EPA-650/2-74-074
July 1974
Environmental Protection Technology Series
E:;:$:|:^
-------
EPA-650/2-74-074
INFLUENCE
OF FLY ASH
COMPOSITIONAL FACTORS
ON ELECTRICAL
VOLUME RESISTIVITY
by
R. £. Bickelhaupt
Southern Research Institute
2000 Ninth Avenue South
Birmingham, Alabama 35205
Contract No. 68-02-0284
ROAP No. 21ADJ-029
Program Element No. 1AB012
EPA Project Officer: L. E. Sparks
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
July 1974
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
ABSTRACT
Twenty-eight fly ash samples were characterized, chemically
analyzed, and fabricated into sintered-disc resistivity
specimens. These ashes represented a broad spectrum of
ash compositions produced by burning coal in commercial
power station boilers. Resistivity and transference
experiments were performed in the temperature range in
which volume conduction prevails.
The results confirmed the conclusions of an earlier investi-
gation: 1) the volume conduction mechanism for fly ash is
ionic and 2) the charge carriers are the alkali metal ions,
principally sodium. It was learned that increasing iron
concentration caused a decrease in resistivity for a given
level of sodium and lithium. No evidence of biased data
or electronic conduction was found. It was rationalized
that in a manner analogous to that for glass the iron
affected the structure of the predominant glassy phase of
the ash thereby inducing the participation of a greater
percentage of the available alkali metal carrier ions.
From these data, empirical equations were developed to
predict the volume resistivity of fly ash as a function
of ash chemistry, temperature and porosity.
This report was submitted in partial fulfillment of
Contract No. 68-02-0284 by Southern Research Institute
under the sponsorship of the Environmental Protection
Agency. Work was completed as of July 1974.
111
-------
iv
-------
TABLE OF CONTENTS
Page
Abstract iii
List of Figures . vi
List of Tables vii
Sections
I Conclusions 1
II Recommendations 2
III Introduction 3
IV Scope 4
V Experimental Procedures 5
VI Results and Discussion 10
VII Prediction of Volume Resistivity 31
VIII Summary 37
IX References 39
-------
LIST OF FIGURES
No. Page
1 Alumina Resistivity Apparatus 7,
2 Typical as Measured Resistivity versus
Reciprocal Absolute Temperature for
Three Ashes 13
3 Resistivity versus Atomic Percentage Lithium
plus Sodium at 1000/T(°K) =1.6 and 40%
Porosity 18
4 Resistivity versus Atomic Percentage Iron
at 1000/T(°K) = 1.6, 40% Porosity and
0.4 a/o Lithium plus Sodium 20
5 Gravimetric Data for Transference
Experiments 24
6 Relative Effectiveness of Lithium and
Sodium as Charge Carriers as a Function
of Iron Concentration 27
7 Correlation Between as Determined and
Predicted Resistivites 36
VI
-------
LIST OF TABLES
No. Page
1 Chemical Analyses of Fly Ashes 12
2 Resistivity, Porosity, Atomic Concentration 15
3 Transference Experiments, Chemical Analyses
of Specimens in Weight Percent 26
vii
-------
viii
-------
SECTION I
CONCLUSIONS
Electrical volume conduction takes place through a
continuous matrix of glassy particles which make up
at least 75 weight percent of the ash.
An ionic mechanism controls the conduction process.
The alkali metal ions are the charge carriers, and
because of the relatively high concentration and
mobility, sodium is the principal participant.
The presence of iron attenuates the resistivity of fly
ash having a given level of alkali metal ion concentra-
tion. The iron concentration does not reduce resistivity
by either a direct ionic or electronic contribution to
the conduction process. It is believed that the iron
in solution in the glassy ash affects a structural
modification that permits the migration of a greater
number of the total alkali metal ions available.
In a regular and reproducible manner/ the volume
resistivity can be related, to: the atomic percent
of lithium and sodium, the atomic percent of iron,
the temperature, and the porosity of the fly ash
specimen. An expression has been developed to yield
a reasonable prediction of volume resistivity in terms
of these parameters.
-------
SECTION II
RECOMMENDATIONS
The concluded research suggests that compounds containing
sodium or iron might be useful as coal additions .to
attenuate high resistivity. Field experience with this
approach would be beneficial.
Since the role of iron is imperfectly understood,
additional laboratory work might be undertaken with
respect to its use as a conditioning agent.
