarily to the duct between the
mechanical collector and the precipitator. As expected,
analysis of these samples showed that the injection of sul-
furic acid did not alter the sulfate content of the ash
collected mechanically; surprisingly, however, these analyses
showed that the injection of sulfuric acid also did not change
the sulfate content of the ash collected by electrostatic pre-
cipitation.
It seems obvious that there was a fundamental difficulty at
Cherokee Unit 3 in injecting sulfuric acid in an effective
form at the intended concentrations. A possible cause of the
difficulty was improper temperature control in the vaporizer
or the manifold leading to the injection nozzles. Excessive
temperatures could have caused sulfur trioxide to decompose
to sulfur dioxide; however, temperatures in excess of 650°C
would have been required for this decomposition process to
occur, and temperatures in this range were not observed. Low
temperatures in the injection manifold could have allowed the
sulfuric acid vapor to condense with water vapor to a liquid;
if this process had occurred, the manifold would have been
flooded, and no such problem was reported. Another possible
cause of ineffective conditioning would be poor dispersal of
the injected gas stream within the flue gas, permitting con-
densation of sulfuric acid and water vapor to occur within
the duct at the relatively low temperature and the increased
partial pressure of water in the flue gas. Condensation of
acid within the duct is a plausible explanation for ineffi-
cient conditioning of the fly ash; however, no evidence of
condensation was reported.
Despite the evident difficulties in injecting sulfuric acid
in an effective form, representatives of the Public Service
Company have reported that the injection of sulfuric acid at
a calculated rate of 15 to 20 ppm has increased the precipita-
tor efficiency from 37.5 to 51.4%.25 These efficiency values
are for the precipitator alone; they do not take into account
the fly ash removed upstream by the mechanical collector.
37
-------
Arapahoe Plant, Unit _4
Unit 4 of the Arapahoe plant (100 MW) was the first of the
four units of that plant to have gas-conditioning facilities.
A study of conditioning at Unit 4 was conducted in October of
1970, immediately after the previously described work at
Cherokee Unit 3 was completed. At the time of the study at
Arapahoe Unit 4, the vapor from 66°B6 sulfuric acid was
injected into the duct between the mechanical collector and
the electrostatic precipitator.
The vaporizer for sulfuric acid at Arapahoe Unit 4 differed
from that at Cherokee Unit 3 in that a stream of ambient air
was electrically heated to 255°C and used to sweep the vapor-
izer and enter the injection manifold at 230°C. Thus, the
vaporizer at the Arapahoe plant had a lower temperature and a
lower background partial pressure of water (less than 0.03 atm
and probably as low as 0.01 atm), and it produced a much
lower vapor ratio of sulfur trioxide to sulfuric acid, only
about 0.3:1.0. These vapors were injected into the flue gas
at a temperature of 135°C.
Low-sulfur coals from Weld County, Colorado, and Hanna County,
Wyoming, are normally burned at Arapahoe Unit 4. The sulfur
and ash percentages are represented by the following data for
coal samples collected in October of 1970:
Sulfur, 0.5%
Ash, 5.9%
The composition of fly ash produced from this coal is shown
in Table 12. Although the results of the coal analysis are
similar to the data for coal used at the Cherokee plant
(pages 27 and 34), the results of the ash analysis are sub-
stantially different, particularly with respect to the base
calcium oxide (page 31). The concentration of calcium oxide
in the ash was far higher at the Arapahoe plant than at the
Cherokee plant.
The results of determinations of the electrical resistivities
of unconditioned and conditioned ash at Arapahoe Unit 4 are
given in Table 13. These data were obtained with a cyclone
probe and an electric field of about 2.5 kV/cm across the ash
samples. The data show clearly that the resistivity of the
ash was quite high without conditioning but was reduced
significantly with conditioning by sulfuric acid.
The results of flue-gas analyses are shown in Table 14. The
data for sulfur trioxide, which are the results of greatest
interest, showed that only minor increases occurred as sul-
furic acid was injected into the gas stream.
38
-------
Table 12. COMPOSITION OF FLY ASH
FROM ARAPAHOE UNIT 4
Component
K20
MgO
CaO
A1203
FejOa
Si02
Weight percentage
0.8
4.2
15.6
19.0
7.3
36.1
Table 13. ELECTRICAL RESISTIVITY OF FLY ASH
AT ARAPAHOE UNIT 4
Injected H2SOi»
concn, ppm
0
6
12
18
Resistivity,3
ohm cm
3.8 x 1012
5.7 x 1010
3.2 x 1010
1.9 x 1010
Temperature, 135°C.
Table 14. CONCENTRATIONS OF FLUE GASES
AT ARAPAHOE UNIT 4
Injected H2SOi,
concn , ppm
0
6
12
18
Concentrations^
SO 3 , ppm
<1
1
2
2
S02 , ppm
387
446
413
430
H20, %
8.9
_
8.8
mm
aFor samples collected at the precipitator inlet
at a temperature of 135°C.
39
-------
Information about the effect of injected sulfuric acid on the
chemical properties of the fly ash is given in Table 15. The
pH data in this table show that slight reductions in the
basicity of the ash occurred with acid injection. The sulfate
contents of the ash show a much more pronounced change with
acid injection, increasing from 1.00% without injection to
2.97% at the highest level of injection. The increases in
the sulfate content of the ash can be compared with the theo-
retical values based on an assumed ash concentration of
1.2 g/m3 (0.5 gr/ft3) between the mechanical collector and
the electrostatic precipitator (the location of the injection
manifold) and a virtually complete pickup of the injected
acid by the ash (little acid remained in the gas phase, as
shown by Table 14). The theoretical increase in sulfate con-
tent at an injected acid concentration of 18 ppm is 4.3%,
whereas the observed increase was 1.5%. Although the ratio
of these values is 3:1, the efficiency of sulfuric-acid condi-
tioning at Arapahoe Unit 4 appears to have been much greater
than that of Cherokee Unit 3 on the bases of both chemical
analyses and resistivity determinations. However, the effi-
ciency data reported by the Public Service Company show a
less striking difference. For Arapahoe Unit 4, the reported
increase in efficiency was from 67.3 to 77.3%; for Cherokee
Unit 3, the increase was from 37.5 to 51.4%.2S
Table 15.
CHEMICAL PROPERTIES OF FLY ASH
AT ARAPAHOE UNIT 4
Injected H2SO«»
concn , ppm
0
6
12
18
Ash propertiesa
pH
11.1
11.1
10.0
10.8
Sulfate, wt %
1.50
2.23
2.50
2.97
For samples collected at the precipitator
inlet at a temperature of 135°C.
Plant 6
The power station referred to as Plant 6 has several units.
One of these units with a rating of 140 MW was being operated
experimentally with a sulfuric acid injection system during
the early part of 1971. The acid injection system functioned
in essentially the same manner as the one at Arapahoe Unit 4.
The sulfuric acid vaporized was 66°Be" commercial acid, and
the vaporizer operated with electrically heated air entering
at 260°C and leaving at 205°C. The vapor ratio of sulfur
40
-------
trioxide to sulfuric acid was slightly lower than that at ,
Arapahoe Unit 4 as the result of the lower temperature of the
gas stream leaving the vaporizer.
The circumstances for conditioning at Plant 6 differed from
those at Arapahoe Unit 4 in several respects. First, Plant 6
had no mechanical collector and thus essentially all of the
fly ash produced in the boiler was subjected to conditioning.
Second, Plant 6 had a higher gas temperature, 160°C rather
than 135°C. Finally, Plant 6 produced a much less basic fly
ash than Arapahoe Unit 4.
Analyses of coal samples collected during the investigation
at Plant 6 led to the following results:
Sulfur, 0.6%
Ash, 11.8%
Analysis of the fly ash produced at Plant 6 gave the results
listed in Table 16. The analytical data show that the sulfur
percentage in the coal burned at Plant 6 (from West Virginia)
was similar to the coals burned at the plants.in the West but
that the fly ash produced from the Eastern coal was distinctly
different from the ash produced from the Western coals, prin-
cipally in having a much lower weight percentage of the basic
constituent calcium oxide.
Table 16. COMPOSITION OF FLY ASH
FROM PLANT 6
Component
Li20
Na20
K2O
MgO
CaO
A1203
Fe20s
SiO2
TiO2
P205
SO 3
Weight percentage
0.09
0.51
3.8
1.3
0.68
30.2
4.9
53.0
2.0
0.17
0.36
41
-------
The results of determinations of the electrical resistivity
of the fly ash and analytical work with the flue gas and the
fly ash are presented in Tables 17, 18f and 19. The princi-
pal points of interest shown by the data in these tables are
as follows:
• Injection of sulfuric acid at increasingly higher
concentrations gradually lowered the resistivity
of the ash from 2 x 10 ohm cm to a value around
1 x 101° ohm cm.
• Most of the injected acid remained in the gas
stream, as indicated by the determinations of
sulfur trioxide.
• Injection of the acid lowered the pH produced by
the fly ash in an aqueous slurry from a weakly
basic value to values in the acid range. How-
ever, it caused a very small increase in the
sulfate content of the ash.
CONDITIONING BY THE INJECTION OF SULFUR TRIOXIDE PRODUCED BY
THE CATALYTIC OXIDATION OF SULFUR DIOXIDE
The catalytic oxidation of sulfur dioxide to the trioxide is
being used with increasing frequence as a method of obtaining
the trioxide for fly-ash conditioning. The process can be
effected by the use of a catalyst of vanadium pentoxide oper-
ating between temperature limits of about 425 and 550°C. A
high temperature is desirable to enhance the effect of the
catalyst on the rate of oxidation of the sulfur dioxide.
However, there is a maximum in the temperature that can be
used, for the oxidation process becomes gradually subject to
thermodynamic hindrance as temperature is increased.35
Studies were carried out with conditioning systems of the
type described above on two occasions at the Bull Run plant
of the Tennessee Valley Authority. Initially, the system in
use at this 900 MW plant treated only 25% of the effluent fly
ash as it passed through one of the four electrostatic precip-
itators. Later, the initial system was replaced with a larger
system capable of treating the fly ash entering all four pre-
cipitators. All of the studies described in this report,
however, were conducted at the same precipitator, referred to
as Collector IB.
Apparent Need for Conditioning at the Bull Run Plant
Research conducted by TVA personnel over a period of several
years showed that a decrease in precipitation efficiency
occurred as the sulfur content of the coal was lowered. In
acceptance tests of Collector IB in 1969 with 2.25%-sulfur
42
-------
Table 17.
ELECTRICAL RESISTIVITY OF FLY ASH
AT PLANT 6
Injected HzSOi,
concn , ppm
0
4
8
16
Type of
resistivity probe3
C
C
C
P
C
P
C
P
C
P
C
P
C
P
C
Resistivity,
ohm cm
2.0 x 1012
2.0 x 1012
1.2 x 1011
0.7 x 10 ll
1.8 x 1011
2.0 x 10 :1
1.5 x 10 ll
1.4 x 10 ll
1.3 x 10 1X
2.1 x 10 ll
1.6 x 10 M
1.1 x 10 10
0.1 x 10 10
2.5 x 10 10
0.2 x 1010
C and P indicate cyclone and point-plant sampling
probes, respectively, which were used at the precip-
itator inlet at a temperature of 160°C. Electric
fields used for the resistivity determinations
averaged about 3 kV/cm.
43
-------
Table 18. CONCENTRATIONS OF FLUE GASES
AT PLANT 6
Injected H2SOi»
concn , ppm
0
4
8
16
Concentrations3
SO 3, ppm
<1
<1
4
6
11
14
12
SO 2 / ppm
373
384
397
391
391
475
488
470
H20, %
7.7
7.3
7.7
7.9
aFor samples collected at the precipitator inlet
at a temperature of 160°C.
Table 19.
CHEMICAL PROPERTIES OF FLY ASH
AT PLANT 6
Injected H2SOi*
concn , ppm
0
4
8
16
Ash properties3
pH
8.1
6.1
5.1
4.4
Sulfate, wt %
0.24
0.32
0.38
0.43
For samples collected at the precipitator
inlet at a temperature of 160°C.
coal as the fuel, the results showed that the manufacturer's
guarantee of 99% efficiency was satisfied under full-load
conditions. In performance tests in 1971 with relatively
low-sulfur coals, substantial decreases in efficiency were
observed.33 With so-called "Bull Run" coal containing 1.2 to
1.6% sulfur, the efficiency of Collector IB was determined
under reduced-load conditions, and it was estimated as only
90% at full load. With "Haddix" coal from Eastern Kentucky,
containing only 0.9% sulfur, the efficiency predicted at full
load was reduced further to only 80%.
44
-------
These marked losses in efficiency led to the trial of condi-
tioning first with sulfur trioxide and then with ammonia in
Collectors IB and 1C, respectively. The greater success
achieved with sulfur trioxide led to an enlargement in the
conditioning facilities for this agent to permit treatment of
fly ash entering all four collectors.
All of the collection of fly ash at the Bull Run plant is
accomplished with electrostatic precipitators. Thus, sulfur
trioxide is injected into the flue gas ahead of each precip-
tator at a location between the air preheater and the precip-
itator. At this location, the mean flue gas temperature is
normally about 125°C.
Results of the First Investigation (1972)
The first investigation of conditioning at the Bull Run plant
was conducted during August of 1972 immediately after the
original facilities for injecting sulfur trioxide were placed
in operation. During this investigation, experimental work
including tasks other than determinations of fly-ash resis-
tivity and analyses of coal, fly ash, and flue gas were under-
taken for the first time with the assistance of TVA personnel,
These additional tasks included determinations of precipita-
tor efficiency with and without conditioning by direct deter-
minations of inlet and outlet fly-ash concentrations and
indirect determinations with light-obscuration devices at the
outlet. They also included systematic determinations of
electrical parameters of the precipitator with and without
conditioning.
Coal and Fly-Ash Compositions—•
TVA's intention was that low-sulfur coal would be burned at
the Bull Run plant during the joint experimental efforts by
TVA and Institute personnel. However, the shortage of this
coal made it necessary to burn "reclaim" coal—that is,
reserve coal consisting of a mixture from various sources.
Thus, the composition of the coal varied more widely than
desired. The sulfur content varied from 0.9 to 1.6%, and the
ash content varied from 14.5 to 15.3%. The average concentra-
tions were as follows:
Sulfur, 1.2%
Ash, 15.3%
Analyses of fly-ash samples collected at the precipitator
inlet in alundum thimbles yielded more consistent results
than analyses of the coal. A representative composition is
given in Table 20. The most striking aspect of this
45
-------
Table 20. COMPOSITION OP FLY ASH
FROM THE BULL RUN PLANT
(August 1972)
Component
Li2O
Na20
K2O
MgO
CaO
A1203
Fe2O3
Si02
TiOa
P20s
Weight percentage
0.04
0.23
2.6
0.82
0.77
30.7
6.8
53.2
1.9
0.20
composition, as in the composition of the fly ash from
Plant 6, is the lower percentage of calcium oxide compared
with the percentages in ash from Western coals.
Sulfur-Dioxide Conversion Efficiency—
Institute personnel determined the efficiencies of conversion
of sulfur dioxide to the trioxide at different flow rates of .
sulfur dioxide through the catalyst. The procedure for deter-
mining the efficiency of conversion at a given flow rate con-
sisted of collecting the effluent sulfur trioxide in an air-
cooled condenser and an impinger-bubbler filled with water
and then collecting the residual sulfur dioxide in a bubbler
filled with aqueous hydrogen peroxide. Both the sulfur tri-
oxide and the sulfur dioxide were thus collected as sulfuric
acid and determined in this form. Conversion efficiencies
were approximately 80 and 70%, respectively, at the flow rates
of sulfur dioxide produced with volume concentrations of 6.5
and 9.2% in the carrier stream of air flowing through the
catalyst. Concentrations of sulfur trioxide thus introduced
into the flue gas were calculated as 32 and 40 ppm at the
estimated flow rate of the flue gas.
Precipitator Efficiency—
TVA personnel completed two determinations of the precipitator
efficiency while sulfur trioxide was being injected and the
plant was operating at full load. The results were 98.7% with
32 ppm of sulfur trioxide injected and 98.4% with 40 ppm
injected. Several days later after the injection had been
discontinued, another determination yielded essentially the
same result, 98.6%.
46
-------
There is a good deal of uncertainty about the value of the
efficiency that would have been representative of the preci^i-
tator performance without conditioning. The value of 98.6%
is suspect because of reports that the effects of sulfur
trioxide conditioning may persist for several days after the
injection is discontinued. The values estimated for full-
load conditions on the basis of previous tests with low-sulfur
coals33 are of doubtful validity because, first, they were
based on extrapolations of results for reduced-load conditions
and, second, the extrapolations were made for a somewhat
higher gas temperature, approximately 135°C rather than 125°C.
TVA's conclusion was that sulfur trioxide injection improved
the efficiency of the precipitator significantly even though
the degree of improvement was difficult to express quantita-,
tively.34 This conclusion was strengthened by the results
obtained with the light-obscuration devices.
Fly-Ash Resistivity and jpther Properties—
The results of determinations of fly-ash resistivity are given
in Table 21. All of these data were obtained with the point-
plane probe and with the electric field in the fly-ash sam-
ples near the breakdown level (between 10 and 20 kV/cra). The
values of resistivity found with the unconditioned ash were
all around 3 x 1010 ohm cm, and the values found with either
32 or 40 ppm of sulfur trioxide injected were approximately
three orders of magnitude lower, around 3 x 107 ohm cm.
Table 21.
ELECTRICAL RESISTIVITY OF FLY ASH
AT THE BULL RUN PLANT
(August 1972)
Injected SO3
concn, ppm
0
32
40
Resistivity,3
ohm cm
4.4 x 10 lo
2.4 x 10 lo
1.2 x 1010
3.3 x 10 10
1.3 x 107
5.0 x 107
3.0 x 107
Temperature, about 125°C.
47
-------
The resistivity data indicate that either of the concentra-
tions of sulfur trioxide injected was excessive. The results
obtained by TVA in later efficiency tests, it is understood,
confirm this conclusion, indicating that the optimum concen-
tration is nearer 20 ppm.
The results of analytical experiments with slurries of fly
ash in water are given in Table 22. The samples of fly ash
investigated in these experiments had been collected in alun-
dum thimbles at the inlet and the outlet of the precipitator
under isokinetic conditions. The pH data show that the uncon-
ditioned ash was acidic and the conditioned ash was even more
acidic. The concentrations of sulfate dissolved from the ash
show that marked increases in the amount of sulfate on the ash
occurred as the result of conditioning. Comparison of data
for samples from the inlet and outlet of the precipitator
indicates that the samples from the outlet were more acidic
and contained more sulfate than those from the inlet. This
difference undoubtedly reflects the difference in surface
area per unit of weight of ash at the two locations. The ash
collected at the outlet was smaller in particle size and thus
had a larger area-to-mass ratio.
Table 22. CHEMICAL PROPERTIES OF FLY ASH
AT THE BULL RUN PLANT (AUGUST 1972)
Injected SO3
concn , ppm
0
32
40
Sampling location
Precipitator inlet
outlet
Precipitator inlet
outlet
Precipitator inlet
outlet
Ash properties3
PH
4'fb
3.9
3.7
4.0
3.6
Sulfate, wt %
0.53
_b
2.24
7.53
2.60
8.83
.For samples collected at a temperature of 125°C.
Not determined because of the inadequate quantity of ash
sampled.
Flue-Gas Concentrations—
Concentrations of sulfur oxides and water vapor in the flue
gas at the inlet to the precipitator are listed in Table 23.
The data in this table that are of the greatest,interest are
the concentrations of sulfur trioxide. The concentrations of
48
-------
Table 23. CONCENTRATIONS OF FLUE GASES
AT THE BULL RUN PLANT (AUGUST 1972)
Injected SO3
concn, ppm
0
32
40
Concentrations^
SO 3 1 ppm
1.4
1.5
2.1
1.3
15.8
13.4
11.2
9.2
9.0
8.6
.11.0
8.6
SO2 , ppm
937
774
792
846
983
994
1074
1010
974
1010
855
892
H20, %
9.1
9.2
-
-
9.2
8.8
—
—
_
-
9.5
***
For samples collected at the precipitator inlet
at a temperature of 125°C.
sulfur trioxide found when there was no injection of the con-
ditioning agent were around 1 to 2 ppm. The concentrations
found when the conditioning agent was injected were higher
but not exactly consistent with the rates of injection. On
one occasion about 9 ppm was found when the injection rate
was 32 ppm, and; on another occasion about 13 ppm was found at
the same injection rate. Approximately 10 ppm was found when
the injection rate was increased to 40 ppm.
Obviously, a substantial fraction of the injected sulfur tri-
oxide—one-fourth to one-third—was always found in the gas
phase. However, the fraction remaining in the gas phase was
lower than that at Plant 6, which had a fly ash of similar
composition (see Tables 16 and 20). Probably the difference
in the relative amounts of sulfur trioxide collected by the
fly ash and allowed to remain in the gas phase was caused by
the difference in temperature, which was lower at the Bull
Run plant by about 35°C. Deposition of sulfur trioxide on
the fly ash at the Bull Run plant could have been forced to
go nearer to completion as the result of condensation with
water vapor as sulfuric acid at the comparatively low flue-
gas temperature. Deposition of sulfur trioxide on the fly
ash at Plant 6 apparently occurred only as a result of an
adsorption process.
49
-------
Precipitator Electrical Data—
The configuration of the electrical sections in Collector IB
of the Bull Run plant is shown in Figure 4. The designation
of the transformer-rectifier set supplying power to each sec-
tion is also given in this figure.
TVA personnel maintained a detailed record of voltages and
currents in all of the eight transformer-rectifier sets of
the precipitator. The data compiled consisted of periodic
readings of meters in the precipitator control room that
registered primary voltages, primary currents, secondary cur-
rents, and spark rates. The data show that erratic variations
occurred in each of these parameters on an hour-to-hour basis
under all conditions of sulfur trioxide injection. They indi-
cate that the circuitry intended to control the power supplies
was unable to maintain constant values of any of the electri-
cal parameters—voltage, current, or spark rate. Hence, the
change in any one of these parameters as a result of a change
in injection conditions was not always easy to discern. Even
so, representative values of primary voltages, primary cur-
rents, and secondary currents in TR Sets 1LB and 1RB (powering
electrical sections adjacent to the inlet of the precipitator)
and in TR Sets 4LB and 4RB (for sections adjoining the outlet)
are listed in Table 24. These values are described as
"representative" inasmuch as each lies near the middle of the
range of the recorded data.
The data in Table 24 indicate that the effects of sulfur
trioxide injection were most clearly manifest by increases
in the values of currents. However, the data show several
anomalies, such as smaller increases in currents in all
transformer-rectifier sets but TR Set 4RB when the concentra-
tion of injected sulfur trioxide was increased from 32 to
40 ppm.
One member of the TVA staff used auxiliary instrumentation to
record secondary current versus secondary voltage in several
of the power supplies with and without sulfur trioxide 4-njec-
tion. The data for an inlet and an outlet set, TR Sets 1LB
and 4LB, are shown by the graphs in Figure 5. The curves in
this figure portray the variations that occurred in current
density (the ratio of measured current to known electrode
area) as manual adjustments were made to alter voltage. The
location of the upper terminus of each curve indicates the
current density and voltage maintained with the power supplies
under automatic control. For each of the curves plotted for
experiments without conditioning, the change in slope to a
negative value corresponds to the onset of moderate to heavy
sparking. The absence of a change in slope for each of the
other two curves indicates that sparking was suppressed dur-
ing conditioning.
50
-------
GAS FLOW
TR 4LB
TR 4RB
TR 3LB
TR 3RB
TR 2LB
TR 2RB
TR 1LB
TR 1RB
Figure 4. Schematic diagram of electrical
sections in Collector IB at the
Bull Run plant
51
-------
Table 24. PRECIPITATOR ELECTRICAL DATA FROM THE
BULL RUN PLANT (AUGUST 1972)
TR
Seta
1LB
1RB
4LB
4RB
Injected SOs
concn, ppm
0
32
40
0
32
40
0
32
40
0
32
40
Primary
voltage, V
310
315
310
320
350
320
345
330
350
385
420
425
Primary
current, A
59
72
65
69
75
62
34
40
37
53 :
68
74
Secondary
current, mA
350
420
380
365
445
405
330b
370b
380b
340
430
465
Locations of the electrical sections powered by these
transformer-rectifier sets are shown in Figure 4.
The recorded values were only one-half of those listed.
However, because one of the two bushings was grounded and
only one-half of the electrode area was in service, the
recorded values were doubled to permit a comparison with
those following for TR Set 4RB.
The current-voltage data shown in Figure 5 are qualitatively
and also semiquantitatively consistent with the observed
decrease in resistivity that was produced with sulfur trioxide
injection. To interpret the data in a quantitative sense,
consider the shift in voltage at a current density of
10 nA/cm2 for TR Set 1LB with the conditioning agent added:
about 2 kV to the left along the voltage axis. Furthermore,
assume that the thickness of fly ash deposited on the elec-
trodes energized by this set was the same for both conditioned
and unconditioned ash. The thickness of the fly-ash deposit
can then be calculated from the equation:
52
-------
Ol
40
I 3°
H
g 20
8 10
0
25
32 PPM
OF S03
TR 1LB
30 35
VOLTAGE, kV
40
80
60
40
20
0
25
\
32 PPM
OF SO 3
TR 4LB
L
30 35
VOLTAGE, kV
40
Figure 5. Current density versus voltage in Collector IB
of the Bull Run plant (August 1972)
-------
t = AV/j(PU-PC)
where t - thickness (cm)
AV » voltage shift (2 x 103 V)
j = current density (10 x 10~9 A/cm2)
p = resistivity of unconditioned ash
u (ca. 3 x 1010 ohm cm)
p = resistivity of conditioned ash
c (ca. 3 x 107 ohm cm)
Solving the equation yields t = 6.7 cm. At first glance, this
appears to be an absurdly high value compared with the wire-
to-plate spacing of 11.4 cm. However/ if the^probable lack
of uniformity in current density and the possible errors in
the resistivity values are taken into account, the calculated
thickness does not appear unreasonably large.
Results of the Second Investigation (1974)
The second investigation of conditioning at the Bull Run
plant was carried out during July of 1974. The enlarged con-
ditioning system treating the fly ash entering all four pre-
cipitators was in operation at that time; however, as during
the preceding investigation in 1972, all of the experimental
work was concerned with the precipitator referred to as
Collector IB.
There were two purposes of the second investigation at the
Bull Run plant. One objective was to determine the extent of
rapping reentrainment with variations in the electrical resis-
tivity of the ash (adjusted by use of the conditioning system)
and with variations in some of the rapping parameters (inter-
val between rapping events, power used to actuate the vibra-
tory rappers, and duration of the rapper vibrations). The
second objective was to determine the change in the resis-
tivity of the ash produced with sulfur-trioxide injection at
a lower level than before (around 15 ppm instead of 32 to
40 ppm) and make a comprehensive study of the distribution of
the injected sulfur trioxide between the flue gas and the fly
ash entering and leaving the precipitator. Attainment of
both experimental objectives was aided by TVA personnel.
Members of the TVA staff determined ash concentrations at the
inlet and the outlet of the precipitator to be used in calcu-
lations of precipitator efficiencies; they also recorded elec-
trical data for various sections of the precipitator. The
results of the studies dealing primarily with rapping
54
-------
reentrainment will be presented in a separate report that is
to be submitted to TVA; the results of the studies more !
directly concerned with the conditioning process are given in
this report.
Coal and Fly-Ash Compositions—
Coal from an unspecified source was burned as a fuel during
the two weeks*'Of the investigation carried out in July of
1974. The sulfur content of this fuel proved to be quite con-
sistent throughout this period, averaging 0.82% by weight
with minimum and maximum values of 0.74 and 0.95%. Analytical
data obtained by TVA for 19 samples of coal collected during
the two-week period are shown by the following averages from
individual analyses:
Sulfur, 0.8%
Ash, 15.9%
Moisture, 7.5%
Heat value, 6,200 cal/g
(11,100 Btu/lb)
Samples of fly ash collected on four occasions with alundum
thimbles at the inlet of the precipitator were analyzed to
determine their overall compositions. Two of the samples
analyzed were collected during the first week of the investi-
gation when there was no injection of sulfur trioxide. The
other two samples were collected during the second week when
sulfur trioxide was injected to produce an approximate concen-
tration of 1.5. ppm. Each analysis showed approximately the
same composition, which is indicated by the average weight
percentages of the several constituents that are given in
Table 25. The composition was very similar to that found dur-
ing the 1972 investigation (Table 20).
Sulfur-Dioxide Conversion Efficiency—
As indicated in the previous paragraph, the investigation dur-
ing the first week was carried out without any injection of
sulfur trioxide, whereas the investigation during the second
week was carried out with the conditioning agent injected to
produce a concentration of about 15 ppm in the flue gas.
This approximate value of the concentration was obtained by
regulating the flow.rate.of sulfur dioxide to the converter,
as indicated by a calibrated monometer. It was assumed that
about 80% of the sulfur dioxide would'be converted to the
trioxide and that the normal flow rate of flue gas to the
precipitator would dilute the sulfur trioxide to the concen-
tration desired.
55
-------
Table 25. COMPOSITION OF FLY ASH
FROM THE BULL RUN PLANT
(July 1974)
Component
LizO
Na20
K20
MgO
CaO
A1203
FeaOa
Si02
TiOz
Pa05
Volatiles
Weight percentage
0.07
0.27
2.7
0.76
0.38
29.8
3.9
52.7
1.88
0.18
7.1
To permit a more accurate estimate of the injected concentra-
tion of the conditioning agent, the actual efficiency of the
conversion of sulfur dioxide was determined experimentally.
The result of the determination confirmed that the conversion
efficiency was about 80%. Other analytical information indi-
cated that at the selected flow rates of sulfur dioxide and
dilution air to the converter the concentration of sulfur
dioxide entering the converter was 4.4% by volume.
Determinations of the flow rate of flue gas at the outlet of
the precipitator (a part of the study of the efficiency of
the precipitator) showed that the flow rate was somewhat
higher than the value originally estimated. Thus, although
the efficiency of conversion of sulfur dioxide to the trioxide
was the expected value, the calculated concentration of sulfur
trioxide injected was somewhat lower than the intended value—
about 14 ppm rather than 15 ppm.
Precipitator Efficiency—
During the first week of the investigation, TVA personnel
made a series of determinations of the precipitator efficiency
without sulfur-trioxide injection but with several changes in
rapping parameters. During the second week, another series
of determinations was made with sulfur-trioxide injection and
with changes in rapping parameters that were similar to those
made during the first week. Data from the efficiency deter-
minations are summarized in Table 26. The information
included in this table consists of the concentration of SO3
used for treating the fly ash, the conditions under which the
56
-------
Table 26. PRECIPITATOR EFFICIENCIES AT THE
BULL RUN PLANT (JULY 1974)
Injected
SOs concn,
ppm
0
14
Test
2E
5E,,
8E
12E
10E
14E
23E
27E
28E
16E
Rapping3
Interval
; N
,• 2N
N
2N
- He
N
2N
2N
2N
- Nc
Power
N
N
N/2
N/2
) rapp:
N
N
N
N/2
3 rapp:
Duration
N
N
N
N
.ng -
N
N
N/6
N/6
ing
Inlet ash
concn/ g/m3
(_gr/ft3)
17.1 (7.43)
16.5 (7.19)
20.4 (8.85)
18.0 (7.81)
19.6 (8.53)
17.0 (7.39)
17.2 (7.48)
20.0 (8.69)
19.3 (8.37)
16.0 (6.97)
Efficiency,
%
69.7
78.4
75.0
77.4
82.0
93.5
97.4
92.0
92.4
96.8
aN indicates normal interval, power, or duration. Multipliers
and divisors of N designate how the intensity of rapping was
reduced.
rappers were operated, the observed concentration of fly ash
at the inlet of the precipitator, and the calculated effi-
ciency of ash collection.
With no injection of sulfur trioxide and with the rappers
operating normally, the observed efficiency was surprisingly
low, only 69.7%; compared with efficiencies determined sev-
eral years earlier. With 14 ppm of sulfur trioxide injected
and the rappers again operating normally, the efficiency
increased to 93.5% but it was still below the level found in
1972 with higher concentrations of the conditioning agent
(see p. 46).
The effects of changes in rapping conditions on precipitator
efficiency will be discussed in detail in another report.
However, in connection with the role of sulfur trioxide as a
conditioning agent, it is noteworthy that extending the inter-
val between rapping events, reducing the intensity of rapper
vibrations, or discontinuing rapping altogether caused a
significant improvement in efficiency in the absence of condi-
tioning. On the other hand, it is evident that none of the
changes in rapping conditions caused a comparable relative
improvement in efficiency when sulfur trioxide was used for
conditioning the fly ash. At least part of the effectiveness
of sulfur-trioxide conditioning appears attributable, there-
fore, to a suppression of rapping losses of collected ash.
57
-------
In further consideration of the role of sulfur trioxide as a
conditioning agent, it is of interest to analyze data showing
how the efficiency of fly-ash collection varied across the
precipitator as an apparent result of the gradient in tempera-
ture. The data in Table 27 are from four representative
efficiency tests in which efficiency values were calculated
for each of the four sampling ports at the precipitator out-
let. Here, the flue-gas temperature ranged from 112 to 135°C.
This temperature gradient was similar to that at the precipi-
tator inlet and was caused by the Ljungstrom air preheater
upstream from the precipitator. In two tests without injec-
tion of sulfur trioxide, there was no clear-cut relationship
between efficiency and gas temperature. In two tests with
injection of sulfur trioxide, on the other hand, there was a
strong indication of decreasing efficiency with increasing
temperature. It is reasonable to conclude that this relation-
ship is the result of decreasing uptake of the sulfur trioxide
by the ash with increasing gas temperature. This conclusion
is supported by the results of fly-ash gas analyses given
later in this report.
Table 27. PRECIPITATION EFFICIENCY AS A FUNCTION OF
GAS TEMPERATURE AT THE BULL RUN PLANT (JULY 1974)
Injected
SO 3 concn,
ppm
0
15
Test
2E
10E
14E
16E
Rapping
Normal
None
Normal
None
Efficiency, %,a vs .
outlet gas temp
112°C
71.4
81.9
96.9
97.6
120°C
75.3
85.4
95.5
97.8
130°C
64.8
78.6
91.8
96.8
135°C
67.7
82.7
90.7
95.0
Calculated by use of the average concentration of fly ash
entering the precipitator.
Indirect indications of dust concentrations evolved from the
precipitator were obtained on a continuous basis with a Lear-
Siegler instrument for determining the opacity of the gas
stream at the outlet of the precipitator. The light-
obscuration data recorded with this device are of interest in
several respects. The recorder-pen fluctuations were greatly
suppressed during sulfur trioxide injection, indicating that
the quantity of ash reentrained during rapping events was
suppressed as a result of conditioning. The range of pen
locations was shifted significantly downscale by injection,
showing that there was a significant reduction in the time-
average concentration of ash evolved from the precipitator as
58
-------
well as a reduction in the concentration immediately following
each rapping event. The rate of pen response as the rate of
injection of sulfur trioxide was altered indicated that both
short- and long-term changes occurred in the ash concentra-
tion. When injection was started, for example, the pen
quickly began to move downscale, but the shift in its location
and the reduction in its oscillations continued over a period
of several hours.
..4-f"-
Fly-Ash Resistivity-—
The point-plane resistivity probe was inserted into one of
the sampling ports at the inlet precipitator where the gas
temperature was near the mean value (about 125°C) and used to
determine the electrical resistivity of the fly ash with and
without conditioning. The results of two series of determina-
tions of resistivity at electric fields near the breakdown
strength of the collected ash are given in Table 28. The
resistivity of unconditioned ash was approximately 2 x 10lo
ohm cm, and the value for ash conditioned with sulfur trioxide
was significantly lower as expected, approximately 3 x 10V
ohm cm. These data may be compared with the results obtained
during the earlier investigation at the Bull Run plant
(page 47). For unconditioned ash, the resistivity values were
essentially the same in both investigations. For conditioned
ash, the resistivity value with 14 ppm of sulfur trioxide
injected was approximately two orders of magnitude higher than
the values found previously with 32 or 40 ppm of the condi-
tioning agent injected. Comparison of the resistivity from
the studies on separate occasions appears justified on the
basis of the similarity of the compositions of the fly ash
(Tables 20 and 25) .
Table 28. ELECTRICAL RESISTIVITY OF FLY ASH
AT THE BULL RUN PLANT (JULY 1974)
Injected S03
concn, ppm
0
14
Resistivity,9
ohm cm
1.2 x 1010
1.1 x 10 10
3-2 x 10 10
3.1 x 109
4.0 x 109
3.2 x 109
aTemperature, about 125°C.
59
-------
Fly-Ash Properties in Aqueous Slurries—
Values of the equilibrium pH and the weight percentage of
soluble sulfate were determined for a series of fly-ash sam-
ples from two sources: (1) alundum thimbles used at the inlet
of the precipitator for the efficiency tests and (2) hoppers
under the precipitator receiving the ash that was removed
from the gas stream during the efficiency tests.
Samples collected in thimbles from each of the sampling ports
during one test without sulfur trioxide injection and during
a second test with injection were analyzed individually. The
results of the analyses are given in Table 29. Other informa-
tion in this table consists of the flue-gas temperatures at
which the samples were collected. The analytical data for
the unconditioned ash show no systematic variation of either
pH or sulfate concentration with sampling temperature or loca-
tion (inlet or outlet of the precipitator). The data for the
conditioned ash, on the other hand, show consistently lower
pH values and higher sulfate concentrations at the outlet
compared with the inlet. Moreover, the sulfate concentrations
for the conditioned ash from the outlet show a marked decrease
as a result of an increase in temperature.
Table 29. CHEMICAL PROPERTIES OF INDIVIDUAL SAMPLES
OF FLY ASH COLLECTED IN THIMBLES AT THE
BULL RUN PLANT (JULY 1974)
Injected SO 3
concn , ppm
0
14
Test
2E
14E
Inlet samples
Temp,
°C
116
122
126
137
112
122
126
136
pH
4.9
5.1
4.8
5.0
4.5
4.5
4.5
4.5
Sulfate,
wt %
0.29
0.22
0.29
0.26
0.47
0.44
0.47
0.44
Outlet samples
Temp,
°C
112
121
130
135
112
121
130
135
pH
5.0
4.9
4.8
4.8
4.2
4.2
4.2
4.2
Sulfate,
wt %
0.26
0.22
0.29
0.29
1.97
1.68
1.35
1.35
Analytical data for composites of samples collected in thim-
bles from several efficiency tests with and without condition-
ing are given in Table 30. Information included in this table
are calculated concentrations of sulfur trioxide in the gas
phase that are equivalent to the concentrations of sulfate in
60
-------
Table 30. CHEMICAL PROPERTIES OF COMPOSITE SAMPLES
OF FLY ASH COLLECTED IN THIMBLES AT THE
BULL RUN PLANT (JULY 1974)
Injected SO3
concnf ppm
4
0
14
Test
,, •
2E
5E
• 8E
12E
14E
23E
27E
Inlet samples^
pH
5.0
4.3
4.7
4.8
4.5
4.3
4.4
Sulfate,
wt %
0.26
0.44
0.22
0.22
0.46
0.58
0.51
Equiv
S03,b
ppm
10.4
16.9
10.4
11.5
18.0
23.2
23.7
Outlet samples3
PH
4.9
4.8
4.8
4.8
4.2
4.1
4.1
Sulfate,
wt %
0.28
0.33
0.33
0.22
1.49
2.02
1.47
Equiy
S03,b
ppm
3.3
2.7
3.9
2.6
3.7
2.3
5.9
Composite samples for a-mean gas temperature of about
,125°C,
Calculated as the concentration of SO3 in the gas phase
that would be equivalent to the concentration of sulfate
in the ash.
the ash. Calculations of these hypothetical concentrations
of sulfur trioxide in the gas phase were made by use of the
observed concentrations of sulfate in the ash and the experi-
mentally determined concentrations of fly ash at the inlet
and outlet of the precipitator. In general, the data in
Table 30 indicate the following:
• pH was lowered substantially by the injection of
sulfur trioxide, particularly for outlet samples.
• Sulfate concentrations were increased markedly by
the injection of sulfur trioxide, again especially
for outlet samples.
• Sulfur trioxide concentrations in the gas phase
equivalent to sulfate concentrations in the ash
were much higher for samples at the inlet during
injection! of the conditioning agent. Thus, they
reflect the expected removal of at least part of
the injected agent from the gas stream by the ash.
• Sulfur trioxide concentrations calculated for out-
let samples do not show much change as a result of
sulfur trioxide injection. The lower concentra-
tions of ash at the outlet during injection offset
the higher concentrations of sulfate in the ash.
61
-------
Analytical data given in Table 30 for the inlet sample from
Test 5E are anomalous. The data indicate much more uptake of
sulfur trioxide than the data for other samples collected
without conditioning. The first analysis of the sample was
repeated to determine whether a serious experimental error
had been made; evidently, there was no significant error.
Perhaps the inlet samples for Test 5E were incorrectly labeled;
the samples analyzed may have been from another test that was
carried out during sulfur trioxide injection.
The results of analyses of composites of the hopper samples
are given in Table 31. The concentrations of sulfur trioxide
in this table are based on the observed concentrations of
sulfate in the ash and the calculated concentrations of ash
that were precipitated and deposited in the hoppers. The
latter values were calculated from the observed concentrations
of ash entering the precipitator and the experimentally deter-
mined efficiencies of precipitation. All of the data in
Table 31 reflect a marked increase in the collection of sulfur
trioxide by the ash during the injection of this compound.
Table 31. CHEMICAL PROPERTIES OF COMPOSITE SAMPLES
OF FLY ASH FROM PRECIPITATOR HOPPERS
AT THE BULL RUN PLANT (JULY 1974)
Injected S03
concn , pj>m
0
14
Test
2E
5E
8E
12E
14E
23E
27E
PH
5.1
5.0
5.1
5.0
4.5
4.5
4.5
Suifate,
wt %
0.16
0.19
0.15
0.19
0.31
0.39
0.43
Equiv SO 3 ,**
ppm
4.4
5.7
5.3
7.7
11.4
14.7
19.4
Calculated as the concentration of SO3 in the gas
phase that would be equivalent to the concentration
of sulfate in the ash.
Data from Tables 30 and 31 are combined in Table 32 to permit
a comparison of the analytical results with the mass conserva-
tion relationship expressed as follows:
The combined quantities of sulfate in ash from out-
let thimbles and precipitator hoppers should be
equal to the quantity of sulfate in ash from inlet
thimbles.
62
-------
Table 32. CONCENTRATIONS OF SULFUR TRIOXIDE EQUIVALENT
TO THE CONCENTRATIONS OF SULFATE IN FLY ASH
AT THE BULL RUN PLANT (JULY 1974)
Injected. SO 3
concn, ppm
0
14
Test
2E
8E
12E
(av)
14E
23E
27E
(av)
Equiv SO 3 concn, ppm
Thimble samples
Inlet
10.4
10.4
11.5
W78
18.0
23.2
23.7
21.6
Outlet
3.3
3.9
2.6
3.3
3.7
2.3
5.9
4.0
Hopper
samples
4.4
5.3
7.7
578
11.4
14.7
19.4
15.2
Difference3
2.7
1.2
1.2
TTT
2.9
6.2
-1.6
2.5
Calculated by subtracting the sum of the outlet and
thimble samples from the value for the inlet sample.
A test of this relationship is provided by the concentrations
of sulfur trioxide equivalent to the quantities of sulfate in
the ash, as summarized in Table 32. If the relationship were
satisfied by the experimental data, all of the values in the
last column of the table would be zero. Obviously, the rela-
tionship is not satisfied precisely. Even so, the precision
in the analytical data appears to be generally acceptable,
especially when examined in terms of the averages in sulfur
trioxide concentrations. Average values of the difference in
sulfate found in the inlet samples and in the combination of
outlet and hopper samples represents 10 to 20% of that found
in the inlet samples. A possible explanation for the discrep-
ancy in the results is that in collecting fly ash in alundum
thimbles part of the sulfur trioxide actually present in the
gas phase within the flue-gas duct is deposited on the col-
lected ash. This source of error would probably be more
significant at the inlet of the precipitator than at the out-
let because of the greater quantity of ash collected.
The data in Table 32 permit estimates to be made of the quan-
tity of the injected sulfur trioxide that was collected by
the fly ash. Based on the differences in the average calcu-
lated concentrations of sulfur trioxide deposited on the ash
with and without injection, the estimates are as follows:
• Inlet samples—(21.6-10.8) =10.8 ppm
• Outlet and hopper samples—(19.2-9.1) = 10.0 ppm
63
-------
The first value represents 77% of the injected sulfur trioxide;
the second value represents 72% of the total.
Flue-Gas Concentrations—
Concentrations of sulfur trioxide, sulfur dioxide, and water
vapor in the flue gas at the inlet and the outlet of the pre-
cipitator were determined at several sampling ports at each
location. As discussed in the following paragraph, the con-
centration of sulfur trioxide depended upon where the sampling
port was located in either of the sampling planes; as expected,
it also depended on whether sulfur trioxide was being injected
to condition the fly ash. The concentrations of sulfur diox-
ide and water, on the ohter hand, were essentially independ-
ent of these factors; they appeared to vary randomly about
average values of 708 ppm and 9.8% by volume, respectively.
The observed concentrations of sulfur trioxide at the inlet
and the outlet of the precipitator are plotted in Figure 6 as
functions of the sampling locations, expressed as fractions
of the distance across the gas stream from one vertical wall
of each duct to the opposite wall. Curves portraying the
temperature profiles across the two ducts are also plotted in
the figure. The concentrations of sulfur trioxide found both
with and without injection obviously decreased as the gas
temperature decreased; thus, concentration profiles are shown
by curves that are, in general, parallel to the temperature
profiles. (In the lower range of gas temperatures, the con-
centrations of sulfur trioxide without injection were less
than 1 ppm and thus were too low for reliable determination;
in the graph, these concentrations are arbitrarily plotted as
0.5 ppm.)