The expressions developed to predict volume resistivity
from the chemical analysis of ash, temperature and
porosity should be evaluated against high temperature
field data as well as laboratory data acquired with
various techniques of resistivity measurement. Labora-
tory work should be undertaken to evaluate the ability
to predict resistivity for laboratory ashed coals.
Research similar to that reported herein should be
conducted in the temperature range in which surface
conduction is operable. An understanding of the surface
mechanism could lead to the prediction of surface
resistivity as a function of certain measurable parameters
and explain the role of conditioning agents.
-------
SECTION III
INTRODUCTION
The purpose of this research was to define the composi-
tional factors that influence the electrical resistivity
of fly ash. This information was desired to attempt to
predict resistivity from fly ash and/or coal compositions.
At the start of the program, it was planned to accomplish
the above purpose with respect to the entire normal
temperature range (125°C to 450°C) for the operation of
electrostatic precipitators. The complex nature of the
research and the limitations of the contract forced
the scope to be limited to the high temperature region
in which the resistivity of the ash is controlled
by a volume conduction process and is not influenced
by environmental characteristics such as water vapor
and SO 3.
Early in 1972, research1 conducted for Calgary Power, Ltd.
of Calgary, Alberta, Canada was completed that showed
the pronounced influence of certain alkali metal ions
on the resistivity of two ashes that were almost identical
in composition with the exception of the amount of sodium.
This work was then expanded to include other ashes of
generally similar composition coming from coals of the
Western part of the United States. This research was
recently published,2 and is summarized below.
Above 250°C the volume conduction process in fly ash is
controlled by an ionic mechanism. The quantity of
electricity passed is proportional to a mass transfer,
and lithium and sodium ions were found to migrate. The
limited migration of potassium was also detected.
Resistivity decreased about two orders of magnitude for
a one order of magnitude increase in the combined
concentration of lithium and sodium.
The research reported herein concerns an attempt to examine
the compositional factors that influence volume resistivity
for a broader spectrum of ash composition. In addition
to qualitatively defining these factors, it was desired
to develop an expression with which one could predict
resistivity from the chemical analyses of ashes.
-------
SECTION IV
SCOPE
Approximately twenty-five fly ashes produced commercially
from Eastern and Western coals were utilized. From this
group, representative ashes were selected for characteriza-
tion regarding particle size, helium pycnometer density,
optical microscopy, and X-ray diffraction. All ashes
were chemically analyzed for the major constituent
elements and were fabricated into sintered-disc specimens
for resistivity determinations between 200 and 450°C.
Transference experiments were conducted for certain ashes
selected on the basis of the resistivity-chemistry
behavior.
-------
SECTION V
EXPERIMENTAL PROCEDURES
RESISTIVITY: SPECIMENS AND PROCEDURES
Electrical resistivity was determined by measuring current
flow under known dc potential for self-supporting discs
of fly ash. Although this type of specimen is unusual
with regard to fly ash resistivity measurements, it is
commonly used in work pertaining to other types of
insulator materials. For the experiments to be conducted
in this research, the self-supporting disc specimen
was desirable. The specimens, experimental equipment,
and procedures followed were generally similar to that
suggested for this type of measurement by ASTM.3
Specimen Preparation
A known quantity of dry fly ash was mixed with 3% of
polyvinyl alcohol (duPont's Elvanol Grade 70-05 used as
10% solids in distilled water) and sufficient distilled
water to produce a paste. After thorough blending, the
mixture was dried to about 4% moisture and rubbed
through a 20-mesh screen. The resultant granulation
was pressed into discs 3.2 cm in diameter and about 4 mm
thick. Double-acting pressing was used with a 2270 kg
load.
The discs were dried overnight at 110°C and sintered in
air for two hours at temperatures selected to produce a
given range of specimen porosity. Although the discs
were weak, sufficient strength was secured so that the
specimens could be handled and measured and would support
themselves in the resistivity apparatus.
Physical measurements and microscopic analysis of the
disc specimens showed that undue densification did not
occur. The specimens were essentially assemblages of
spherical particles having a porosity of 31 to 47%.
Comparative chemical analyses were made for "as received"
ash and respective sintered discs. No alteration in
overall chemistry could be detected due to specimen
preparation.
-------
Using a gold alloy paste, Engelhard #9696, electrodes
1.9 cm in diameter were painted on opposite faces of the
disc at the center. The electrodes were fired at 900°C
for 30 to 40 minutes in air.
Resistivity Apparatus
A specimen holder was built exclusively from 99.5% alumina
and was used to position the fly ash disc between
contactors and leads under about 1 kg of load. A specimen
holder similar to that used in this work is shown in
Figure 1.