Areas under the curves portraying the concentrations of sul-
fur trioxide indicate that the average concentrations of sul-
fur trioxide were as follows:
Without injection—
Inlet, 1.2 ppm
Outlet,1.0 ppm
With injection—
Inlet, 5.8 ppm
Outlet, 5.6 ppm
The apparent differences between inlet and outlet concentra-
tions with or without injection are not large enough to
represent real differences. Thus, it is reasonable to con-
clude that the concentration of sulfur trioxide was essen-
tially the same on both sides of the precipitator (about 1 ppm
64
-------
u
o
w
tf
D
i
W
Pk
U
o
a
w
w
150
140
130
120
110
150
140
130
120
110
T I IT
II
OUTLET
SO3 WITH
INJECTION
SO 3 WITHOUT
INJECTION
I i I
10
4 §
o
en
i i
I I
O
INLET
. SO3 WITH
\ INJECTION
SO 3 WITHOUT
INJECTION
a ~
I I
10
z;
o
«.§
0 0.2 0.4 0.6 0.8 1.0
FRACTION OF DISTANCE ACROSS GAS DUCT
Figure 6. Concentration of sulfur trioxide as
a function of gas temperature at the
Bull Run plant (July 1974)
65
-------
without injection or about 6 ppm with injection) and that the
concentration in either location was increased by essentially
the same amount as a result of injection (about 5 ppm) . The
concentration of sulfur trioxide not collected by the fly ash
and thus allowed to pass from the precipitator to the stack
and thence to the atmosphere represents about 36% of the con-
centration injected upstream from the precipitator (calculated
as 14 ppm).
Analyses of fly ash discussed previously in this report indi-
cate that 10 to 11 ppm of the injected sulfur trioxide was
collected by the fly ash. Thus/ combined analyses of the ash
and the gas accounted for 15 to 16 ppm of injected sulfur
trioxide, slightly in excess of the 14 ppm calculated as the
total concentration injected. The discrepancy seems gratify-
ingly small in view of the numerous potential sources of
analytical error.
Precipitator Electrical Data—
Electrical data for the power supplies of the precipitator
under automatic control are compiled in Table 33. The data
consist of values of primary voltage, primary current, secon-
dary current, and spark rate; the data were computed by aver-
aging meter readings during a series of efficiency tests,
first without sulfur trioxide injection and then with injec-
tion. With a few exceptions, the injection of sulfur trioxide
increased voltages and currents and decreased the spark rate.
These changes are all consistent with the observed decrease
in fly-ash resistivity, but they are not necessarily indica-
tive of this effect alone as discussed below.
Electrical data obtained with the power supplies under manual
control and with the primary voltage thus variable are shown
in Figure 7. Here, the data show current densities (calcu-
lated from recorded values of secondary current) as functions
of secondary voltage in an inlet and an outlet electrical
field of the precipitator. The segments of the curves with
positive slopes portray the data obtained with no sparking or
very light sparking; the segments with negative slopes in
regions of high current density represent the experimental
results with moderate to heavy sparking. The short horizontal
lines intersecting each curve indicate the average values of
current density observed with the power supplies under auto-
matic control (calculated from the values of secondary current
given in Table 33).
The aspect of the data in Figure 7 that appears to be of
major significance is the indication that the injection of
sulfur trioxide permitted both higher current densities and
higher voltages to be reached without the occurrence of exces-
sive sparking, with the reason being some effect other than
66
-------
Table 33. PRECIPITATOR ELECTRICAL DATA FROM THE
BULL RUN PLANT (JULY 1974)
TR
Set
1LB
1RB
2LB
2RB
3LB
3RB
4LB
4RB
Injected SOa
concn, ppm
0
14
0
14
0
14
0
14
0
14
0
14
0
14
0
14
Primary
voltage, V
215
285
175
225
250
320
275
335
285 -
325
320
385
260
300
280
360
Primary
current, A
16
32
14
16
22
37
33
29
19
23
30
34
13
17
15
17
Secondary
current, mA
150
165
160
140
150
210
200
175
110
130
145
165
100
110
120
105
Sparks
per min
230
105
265
150
180
150
120
60
210
150
140
80
220
165
270
70
the observed lowering of fly-ash resistivity. The shifts in
the current-voltage curves were to the right along the volt-
age axis, whereas shifts to the left would conform to the
lowering of resistivity as the only effect of the condition-
ing process. Shifts in the voltage curves to the right along
the voltage axis at least suggest the possibility of a space-
charge effect resulting from the introduction of less mobile
charge carriers in the gas stream. One possibility is that
the added concentration of sulfur trioxide (or the vapor of
sulfuric acid) assumed most of the ionic space charge and the
new ions thus introduced carried current with a lower mobil-
ity than the normally occurring ions produced from oxygen,
water vapor, and sulfur dioxide. An alternative possibility
is that part of the added sulfur trioxide was condensed as a
fine mist of sulfuric acid and then electrically charged,
causing a very pronounced shift in charge carriers from gas-
eous ions to relatively immobile acid particles.
67
-------
35
CO
e
>
B
H
CO
2
D
u
30
25
20
15
10
_x
\
v WITH INJECTION
\ (14 PPM OF S03)
24 28 32
VOLTAGE, kV
36
\*
\WITH
\ INJECTION
(14 PPM —I
\ OF S03)
WITHOUT
INJECTION
O
24 28 32
VOLTAGE, kV
Figure 7. Current density versus voltage in Collector IB
of the Bull Run plant (July 1974)
-------
There is a decided contrast between the effects of sulfur-
trioxide injection at the Bull Run plant during the first
investigation in 1972 and the later investigation in 1974
(compare Figures 5 and 7). During the first investigation,
sulfur trioxide injection caused a shift in the current-
voltage curves to the left on the voltage axis in a manner
consistent with a decrease in resistivity. During the second
investigation, the shift observed was to the right for dif-
ferent reasons'''such as the one described above. Another dif-
ference between the results of the two investigations was
that sparking required much higher current densities and
voltages in the first investigation than in the second. A
possible explanation for this difference is that the align-
ment of the corona wires and collection electrodes deterio-
rated significantly during the time between the two investi-
gations.
69
-------
SECTION VI
RESULTS OF STUDIES OF CONDITIONING WITH AMMONIA
The studies of conditioning with ammonia in TVA power plants
are classified in the following discussion as to the sulfur
content of the coal from which the fly ash originated. In
two power plants, ammonia was investigated as a^conditioning
agent for ash from low-sulfur coals. Furthermore, in one of
these two plants and in a third plant, ammonia was investi-
gated as a conditioning agent for ash from high-sulfur coals,
as in the original study of Reese and Greco.5 Both types of
coal were from the Eastern mines that supply the TVA system.
The fly ash produced by the low-sulfur coals differed signifi-
cantly from the ash evolved from the low-sulfur Western coals
that were involved in several of the studies of sulfur tri-
oxide conditioning discussed in Section V of this report; the
ash was more similar to that from the low-sulfur Eastern coals
burned during the studies of sulfur trioxide at Plant 6 and
at TVA's Bull Run plant. The major differences in ash prop-
erties were lower electrical resistivities and lower calcium-
oxide contents (and hence lower basicities). Associated with
these differences in ash properties were higher naturally
produced concentrations of sulfur trioxide in the flue gas
compared with the concentrations produced from low-sulfur
Western coals.
It might be expected that ammonia conditioning would occur by
substantially different mechanisms with the fly ash from low-
and high-sulfur coals. However, the nature of the combustion
products described above for the low-sulfur coals led to the
occurrence of similar conditioning phenomena with ash from
both low- and high-sulfur coals.
POWER PLANTS BURNING LOW-SULFUR COALS
Although TVA's initial experience with ammonia conditioning
had only involved fly ash from high-sulfur coal, the utility
undertook an evaluation of ammonia conditioning of ash from
70
-------
low-sulfur coals during 1972. The first investigation with
a low-sulfur coal was made at the Widows Creek plant, where
Reese and Greco's original investigation had been carried
out.5 The second investigation was made at the Bull Run
plant shortly after the initial trials of sulfur trioxide
conditioning were started at this plant (described in
Section V).
Widows Creek Plant, Unit 7
During brief occasions during June and July of 1972, low-
sulfur coal instead of the usual high-sulfur coal was burned
on a trial basis at the Widows Creek plant in Unit 7 (550 MW)»
In some of the experimental work, no ammonia was used. In
comparative experiments, however, ammonia injection downstream
from the air preheater was used to condition the ash. (Orig-
inally, ammonia was injected upstream from the air preheater;5
however, problems with preheater pluggage necessitated reloca-
tion of the injection manifold.)
TVA personnel determined the precipitation efficiency of fly-
ash collection with and without conditioning. They also
investigated the electrical behavior of the precipitator with
and without conditioning. Members of the Institute staff
obtained data on the properties of the fly ash and the flue
gas. The first experiments were conducted during June with
the precipitator designated as Collector 7A, but they were
discontinued before completion because of a boiler-tube leak.
The later more extensive experiments were conducted during
July with one of the other precipitators of identical design,
Collector 7B.
Coal and Fly-Ash Compositions—
The results of coal analyses for the experiments in June and
July were very similar. The following data are representative
of the coal burned during the two series of experiments:
Sulfur, 0.9%
Ash, 17.6%
Moisture, 5.4%
Heat value, 6,4.00 cal/g
(11,600 Btu/lb)
The results of a comprehensive analysis of the fly ash (col-
lected in an alundum thimble at the inlet of Collector 7B)
are given in Table 34. This analysis indicates that the mate-
rial followed the usual pattern of fly ash from Eastern coals
in containing little calcium oxide.
71
-------
Table 34. COMPOSITION OF FLY ASH FROM A
LOW-SULFUR COAL AT WIDOWS CREEK UNIT 7
(June-July 1972)
Component
Li2O
Na20
K20
MgO
CaO
A1203
Fe2O3
Si02
Ti02
P205
SO 3
Weight percentage
0.09
0.21
3.2
1.3
0.79
30.2
6.0
54.2
1.3
0.60
0.29
Precipitator Efficiencies—
Two experiments at Collector 7B with and without ammonia pres-
ent as a conditioning agent yielded efficiency data that were
considered reliable by the supervisor of the TVA research
team. One of these experiments was carried out with a gas
temperature of 143°C; the efficiencies calculated from inlet
and outlet fly-ash concentrations were 93.9% with no ammonia
injection and 98.5% with 13 ppm of ammonia injected. The
second experiment showed an even greater improvement in effi-
ciency at 132°C; the efficiency increased from 89.9% to 98.3%
with 10 ppm of ammonia injected.35
A^Bailey bolometer equipped with a recorder provided a real-
time record of the changes in the concentration of fly ash
emitted from the precipitator. Usually, the recorder
responded rapidly to changes in the rate of ammonia injection,
The recording for one day of experimentation yielded the fol-
lowing results:
• No ammonia injection. The pen fluctuated rapidly and
irregularly between scale positions of approximately
22 and 48 on a range of 100.
• Ammonia injected for roughly 20 min intervals, first
at approximately 7.5 ppm and later at approximately
10.0 ppm. Injection at the lower concentration
caused a rapid change in oscillation of the pen,
lowering the range of scale positions from ,22-48 to
20-38 and regulating the time interval between the
72
-------
high values (appearing as spikes) to about 5 min.
Subsequent injection at the higher concentration
caused further suppression of the range of scale
positions to 20-30.
• Ammonia injected for several hours with the concen-
tration increased from 10.0 to 13.1 ppm. The main
effect was further suppression of the range in pen
positions to 20-24.
• Ammonia injection discontinued. A rapid return of
the pen oscillation to the range and frequency first
noted was observed.
Close observation of the pen movement giving the spikes spaced
at 5-min intervals during ammonia injection showed that
approximately five surges in pen location produced each of the
spikes evident on the slowly moving recorder chart. Each of
these surges appeared to occur at the same time that the
collector electrodes in the outlet field of the precipitator
were rapped. Thus, one of the conclusions based on the bolom-
eter recordings is that the injection of ammonia was more
effective in aiding fly-ash removal in the inlet and middle
fields of the precipitator than in the outlet field, primarily
because of the rapping losses that occurred in this field.
Fly-Ash Resistivity and Other Properties—
The results of determinations of the resistivity of fly ash
at the inlets of Collectors 7A and 7B are given in Table 35.
All of these data were obtained by the use of the point-plane
probe and by the application of an electric field near the
breakdown value in each collected sample. There is no indica-
tion that the injection of ammonia at any concentration (up
to 14 ppm) caused a change in resistivity that was larger
than the uncertainty in each measurement.
The results of analytical experiments with samples of the fly
ash in aqueous slurries are given in Table 36. The pH data
indicate that each sample produced a mildly acidic slurry but
that the samples collected during ammonia injection were less
acidic than those collected without injection. The ammonia
concentrations appear to confirm the postulate that ammonia
was present either in the form of a discrete particulate (such
as ammonium sulfate) or a surface deposit on the fly ash. The
sulfate concentrations apparently varied randomly as changes
were made in ammonia injection.
73
-------
Table 35. ELECTRICAL RESISTIVITY OF FLY ASH FROM A
LOW-SULFUR COAL AT WIDOWS CREEK UNIT 7
(JUNE-JULY 1972)
Precipitator
7A
7B
7B
7B
Gas temp,
°C
138
132
143
149
Injected NH3
concn , ppm
0
14
0
10
0
13
0
9
Range of resistivity
values, ohm cm
(1.6-2.7) x 1011
(1.3-8.0) x 1011
(3.6-3.7) x 1011
(3.6-4.6) x 1011
(2.2-3.9) x 1011
(1.7-2.6) x 1011
(2.8-5.9) x 1011
(2.4-3.2) x -10 M
Table 36. CHEMICAL PROPERTIES OF FLY ASH FROM A
LOW-SULFUR COAL AT WIDOWS CREEK UNIT 7
(JULY 1972)
Gas
temp,3 °C
132
143
149
Injected NH3
concn, ppm
0
10
0
13
0
9
Sampling
location
Inlet
Outlet
Inlet
Outlet
Inlet
Inlet
Inlet
Outlet
Inlet
Outlet
Asn properties
pH
5.1
4.4
5.4
4.6
4.9
6.0
5.0
4.6
5.3
4.9
Ammonia ,
wt %
0.004
-
0.040
-
0.005
0.061
0.007
0.008
0.035
0.047
Sulfate,
wt %
0.27
0.28
0.27
0.44
0.35
0.27
0.16
0.41
0.25
0.44
Precipitator 7B.
74
-------
Flue-Gas Concentrations—
Concentrations of ammonia, sulfur oxides, and water vapor
found at the inlets to Collectors 7A and 7B are listed in
Table 37. The data for ammonia and sulfur trioxide are of
primary interest, as discussed below.
Table 37. CONCENTRATIONS OF FLUE GASES FROM A
LOW-SULFUR COAL AT WIDOWS CREEK UNIT 7
(JUNE-JULY 1972)
Precipitator
7A
7B
7B
7B
Gas
temp,
°C
138
132
143
149
Injected
NH3 concn,
ppm
0
14
0
10
0
13
0
9
Concentrations
NH3,
ppm
0.0
0.0
1.0
0.4
0.0
0.6
-
0.9
0.7
-
1.4
0.6
0.4
0.9
0.5
0.6
0.5
0.5
0.4
—
S03,
.EEB
2.9
1.7
0.2
0.3
-
4.0
5.0
0,8
0.9
0.6
7.2
5.0
5.9
1.5
1.3
—
5.1
4.2
1.5
0.9
S02,
ppm
617
621
556
626
-
565
560
554
549
528
676
670
543
638
679
—
578
577
578
599
H20,
%
«•
-
8.1
8.3
-
_
—
8.4
-
-
—
-
8.9
8.7
-
_
9.5
-
-
"•
Small concentrations of ammonia (less than 1 ppm) and somewhat
larger concentrations of sulfur trioxide (3 to 6 ppm) were
usually found without ammonia injection. The small concen-
trations of ammonia are attributed to slow bleeding of the
conditioning agent into the flue gas even when the valves in
the injection system were closed. The source of the ammonia
may have been residual gas in the injection lines between the
valves and the nozzles, or it may have been the cakes of
ammonia-containing fly ash encrusted around the injection
75
-------
nozzles (these aggregates sometimes become very large and are
known as "hornet nests"). The concentrations of sulfur triox-
ide found in the absence of deliberately injected ammonia are
in a range of surprisingly high values when they are compared
with the concentrations found in the flue gas from low-sulfur
Western coals (Section V). Undoubtedly, the lower basicity
of the ash at the Widows Creek plant was responsible for the
comparatively high background concentrations of sulfur triox-
ide.
In each instance, the injection of ammonia lowered the concen-
tration of sulfur trioxide. This observation is consistent
with the occurrence of increased adsorption of sulfur trioxide
on the surface of the fly ash as the acid on the surface is
neutralized by ammonia. It is also consistent with the reac-
tion of ammonia and sulfur trioxide (actually present as sul-
furic acid) to produce particles of ammonium sulfate or
ammonium bisulfate, as shown by the chemical equations on
page 17.
Precipitator Electrical Data—
The configuration of the electrical sections of Collector A
and the designations of the transformer-reactifier sets that
power each section are shown schematically in Figure 8. The
configuration of Collector B is identical to that of Collec-
tor A. The power supplies for Collector B are designated in
essentially the same manner as those for Collector A; the
only exception is the use of the term B in place of A.
During the brief study of Collector A, one member of the TVA
technical staff investigated the waveforms of secondary volt-
age and secondary current produced by TR Set 7A2 in one of the
two inlet sections of that precipitator. Peak values of each
electrical parameter between sparks were first determined
while ammonia was being injected at a concentration of 14 ppm;
the results were: maximum voltage, 60 kV; maximum current,
1.00 A. The peak values of each parameter were again deter-
mined within a few minutes after ammonia injection had been
discontinued; in this instance, the results were: maximum
voltage, 58 kV; maximum current, 1.25 A. The results of the
two determinations indicate that discontinuing ammonia injec-
tion rapidly brought about a decrease in the peak voltage and
an increase in the peak current. Both effects were consist-
ent with other observations of the electrical behavior of
Collectors A and B with and without ammonia injection, as
described subsequently.
During the study of Collector A, another type of experiment
yielded the relationships between the time-average-secondary
current and voltage produced by TR Set 7A2 under manual
76
-------
GAS FLOW
1
TR 7A5
TR 7A4
TR 7A3
TR 7A1
TR 7A2
Figure 8. Schematic diagram of electrical
sections of Collector 7A at the
Widows Creek plant
77
-------
control. The results of this experiment are presented in
Figure 9, with the observed values of current converted to
the corresponding values of current density. Here, with dust-
coated electrodes, the effect of ammonia injection was compa-
rable to the effect on peak current and voltage that was
observed with the set under automatic control—namely, a
suppression of current density at a given voltage or an
enhancement of voltage at a given current density.
During the subsequent study of Collector B, the information
obtained with and without ammonia injection included the
following: wave-form data for secondary voltages and currents
of TR Sets 7B1 and 7B5 (powering electrical sections adjacent
to the inlet and the outlet), time-average values of primary
voltages and currents in all five sets, and time-average
values of secondary voltages and currents of Sets 7B1 and ?B5
with clean electrodes. The data for these parameters are
given in Tables 38 and 39 and in Figure 10. Here, the experi-
mental data again indicate that ammonia injection tended to
suppress current and enhance voltage with the effect being
more pronounced in the inlet section of the precipitator.
Either of two interpretations may be given for the electrical
data. One interpretation is that the electrical resistivity
of fly ash on the collecting electrodes was higher with the
injection of ammonia than without injection. This interpreta-
tion, however, cannot be reconciled with either the rapid
change in the voltage and current waveforms or the shift in
the location of the current-voltage curve for clean electrodes
when injection was discontinued. The second interpretation,
which would be consistent with all of the data, is that the
electrical behavior of the gas stream rather than the depos-
ited fly ash was altered by changes in the injection of
ammonia. More explicitly, the second interpretation is that
a fine particulate produced by the reaction of ammonia with
the sulfur trioxide and water vapor normally present was elec-
trically charged in the precipitator. Transfer of charge
from molecular ions to small particles would lower the effec-
tive electrical conductivity of the gas stream, lowering the
current produced by a given voltage. Alternatively, charging
the small particles and simultaneously maintaining the origi-
nal current would enhance the electric field in the gas stream
through a space-charge effect.
Bull Run Plant, Collector C
The discussion of the initial trials of sulfur trioxide as a
conditioning agent at the Bull Run plant in Section V of this
report pointed out that trials of ammonia as a conditioning
agent were also made. Whereas sulfur trioxide was initially
injected into the flue gas entering Collector IB, ammonia was
78
-------
40
§ 30
w
Q
3
2o
10
25
NO NH3
INJECTED
NH3 INJECTED
TR 7A2
30 35
VOLTAGE, kV
40
45
Figure 9. Current density versus voltage in
Collector 7A of the Widows Creek plant
(low-sulfur coal, June 1972)
79
-------
Table 38. PEAK VALUES OF PRECIPITATOR SECONDARY VOLTAGES
AND SECONDARY CURRENTS AT WIDOWS CREEK UNIT 7
(LOW-SULFUR COAL, JULY 1972)
Gas
temp, °C
132
143
149
Injected NH3
concn , ppm
0
10
0
13
0
9
TR Set 7B1
Voltage,
V
54
58
55
57
55
56
Current,
A
1.10
0.90
1.15
0.80
1.15
0.95
TR Set 7B5 _
Voltage,
V
62
60
60
59
57
58
Current,
A
2.00
1.88
2.00
2.00
1.63
2.00
Table 39. PRIMARY VOLTAGES AND CURRENTS SUPPLIED BY THE
TRANSFORMER-RECTIFIER SETS AT WIDOWS CREEK UNIT 7
(LOW-SULFUR COAL, JULY 1972)
TR
Set
7B1
7B2
7B3
7B4
7B5
Injected NH3
concn,3 ppm
0
13
0
13
0
13
0
13
0
13
Voltage, V
260
280
230
270
230
245
300
295
270
270
Current, A
70
56
67
60
110
105
75
60
110
110
Temperature, 143°C.
80
-------
oo
H1
e
o
w
Q
2
W
S
D
U
50
40 —
•S
. 30
E-"
20
10
TR 7B1
NO NH3
INJECTED
.13 PPM
OF NH3
INJECTED/
/
30 35
VOLTAGE, kV
40
100
80 —
60
40
20
TR 7B5
25
NO NH3
INJECTED/ /
13 PPM
OF NH3
INJECTED
30 35
VOLTAGE, kV
40
Figure 10. Current density versus voltage in Collector 7B
of the Widows Creek plant (low-sulfur coal, July 1972)
-------
injected into the gas entering Collector 1C. The two precip-
itators are identical, having the sectionalization previously
shown in Figure 4. The transformer-rectifier sets energizing
the electrical sections of Collector C are denoted by the
same terms as those in Collector B except for the use of the
letter C in place of B.
In the investigation of ammonia conditioning at Collector 1C
of the Bull Run plant, all of the experimental approaches
that have been discussed in connection with the Widows Creek
plant were used. A lesser effort was put forth in completing
some types of experimental work, however, in order to devote
attention to a new procedure that was expected to confirm
that ammonia conditioning occurred through a space-charge
mechanism. This procedure involved the use of a condensation-
nuclei counter in conjunction with diffusion batteries for
detecting ultrafine particulate and classifying this material
by size. 6
Coal and Fly-Ash Compositions—
The experimental study of ammonia conditioning at the Bull
Run plant was conducted on two occasions, first in September
of 1972 and then in October of the same year. The experiments
during September were terminated sooner than expected (because
of a leaking boiler tube) before any sample of the coal had
been collected. During the initial experiments in September,
it was known that reclaim coal was used as the fuel; during
the later experiments the same month, Haddix coal was burned.
All of the experiments in October were conducted with Haddix
coal as the fuel.
Previous analyses showed that the reclaim coal had a sulfur
concentration varying between 0.9 and 1.6% and an ash concen-
tration varying between 14.5 and 15.3% (page 45). During
October, analyses of the Haddix coal yielded the following
data:
Sulfur, 1.2%
Ash, 17.0%
Moisture, 5.7%
Heat value, 6,170 cal/g
(11,100 Btu/lb)
Compositions of fly ash produced from reclaim coal and Haddix
coal are given in Table 40. The composition of the ash from
reclaim coal is a repetition of the composition originally
presented in Table 20. The composition of the ash from Haddi*
coal was determined by analyzing samples of ash from the
hoppers of Collector 1C during the experiments in October.
The two compositions are virtually identical.
82
-------
Table 40. COMPOSITIONS OF FLY ASH
FROM THE BULL RUN PLANT
(Ammonia Conditioning, September-October 1972)
Component
Li20
Na20
K20
MgO
CaO
A1203
Fe203
SiO2
Ti02
P205
Weight percentaae
Reclaim coal
0.04
0.23
2.6
0.82
0.77
30.7
6.8
53.2
1.9
0.20
Haddix coal
0.04
0.27
2.4
0.69
0.88
29.6
5.7
54.9
2.2
0.25
Precipitator Efficiencies—
Apparently, no comparison was made of the efficiencies of
Collector 1C with and without ammonia injection. A comparison
had been planned for the investigation during September, but
it had to be cancelled because of the unscheduled plant'outacre
Jf a comparison was made later, the results are not available "
for inclusion in this report.
During the work in both September and October, efforts were
made to obtain indirect indications of the effect of ammonia
injection on the precipitator performance by use of a Bailey
bolometer and recorder. The performance of the bolometer was
not altogether satisfactory because imperfect maintenance per-
mitted fly ash to accumulate on the lenses and thus to obscure
changes occurring in ash concentrations in the precipitator
Affluent. Even so, the recorder usually indicated that start-
ing ammonia injection decreased the opacity of the effluent
and discontinuing injection increased the opacity, with
response times being fairly brief after either change.
Fj-y-Ash Resistivity and Other Properties—
Fly-ash resistivity data were obtained with the point-plane
Probe and with the collected samples under approximately the
slectric field required for breakdown. All of these data
Were obtained during the initial experiments in September
While Haddix coal was used as the fuel. The data are listed
in Table 41. They indicate that the resistivity of the uncon-
ditioned ash was comparable to that of unconditioned ash from
83
-------
Table 41. ELECTRICAL RESISTIVITY OF FLY ASH
AT THE BULL RUN PLANT
(AMMONIA CONDITIONING, SEPTEMBER 1972)
Injected NH3
concn, ppm
0
7
Resistivity , a
ohm cm
3.1 x 10 10
2.2 x 1010
3.8 x 10 10
Temperature, about 127°C.
reclaim coal (Table 21), which is not surprising because of
the similarity in compositions of ash from the two coals.
The data also indicate that 7 ppm of injected ammonia failed
to produce a detectable change in the resistivity.
Analytical experiments with fly-ash samples in aqueous slur-
ries yielded the data given in Table 42 for pH, soluble
ammonia, and soluble sulfate. As indicated in the second
column of this table, the ash samples were collected by three
different methods: (1) insertion of an alundum thimble of
the type employed for grain-loading measurements in the inlet
duct of the precipitator; (2) insertion of a sampling train
consisting of a cyclone, a cascade impactor, and a filter in
series in the inlet and the outlet ducts; and (3) removal of
material deposited in the precipitator hoppers. All of the
pH values were in the acidic range; none of these data shows
any effect of ammonia on the acidity of the ash. The concen-
trations of ammonia in the ash show increases as ammonia was
injected (thimble samples), as the sampling location was
moved toward the outlet of the precipitator (all types of
samples), and as the particle size of the ash decreased
(cyclone, impactor, and filter samples). The variation in
the observed ammonia concentrations with each of the varia-
tions in sampling conditions is consistent with the assumed
reaction of ammonia with naturally occurring sulfur trioxide
and water vapor to produce a fine particulate of either
ammonium sulfate or ammonium bisulfate.
Before the samples collected on filters were treated with
water, they were examined by x-ray diffraction to determine
whether either ammonium salt could be identified T£O «™io
collected during ammonia injection at a concentration of
15 ppm produced weak diffraction bands in the location*
expected for ammonium sulfate, but the diffraction
was not strong enough to be clearly identified
84
-------
Table 42. CHEMICAL PROPERTIES OF FLY ASH
AT THE BULL RUN PLANT
(AMMONIA CONDITIONING, OCTOBER 1972)
Injected NH
concn, ppm
15
15
Sample
collecto
Thimble
Thimble
Cyclone
Impactor
Filter
Cyclone
Impactor
Filter
Hoppers
Ash properties
Ammonia, j Sulfate",
wt %
Sampling
location3
Inlet
Outlet
4.5 <0.001
4.4 0.001
0.37
0.67
Inlet
Outlet
0.048
0.101
0.40
0.73
Inlet
Inlet
Inlet
0.04
0.15
0.30
Outlet
Outlet
Outlet
0.15
0.32
7.20
0.033
0.048
Inlet
Outlet
0.20
0.37
The terms "inlet" and. "outlet" for samples other than
the hopper samples refer to ducts entering and leaving
the precipitator, respectively. These terms for hopper
samples refer to locations of the hoppers under the
electrical sections of the precipitator. Gas tempera-
ture, about 127°C.
Flue-Gas Concentrations—
Concentrations of flue gases at the precipitator inlet with
and without ammonia injection are given in Table 43. These
analytical data show the same effect of ammonia injection as
the gas concentrations previously recorded for the Widows
Creek plant (page 75): a decrease in the concentration of
sulfur trioxide as a result of ammonia injection.
Precipitator Electrical Data—
Values of primary voltages, primary currents, secondary cur-
rents, and spark rates in the eight transformer-rectifier
sets supplying power to Collector 1C under automatic control
are given in Table 44. The data were recorded during one day
of experimentation: first with no ammonia injected, then with
7 Ppm of ammonia injected, and finally with ammonia injection
85
-------
Table 43. CONCENTRATIONS OF FLUE GASES AT THE
BULL RUN PLANT
(AMMONIA CONDITIONING, SEPTEMBER 1972)
Injected NH3
concn, ppm
0
7
Concentrations9
NH 3 , ppm
<0.3
<0.3
<0.3
SO 3, ppm
1.3
1.8
2.3
0.8
SO 2 / ppm
640
680
725
711
H20, %
11.1
-
Temperature, about 127°C.
discontinued. These data show that after ammonia injection
was started both primary and secondary currents decreased and
that spark rates increased. Further, they show evidence of
these changes being reversed after ammonia injection was dis-
continued .
Relationships between secondary voltages and secondary cur-
rents (computed as current densities) for the two power
supplies at the precipitator inlet are shown in Figures 11
and 12. The data points shown as open symbols were recorded
with the power supplies under manual control; the data points
shown as closed symbols, on the other hand, were recorded
with automatic control. Figure 11 shows the data recorded
without any modification in the control circuitry, it shows
that ammonia displaced the voltage-current curves"to the
right along the voltage axis, as expected from a space-charge
effect, but it also shows curves with negative slopes that
resulted from the heavy sparking induced by ammonia iniection
Figure 12 shows the data recorded with a modification in the
control circuitry that allowed data to be recorded at lower
voltages where sparking did not occur even during ammonia
injection. This figure shows simply a dispIaceSL^n the
voltage-current curve along the voltage axis, similar to the
displacements found at the Widows Creek plani (pages 79 and 81) •
Ultrafine Particle Concentrations—
As mentioned in the preliminary discussion of ammonia condi-
tioning at the Bull Run plant, a condensation-nu^ei center
equipped with diffusion batteries was used to detect the
ultrafine particulate that was produced *»ri~ aerect.tne .
tion.36 The counter operates through the foil a?nonia in^ec"
a gas stream containing suspended particles is^iluteTwitr6'
air, saturated with water vapor, cooled adiabatically to
86
-------
Table 44. PRECIPITATOR ELECTRICAL DATA FROM THE
BULL RUN PLANT
(AMMONIA CONDITIONING, SEPTEMBER 1972)
TR
Set
1LCC
NHa Time,
injection3 mink
2LC
Off
On
Off
Off
On
Off
5
10
60
5
35
5
10
60
5
Primary Primary
voltage,I current,
V A
320
290
305
300
300
320
355
370
370
380
370
58
50
45
34
48
52
98
94
88
68
84
Secondary
current,
mA
290
260
230
200
250
260
680
600
560
500
500
Sparks
er min
125
300+
300+
300+
300+
300+
50
300+
300+
300+
300+
3LC
4LC
Off
On
Off
Off
On
Off
\«/ ~4m 4.
35
5
10
60
5
35
5
10
60
5
35
390
370
360
350
360
360
380
370 |
370
370
360
350
370
90
76
58
54
1 52
52
63
62
52
50
48
48
56
560 275
380 175
360 [ 300+
340
350
360
360
350
310
290
300
300
320
300+
300+
300+
300+
55
160
300+
300+
300+
275
1 I L J "
^Concentration in gas duct during injection, approximately
leadings were made before NH3 injection was started 5 to
60 min after it was started, and 5 to 35 mm after it was
neoo bushings was grounded, and thus the current was
only about one-half of the normal value.
87
-------
00
00
40
30
20
w
Q
10
NO NH3
INJECTED,
5 PPM
of NH3
INJECTED
TR 1LC
1
26 30
VOLTAGE, kV
34
40
30
20
10
26
NO NH3
INJECTED
TR IRC
I
30 34
VOLTAGE, kV
38
Figure 11. Current density versus voltage in Collector 1C
of the Bull Run plant
(ammonia conditioning, September 1972)
-------
00
vo
8
NO NH3
INJECTED
5 PPM
OF NH3
INJECTED
TR 1LC
26 30 34
VOLTAGE, kV
16
8
22
NO NH3
INJECTED
5 PPM
OF NH3
INJECTED^
TR IRC
26 30 34
VOLTAGE, kV
38
Figure 12
Current density versus voltage in Collector 1C
of the Bull Run plant
(ammonia conditioning, October 1972)
-------
induce condensation of water vapor and growth of the parti-
cles, and then passed through a photoelectric counter that is
able to detect the enlarged particles. The diffusion batter-
ies permit particle-size classification; when one of the
diffusion batteries is placed in the sampling line of the
condensation-nuclei counter, it removes particles smaller
than a characteristic size.
When the condensation-nuclei counter was used at the Bull Run
plant, its sampling line included a cyclone and several stages
of a Brink impactor to remove particles larger than 1.0 ym.
Thus, the range of sizes of particles detected had an upper
limit of 1.0 ym and a lower limit governed by the diffusion
battery selected. The lower limit was estimated as about
0.005 urn when no diffusion battery was used; this is the size
before growth in the presence of water vapor above the satura-
tion level.
The results obtained with the condensation-nuclei counter and
diffusion batteries are presented in Table 45. An important
matter to be noted in examining the data in this table is
that the data in each group were recorded during a brief time
interval, usually 15 min or less. Emphasis should therefore
be placed on the relative values of particle concentrations
within each group as only one experimental parameter was
altered {usually the ammonia concentration). Comparison of
concentrations in different groups of data may be misleading
because of varying sampling efficiencies and variations in
the background concentrations of small fly-ash particles.
One observation to be made from the data in Table 45 is that
changes in the rate of ammonia injection usually made signifi-
cant changes in the concentration of particles counted.
Although not shown by the table, these changes were observed
very rapidly—in a matter of seconds—as the rate of injection
was altered. The smallest effects of adding ammonia were
observed when the flue-gas temperature was abnormally low
only 91°C rather than 124 to 127°C—which would have caused
condensation of most of the sulfur trioxide vapor expected at
the higher temperatures and thus would have permitted a much
smaller amount of particulate to form as ammonia was injected*
In the experiments at the precipitator inlet, the increase in
particle concentration did not appear to vary substantially
with variations in the ammonia concentration from about 5 to
10 PPm. This observation gives weight to the assumption that
only the particulate formed between sulfur trioxide and
ammonia was detected in addition to the fly-ash particulate
For a given concentration of sulfur trioxide reactant, twice
this value for the concentration of ammonia iniected could
theoretically convert all of the sulfur trioxide to ammonium
90
-------
Table 45. CONCENTRATIONS OF SUBMICRON PARTICLES
AT THE BULL RUN PLANT
(AMMONIA CONDITIONING, OCTOBER 1972)
Gas
temp, °C
127
Sampling
location ,
Inlet
Injected NH3
concn , ppm
0
6 , i
Minimum
particle
size, ym
1
<0.010
<0.010
| Particle "
concn,
no. /cm3
4.1 x 106
13.7 x 106
91
Inlet
<0.010
<0.010
<0.010
1.3 x 10s
1.9 x 106
2.2 x 10s
124
124
124
124
124
Inlet 0
5
°
Inlet 5
5
5
Inlet 5
Outlet 0 „
Max.0
°
Outlet 0
5
10
20
15
°
<0.010
<0.010
<0.010
<0.010
<0.010
0.014
0.050
<0.010
<0.010
<0.010
<0.010
<0.010
<0.010
<0.010
<0.010
<0.010
<0.010
<0.010
1.8 x 106
5.9 x 106
5.8 x 10 6
1.5 x 106
10.0 x 106
8.9 x 10 6
3.1 x 106
9.5 x 106
3.2 x 10 $
0.9 x 10s
7.6 x 106
1.0 x 106
1.0 x 10 6
3.1 x 106
3.4 x 106
6.9 x 106
5.9 x 106
1.1 x 106
alf other than <0.010 pm, controlled with a diffusion bat-
bAbnormally low during the gradual increase in unit load
CMfximumnattainIble with the injection system but not deter-
mined, owing to the fact that the NH3 flowmeter was mad-
vertently bypassed.
91
-------
sulfate, and any greater concentration of ammonia injected
could be unable to react with sulfur trioxide. Based on the
sulfur trioxide concentrations given previously in Table 43,
an ammonia concentration of about 5 ppm would have caused com-
plete conversion of sulfur trioxide to ammonium sulfate par-
ticulate.
In comparatively limited work at the precipitator outlet,
ammonia concentrations in the 15- to 20-ppm range appeared to
produce more particulate than concentrations in the 5- to
10-ppm range. This finding may indicate that the concentra-
tion of ammonium sulfate particulate was governed not only by
stoichiometric factors but by reaction kinetics and that the
concentration was increased by the increased reaction time
during gas flow through the precipitator.
A second observation to be made from the data in Table 45,
based on a comparison of particle concentrations detected
with and without diffusion batteries in the sampling line, is
that at the precipitator inlet more than 50% of the particu-
late formed by ammonia injection was in the size range below
0.05 ym. Time did not allow any study of particle size at
the precipitator outlet. It is possible, however, that the
size range was even smaller at the outlet than at the inlet
This possibility is suggested by the fact that the experi-
mental data give no indication of a reduction in number con-
centration across the precipitator, despite other evidence
cited later indicating that a reduction in mass concentration
did occur at other power plants.
POWER PLANTS BURNING HIGH-SULFUR COALS
Widows Creek Plant, Unit 7
During November of 1972, a study was made of ammonia condi-
tioning at Unit 7, Collector 7A, of the WidowsTreek plant
while the customary high-sulfur coal was being burned as the
fuel. Analyses of coal samples over a period of several davs
indicated that the following properties of ?he coaTwere
representative of the properties prevailing during the inves-
tigation:
Sulfur, 3.5%
Ash, 13.8%
Moisture, 6.5%
Heat value, 6,330 cal/g
(11,400 Btu/lb)
92
-------
At m? Y aS^ ?ave the comP°sition shown bv
46 This composition differs from the composition
fly ash from other Eastern coals in having a hiahfr S °f
of calcium oxide, but the percentage of calcium
, cum oxr?
lower than the values for ash from the Wester^oalfV3 Stil1
in Section V. coals discussed
Table 46. COMPOSITION OF FLY ASH FROM A
HIGH-SULFUR COAL AT WIDOWS CREEK UNIT 7
(November 1972)
Component
Li20
Na20
K20
MgO
CaO
A1203
Fe203
SiO2
Ti02
P20S
SO 3
Weight percentage
0.02
0.49
2.6
0.94
3.7
19.0
22.0
48.1
1.0
0.26
1.0
No determination of the precipitator efficiency was made dur-
ing the study at the Widows Creek plant in November, 1972
However, the results from earlier efficiency tests by TVA*5
ft Collector B of Unit 7 (identical to Collector A) are given
in Table 47. These results were obtained while high-sulfur
coal was being burned as a fuel, and the data obtained with
ammonia conditioning presumably indicate the effects on effi-
ciency that were produced by ammonia conditioning during the
investigation discussed in this report.
During this investigation, an effort was made to use a Bailey
Bolometer to obtain indirect indications of the effects on
Precipitator efficiency that were produced by ammonia condi-
tioning, in several experiments, the bolometer did not func-
tion satisfactorily because of insufficient efforts to keep
lenses free of dust accumulations. When the performance of
the instrument was relatively satisfactory, the bolometer
indicated that the injection of ammonia caused a gradual
suppression in both the time-average particulate concentration
m the precipitator effluent and the intensity of rapping
Puffs emitted by the precipitator during rapping. On the
93
-------
Table 47. PRECIPITATOR EFFICIENCIES WITH HIGH-SULFUR COAL
AT WIDOWS CREEK UNIT 7 (JUNE-JULY 1970)a
Gas temp,
°C
ca. 145
ca. 130
ca. 130
Injected NH3
concn , ppm
0
0
ca. 11
Gas velocity,
relative valueb
0.95
1.00
1.05
0.95
1.00
1.05
0.95
1.00
1.05
Efficiency,
%
96
94
92
87
84
78
>99
98C
97C
Assumed to indicate approximate efficiencies during the
.investigation.
Relative to the design value of 2.42 m/sec.
cEstimated by extrapolation of data at lower gas veloci-
ties. Believed to be conservative estimates from the
available data.
other hand, the instrument indicated that the discontinuance
of ammonia injection caused a relatively rapid reversal of
the initial effects.
Overall, the investigation of ammonia conditioning at the
Widows Creek plant with high-sulfur coal as the fuel consisted
of the same experimental approaches as the previous studies
at this plant and the Bull Run plant with low-sulfur coals
The experimental data are given in detail in Tables 48 through
52 and Figure 13, and are presented on the following paaes
The principal observations to be made from an inspection of
the experimental data are as follows:
• Fly-ash resistivity (Table 48). NO increase in the
resistivity of the ash was detected as the result of
ammonia infection at either of the two flue-gas tem-
peratures investigated, 132 or 154°C, contrarv to the
prediction based on the earlier work of Reese and
Greco.5 A decrease in resistivity was apparent at
the higher temperature. However, in view of the
moderate resistivity of the unconditioned ash, this
change (if real) could not have been responsible
for an improvement in precipitator performance
94
-------
Table 48. ELECTRICAL RESISTIVITY OP FLY ASK
FROM A HIGH-SULFUR COAL AT
WIDOWS CREEK UNIT 7 (NOVEMBER 1972)
temp, °C
132
154
concn, ppm
0
20
0
22
Resistivity,
ohm cm
2.8 x 108
1.0 x 108
1.1 x 10a
1.0 x 10 8
3.0 x 108
3.9 x 1010
1.3 x 10 10
7.1 x 10 8
4.2 x 10 8
Table 49. CHEMICAL PROPERTIES OF FLY ASH FR'OM A
HIGH-SULFUR COAL AT WIDOWS CREEK UNIT 7
(NOVEMBER 1972)
Injected NHs
concn, ppm
0
20
0
20
Sample
collector
Thimble
Thimble
Hoppers
Sampling
location3
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Hoppers Inlet
Outlet
pH
10.3
10.6
10.0
10.0
10.8
10.7
10.7
10.3
Ash properties
Ammonia
wt %
<0.001
0.009
0.117
0.127
0.010
0.016
0.044
0.074
suifate,
wt %
1.49
1.39
1.57
2.15
0.76
0.94
0.92
1.42
3The terms "inlet" and "outlet" for samples other than
the hopper samples refer to ducts entering and leaving
the precipitator, respectively. These terms for hopper
samples refer to locations of the hoppers under the
electrical sections of the precipitator. Gas tempera-
ture, about 132°C.
95
-------
Table 50. CONCENTRATIONS OP FLUE GASES FROM A
HIGH-SULFUR COAL AT WIDOWS CREEK UNIT 7
(NOVEMBER 1972)
Gas
temp, °C
132
154
Injected NHa
concn , ppm
0
11
20
0
22
NHs , ppm
0.14
0.06
0.07
0.09
0.09
—
0.06
0.04
0.06
-
—
—
0.37a
"^
^*
Concentrations ' ''' lll~"
SO 3 , ppm
10.5
13.1
9.5
6.9
6.5
6.3
9.1
5.0
3.2
8.7
9.8
12. 3a
10. 9a
2.9a
1 . Oa
0.5a
so2 , ppm
2410
2500
2620
2400
2430
2580
2420
2370
1820
1410
1800
2450a
2200a
2510a
2430a
2570a
H2O, %
7.8
6.8
8.0
8.3
^
8.2
-
8.5
7.4a
7.7a
^m
-
These concentrations were determined at the orecinitator
outlet. All others in this table were de?erminedPat the
inlet.
96
-------
Table 51. PRECIPITATOR ELECTRICAL DATA FROM
WIDOWS CREEK UNIT 7
(HIGH-SULFUR COAL, NOVEMBER 1972)
TR
Se
7A
7A2
7A3
7A4
7A5
Injected NH3
concn , ppm
0
23
0
23
0
23
0
23
0
23
Primary data3
Voltage, Current,
V 1 A
295
325
290
300
275
285
195
210
275
275
i
70
66
72
67
200
185
90
110
180
175
secondary da>*a "'-
Voltage
kV
-
43.7
47.3
—
-
40.5
40.9
Current,
mA
-
234
248
^^
mf
685
665
Each of the electrical values listed with ammonia
injection was essentially at steady-state within a
few minutes after injection was started.
97
-------
VD
00
55
W
70
60
50
40
30
0 20
10
I I
TR 7A2
20 PPM
I I
30 35 40 45
VOLTAGE, kV
50
70
60 —
50
40
30
20
10
— NH
~1 T
TR 7A5
PPM
I I I
25 30 35 40
VOLTAGE, kV
45
Figure 13. Current density versus voltage in Collector 7A
of the Widows Creek plant
(high-sulfur coal, November 1972)
-------
Table 52. CONCENTRATIONS OF SUBMICRON PARTICLES
AT THE WIDOWS CREEK PLANT
(HIGH-SULFUR COAL, NOVEMBER 1972)
Sampling
location5
Injected NH3 Minimum size
concn, ppm detected, ym
Particle
concn,
no./cm
No. % above
Inlet
Inlet
Inlet
Inlet
Outlet
Outlet
- — - — L
0 0.005
0.014
0.050
11 0.005
0.014
0.050
0 0.005
0.014
0.050
23 0.005
0.014
0.050
0 0.005
0.014
20 0.005
0.014
0.050
6.5 x 10(
5.4 x 10e
2.9 x 106
19.5 x 106
16.7 x 106
11.2 x 106
12.0 x 106
10.3 x 10s
6.3 x 10s
30.9 x 106
29.0 x 106
17.0 x 106
0.43 x 106
0.35 x 106
1.40 x 106
0.98 x 106
0.79 x 106
... i -••.••• %«*««i iQ J.