The assembled holder was inserted into a vertical,
platinum-resistor tube furnace having a uniform tempera-
ture zone (± 3°C) 8 cm in length. The holder was aligned
vertically so that the specimen was at the center of this
zone, and horizontally so that the apparatus did not
contact the furnace. The leads, insulated with 99.5
alumina, were directed out the top and bottom of the
furnace to the electrical terminals.
Specimen temperature was determined using a chromel-alumel
thermocouple located 3 mm from the positive face of the
specimen and 1 cm from the center of the specimen. The
thermocouple output was measured using a Rubicon Model
2745 potentiometer with cold junction correction.
In addition to the specimen and related leads, the test
circuit contained a dc voltage source, Keithley Model
240A high voltage supply, and a current measuring device,
Keithley Model 610 electrometer.
Test Procedure
With the specimen holder in place, the power to the furnace
was started. The furnace was allowed to heat and thermally
equilibrate overnight at about 430°C. The dc potential
(200 volts) was connected to the specimen and time recorded.
The combination of the noble metal electrodes and direct
current applied voltage resulted in a condition of elec-
trode polarization. Electrometer readings decreased with
time after the initiation of the applied voltage. It was
-------
Figure 1. Alumina resistivity test apparatus
-------
found that after 30 minutes, the rate of decrease in
current was negligible with reference to the time
required to obtain the resistivity data. Within the
objectives of this research, this procedure was
acceptable. After the 30 minute period, the power to
the furnace winding was turned off, and the specimen
temperature and current readings were simultaneously
taken at about 30°C intervals as the temperature decreased.
The test was conducted in a normal laboratory environment
containing nominally 1% water by volume.
Resistance was .calculated from:
R = V/I (1)
where
R = resistance, ft,
V = dc voltage, volts, and
I = measured current, amperes.
Resistivity was calculated from:
p = R x A/1 (2)
where
p = resistivity, ft-cm,
R = resistance, ft
A = area of electrodes, cm2, and
1 = specimen thickness, cm.
Resistivity was plotted versus reciprocal absolute
temperature.
TRANSFERENCE MEASUREMENTS
The principles and techniques used in making transference
measurements are reviewed in the literature, for example
by Kingery.1* An unsophisticated approach to this
measurement was used because of the complex chemistry of
fly ash. In violation of experimental boundary conditions
for a transference experiment, blocking electrodes were used.
The blocking electrodes, noble metals in this case, are
incapable of accepting or introducing charge carriers to
support ionic conduction. Therefore, one cannot obtain
unequivocal quantitative data regarding the conduction
mechanism.
8
-------
Three disc specimens were positioned in axial alignment
with disc faces contiguous. Electrodes were affixed
only on those faces in contact with the positive and
negative contactors and leads. After the furnace had
equilibrated at about 600°C, the dc voltage (1000 volts/cm)
was applied and the current-time relationship was
recorded for approximately 200 hr.
The current-time plot was graphically integrated and
converted to coulombs of electricity. This measured
quantity of electricity, the measured weight change of
the specimens, and pre- and post-test chemical analyses
were used to illustrate a mass transfer and to qualitatively
identify the charge carriers.
-------
SECTION VI
RESULTS AND DISCUSSION
ASH CHARACTERIZATION
Several ashes representing both Eastern and Western coals
were subjected to a characterization study. Optical and
electron microscopy revealed that the ash particles were
mainly spherical in shape over the entire size spectrum
for all the ashes examined. Each ash possessed a wide
variation in color and light transmission. Density
measured by helium pycnometer on dried "as received"
ashes varied from 2.2 to 2.7 grams/cc. It was noted
that the ashes having the greater amounts of iron and/or
calcium present possessed the higher densities; however,
it should also be stated that unburned carbon and ceno-
spheres greatly affect the density value. The mass-mean-
diameter of the ashes varied from 7 to 20 microns. No
particular significance is attached to these size data
since the sample history was not known in all cases.
The ashes were also examined by microprobe analysis and
X-ray diffraction. The microprobe analyses showed that,
within a given particle and among particles of a given
ash, chemical heterogeneity may exist. X-ray diffraction
patterns for the Western ashes revealed the presence of
the crystalline compounds: quartz, mullite, and a calcium
silicate. Calcium sulfate was also found in certain ashes.