100
80
45
100
85
57
100
89
52
100
94
55
100
81
100
70
56
Gas temperature, 132°C,
99
-------
• Fly-ash properties in aqueous slurries (Table 49).
Samples of fly ash were collected from the gas ducts
entering and leaving the precipitator and from the
hoppers of the precipitator. Analyses of the ash
in aqueous slurries showed that the concentration
of ammonia in the ash increased markedly when the
conditioning agent was injected into the gas stream.
• Flue-gas concentrations (Table 50). The injection of
ammonia always lowered the concentration of sulfur
trioxide in the gas stream entering the precipitator.
• Precipitator electrical data (Table 51 and Figure 13).
The injection of ammonia caused a rapid change in the
electrical behavior of the precipitator, most notably
a shift in the secondary voltage-current curve toward
higher voltages for a given current density or toward
a lower current density for a given voltage.
• Ultrafine particulate concentrations (Table 52). The
condensation-nuclei particle counter showed rapid
increases in fine-particle concentrations as ammonia
injection was started and rapid reversal of this
effect as injection was stopped. The effects were
less pronounced at the outlet of the precipitator
than at the inlet, suggesting that a large fraction
of the small particulate introduced by ammonia
injection was collected in the precipitator.
All of the above observations are consistent with the occur-
rence of ammonia conditioning through the space-charge effect
which has previously been hypothesized as the explanation for'
ammonia conditioning at the Widows Creek and Bull Run plants
with low-sulfur coals. ^
Gallatin Plant, Unit 4
During the first part of 1973, TVA made a preliminary instal-
lation of ammonia conditioning facilities at the Gallatin
plant. The initial installation permitted treatment of one-
third of flue gas emitted from Unit 4 (290 MW), which passed
through one of the three electrostatic precipitators for this
unit (Collector 4C). The success achieved in treatina flue
gas from high-sulfur coal led to an enlargement of the ammonia
conditioning system to treat all of the flue gas from each of
the four units at the Gallatin plant. The enlargement of the
system was completed early in 1974, at which time the
Institute made a study of the conditioning process at Collec-
tor 4C.
100
-------
Coal and Fly-Ash Properties—
The coal burned in Unit 4 of the Gallatin plant durinq the
investigation of ammonia conditioning was high in sulfur co
tent. Representative properties of the coal samples collect H
during this investigation are shown below:
Sulfur, 4.0%
Ash, 14.2%
Moisture, 7.0%
Heat value, 6,300 cal/g
(11,300 Btu/lb)
Samples of fly ash collected from the hoppers of Collector 4c
varied significantly in the concentrations of several compo-
nents, principally calcium oxide and sulfur trioxide. in the
composition shown in Table 53, the ranges of concentrations
of these two components are given with the average concentra-
tions of the remaining components.
Table 53. COMPOSITION OF FLY ASH
FROM GALLATIN UNIT 4
Component
Li2O
Na20
K20
MgO
CaO
A1203
Fe203
SiO2
Ti02
P205
SO 3
Weight percentage
0.02
0.62
2.6
0.97
1.2-2.9
21.6
19.9
46.9
2.0
0.29
1.2-2.3
Precipitator Efficiencies—
There are no data available that would permit a comparison of
the absolute efficiencies of Collector 4C with and without
ammonia injection. The manifold for injecting ammonia was
Placed in the original sampling ports that were designed for
grain-loading measurements upstream from the precipitator.
Thus, since this manifold was installed, the only measurements
relative Ic efficiency have consisted of measurements at the
outlet of the precipitator.
101
-------
The results of TVA's grain-loading measurements at the outlet
of Collector 4C37 are given in Table 54. These data were
obtained during February of 1973 with the original condition-
ing facilities in operation. The data indicate that 18 ppm
of ammonia lowered the emission level by about 75% at 143°C
(the usual flue-gas temperature in the precipitator) and
32 ppm lowered the emission level by about 65% at 154°C.
Table 54. FLY-ASH EMISSION FROM
GALLATIN UNIT 4 (COLLECTOR 4C)
Gas temp ,
°C
143
154
Injected NH3
concn, ppm
0
18
0
32
Outlet dust concn,
g/m3 (gr/ft3)
0.338 (0.147)
0.081 (0.035)
0.299 (0.130)
0.108 (0.047)
Use of an Anderson impactor during the investigation of
February, 1974, yielded values of the total particulate con-
centration plus the particulate concentrations in various
ranges of effective diameter between the limits of 0.38 ym
and 10.8 ym. The data obtained with this impactor do not
necessarily reflect the overall emission levels, in contrast
with the data obtained by TVA, for the sampling'nozzle of the
impactor was always at the same location within the outlet
duct, at an insertion of about 1 m into the duct at one of
the six ports used by TVA with multiple insertions in all
ports. Moreover, one series of experiments was conducted
with the rappers in operation and another series was conducted
with the rappers deenergized, whereas the tests by TVA were
all performed with the electrode rappers in operation.
The results obtained with the Andersen impactor are presented
in Table 55. Inspection of the total concentrations found
with the rappers in operation indicates that the injection of
20 ppm of ammonia lowered the concentration from 0 227 <-o
V4? ^"'-JiS-X87 t0 °'°617 g^/ft3)' Comparison of the data
obtained with the rappers not in operation indicates that
20 ppm of ammonia lowered the concentration from 0 135 to
°;01S S/m3J°:05f to 0.0069 gr/ft3). Further comparison of
the data obtained with and without rapping, either with or
without ammonia injection, indicates that reentrainment of
particulate as the result of rapping contributed markedly to
the level of particulate emission.
102
-------
Table 55. EMISSION OP FLY ASH IN VARIOUS SIZE
AT GALLATIN UNIT 4 (COLLECTOR 4C)a
Rappers
Particle
size/ ym
On I >10.8
10.8-6.7
6.7-4.5
4.5-3.1
3.1-2.0
2.0-0.96
0.96-0.57
0.57-0.38
<0.38
Total
Off | >10.8
10.8-6.7
6.7-4.5
4.5-3.1
3.1-2.0
2.0-0.96
0.96-0.57
0.57-0.38
<0.38
Total
Outlet dust concentration1
No NH3
injected
*v ppm or
NH3 injected
0.1520
0.0136
0.0060
0.0025
0.0018
0.0039
0.0133
0.0147
0.0191
(0
(0
(0
(0
(0,
(0.
(0.
(0.
.0661)
,0059)
.0026)
,0011)
0008)
0017
0058)
0064)
0.0764
0.0248
0.0173
0.0099
0.0053
0.0039
0.0012
0.0002
0.0030
0.2270 (0.0987) I 0.1419
(0.0083)
(0.0332)
(0.0108)
(0.0075)
(0.0043)
(0.0023)
(0.0017)
(0.0005)
(0.0001)
(0.0013)
7076617)'
0.1104
0.0069
0.0046
0.0030
0.0030
0.0030
0.0014
0.0002
0.0030
(0.0480)
(0.0030)
(0.0020)
(0.0013)
(0.0013)
(0.0013)
(0.0006)
(0.0001)
(0.0013)
0.0087
0.0009
0.0002
0.0002
0.0002
0.0007
0.0002
0.0002
0.0044
0.1355 70.0589)
(0.0038)
(0.0004)
(0.0001)
(0.0001)
(0.0001)
(0.0003)
(0.0001)
(0.0001)
(0.0019)
(0.0069)
a
Gas temperature, 138°C.
The injection of ammonia was evidently helpful in lowering the
total emission of particulate with or without rapping. The
effect of ammonia with rapping may be attributed to either or
both of two mechanisms of conditioning: (1) an increase in
the cohesiveness of fly-ash particles and (2) an increase in
the space-charge component of the electric field. The effect
of ammonia without rapping probably can be attributed to the
space-charge effect alone.
The data in Table 55 for the concentrations of particulate in
various particle-size ranges indicate that with rapping the
injection of ammonia actually increased the concentrations in
all diameter ranges between 2.0 and 10.8 ym while causing a
decrease in the total concentration. The particle-size data
obtained without rapping do not show this effect from the
injection of ammonia: It seems reasonable to draw from these
observations thT conclusion that the injection of ammonia dur-
ing rapping increased the size of agglomerates of fly ash that
103
-------
were reentrained, an effect that might well be expected as
the result of conditioning through the mechanism of increased
cohesiveness of individual particles.
Information about the concentration of fly ash emitted from
the precipitator were also obtained with a Bailey bolometer,
as discussed in detail later in this report (pages 110 through
113}. The bolometer confirmed the findings with the Andersen
impactor indicating that the injection of 20 ppm of ammonia
markedly lowered the level of particulate emission from the
precipitator with the rappers in service. Also, this instru-
ment confirmed the conclusion from the impactor data indicat-
ing that, either with or without the injection of ammonia,
the discontinuance of rapping sharply lowered the level of
emission (it was unable, however, to show the difference in
the levels of emission with and without ammonia injection
that were found with the impactor while the rappers were not
in service).
Fly-Ash Resistivity—
The results of determinations of the fly-ash resistivity with
the point-plane resistivity probe are given in Table 56.
There is an indication from some of the data that the injec-
tion of ammonia lowered the resistivity of the ash. However,
the poor reproducibility of the data make it uncertain that
the apparent effect was real. In any event, there is no
indication of an increase in resistivity, the phenomenon ori-
ginally suggested by earlier work on ammonia conditioning of
fly ash from high-sulfur coal.
Voltage-current data presented later in this report indicate
that the injection of ammonia either increased the resistivity
of the ash on the collecting electrodes or altered the elec-
trical properties of the gas stream through a space-charge
effect. The rapidity of the changes in the electrical data
and the stability of the adjusted values with time after
injection was started or stopped strongly indicates that the
effect was on the gas stream, not on the deposited ash.
Evidence of Conditioning through a Space-Charge Effect
In the studies of ammonia conditioning of the Widows Creek
and the Bull Run plants, the evidence for conditioning through
a space-charge effect consisted of the following- (1) fly-
ash analyses showing the presence of ammonia as a component of
particulate material collected either from the inlet and out-
let ducts or from the hoppers of the precipitator; (2) flue-
gas analyses showing loss of sulfur trioxide from the eras
phase when ammonia was injected; (3) rapid changes in the
electrical behavior of the precipitator power supplies; and
(4) rapid appearance and disappearance of ultrafine
104
-------
Table 56. ELECTRICAL RESISTIVITY OF FLY
AT GALLATIN UNIT 4
Gas
temp, °C
138
132
Injected NHs
concn, ppm
0
10
20
0
20
Resistivity, ohir'
14.0 x 106
3.8 x 10e
92.0 x 108
7.9 x 108
0.9 x 108
0.5 x 108
8.2 x 108
2.1 x 108
50.0 x 108
10.0 x 108
16.0 x 108
2.9 x 10a
1.3 x 108
14.0 x 108
1.9 x 108
0.3 x 108
0.1 x 108
4.2 x 108
0.8 x 108
1.1 x 108
2.9 x 108
5.3 x 108
Particulate detected with the condensation-nuclei counter
when ammonia injection was started or stopped.
In the studies of ammonia conditioning at the Gallatin plant
each type of evidence for the space-charge effect was once '
again obtained. The experimental data indicating the occur-
rence of the space-charge effect are presented in Tables 57
through 60 and Figures 14 and 15 on the pages immediately
following. The data are self-explanatory in nearly all in-
stances, in view of their similarity to similar data that have
keen discussed previously in this report. Thus, no detailed
commentary on the data seems necessary. However, attention
*s directed particularly to Figure 14, which shows very clear-
ly how rapidly the electrical behavior of the precipitator
Desponded to the injection of ammonia.
lyldence of Conditioning through an Apparent Change in the
SoKesiveness of Fly-Ash Particles-—
?he possibility that ammonia conditioning involved a change
**> the cohesive forces between fly-ash particles as well as
Jhe space-charge effect was considered for several reasons.
first, TVA personnel with experience in ammonia conditioning
Believe that an increase in cohesiveness occurs? they base
this belief on observations of phenomena such as (1) the
formation of large agglomerates of particulate known as
hornet nests" around the ammonia injection probes, (2) the
105
-------
Table 57. CHEMICAL PROPERTIES OF FLY ASH
AT GALLATIN UNIT 4a
Injected NH3
concn, ppm
0
20
Ash properties
pH
8.6
8.6
Ammonia,
wt %
<0.01
0.21
Sulfate,
wt %
1.2
1.7
Collected at the precipitator inlet in an
alundum thimble at a gas temperature of
138°C.
Table 58. CONCENTRATIONS OF FLUE GASES AT
GALLATIN UNIT 4
Gas
temp, °C
138
132
Injected NH3
concn , ppm
0
10
20
0
20
NH3 , ppm
<0.1
<0.1
<0.1
<0.1
0.9
2.0
-
0.5
3.5
concenfr
~e7\ •
SO 3 , ppm
6.9
9.2
0.9
0.9
1.8
<0.9
7.4
7.4
<0.9
<0.9
• —
ations
SO2 , ppm
2630
2980
2580
2760
3310
3180
3390
3450
3500
3480
"
•••••••SBMMKBBOnc
H20, %
8.8
8.2
:
9.7
8.9
-
106
-------
Table 59. PRECIPITATOR ELECTRICAL DATA FROM
GALLATIN UNIT 4
TR
Seta
4C1
4C2
4C3
4C4
Injected
NH3 concn
ppra
0
20
0
20
0
20
0
20
Primar
Voltage
V
378
382
322
363
313
326
313
320
v data
Current,
A
70
37
73
72
77
77
76
76
Voltage
1 kV
38.0
45.6
33.5
39.9
30.7
33.1
29.6
30.1
Current
density,
nA/cm
42.3
19.4
42.8
42.3
46.3
46.9
45.5
45.8
Sparks
per
min
0
70
0
26
0
0
0
0
^Listed in the sequence from the inlet to the outlet of the
precipitator.
Table 60. CONCENTRATIONS OF SUBMICRON PARTICLES
AT GALLATIN UNIT 4
Sampling
location
Inlet
Outlet
Uncertain
system.
Injected NH3
con en r ppm
0
>20a
o
20
Particle concn, no./1 cm , "
as a function of minimum size
0.005 ym
22 x 106
61 x 106
U.014 urn
17 x 106
26 x 106
0.5 x 106
1.5 x 106
0.064 urn
0.2 x 106
0.9 x 106
as a result of a broken flowmeter in the injection
107
-------
50
40
> 30
w
3
~n
20
10
NH3 ON
(20 PPM)
NH3 OFF
TR 4C1
1000
1100
HOUR
1200
Figure 14. Rapidity of the effect of ammonia
injection on the voltage supplied to
the inlet electrical field of
Gallatin Precipitator 4C
108
-------
50
o
vo
301
201
10
SET 4C4—
NO NH3
SET 4C1—
NO NH3
SET 4C1—
20 PPM
Of NH3
15
20
25
30 35
VOLTAGE, kV
40
45
50
Figure 15. Current density versus voltage in
Precipitator 4C of the Gallatin plant
-------
accumulation of ash deposits on exhaust fans between precipi-
tators and stacks, and (3) the apparent "stickiness" of ash
in systems for removing ash from precipitator hoppers.
Second, Institute personnel using a Brink impactor at the
Bull Run plant noted a distinct difference in the compactness
of ash deposited within the impactor when ash samples were
collected with and without ammonia conditioning. Third,
representatives of the Koppers Company who have had consid-
erable experience with ammonia conditioning believe that
ammonia increases the cohesiveness of fly ash; moreover, they
have taken photomicrographs of ash particles that reportedly
reveal the agglomeration of spheres of fly ash through
bridges of a feather-like material thought to be ammonium
sulfate.38 Finally, Dalmon and Tidy15 have reported that
the use of ammonium sulfate as a conditioning agent may in-
crease the efficiency of collection of fly ash, especially
if the effective electrical resistivity of the ash is low
as the result of the presence of a substantial amount of low-
resistivity carbon particles. Increasing the cohesiveness
of low-resistivity fly ash would, of course, provide a
mechanical force to hold the ash on the collecting electrodes
of a precipitator, taking the place of the electric force
associated with high-resistivity ash and thus minimizing
reentrainment.
The evidence for an increase in the cohesiveness of fly-ash
particles consisted of the data obtained with two types of
photoelectric measurements at the precipitator outlet. One
type of measurement was made with a Bailey bolometer to
detect changes in the obscuration of a light beam across the
outlet duct. The other type of measurement was made with a
light-scattering photometer known as a Climet Particle
Analyzer (Model No. CT-201).36 Both the Bailey bolometer
and the Climet counter gave real-time indications of the rate
of change in the level of particulate emission as a change
in ammonia injection was made.
The results obtained with the Bailey bolometer during one
day of experimentation are shown by a reproduction of the
recorder chart in Figure 16. During the first period of
about 4 hr before the injection of ammonia was started, the
bolometer gave evidence of wide variations in the degree of
light obscuration; the recorder chart showed numerous spikes
of high intensity that were presumably caused by puffs of
fly ash reentrained during electrode rapping. During the
following period of about 4 hr with 20 ppm of ammonia in-
jected, the spikes were continued at about the same magnitude
for about 30 min but then were gradually suppressed during
the remaining time of ammonia injection. After ammonia in-
jection was discontinued, the spikes remained suppressed for
110
-------
1300
1400
1200
1500
UOO
1600
1000
1700
O90O
0800
I80O
1900
100 80 60 40 20
20 40 60 80 100
RELATIVE VALUE OF
LIGHT OBSCURATION
Figure 16. Effects of changes in ammonia injection on the emission
of particulate from Gallatin Precipitator 4C
as indicated by the Bailey bolometer
(20 ppm of NH3)
-------
about 1 hr but then gradually increased in magnitude until
they approached the original magnitude. It is evident that
changes in the intensity of the rapping spikes occurred over
extended times following each change in ammonia injection,
in contrast to the changes in electrical data which were very
rapid.
The results obtained with the bolometer during another day of
experimentation as shown in Figure 17. This reproduction of
the recorder chart shows the information obtained with and
without electrode rapping, first with no ammonia injected and
then with 20 ppm of ammonia injected. The deenergization of
the rappers caused the disappearance of rapping spikes, with
or without ammonia injection, and the recorder pen traced es-
sentially a smooth curve that was indicative of minimal ob-
scuration. The beginning of ammonia injection with the rappers
energized caused a gradual rather than an immediate suppres-
sion of the rapping spikes similar to that described above.
The data obtained with the Climet particle counter are given
in Table 61. These data consist of the concentrations by
number of particles in two size ranges: 0.5-2.0 and 1.0-
2.0 ym. Although the table shows concentrations at discrete
intervals of time, the recording of data was in real time,
permitting the detection of rapping puffs at intervals of
about 6 sec every 36 sec (timed for the outlet section of
the precipitator) and the detection of concentration changes
after changes were made in ammonia injection. In Table 61,
the concentrations listed as minimum values were observed
between rapping events, and the concentrations listed as
maxima were observed during rapping.
The data constituting the third group of experimental data
in Table 61 illustrate the delay in the change in fly-ash
concentrations emitted from the precipitator after the in-
jection of ammonia was started. These data indicate that
the injection of ammonia gradually brought about decreases
in the amount of ash emitted from the precipitator during
time intervals both between and during rapping events.
The data in Table 61 make it apparent that either with or
without ammonia injection the rapping of the electrodes
made a highly important contribution to the time-average
emission of particulate. Consider the data for particles
in the 1.0-2.0 urn size range recorded at 0905 hr without
ammonia injection. Taking the minimum concentration of
44 particles/cm3 to be representative for the 30-sec inter-
val between rapping events and the maximum concentration of
2600 particles/cm to be representative for the 6-sec inter-
val during rapping, one can calculate that the time-average
concentration was approximately 400 particles/cm3 (roughly
112
-------
RAPPERS
OFF
RAPPERS
OFF
RAPPERS
OFF
RAPPERS
POWER
1300 / OFF
O8OO
.1500
0700
u>
0600
INK FLOW
INTERRUPTED
1600
1700
100 80 60 40 20
20 40 60 80 100
RELATIVE VALUE OF
LIGHT OBSCURATION
Figure 17. Effects of changes in ammonia injection and electrode rapping
on the .emission of particulate from Gallatin Precipitator 4C
as indicated by the Bailey bolometer
(20 ppm of NH3)
-------
Table 61. EMISSION OF FLY ASH IN VARIOUS SIZE RANGES
AS A FUNCTION OF ELECTRODE RAPPING AND
AMMONIA INJECTION AT GALLATIN UNIT 4a
Injected
NH3
concn,
ppm
0
0
20
20
Electrode
rapping
On
Off
On
Offd
Duration
of NH3
injection,
min
-
-
15
26
38
62
85
Particle concn, no. /cm3,
vs . size ranae
0.5-2.0 ym
Minb
1200
710
610
380
270
250
250
Maxb
>4400
-
2700
2300
420
-
1.0-2.0 urn
Minb
44
43
80
<4
<4
<4
<4
Maxc
2600
-
1100
38
38
-
^Determined with the Climet particle counter.
Recorded during intervals between rapping events in the out-
let electrical section.
GRecorded during intervals of rapping in the outlet electri-
,cal section, repeated approximately every 30 sec.
Discontinued 60 min after the injection of ammonia was
started.
10 times the value found without rapping). Similar considera-
tion of data for the 1.0-2.0 ym range recorded at 1138 hr
with injection leads to the calculation of a time-average
concentration of at least 6 particles/cm3 (at least 1.5 times
the value found without rapping). (The indicated ratio of
the time-average concentrations without and with ammonia
injection—about 100:1—is not apparent in the Andersen im-
pactor data given in Table 55. There is no apparent explan-
ation for the discrepancy in the Climet and Andersen data).
It was somewhat surprising to find evidence that rapping
caused reentrainment of particles in size ranges as low as
0.5-2.0 and 1-0-2.0 ym. It was not expected that the force
of rapping could break up agglomerates of particles on the
precipitator electrodes to the degree that^ppreciab^e con-
centrations of this fine material would be produced.
Admittedly, there is some uncertainty in the actual size
limits of the particles counted, particularly the uooer size
limit. This limit wag governed by the properties^ of the
cyclone that was used in the sampling iL^to prevent Urge
particles from entering the Climet counter. The specified
114
-------
upper limit of 2.0 \im may not be precisely correct, but it
should not be in error to a degree altering the conclusion
that very fine particles are dispersed and reentrained by
electrode rapping.
115
-------
SECTION VII
DISCUSSION OF CONDITIONING WITH SULFUR TRIOXIDE
Three aspects of the conditioning of fly ash with sulfur
trioxide are considered in this section. The first is a
practical question: What are the parameters of a power-
plant emission system that determine the effectiveness of
the conditioning process? The second aspect of sulfur tri-
oxide conditioning involves two theoretical questions: What
is the mechanism by which sulfur trioxide is collected by
fly-ash particles after the conditioning agent is injected
into a flue-gas stream? And what are the mechanisms by which
the collected agent conditions the ash and improves its
electrostatic precipitation? (From the results of the
research described previously in this report, it is apparent
that the process of lowering the electrical resistivity is
not the only important phenomenon, despite the usual concept
to the contrary.) The third subject is posed by this ques-
tion: What undesirable effects, if any, are encountered
during the use of sulfur trioxide for conditioning?
PARAMETERS OF A POWER-PLANT EMISSION SYSTEM THAT AFFECT THE
EFFICIENCY OF SULFUR TRIOXIDE CONDITIONING
Source of the Sulfur Trioxide Injected
Three sources of sulfur trioxide as a conditioning agent are
discussed in this report:
(1) Anhydrous sulfur trioxide in a stabilized liquid
form ^
(2) Concentrated sulfuric acid as a liquid
(3) A mixture of sulfur dioxide with air passing
^C^alItiC °*idizer that converts ?he
to the trioxide
116
-------
There is a fourth source that is currently receiving consid
ation: elemental sulfur that is burned to produce the dio rt~
which is then catalytically converted to the trioxide in e'
a mixture with excess air. However, the research discussed
in this report did not include any experience with sulfur
trioxide from elemental sulfur as the ultimate source mate-
rial.
Other publications have discussed the advantages and disad-
vantages of the various methods of injecting sulfur trioxide 3/1°
These publications have dealt with such matters as equipment*
requirements, cost factors, and safety considerations. This
report, therefore, is limited to a consideration of the
effectiveness of the sulfur trioxide from injection systems
based on the first three source materials listed above.
The experimental data give no apparent basis for a choice
among the different types of injection systems. The resis-
tivity data indicate that each type of system produced the
expected change in the properties of the fly ash. Only in
one instance was there evidence of the failure of an injec-
tion system to lower the resistivity of fly ash. This
failure occurred at Unit 3 of the Cherokee plant, where use
was made of one of the two designs of an injection system
based on volatilization of sulfuric acid. The failure is
attributed to an unidentified flaw in the installation or
the operation of the injection system rather than to a basic
problem with the design concept.
Experience by personnel of Southern Research Institute in
confidential studies of sulfur trioxide conditioning for pri-
vate industry has shown failures of sulfur trioxide condi-
tioning to produce the expected reductions in fly-ash re-
sistivity in power plants other than the Cherokee plant.
Details of these studies cannot be included in this report.
However, one of the principal conclusions reached in these
studies can be stated here. It is evident that, regardless
of the tvoe of injection system used, careful attention must
be given to the design of the interface of an injection
system with the flue-gas stream. The design must permit
efficient mixing of the stream of concentrated sulfur tri-
oxide in carrier gas with the large excess of flue gas. if
the mixing is inefficient, sulfur trioxi de will co nfcine with
water vapor and undergo condensation as a mist of sulfuric
-
i s
installations of injection systems than in others.
117
-------
Site of Sulfur Trioxide Injection
Three sites of sulfur trioxide injection in the train lead-
ing flue gas from the boiler of a power plant to the stack
have been investigated. One site is upstream from the
mechanical collector (cyclone) in a system for gas cleaning
comprised of both the cyclone and an electrostatic precipita-
tor. Another site is the duct immediately upstream from the
precipitator in a plant where no mechanical collector is used.
The third site is the duct between a mechanical collector
and a precipitator.
Conditioning of fly ash apparently can be successful in
plants with all of the locations of sulfur trioxide injec-
tion. Deposition of part of the sulfur trioxide on the
larger fly-ash particles that were subject to mechanical
removal did not prevent sufficient deposition on the smaller
particles removed by electrostatic precipitation. Presumably,
the smaller surface area of the larger particles was the
predominant factor in ensuring adequate conditioning of the
smaller particles.
Another site of sulfur trioxide that might be used is the gas
duct upstream from the air preheater, in which the tempera-
ture of the flue gas is lowered from about 300°C to the
customary range around 150°C at the inlet of the electrostatic
precipitator. This site upstream from a preheater has been
seldom used, although it offers the theoretical advantages
of a temperature high enough to avoid loss of sulfur trioxide
by acid condensation and a longer time of contact between
the conditioning agent and the fly ash (more nearly simulating
the conditions of contact between the naturally produced gas
and the ash). Injection of sulfur trioxide upstream from
the air preheater may lead to acid corrosion of the cold-side
surfaces of the preheater, but it is questionable whether
the injection of the gas would cause a greater problem than
the natural occurrence of the gas at concentrations associated
with medium- and high-sulfur coals. Watson and Blecher11*
are among the few investigators who have used high-tempera-
ture injection upstream from an air preheater. Their results
were unfavorable, but the reason for the lack of success is
not known.
Concentration of Sulfur Trioxide Injected
A graph showing the results of resistivity determinations as
a function of the concentration of sulfur trioxide iniected
in various power plants is given in Figure 18. The several
power plants are denoted in this figure by the numbers that
118
-------
were listed previously in Table 1.* Also, the gas tempera
tures at which the precipitators operated are shown in th
figure. Information about still other parameters that are
assumed to be pertinent to the relationship between resist!
ity and injected concentration of sulfur trioxide is list d"
below:
Plant Gas Cleaning pH of ash
2 Mechanical and electrostatic 7.0
3 Electrostatic only 11.0
5 Mechanical and electrostatic 11.1
6 Electrostatic only 8.1
7 Electrostatic only 4>8
The pH values listed above are equilibrium values of ash-
water slurries.
Different curves are plotted in Figure 18 for the five plants
represented. Only two curves portraying the data for one
group consisting of Plants 3, 5, and 6 and another group con-
sisting of Plants 2 and 7 could be constructed to represent
the experimental data if the uncertainty in the measured
values of resistivity were one order of magnitude, which may
be justifiably assumed. However, any effort to limit the
number of correlations for the different plants is probably
not warranted because of the wide variations in the parameters
listed.
One of the principal conclusions warranted from Figure 18 is
that for fly ash with a resistivity of around 1 x 1012 ohm cm
prior to conditioning an injected concentration of 15 to
20 ppm of sulfur trioxide was sufficient to lower the resis-
tivity to an acceptable value of 1 x 10 ohm cm. Another
major conclusion is that for fly ash with a normal resistiv-
ity of about 1 x 10u ohm cm or less, an injected concentra-
tion of only 5 ppm of sulfur trioxide would have been adequate.
It is debatable whether the fly ash at Plant 7 (the Bull Run
plant), which had the lowest resistivity prior to conditioning
(between 1 x 101° and 1 x 10l: ohm cm), was in need of a
reduction in resistivity to improve its electrostatic
*Those olants for which names can be specified are as follows:
Plant 2!cSerokee Shit 2; Plant 5, Arapahoe Unit 4; Plant 1,
Bnii D,,« n«i** i (Collector IB).
^j-cinr /, cneroKee uiu-v- ~ • ~ \
Bull Run Unit 1 (Collector IB).
119
-------
1 3
Ni
O
10
I
• PLANT 2
A PLANT 3
• PLANT 5
PLANT 6
PLANT 7
PLANT 5 (135°C)
PLANT 2 (143°C)
PLANT 3 (110°C)
PLANT 6 (160°C)
PLANT 7 (125°C)
5 10 15 20 25 30 35
CONCENTRATION OF SO3 INJECTED, PPM
Figure 18. Resistivity as a function of the
concentration of injected sulfur trioxide
40
-------
was
precipitation. If the values of resistivity of uncondit-i™ *
ash that were measured at this plant on two separate oi~ ed
are valid, they are low enough to indicate that the facto°nS
limiting precipitation efficiency was something other than
the resistivity. If, on the other hand, the precipitate
electrical data obtained on one occasion (Figure 5) ar
into account, one concludes that lowering of the resisti ^£en
was helpful. However, a discrepancy between two sets ofV1
electrical data (Figures 5 and 7) further complicates the
interpretation of the effects of the conditioning process
(An effort to interpret the effects of conditioning of the
Bull Run plant in terms of a mechanism other than the lower-
ing of resistivity is given later on pages 128 through 131)
Chemical Properties of the Fly Ash
Even though the chemical composition of the fly ash treated
in various power plants varied widely, all of the fly ash wa
susceptible to conditioning with sulfur trioxide. This con-
clusion is apparent from the resistivity data plotted in
Figure 18. Still, the conclusion given earlier, that the re-
quired concentration of sulfur trioxide varies from one ash
to another, must be kept in mind.
One aspect of the chemistry of fly ash that appears to be
related to the conditioning process is the change in the con-
centration of sulfate on the surface of the ash with condi-
tioning. Resistivity data for conditioned and unconditioned
ash are shown in Figure 19 as functions of the concentration
of sulfate dissolved in aqueous slurries. The data in this
figure can be used to discuss two questions: (1) Does the
quantity of sulfate on unconditioned ash give an indication
of the magnitude of resistivity and the need for conditioning?
(2) Does the quantity of sulfur trioxide added to the ash
during conditioning give an indication of the success in
lowering resistivity?
The apparent answer to the first question is negative. The
two ashes with the highest concentrations of sulfate prior
to conditioning had resistivities of about 2 x 10ll and
4 x 1012 ohm cm. The fact that these ashes were the only
ones collected between mechanical and electrostatic dust col-
lectors, which caused them to have the highest area-to-mass
ratios, is the probable cause of their high sulfate concentra-
tions.
The apparent answer to the second question is also apparently
negative The ash from Plant 5 (Arapahoe Unit 4) required
a much larger increment in sulfate concentration than any
of the other materials to reach a resistivity value of around
1 x 1010 ohm cm? The behavior in this ash is attributed to
121
-------
to
to
10
1 3
10
10
I 2
1 1
s
u
O 10
1Q
JH
H
H
>
g
M
W
w
10
10
^ PLANT 6
(160°C)
PLANT 2 (143°C)
PLANT 5
(135°C)
PLANT 7 (125°C)
0
0.5 1.0 1.5 2.0 2.5 3.0
CONCENTRATION OF SULFATE IN ASH, WT %
3.5
4.0
Figure 19. Resistivity as a function of the
sulfate concentration in fly ash
-------
the highly basic nature of the material, which resulted '
conversion of much of the collected sulfur trioxide to ln
(perhaps calcium sulfate) that was not of low intrinsic3 S t
resistivity.
Temperature of the Flue Gas
Fly ash was successfully conditioned with sulfur trioxide
gas temperatures ranging from a maximum of 160°C to a mini
of 110°C. It is noteworthy that successful results were ob^™
tained at 160°C despite the fact that this temperature was
too high to permit the collection of sulfur trioxide by the
process of condensation and the neutral quality of the ash
was not favorable to base-induced adsorption of sulfur triox-
ide. It is noteworthy also that at the other extreme of tem-
perature, 110°C, the addition of sulfur trioxide was effective
even though at such a low temperature the conditioning of fiv
ash by the water vapor in the flue gas is greatly facilitated
In other words, at the low temperature the effect of a small '
concentration of added sulfur trioxide was readily apparent in
competition with the effect of the far higher concentration
of water vapor.
MECHANISMS OF COLLECTION OF SULFUR TRIOXIDE ON FLY-ASH
PARTICLES
There are at least two distinctly different mechanisms by
which sulfur trioxide in flue gas can be collected on the
surfaces of fly-ash particles. One mechanism is the condensa-
tion of a mixture of sulfur trioxide and water vapor, with
fly-ash particles serving as condensation nuclei. This phen-
omenon results in the formation of a liquid layer on the
particles; however, it can occur only if the temperature is
below the dew point of the vapors existing in the duct.
Other mechanisms involve the adsorption of sulfur trioxide
and probably the concurrent adsorption of water as well. AS
discussed subsequently, different mechanisms of adsorption
may exist, depending upon the sequence in which sulfur /tri-
oxide and water are adsorbed.
Acid Condensation
The phenomenon of acid condensation can be critically analyzed
only if reliable thermodynamic data exist for predicting
dew points of mixtures of sulfur trioxide and water vapor or
if reliable experimental data exist for showing dew points of
adequately analyzed vapor mixtures.
Two notable attempts to secure the necessary thermodynamic
data for predicting dew points have been made The first of
these two efforts was reported by Muller in 1959 and was
123
-------
based on data previously compiled by Greenewalt39 and Abel? "*°
the second effort, reportedly with the advantage of more
accurate background data, was reported by Gmitro and Vermeulen
in 1963 31 and 19641*1 and later summarized by Snowdon and Ryan
in 1969.42 Dew points predicted from the original work of
Greenewalt and Abel and the later work of Gmitro and
Vermeulen are shown in Figure 20 for various concentrations
of sulfuric acid at two different concentrations of water, 8
and 10% (covering the usual range found in flue gases pro-
duced in coal-burning plants) for an assumed total pressure
of 1 atm.* The curves in Figure 20 show that at a given set
of sulfur trioxide and water vapor concentrations the dew
point predicted by the data of Greenewalt and Abel is about
15°C higher than the value predicted by the data of Gmitro
and Vermeulen. They also show that at a given sulfur triox-
ide concentration the dew point predicted by either source of
data decreases about 3°C as the concentration of water vapor
is lowered from 10 to 8%.
Numerous experimental efforts have been made to determine
dew points of mixtures of sulfur trioxide and water vapor of
known compositions. One of the most recent and perhaps the
most reliable study was reported by Lisle and Sensenbaugh. *3
The results of this study indicate that the dew-point curves
based on the work of Greenewalt and Abel are more accurate
than those based on the work of Gmitro and Vermeulen/ despite
the access of the latter authors to more up-to-date reference
data.
To predict the possibility that acid condensation may have
occurred in the several plants where sulfur trioxide was used
as a conditioning agent, the range of reported concentrations
of the agent in each plant is shown by data points along a
horizontal dashed line in Figure 20. The conclusion reached
by comparing the locations of the dew-point curves is that
only the temperature at Plant 3 was clearly below the dew
point at all injected concentrations of conditioning agent.
The possibility that the temperature at Plant 5 or Plant 7
was below the dew point at the two higher concentrations of
conditioning agent cannot be excluded; however, the tempera-
ture at Plant 5 was definitely above the dew point at the
lowest concentration of the agent, which gave evidence of a
marked reduction in resistivity. Finally, the possibility
that the temperature in either Plant 2 or Plant 6 was ever
*Recalculation of dew points from the data of Greenewalt and
Abel was made, since the results of Muller's calculations
these data cannot be easily seen from Muller's graphical
summary.
124
-------
to
en
170
160
TEMP, PLANT 6
150
140
TEMP, PLANT 2
TEMP, PLANT
120
110
TT
iiir~r
DEW POINTS—
% H2O
DEW POINTS—
% H2O:
TEMP, PLANT 3~
SOURCES OF
DEW-POINT DATA;
REF. 31, 41 -
REF. 39, 40
TEMP. PLANT 5
100
I I I I I I I I I I
I 1 I I I I I I
4 6 8 10 20
CONCENTRATION OF SO3, PPM
40
60 80 100
Figure 20. Dew points of vapor mixtures of
sulfur trioxide and water
-------
below the dew point can be completely excluded, assuming that
the possible locations of the dew points cannot be outside
the range indicated by the curves in the figure.*
The above conclusions indicate, at least, that condensation
of acid is not necessary for the conditioning of fly ash by
sulfur trioxide. Although condensation may have been involved
in the conditioning process at Plant 3, it did not necessar-
ily occur; if chemisorption of the sulfuric acid on the highly
basic ash occurred more rapidly than the condensation process,
the concentration of sulfuric acid vapor could have been
lowered to levels below the minimum required for condensation.
Acid Adsorption
The above consideration of the condensation mechanism leads
to the conclusion that acid adsorption is at least a
sufficient mechanism, if not a necessary mechanism, for fly-
ash conditioning in the presence of sulfur trioxide.
The pioneering work by Chittum8 led to the conclusion that
conditioning of fly ash by an acidic vapor involves, first,
the chemisorption of this vapor on the fly ash and then
enhanced adsorption of water vapor on the chemically altered
surface of the fly ash. Chittum argued that a strongly basic
ash would have a greater affinity for an acidic vapor (such
as sulfur trioxide) than a neutral or acidic ash and, thus,
a strongly basic ash should be most easily conditioned.
The results of our work appear to be inconsistent with
Chittum1s hypothesis. This point can be supported by the
data obtained in the studies at Plants 2 and 5. The ash at
Plant 5 was more basic than that at Plant 2 and, moreover,
was conditioned at a slightly lower temperature. In view of
the differences in the conditioning parameters, the Chittum
hypothesis would predict more efficient conditioning of the
ash at Plant 5. The experimental data show that more effi-
cient conditioning occurred with the ash at Plant 2, whether
the criterion of effectiveness is based on the lowering of
resistivity observed with a given rate of injection of condi-
tioning agent or the lowering of resistivity observed with a
given amount of agent collected as sulfate on the ash.
*In the above discussion, the fact that the total pressure
was not precisely 1 atm in all of the plants is ignored. In
Plants 2 and 5 (located in the Denver area), the total pres-
sure was only about 0.83 atm. However, the conclusions
reached above would not be significantly altered by making a
correction for the difference in pressures.
126
-------
Additional data contrary to the Chittum hypothesis were
obtained at Plant 6, where the resistivity was lowered mark-
edly with either a low rate of injection of conditioning
agent or a small amount of agent collected on the ash desoH-
the near neutrality and high temperature of the unconditioned
ash (two factors that were unfavorable for chemisorption) .
The studies described in this report indicate that a highly
basic ash will require more acid conditioning agent than a
neutral ash, in terms of either the concentration injected in
the duct or the concentration collected in the ash, for the
following reasons: (1) the acid initially collected is not
simply adsorbed but it undergoes reaction with basic compo-
nents of the ash, such as calcium oxide, to form sulfate
salts, such as calcium sulfate, that may have little affinity
for water; (2) the amount of acid collected must increase to
the point that an adequately thick protective layer of sul-
fate salts covers the remaining basic components of the ash
and the amount of acid collected thereafter can exist on the
surface of the ash in the form of sulfuric acid rather than
as sulfate salts; (3) the layer of sulfuric acid ultimately
produced exhibits a strong affinity for water and, thus, a
layer consisting of both sulfuric acid and water with the
necessary electrical conductivity is finally produced on the
surface of the ash.
The fact that only a small amount of collected sulfate was
apparently required to condition the highly basic fly ash at
Plant 3 appears, on initial consideration, to be inconsistent
with the theoretical concepts outlined in the preceding para-
graph One possible rationale for the apparent discrepancy
was that conditioning in Plant 3 occurred at an unusually
low temperature, where the rate of diffusion of sulfuric acid
through the initially produced layer of sulfate salts would
be relatively low and, thus, the required thickness of the
protective layer of sulfate salts would be reduced.
MECHANISM OF SURFACE CONDUCTION ON FLY ASH CONDITIONED WITH
SULFUR TRIOXIDE
It is reasonable to assume that the predominant conductive
iMteriaTon fly ash conditioned with sulfur trioxide is sul-
Sric"cid? occurring as a surface film that may be only a
few molecular layers thick. The assumed predominance of sul-
fmMJ i^?2 II the conductive material is consistent with the
furic acid as the conauc ^ component of bulk aqueous
behavior of Jhis substance a Jur±c acid ±s ^
solutions. The as ™
made by southern Research
PH behavior of fly ash in aqueous
slurries.
127
-------
In aqueous solutions, the conductivity of sulfuric acid and
other strong acids is attributable to the high apparent mobil-
ity of hydrogen ions that are formed by ionization of the
acids. The mobility of hydrogen ions is so great in compari-
son with the mobility of other singly-charged cations that
the electrical migration of hydrogen ions is explained in
terms of a unique phenomenon termed the "proton-jump" process.
In this process, the transfer of charge occurs along a chain
of hydrogen ions and water molecules, rather than by the
direct migration that is required for other cations such as
sodium and potassium ions.
For the studies of the pH behavior of slurries of fly ash
and water discussed in this report, the major emphasis has
been placed on the pH values reached at equilibrium or steady
state. However, auxilliary experiments on the pH behavior
of fly-ash slurries, as described in a previous report,22
revealed that equilibrium pH values often do not give a true
indication of the acid-base composition of substances resid-
ing on the outermost surfaces of fly-ash particles. A fairly
common experimental result was the finding that, immediately
after contact was made between fly ash and water, the pH
decreased to values in the acid range at least momentarily,
even though the pH sometimes increased subsequently to equil-
ibrium values in the alkaline range. This result was attrib-
uted to the existence of (1) a discrete surface layer of
sulfuric acid, which dissolved rapidly to give the initial
acidic pH values, and (2) an excess of soluble base toward
the interior of the fly-ash particles, which neutralized the
surface acid by a slower process of dissolution. Additional
experiments involving the treatment of fly ash with ethanol
rather than water gave further evidence of the occurrence of
sulfuric acid as a surface material.22 Treatment with ethanol
rather than water was apparently a successful means of dis-
solving the surface acid without dissolving the interior base.
It may be necessary to revise the present concept of sulfuric
acid as the predominant conductor of electricity on condi-
tioned fly ash on the basis of an investigation of surf,ace-
conduction processes being conducted by Bickelhaupt of
Southern Research Institute.1*5 in a preliminary report,
Bickelhaupt has stated that alkali-metal ions, especially
sodium ion, carry a large fraction of the current over the
surfaces of fly-ash particles. His conclusion is based on
results obtained with a laboratory environment, where tempera-
tures range from about 100°C to 150°C and water vapor is main-
tained at concentrations of 9%-volume in an atmosphere of air
surrounding the fly ash. In later work, sulfur trioxide will
be added to the gaseous environment and the contribution of
alkali-metal ions will be reexamined. However, on the basis
of preliminary data, Bickelhaupt has suggested that sulfur
128
-------
trioxide conditioning may lead primarily to a process in
which alkali-metal ions are released from the glassy mat '
of fly ash and the numbers of these ions available to
electricity are thereby increased.
MECHANISM OF CONDITIONING BY SULFUR TRIOXIDE THROUGH
OTHER THAN LOWERING OF ELECTRICAL RESISTIVITY
Increasing the Gohesiveness of Fly-Ash Deposits on P
tor Electrodes —.
Different investigators15'46 have emphasized the importance
of the cohesiveness of fly-ash particles in maintaining the
physical integrity of the deposits and avoiding excessive
dispersal and reentrainment of individual particles and
agglomerates during electrode rapping. For fly ash of high
electrical resistivity, cohesive forces are not essential
for the high electric field that can be maintained through
deposits of this material will restrain losses by rapping
reentrainment. For fly ash of low resistivity, on the other
hand, cohesive forces are essential, for the restraining
force of the electric field is not adequate.
Dalmon and Tidy15 have conducted experiments in which the two
mechanisms of sulfur trioxide conditioning (decreasing resis-
tivity and increasing cohesiveness) were apparent. The
results obtained by these investigators showed that the elec-
trostatic precipitation of fly ash with an inherent high
resistivity was improved by sulfur trioxide through its
effect on resistivity. The results showed, on the other hand,
that the precipitation of fly ash with a low resistivity
(caused by the presence of a large amount of conductive carbon
particles) was improved by sulfur trioxide through the
increased cohesiveness of fly-ash particles.
Studies of sulfur trioxide conditioning at the Bull Run plant,
as described in this report, lead to the conclusion that the
predominant mechanism of conditioning was increased cohesive-
ness The resistivity of the unconditioned ash was not
abnormally high. Furthermore, the fact that the resistivity
was lowered by conditioning is not necessarily proof of the
importance of this process; indeed, the lowering of resistiv-
ity as the sole effect of conditioning might have led to a
lower efficiency of precipitator operation through enhanced
reentrainment losses. The evidence for conditioning through
inc?eased^hes?veness is admittedly indirect; even so, this
evident esoecially as obtained during the second of the two
investigates at the Bull Run plant (July 1974) , is worthy
of consideration. First, precipitation efficiency was in-
creased despite the possibility that increased reentrainment
129
-------
losses might have resulted from lowered resistivity. Second,
light-obscuration data obtained with the Lear-Siegler instru-
ment gave distinct evidence of reduced rapping-puff intensity.
Third, several visual observations were indicative of in-
creased cohesiveness, as follows:
• With conditioning, the ash collected in alundum
thimbles during precipitator efficiency determina-
tions adhered more or less uniformly to the entire
gas-filtration area; without conditioning, the ash
did not adhere to the filtration area but fell into
the bottom of the thimble.
• With conditioning, sampling pipes and other experi-
mental devices inserted into the gas duct were
densely coated with fly ash on surfaces facing the
gas flow. The devices inserted in the colder side
of the duct (where the uptake of sulfur trioxide
by the ash was more efficient) were more densely
coated with ash than the devices inserted at the
higher gas temperatures. The ash was held tena-
ciously to all collection surfaces, not being
removed simply by brushing but being removed only
by vigorous scraping. Without conditioning, coatings
were comparatively light and more easily removed.
• Thicknesses of ash deposits on the precipitator
electrodes are characteristically greater with con-
ditioning than without conditioning. (This state-
ment is based on observations made by TVA personnel
during periodic outages, which have allowed direct
inspections of the electrodes.)
Finally/ as evidence of increased cohesiveness, there is the
occurrence of the "snow-flake" phenomenon: the deposition
of large agglomerates of fly ash in the area underneath the
plume. (The snow-flake phenomenon is discussed later in
this report in the context of undesirable manifestations of
sulfur trioxide conditioning.)
Altering the Conductive Properties of Flue Gas
The pronounced effect of ammonia conditioning on the conduc-
tive properties of flue gas (the "space-charge" effect) has
been demonstrated in several power plants, as previously
described in this report. A similar effect of sulfur trioxide
conditioning may also occur through one of the following two
processes:
• lonization of the molecules of sulfuric acid intro-
duced by sulfur trioxide injection
130
-------
• lonization of small droplets of condensed
acid introduced when the injected sulfur trioxide
is below the dew point.
No direct evidence for either process was obtained durino +h
studies described in this report. However, some of the
results obtained at the Bull Run plant point to the possible
significance of one or both of these effects. These results
are discussed below; however, the two phenomena are discuss^
first from a theoretical point of view.
It is possible that the ionization of molecules of sulfuric
acid may take precedence over the ionization of molecules of
other electronegative gases. It has been clearly established
by studies of the ionization of air47 that the addition of
electronegative gases such as water vapor or sulfur dioxide
in the region of a negative corona causes a transfer of elec-
trons to the more electronegative species, producing less
mobile charge carriers and increasing the voltage required
for electrical breakdown of the gaseous mixture. The occur-
rence of charge transfer from oxygen molecules to sulfur
dioxide molecules, for example, depends upon the relative
affinities of the two gases for electrons and the relative
concentrations of the two gases.
It is not certain that the electronegativity of sulfuric acid
molecules would be sufficiently great to permit an extensive
transfer of charge to these molecules from other types of
molecules. With the concentration of sulfuric acid at a
typical value of about 10 ppm and the concentration of sulfur
dioxide at a typical concentration in excess of 500 ppm, the
electronegativity of sulfuric acid would have to be very high
indeed, relatively speaking, to permit the transfer of charge
to occur extensively. Unfortunately, there do not appear to
be any experimental data available at present from which the
efficiency of process can be estimated.
At least the feasibility of selective ionization of sulfuric
acid molecules can be evaluated, however, by comparing the
concentration of these molecules with the concentration of
gaseous ions occurring in an electrostatic precipitator. A
concentration of 10 ppm of sulfuric acid corresponds to a
molecular concentration of 1.73 x 101* molecules/cm3 (calcu-
lated for 150°C and 1 a tin) . A corona current density of
20 nA/cm2 at an electric field of 3 kV/cm in the gas stream
within a precipitator corresponds to an ionic concentration
of 1 97 x 107 ions/cm3 (calculated for the same temperature
and Measure with an assumed ion mobility of 2.2 cm2/(volt
sec))! Thus although the concentration of sulfuric acid is
low in comparison with the concentrations of other flue-gas
131
-------
components, its ratio to the concentration of gaseous ions is
of the order of 1 x 107. Hence, it is at least theoretically
possible for the sulfuric acid molecules to assume all of the
charge carried by ions and, as a result of their mass and
size/ to reduce substantially the effective mobility of
gaseous ions. One of the practical effects of sulfur trioxide
conditioning, therefore, may be a space-charge effect analo-
gous to that produced by ammonia conditioning.
It may be more easily seen how a mist of sulfuric acid can
produce a space-charge effect. If injected sulfur dioxide
is partly condensed as an acid aerosol of small particle
size, transfer of charge from gaseous ions to be aerosol
particles would have a pronounced space-charge effect,
analogous to that attributed to ammonium sulfate or bisulfate
particles during ammonia conditioning.
What practical reason is there for presenting the foregoing
arguments about a space-charge effect from sulfur trioxide
conditioning? During the studies described in this report,
only one type of experimental information was obtained to
suggest the need for considering the possibility of a space-
charge effect. This information was obtained during the
second investigation of sulfur trioxide conditioning at the
Bull Run plant (July 1974) when the precipitator voltage
currents with and without sulfur trioxide injection were com-
pared as shown in Figure 7. The effects of injection—
increasing the voltage required for a given current density
or decreasing the current density at a given applied voltage—-
at least suggest the occurrence of a space-charge effect in
view of the fact that no apparent change occurred in fly-ash
resistivity.
DELETERIOUS EFFECTS OF SULFUR TRIOXIDE CONDITIONING
Stack Losses of Sulfur Trioxide
It is sometimes claimed that injected sulfur trioxide i's
quantitatively collected on fly ash and none of the vapor
escapes to the stack and thence to the atmosphere.10 Under
some circumstances, this claim appears to be justified. For
instance, with high-alkalinity ash as studied at the Arapahoe
plant, little of the injected conditioning agent was found
in the flue gas entering the precipitator, and little if
any would be expected at the outlet. Under other circum-
stances, the claim of no stack loss of injected sulfur
trioxide must be challenged. For example/ with low-alkalinity
ash at Plant 6, a substantial fraction of the agent was found
as vapor at the precipitator inlet. In view of the finding
132
-------
by Cuf fe ejt aj^. * 8' * 9 of roughly equal concentrations of
naturally occurring sulfur trioxide at precipitator inlets
and outlets, no appreciable removal of the injected vapor
would be expected within the precipitator at Plant 6 or
other plants with similar circumstances. Finally, with
weakly acidic ash at the Bull Run plant, the study in July
1974 showed that roughly one-third of the injected sulfur
trioxide remained in the gas stream at both the inlet and
the outlet of the precipitator.
In view of the increasing concern about the toxicity of
sulfate particulates, it is important to regulate the rate
of sulfur trioxide injection to minimize loss to the stack
One state (West Virginia) actually prohibits sulfur trioxide
conditioning to avoid the problem. This extreme position
may not be justifiable. Even so, it is clear that further
research is needed to define the acceptable amount of stack
loss. Also, further research is needed to define the concen-
trations of injected sulfur trioxide under various power-plant
circumstances that will permit the desired degree of condi-
tioning to occur without allowing an unacceptable stack loss.
Fallout of Large Fly-Ash Aggregates from a Power-plant Plume
Currently, this phenomenon is the cause of considerable annoy-
ance at the Bull Run plant, where it is referred to as "snow
flaking." With injected concentrations of sulfur trioxide of
the order of 20 ppm, which are found to give the best improve-
ment in precipitator performance, fallout of large fly-ash
aggregates in the vicinity of the plant causes complaints from
the plant personnel and the residents of the area around the
plant.
The snow-flake problem was observed during the investigation
at the Bull Run plant in July 1974, and an effort was made
to determine the factors that contribute to the problem.
Direct observations and information obtained from the plant
personnel indicate that the fallout of slow flakes is more
severe under conditions of high ambient humidity, which
occur at night or in early morning. Consideration was given
to possible origins of the material: (1) reentrainment from
the orecipitator electrodes or the walls of the outlet duct
and the stack and (2) formation of aggregates of fly ash
within the plume issuing from the stack. The apparent rela-
tionshio of the problem to ambient conditions suggests forma-
tion ?n fhe plume. Also, observations that the surfaces of
individual aggregates are rounded rather than flat before
impact of the material on the ground suggest formation in
the DluSe If reentrainment were the origin of the material,
flit surfaces prior to impact would be expected. Actually,
the deposited particles always have flat surfaces against*'
133
-------
the plane of impact, but such surfaces appear to be produced
only during impact. The final conclusion was that aggrega-
tion occurred in the plume as the direct result of condensa-
tion of excess sulfur trioxide in the presence of high
moisture levels, with the resulting increase in cohesive
forces between fly-ash particles.
It is ironic that the flake problem becomes severe when the
concentration of injected sulfur trioxide reaches the optimum
level for good precipitator performance. Unfortunately, it
is not possible to recommend any measure that would overcome
the problem, at least not on the basis of information pres-
ently available. A worthwhile task in further research would
be to survey power plants using sulfur trioxide conditioning
to determine how common the problem is and to make an effort
to deduce more definitely what factors contribute to the
difficulty.
134
-------
SECTION VIII
DISCUSSION OF CONDITIONING WITH AMMONIA
As in the instance of fly-ash conditioning with sulfur trio
ide, there are both practical and theoretical questions a««
ciated with the use of ammonia as an alternative agent th
require discussion. These questions are similar to those
dealt with in the preceding discussion of sulfur trioxide
What are the circumstances in a power plant that permit ail
nia to be used effectively? What are the mechanisms of
conditioning by ammonia? What disadvantages are associated
with ammonia conditioning?
PARAMETERS OP A POWER-EMISSION SYSTEM THAT AFFECT THE EFFT
CIENCY OF AMMONIA CONDITIONING
Fly Ash from Low-Sulfur Coals
For treating fly ash from low-sulfur coals, ammonia condition-
ing was investigated under more limited circumstances than
sulfur trioxide conditioning. Thus, it is difficult to qive
a thorough discussion of the power-plant parameters that
affect the efficiency of ammonia conditioning. Moreover it
is difficult to compare ammonia and sulfur trioxide as condi-
tioning agents in plants burning low-sulfur coals because of
the lack of opportunities to make direct experimental compari-
sons under essentially identical circumstances.
On the basis of direct experience, it can be stated that
ammonia conditioning produced desirable results in two power
plants burning Eastern coals containing about 1% sulfur, in
these plants, the fly ash had moderate to moderately high
electrical resistivities (about 3 x 1010 to 3 x 101"1 ohm cm)
at gas temperatures around 130°C. Moreover, the fly ash was
mildly acidic, and the flue-gas stream contained naturally
produced sulfur trioxide at concentrations of about 2 to
5 ppm. Under these conditions, ammonia conditioning had no
apparent effect on the resistivity of the fly ash, but it did
produce a marked enhancement of the electric field in the
interelectrode space of the precipitator through a space-
charge effect.
As demonstrated by precipitator efficiency data obtained with
and without conditioning, the practical results of ammonia
135
-------
conditioning were decidedly favorable in one plant (Widows
Creek Unit 7). Practically speaking, the results in the
second plant (Bull Run Collector 1C) were not fully assessed,
for no comparative determinations of precipitator efficiencies
were made with and without conditioning. However, TVA
regarded the results of ammonia conditioning as relatively
unsuccessful on the basis on light-obscuration data obtained
with the Bailey bolometer and on the basis of the electrical
behavior of the precipitator, which was characterized by
intensive sparking during the injection of ammonia.
Because of the similarity in the circumstances at the two
plants and the parallel in the mechanistic aspects of ammonia
conditioning in both plants, it is important to attempt to
give an explanation of the apparent difference in precipitator
performances. The space-charge effect of ammonia at the Bull
Run plant was undoubtedly the direct cause of intensive spark-
ing; however, if the power-supply controls had operated more
satisfactorily, they should have been able to suppress the
intensity of sparking. Another factor of possible importance
at the Bull Run plant in addition to unsatisfactory perfor-
mance of the power supplies was electrode misalignment. In
the inlet section of the Bull Run precipitator, the onset of
sparking during ammonia injection occurred at a voltage around
35 KV and an average current density below 10 nA/cm2. In the
corresponding section of the Widows Creek precipitator, the
onset of sparking during injection did not occur until the
voltage exceeded 40 kV and the average current density
approached 40 nA/cm2. The Bull Run precipitator admittedly
has a narrower designed wire-to-plate spacing (11.4 cm) com-
pared with the Widows Creek precipitator (12.7 cm). However,
the difference in design spacings seems an unlikely explana-
tion for the pronounced difference in electrical data, and a
significant distortion from the design spacing in the Bull
Run precipitator seems to be a more likely explanation. In
summary, it does not appear reasonable to attribute the lack
of success of ammonia conditioning at the Bull Run plant to a
fundamental shortcoming in the mechanism of conditioning.
Studies at the Bull Run plant provided the only data for
ammonia and sulfur trioxide as alternative conditioning agents
under essentially the same circumstances, it does not appear
that even at this plant a meaningful comparison of the two
agents can be made, largely because of the practical problem
of sparking with ammonia conditioning as discussed in the pre-
ceding paragraph. Clearly, however, the two agents performed
by quite different mechanisms: ammonia through the space-
charge effect in the interelectrode space and'sulfur trioxide
by lowering the resistivity of the fly ash (and perhaps both
by increasing the cohesiveness of the ash). The conclusion
reached by TVA was that more successful results were obtained
136
-------
with sulfur trioxide; hence, the decision was made to disman
tie the ammonia system and install an enlarged sulfur trioxirt
system. x±ae
It is possible only to make conjectures about how ammonia
would perform as a conditioning agent for fly ash from low
sulfur Western coals, for which sulfur trioxide was directlv
investigated in a number of power plants. Circumstances in
these plants differ from those in which ammonia was studied
experimentally in the following ways:
• Sulfur in the coal, typically nearer 0.5% than
1.0%
• Resistivity of the fly ash, usually higher than
the values for ash treated with ammonia
• Alkalinity of the fly ash, usually higher than
that of the ash treated with ammonia (creating a
favorable environment for the collection of sulfur
trioxide but a hostile environment for the adsorp-
tion of ammonia)
• Sulfur trioxide concentration produced naturally
in the flue gas, lower than that in the plant
where ammonia was injected (making the reaction of
ammonia to produce the space-charge effect more
unlikely)
Although any conjectures that can be made about the effects
of ammonia conditioning under the different circumstances out-
lined above may be misleading, a tentative estimation of the
utility of ammonia conditioning under these circumstances is
given in a later discussion of conditioning mechanisms (pages
138 through 141) .
Fly Ash from High-Sulfur Coals
As an aid in improving the collection of fly ash from high-
sulfur coals, ammonia conditioning clearly appears to be of
value as demonstrated by investigations at the widows Creek
and Gallatin plants. The mechanism does not appear to involve
the resistivity of the ash, but it appears instead to consist
of a space-charge effect and an increase in the cohesiveness
of the ash. The latter two conditioning mechanisms may lead
not only to improved collection of fly ash but to elimination
of part of the sulfur trioxide gas that is condensed to sul-
furic acid mist when flue gas is evolved to the atmosphere
and thus cooled to a temperature below the acid dew point.
The theoretical aspects of the two alternative conditioning
mechanisms and the environmental impact of ammonia
137
-------
conditioning in power plants burning high-sulfur coals are
discussed further in the remainder of this report.
MECHANISMS OF AMMONIA CONDITIONING
Alteration of the Electrical Resistivity of Fly Ash
Data on the electrical resistivity of fly ash presented
earlier in this report indicate that ammonia conditioning had
no clear-cut effect on this property of the ash under the
Particular circumstances investigated. Treatment of fly ash
in two plants burning low-sulfur coals did not produce a
measurable reduction in resistivity, contrary to the result
tentatively expected. Treatment of ash in two plants burning
high-sulfur coals did not produce a measureable increase in
resistivity, contrary to the hypothesis suggested previously.5
(Some of the data for the ash from high-sulfur coals indicated
a possible lowering of resistivity—an unwanted effect—but
the evidence for lowered resistivity was not conclusive).
As stated previously in the summary of other work on ammonia
conditioning (Section III), some investigators in the past
have failed to observe changes in resistivity with ammonia
conditioning. On the other hand, other investigators have
reported significant changes in resistivity (sometimes
increases and sometimes decreases, depending upon circum-
stances). Whether the observed changes were real effects or
spurious experimental phenomena cannot now be ascertained.
It is noteworthy, however, that even in the absence of mea-
sured changes in resistivity the electrical data for some pre-
cipitators have shown strong evidence of the suppression of
back corona, a common manifestation of excessive resistivity.
It is necessary, therefore, to conclude that a lowering of
resistivity actually occurred in these investigations or to
offer an alternative explanation for the electrical data.
The discussion immediately following offers some hypotheses
about the circumstances under which a lowering of high resis-
tivity may occur. The discussion under the subsequent head-
ing of the space-charge effect gives a different rationale
for suppression of back corona (more exactly, suppression of
the outward electrical manifestation of back corona).
The surface of fly ash should not, in many instances, consti-
tute a favorable environment for the adsorption of ammonia.
Moreover, the physical properties of ammonia—volatility, in
particular—are such that the adsorption of the compound
should not be readily accomplished under typical flue-gas
conditions: a very low partial pressure of ammonia (of the
order of 10 ppm) and a reasonably high temperature (around
150°C). Under what circumstances, therefore, can surface
conditioning of fly ash by ammonia occur?
138
-------
Fly ash with acidic surface material may be receptive to th
adsorption of ammonia. However, surface material with
strongly acidic properties is most likely to be adsorbed
sulfur trioxide. It seems improbable that the resistivity
of the normal occurring adsorbed material would be measurablv
lowered by the additional adsorption of ammonia, which would
probably have the chemical effect of converting the acid to
an ammonium salt with an increased resistivity.
Fly ash with alkaline surface material should be hostile to
the adsorption of ammonia, except perhaps with simultaneous
adsorption of some other gas that could mask the inherent
alkalinity of the original surface material. One gas that
could intervene in the manner described is water vapor.
Indeed, there have been reports that the effectiveness "of
ammonia conditioning is increased by simultaneous injection
of ammonia and water vapor; however, one of the anomalies
in these reports is that the quantity of water vapor added
would not raise the concentration of water vapor substan-
tially from its normal level in flue gas. Other gases that
could intervene to aid the adsorption of ammonia are sulfur
trioxide and sulfur dioxide. Sulfur trioxide might be pres-
ent at a very low concentration and thus might not be effec-
tively adsorbed; however, with the possibility of sulfur
trioxide and ammonia reacting to form a nonvolatile salt on
the fly-ash surface, the effectiveness of adsorption of both
gases might be enhanced. Sulfur dioxide, although always
present in flue gas at much higher concentrations than
sulfur trioxide, should be difficult to collect on a fly-
ash surface, either alone or in combination with ammonia
(ammonium sulfites are much less stable than ammonium
sulfates). There is the possibility, however, that the com-
bination of sulfur dioxide and ammonia might be rapidly
oxidized to produce an ammonium sulfate as a stable product
on the fly-ash surface, provided they undergo combination to
form at least a small amount of sulfite as an intermediate.
The rationale for this assumption is the evidence that the
oxidation of sulfur dioxide to sulfur trioxide in the atmos-
phere is catalyzed by ammonia, with ammonium sulfate as a
stable reaction product.
of the Space-Charge Component of the Electric
Field ' "
As previously indicated in this report, the space-charge
effect is attributed to the reaction between injected ammonia
and normally occurring sulfur trioxide in the presence of
water vapor to form a fine parti culate of ammonium sulfate or
ammonium bisulfate. lonization of the reaction product within
an electrostatic precipitator, by means of charge transfer
from gaseous ions, produces carriers of electricity that are
139
-------
less mobile on the average than the carriers otherwise present,
As a consequence, a higher voltage is required in the precipi-
tator to maintain a given current density. Insofar as the
collection of fly-ash particles is concerned, the benefit
derived from the space-charge effect is a higher electric
field in the interelectrode space for charging the fly ash
and driving the particles to the collection electrodes.
The occurrence of the space-charge effect was supported
experimentally by several types of observations:
• Rapid electrical changes in the precipitator
as ammonia injection was either started or
stopped
• Rapid appearance of fine particulate (median size
by number of the order of 0.05 ym) as ammonia
injection was started or disappearance of the
material as injection was discontinued
• Absence of significant concentrations of ammonia
in the gas phase
• Presence of ammonia as a constituent of particulate
samples collected on filters and the precipitator
hoppers
• Loss of sulfur trioxide from the flue gas during
ammonia injection
Each of these observations, with one exception, was observed
in all of the power plants using ammonia conditioning. The
one exception was that involving detection of the appearance
and disappearance of fine particulate by means of condensation-
nuclei counting. This experimental technique was not employed
in the first power-plant study, which was conducted with low-
sulfur coal at the Widows Creek plant. There is no reason,
however, for suspecting that the use of this experimental
method would have failed to yield information similar to that
found elsewhere.
In each of the power-plant studies conducted with ammonia con-
ditioning, the flue-gas environment was conducive to the
occurrence of the space-charge effect. Specifically, adequate
concentrations of sulfur trioxide were present as a result of
fuel combustion, and flue-gas temperatures were in a range
suitable for the sulfur trioxide to react with the injected
ammonia. In addition, the resistivity of the fly ash was not
excessive, lying in the range from about 3 x 108 to 3 x 1011
ohm cm. Thus, the precipitators were able to operate at
acceptable current densities without excessive sparking,
140
-------
except in the instance of the Bull Run plant, even with the
added voltage drop imposed with ammonia conditioning.
(Reasons for the exceptionable behavior at the Bull Run plant
have been attributed in previous discussion to the probability
of electrode misalignment.)
The necessary conditions for the space-charge effect were
satisfied in all of the studies even though coals ranging
from.about 1% to 4% in sulfur content were burned as fuels.
The availability of sufficient naturally-produced sulfur
trioxide, despite the range in sulfur concentrations in the
coals, apparently can be attributed to the chemical composi-
tions of the fly ash. The compositions of the ash from the
low sulfur coals are to be noted especially; they were low
enough in alkaline components to prevent removal of sulfur
trioxide from the flue gas and, thus, they were able to allow
enough of this compound to remain in the gas phase to undergo
reaction with the injected ammonia.
Increasing the Cohesiveness of Fly-Ash Particles
The clearest evidence of this effect was obtained at the
Gallatin plant, where high-sulfur coal was used as the fuel
and the fly-ash resistivity was low (about 3 x 108 ohm cm).
It was apparent that ammonia conditioning was effective in
overcoming loss of collected fly ash by rapping reentrainment,
a problem expected to be more severe with low-resistivity ash
than with high-resistivity ash. With low-resistivity ash,
the electric field in the deposited ash may be insufficient
to prevent reentrainment and it may even produce a positive
force to intensify reentrainment; *** with this material, there-
fore, the action of surface effects between particles to main-
tain the physical integrity of aggregates on the collection
electrodes of a precipitator may be needed. With high-
resistivity ash, the electric field may provide the only
restraint needed to prevent reentrainment.
The exact nature of the process by which ammonia conditioning
increases the cohesiveness of fly ash has not been identified.
It is assumed, however, to involve a.chemical reaction between
the injected ammonia and the naturally occurring sulfur triox-
ide. Of the two possible products, ammonium sulfate and
ammonium bisulfate, the latter compound can more reasonably
be expected to increase cohesiveness because of its occurrence
as a liquid at temperatures below 144°C.28
141
-------
DELETERIOUS EFFECTS OF AMMONIA CONDITIONING
Enhancement of Sparking
This difficulty was encountered at the Bull Run plant during
ammonia conditioning and was assumed to be a manifestation
of the space-charge effect. Under the particular circumstances
of this plant, however, the enhancement of sparking was
attributed to electrode misalignment and possibly to poor
power-supply response. With the fly ash having only a moder-
ately high resistivity, the problem would not have been
expected otherwise.
The difficulty of enhanced sparking may, however, be serious
in power plants with well-aligned precipitator electrodes if
the resistivity of the fly ash is normally high and not
lowered by ammonia injection. In particular, the problem may
be quite serious in power plants burning low-sulfur Western
coals. In such plants, the net effect of ammonia injection
may be deleterious unless the conditioning agent lowers the
resistivity of the ash as well as creating the space-charge
effect. Further research is needed to assess the net effect
of ammonia conditioning under the circumstances indicated.
Increase of Fine-Particle Emission
In view of the present concern about the emission from elec-
trostatic precipitators of fine particles (smaller than 1 or
2 urn in effective diameter), the generation of fine particles
by ammonia injection would appear to be an objectionable
aspect of this conditioning process. However, there are at
least two observations that tend to minimize concern about
the atmospheric impact of fine-particle generation. One
observation is that a significant fraction of the particulate
produced by ammonia conditioning is removed in a precipitator.
Another observation is that a large part of the sulfur tri-
oxide present in flue gas is converted to a collectible par-
ticulate; this means that the concentration of sulfur trioxide
that can appear in the plume is lowered and that the con-
centration of sulfuric acid mist formed by condensation in
the plume is lowered. The net result, therefore, may be an
effective reduction of fine particles in the plume when
ammonia conditioning is employed. Further research is needed
to clarify this matter.
142
-------
SECTION IX
REFERENCES
1. White, H. L. Industrial Electrostatic Precipitation.
Reading (Mass.), Addison-Wesley Publishing Company, inc
1963. p. 294-330.
2. Oglesby, S., and G. B. Nichols. A Manual of Electro-
static Precipitator Technology: Part I—Fundamentals.
Southern Research Institute, Birmingham, Ala. Contract
CPA 22-69-73. The National Air Pollution Control Admin-
istration, Cincinnati, Ohio. August 25, 1970. p. 166-
186.
3. Archer, W. E. Electrostatic Precipitator Conditioning
Techniques. Power Eng. 76:50-53, December 1972.
4. Dismukes, E. B. Conditioning of Fly Ash with Sulfamic
Acid, Ammonium Sulfate, and Ammonium Bisulfate. Southern
Research Institute, Birmingham, Ala. Contract 68-02-1303.
Environmental Protection Agency, Research Triangle Park,
N. C. October 1974. Publication No. EPA-650/2-74-114.
51 p.
5 Reese, J. T., and J. Greco. Experience with Electro-
static Fly-Ash Collection Equipment Serving Steam-
Electric Generating Plants. J. Air Pollut. Contr. Assoc.
18:523-528, August 1968.
6 Baxter, W. A. Recent Electrostatic Precipitator Experi-
ence with Ammonia Conditioning of Power Boiler Flue
Gases. J. Air Pollut. Contr. Assoc. 18:817-820,
December 1968.
7 qtull D R , and H. Prophet (ed.). JANAF Thermochemical
Tables 'Washington, National Bureau of Standards, 1971.
Unnumbered pages listed in this alphabetical order: H2O,
, and 03S.
J F. Western Precipitation Corporation, Los
Anaee Calif. Unpublished data from studies in 1942-
19457 (For excerpts, see Reference 1).
« ,. v r T and K. Darby. Efficiency of Electro-
' ipitates as Affected by the Properties and
Coal. J. In.t. Fuel (London). 36:184-197,
May 1963.
143
-------
10. Busby, H. G. T., C. Whitehead, and K. Darby. High Effi-
ciency Precipitator Performance on Modern Power Stations
Firing Fuel Oil and Low Sulphur Coals. Lodge-Cottrell,
Ltd. (Presented at Second International Clean Air Con-
gress of the International Union of Air Pollution Control
Association. Washington. December 6-11, 1970.) 56 p.
11. Darby, K., and D. O. Heinrich. Conditioning of Boiler
Flue Gases for Improving Efficiency of Electrofilters.
Staub Reinhaultung Luft (English edition). 26:12-17,
November 1966.
12. Coutaller, J., and C. Richard. Amelioration du
Depoussierage Electrostatique par Injection de SO3
[Improvement of Electrostatic Precipitators by Injection
of SOsl. Pollut. Atmos. (Paris). 9:9-15, January-March
1967.
13. Schrader, K. Improvement of the Efficiency of Electro-
static Precipitation by Injecting SO3 into the Flue Gas.
Combustion. 42(4):22-28, October 1970.
14. Watson, K. s., and K. J. Blecher. Further Investigation
of Electrostatic Precipitators for Large Pulverized Fuel-
Fired Boilers. Air Water Pollut. Int. J. (Oxford,
England). 10:573-583, September 1966.
15. Dalmon, J., and D. Tidy. The Cohesive Properties of Fly
Ash in Electrostatic Precipitation. Atmos. Environ.
(Oxford, England). 6:81-92, February 1972.
16. Dalmon, J., and D. Tidy. A Comparison of Chemical Addi-
tives as Aids to the Electrostatic Precipitation of Fly-
Ash. Atmos. Environ. (Oxford, England). 6:721-734,
October 1972.
17. Voltz, S. E., and S. W. Weller. Effects of Ammonia on
the Electrical Resistivity of Silica-Alumina Catalysts.
J. Phys. Chem. 62:574-578, May 1958.
18. Tassicker, 0. J. Wollongong University College, The
University of New South Wales, Australia (present affil-
iation, Electric Power Research Institute, Palo Alto,
Calif.). Private communication, July 1972.
19. Saponja, W. Calgary Power Ltd., Calgary, Alberta. Pri-
vate communication. May 1972.
20. Bias, J. G. Termicus Asturianus, Oveido, Spain. Private
communication. June 1973.
144
-------
21. Dalmon, J. Central Electricity Research Laboratories,
Leatherhead, Great Britain. Private communication.
July 1973.
22. Dismukes, E. B. A Study of Resistivity and Conditioning
of Fly Ash. Southern Research Institute, Birmingham,
Ala. Contract CPA 70-149. Environmental Protection
Agency, Research Triangle Park, N. C. February 1972
Publication No. EPA-R2-72-087 (PB 212607) . 138 p.
23. Dismukes, E. B. Unpublished data. Southern Research
Institute, Birmingham, Ala. Research Agreement TV36921A
with the Tennessee Valley Authority. June 1972 through
July 1974.
24 SO 3 Injection to Aid Stack Cleanup? Electric World.
173:22-24, June 1970.
25 Green, G. P., and W. S. Landers. Operating Experience
with Gas Conditioned Electrostatic Precipitators. (Pre-
sented at Symposium on Control of Fine-Particulate Emis-
sions from Industrial Sources under Sponsorship of the
U.S.-U.S.S.R. Working Group on Stationary Source Air
Pollution Control Technology. San Francisco. January
15-18, 1974.) 19 p.
26 Cohen, L./ and R. W. Dickinson. The Measurement of the
* Resistivity of Power Station Fine Dust. J. Sci. Instrum.
40:72-75, 1963.
27 Nichols, G. B. Techniques for Measuring Fly Ash Resis-
tivity. Southern Research Institute. Contract 68-02-0284
Environmental Protection Agency, Research Triangle Park,
N.C. Publication EPA-650/2-74-079. August 1974. 43 p.
2R Kellev, K. K., C. H. Shomate, F. E. Young, B. F. Naylor,
A E Salo, and E. H. Huffman. Thermodynamic Properties
of Ammonium and Potassium Alums and Related Substances,
th Reference to Extraction of Alumina from Clay and
Alunite Bureau of Mines, Washington, D. C. Technical
Paper 688. 1946. p. 66-69.
« T«=r, K J Environmental Protection Agency, Research
T?ianSle Park, N. C. (present affiliation, Wollongong
university College, The University of New South Wales,
Australia! Private communication, May 1971.
p Beitrag zur Frage der Einflusses der
Swelfeisaure auf die Rauchgas-Taupunktemperatur. Chem.-
Ina.-Techn. 31:345-350, 1959.
30.
Ing.-Techn.
145
-------
31. Gmit.ro, J. I., and T. Vermeulen. Vapor-Liquid Equilibria
for Aqueous Sulfuric Acid. University of California,
Lawrence Radiation Laboratory, Berkeley, Calif. Contract
W-7405-eng-48. June 24, 1963. 81 p.
32. Stull, D. R., and H. Prophet (Reference 1). Unnumbered
pages listed in this alphabetical order: H2O, HaOitS,
and 02S.
33. Electrostatic Fly-Ash Collector Performance Tests; Bull
Run Steam Plant Unit 1; March 16-18, 1971. Results
Report No. 69. Tennessee Valley Authority, Chattanooga,
Tenn.
34. Electrostatic Fly-Ash Precipitator Performance Tests
with SOs Gas Conditioning; Bull Run Steam Plant Unit 1,
Precipitator B; August 1-31, 1972. Results Report
No. 77. Tennessee Valley Authority, Chattanooga, Tenn.
35. Widows Creek Steam Plant Unit 7 Electrostatic Fly-Ash
Collector Efficiency Tests While Burning Low-Sulfur Coal;
June 26-27 and July 18-21, 1972. Results Report No. 72.
Tennessee Valley Authority, Chattanooga, Tenn.
36. McCain, J. D., K. M. Gushing, and W. B. Smith. Measure-
ment of the Fractional Efficiency of Pollution Control
Devices. Southern Research Institute, Birmingham, Ala.
(Presented at the Air Pollution Control Association
Meeting. Denver. June 9-12, 1974). 29 p.
37. Electrostatic Fly-Ash Collector Performance Tests;
Gallatin Steam Plants Units 2, 3, and 4; December 5,
1972-February 8, 1973. Results Report No. 74. Tennessee
Valley Authority, Chattanooga, Tenn.
38. Zarfoss, J. R. Environmental Elements Corporation (Sub-
sidiary of the Koppers Company), Baltimore, Md. Private
communication, May 1974.
39. Greenewalt, C. W. Partial Pressure of Water out of
Aqueous Solutions of Sulfuric Acid. Ind. Eng. Chem.
17:522-523, May 1925.
40. Abel, E. The Vapor Phase above the System Sulfuric Acid-
Water. J. Phys. Chem. 50:260-283, 1950.
41. Gmitro, J. I., and T. Vermeulen. Vapor-Liquid Equilibria
for Aqueous Sulfuric Acid. Am. Inst. Chem. Eng. J. 10:
740-746, September 1964.
146
-------
42. Snowdon, P. N., and M. A. Ryan. Sulphuric Acid Condensa
tion from Flue Gases Containing Sulphur Oxides J Tn +.
Fuel (London). 42:188-189, May 1969. " * nst<
43. Lisle, E. S., and J. D. Sensenbaugh. The Determination
of Sulfur Trioxide and Acid Dew Point in Flue Gases.
Combustion. 36:12-16, January 1965.
44. Robinson, R. A., and R. H. Stokes. Electrolyte Solu-
tions. London, Butterworths Scientific Publications
1955. p. 116-117.
45. Bickelhaupt, R. E. Surface Resistivity and the Chemical
Composition of Fly Ash. Southern Research Institute,
Birmingham, Ala. (Presented at the Symposium on Electro-
static Precipitators for the Control of Fine Particles
Pensacola Beach, Fla. September 30-October 2, 1974 ) '
20 p.
46. Penney, G. W., and E. H. Klinger. Contact Potentials
and the Adhesion of Dust. Trans. AIEE 81(1):200-204,
July 1962.
47. Oglesby, S., and G. B. Nichols (Reference 2). p. 23-56.
48. Cuffe, S. T., R. W. Gerstle, A. A. Orning, and C. H.
Schwartz. Air Pollutants from Coal-Fired Power Plants;
Report No. 1. J. Air Pollut. Contr. Assoc. 14:353-362,
September 1964.
49. Gerstle, R. W., C. T. Cuffe, A. A, Orning, and C. H.
Schwartz. Air Pollutants from Coal-Fired Power Plants7
Report No. 2. J. Air Pollut. Contr. Assoc. 15:59-64,
February 1965.
50 Carabine, M. D. Interactions in the Atmosphere of Drop-
lets and Gases. Chem. Soc. Rev. (London). 1:411-429,
1972.
51 Fritz/ J. S., and S. S. Yamamura. Rapid Microtitration
of Sulfate. Anal. Chem. 27:1461-1464, September 1955.
52 Fielder, R. S., and C. H. Morgan. An Improved Titri-
" metric Method for Determining Sulphur Trioxide in Flue
Gas Anal. Chim. Acta. 23:538-540, 1960.
147
-------
SECTION X
APPENDIX. EXPERIMENTAL METHODS
ANALYSIS OF COAL
Samples of coal were analyzed systematically for sulfur and
ash contents by procedures included in ASTM Method D-271.
Some of the samples were also analyzed for moisture and heat
value by other procedures in this Method.
DETERMINATION OF THE ELECTRICAL RESISTIVITY OF FLY ASH
All of the data on the electrical resistivity of fly ash
presented in this report were obtained by collecting fly ash
at the inlets of electrostatic precipitators and measuring
the resistivity with the samples in sitia. Three different
devices were used for this purpose. Two that were employed
only in some of the initial studies of sulfur trioxide con-
ditioning were of the cyclone type. The third device, which
later became the standard apparatus for determining resistiv-
ity, was of the point-plane type.
Cyclone Resistivity Probes
Apparatus Developed by Cohen and Dickinson26—
The essential features of this apparatus are shown in Figure
21. Furthermore, the details of the operating procedure
are given in the publication by Cohen and Dickinson.26 The
principal steps in the operation of this apparatus, however,
are summarized below:
1. The assembly shown in the figure is placed near
the flue-gas duct at the inlet of a precipitator
and brought to the same temperature as the flue
gas.
2. A stream of flue gas is pumped from the duct to
the cyclone, where fly ash is deposited under
centrifugal force.
3. The contents of the cyclone are rapped into the
resistivity cell, which consists of two concen-
tric cylindrical electrodes.
148
-------
EXHAUST FROM
CYCLONE
THERMOMETER
INLET FROM
SAMPLING
PROBE
CYCLONE
HEATER
RESISTIVITY
CELL
CONNECTION
TO MEGOHMETER
Piaure 21. Resistivity apparatus using a
mechanical cyclone dust collector
(Cohen and Dickinson*°)
149
-------
4. A known voltage is applied across the fly-ash
sample and the resistivity of the ash is calcu-
lated from the applied voltage, the measured cur-
rent/ and the geometrical features of the resis-
tivity cell.
Nichols2 7 has discussed various problems encountered with the
use of this apparatus. One of the most serious disadvantages,
which led to the use of other resistivity probes during this
investigation, is ensuring a constant temperature as the fly-
ash sample is withdrawn from the duct and deposited in the
resistivity cell.
Apparatus Designed at Southern Research Institute27—
A modification of the apparatus of Cohen and Dickinson that
permitted the cyclone collector and resistivity to be placed
within a flue-gas duct is shown in Figure 22. This apparatus
was constructed at Southern Research Institute to overcome
the previously mentioned problem of temperature control with
the cyclone and the cell in an external chamber. A major
problem in the use of this probe, as well as the apparatus
of Cohen and Dickinson, was the uncertainty in knowing whether
the resistivity cell had been filled with fly ash before the
determination of resistivity was made.
Point-Plane Probe
2 7
A point-plane resistivity probe designed at Southern Research
Institute for insertion into a flue-gas duct is illustrated
in Figure 23. The primary difference in the operation of
this apparatus and either of the cyclone devices was that
sample collection occurred under the influence of a corona
and an electric field rather than a centrifugal force. The
"stationary point" illustrated in the figure served as the
source of a negative corona, and a circular electrode sur-
rounded by the "grounded ring" served as the collection elec-
trode for fly-ash particles charged by the corona. Values of
resistivity were measured in two ways: (1) by comparing the
values of current with different voltages applied between the
corona point and the collection electrode, before and after
ash was collected, and (2) by moving the "shaft" to place a
second circular electrode on the upper surface of the col-
lected ash and determining the voltage-current relationship
in the sample between the adjustable and fixed electrodes.
In each procedure, the calculation of resistivity was based
on the thickness of the sample as determined by lowering
the adjustable electrode the required distance to make con-
tact with the sample.
150
-------
ELECTRICAL
CONNECTION
THERMOCOUPLE
TEFLON TIP
TO r
VACUUM l .
PUMP I
OUTER
ELECTRODE
PIPE
en
CYCLONE'
COLLECTOR
TEFLON CELL / CENTER
ELECTRODE
GAS XINLET STAINLESS'STEEL LINER
VIBRATOR
Figure 22. Cyclone probe inserted in duct
(Nichols27)
-------
HIGH VOLTAGE
CONNECTION
DIAL INDICATOR
PICOAMMETER
CONNECTION
MOVABLE SHAFT
STATIONARY POINT
GROUNDED RING
Figure 23. Point-to-plane resistivity probe
equipped for thickness measurement
(Nichols27)
152
-------
ANALYSIS OF FLY ASH
Overall Composition
The overall composition of a sample of fly ash was determined
as follows:
• One portion of the sample was dissolved in a con-
centrated mixture of hydrofluoric and sulfuric acids
The resulting solution was then analyzed by atomic
adsorption spectroscopy for lithium, sodium, potas-
sium, magnesium, calcium, iron, and titanium. A
separate portion of the solution was analyzed color-
imetrically for phosphorus .
• A second portion of the sample was fused with
sodium hydroxide. The fused mixture was then dis-
solved in water and analyzed colorimetrically for
aluminum and silicon.
• A third portion was fused with sodium carbonate.
Sulfur was determined turbidimetrically as sul-
fate in an aqueous solution of the fused sample.
Weight percentages of the stable oxides in the original fly
ash were then calculated from the concentrations of the
several elements.
Water-Soluble Components
A slurry was prepared of each fly-ash sample, consisting of
0.1 g of fly ash and 30 ml of distilled water. The slurry
was stirred for 20 to 30 min to dissolve the soluble constit-
uents of the ash. Then, the following steps were followed:
1. The pH of the slurry was measured with a glass
electrode to determine whether soluble acid or
soluble base was in excess.
2. A portion of the liquid phase was treated with a
cation-exchange resin to replace interfering
cations with hydrogen ion. The sulfate in the
treated liquid was then titrated with barium per-
chlorate to the end point indicated by Thorin.
(This procedure was based on analytical methods
for sulfate described by Fritz and Yamamura5 l and
Fielder and Morgan.
3. Another portion of the liquid was treated with
sodium hydroxide, and the concentration of ammo-
nia was determined with a membrane electrode
(Model 95-10 of Orion Research Incorporated).
153
-------
DETERMINATION OF FLUE GASES
Sulfur Oxides
Sulfur oxides were collected in a gas-sampling train consist-
ing of (1) a heated sampling probe with a quartz-wool filter
(to remove fly-ash particles), (2) a condenser maintained at
about 80°C to collect sulfur trioxide as sulfuric acid, (3) a
bubbler containing aqueous hydrogen peroxide to collect sulfur
dioxide as sulfuric acid, (4) a Drierite tube to remove water
vapor, and (5) a flow meter to determine the total volume of
gas sampled in the dry state. The collected sulfur oxides
were titrated with barium perchlorate by the precedure used
for titrating soluble sulfate in fly ash. Further details on
the analytical method are given in a previous report from
Southern Research Institute22 and by other investigators1*3
whose design of the sulfur trioxide condenser was adopted.
With the result of an independent determination of water vapor,
as described below, concentrations of the sulfur oxides were
calculated for the moist flue gas as sampled.
Ammonia
Ammonia was sampled through the heated sampling probe men-
tioned above and absorbed in a bubbler of dilute sulfuric
acid. The bubbler solution was then made alkaline with sodium
hydroxide, and ammonia was determined with the Orion membrane
electrode.
Water Vapor
Water vapor was collected from a measured volume of flue gas
in a preweighed cartridge of Drierite and determined gravi-
metrically.
DETERMINATION OF FINE-PARTICLE CONCENTRATIONS
Use was made of optical, diffusional, and inertial devices
for determining the concentrations of fine particles. The
specific items of instrumentation employed were a General
Electric condensation-nuclei counter with diffusion batteries,
a Climet photoelectric particle counter, and cascade impactors
of the Brink and Anderson types. Reference is made to a paper
by McCain et al^. 36 for details on the operation of these
instruments.
154
-------
DETERMINATION OF PRECIPITATOR ELECTRICAL PARAMETERS
Some of the precipitator electrical data described in this
report were obtained from readings of the manufacturer's
installed meters for primary voltage and current, secondary
current, and spark rate. The remaining data—principally on
secondary voltage but including auxiliary information on
secondary currents—were obtained with special instrumentation
installed and operated by Mr. Gerald D. Whitehead, a member
of the TVA technical staff.
155
-------
TECHNICAL REPORT DATA
(Please read ftiuniclwns on the reverse before completing)
i. REPORT NO. TVA F75PRS-5
EPA-600/2-75-Q15
4. TITLE ANDSUBTITLE
Conditioning of Fly Ash with Sulfur Trioxide and
Ammonia
5. REPORT DATE
August 1975
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Edward B. Dismukes
3. RECIPIENT'S ACCESSION'NO.
J. PERFORMING ORGANIZATION REPORT NO
3ORI-EAS-75-311
Project2932-3-F
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 Ninth Avenue South
Birmingham, Alabama 35205
10. PROGRAM ELEMENT NO.
1AB012; ROAP 21ADJ-029
11. CONTRACT/GRANT NO.
TVA--TV36921A; EPA--
CPA 70-149 and 68-02-1303
12. SPONSORING AGENCY NAME AND ADDRESS
TVA, Power Research Staff
Chattanooga, Tennessee 37401 AND
EPA, Industrial Environmental Research Laboratory
Research Triangle Park North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final: 1970 - 1975
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT rj.ne rep0r£ summarizes research on the conditioning of fly ash in coal-
burning electric power stations with two flue-gas additives--sulfur trioxide and
ammonia. It presents experimental data on the use of these additives to improve the
efficiency of electrostatic precipitation of fly ash by adjusting the electrical resisti-
vity of the ash and by other less widely recognized mechanisms. The report shows
that the primary role of sulfur trioxide is lowering resistivity from the excessive
values found with ash from low-sulfur coals. It also indicates that the role of ammonia
does not involve a change in resistivity, despite findings to the contrary by other
investigators. At least for the specific circumstances investigated, the research data
indicate that conditioning by ammonia involves a space-charge enhancement of the
electric field in the interelectrode space of a precipitator and, sometimes additionally,
an increase in the cohesiveness of the collected ash. The report addresses both the
theoretical aspects of conditioning mechanisms, and such practical matters as the
effectiveness of each agent as a function of the concentration added, the facilities used
for adding the agent, the chemical composition of the ash treated, and the temperature
of the ash during conditioning and precipitation.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
l>.IDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Air Pollution
Fly Ash
Treatment
Sulfur Trioxide
Ammonia
oal
Combustion
Flue Gases
Additives
Electrostatic
Precipitation
Utilities
Air Pollution Control
Stationary Sources
13 B
21B
07 B
11G
13 H
21D, 08G
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (Thispage)
Unclassified
21. NO. OF PAGES
169
22. PRICE
EPA Form 2220-1 (9-73)
157
------- |