The Eastern ashes contained: quartz, mullite, an iron
oxide, and an iron-alumina-silicate. Although the
various compounds produced a large number of diffraction
lines, it was noted that the line intensities were quite
low. Furthermore, a pair of halos at the low two theta
position, usually an indication of amorphous material in
an X-ray diffraction pattern, were observed. A quanti-
tative analysis5 was made on two ashes to determine the
amount of silica present. Silica was the most prominent
crystalline compound in these ashes. Less than four
percent silica was determined. From the observations, it
was concluded that the ashes contained unburned carbon,
minor amounts of several crystalline compounds, and large
percentages of glassy solids.
10
-------
The above statements of characterization are in contrast
with the observations of certain European investigators.6
It has been observed that a European ash may microscopically
display crystalline morphology and reveal a high percentage
of crystalline compounds by X-ray diffraction. It is
believed that this condition is caused by a combination of
low boiler temperature and the use of coals producing
very high percentages of ash.
All the ashes used in this research were chemically analyzed
for the major elements. These data expressed in weight
percent as oxides are shown in Table 1. Ashes 1 through
15 come from Western coals, while ashes 16 through 28 come
from Eastern coals. Excluding one or two unusual ashes,
the comparison between chemical analyses for Western and
Eastern ashes may be summarized as follows: 1) The sodium
oxide concentration of Western ashes ranges from less than
that of a typical Eastern ash to six times as much, 2)
Eastern ashes contain more lithium and potassium oxide
than the Western ashes, and 3) The combined concentration
of MgO plus CaO plus Fe20a is about 10 to 30 percent for
all ashes. About 60 to 80% of this total is MgO plus
CaO for Western ashes, while the large percentage is Fe203
for Eastern ashes.
RESISTIVITY DATA
Resistivity data were obtained in the temperature range
of 200 to 450°C for the twenty-eight ashes shown in Table 1.
Typical data are shown graphically in Figure 2 with
resistivity plotted on a log scale ordinate versus the
reciprocal of absolute temperature. For a given tempera-
ture, a wide range in resistivity was encountered (about
3 orders of magnitude) for all the ashes studied.
The linear curves indicate the data can be interpreted
in terms on an Arrhenius equation, in logarithmic form,
log p = log po + [(9/k) log e] (1/T) (3)
where
p = resistivity
po = a complex material parameter
0 = experimental activation energy
k = Boltzmann's constant
T = absolute temperature.
11
-------
Table 1. CHEMICAL ANALYSES OF FLY ASHES
ASH
NUMBER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
EXPRESSED IN WEIGHT PERCENT AS OXIDES
Li20
0.04
0.01
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.01
0.01
0.01
0.02
0.01
0.01
0.05
0.04
0.05
0.02
0.02
0.01
0.04
0.03
0.07
0.05
0.04
0.05
0.06
Na20
0.09
0.24
0.25
0.34
0.32
0.53
1.30
1.07
1.77
3.10
2.10
2.40
0.38
1.47
1.84
0.48
0.49
0.32
0.23
0.51
0.35
0.33
0.38
0.40
0.16
0.51
1.55
0.42
K20
0.30
0.89
0.89
0.77
0.90
1.10
0.72
0.70
1.13
0.80
1.00
1.00
2.36
0.68
0.20
2.75
3.12
2.30
2.80
2.80
2.36
3.88
3.34
3.10
2.60
3.80
2.80
3.70
MgO
5.40
3.30
1.88
1.60
1.80
1.70
2.50
2.23
1.93
0.90
0.99
0.93
2.59
1.73
12.75
0.88
1.09
0.98
0.86
0.93
1.66
1.57
1.29
1.20
0.93
1.30
1.15
1.30
CaO
21.50
23.50
11.10
11.60
12.70
12.20
8.60
8.30
6.36
10.40
12.30
13.60
11.41
7.33
31.00
0.87
2.48
4.60
2.20
0.26
3.72
0.77
1.04
1.40
0.87
0.68
2.50
1.10
Fe203
4.30
5.53
3.71
3.60
3.90
4.20
4.65
4.66
4.61
3.40
4.60
4.80
6.03
5.33
11.20
5.50
13.24
23.70
23.00
17.90
16.10
10.01
9.70
8.80
8.40
4.90
3.85
9.90
AljOj
22.80
21.20
23.60
25.30
25.80
25.10
18.20
17.70
24.60
28.20
26.50
26.20
19.64
23.10
14.80
27.80
26.40
21.20
21.00
21.90
17.80
27.50
25.90
29.00
31.00
30.20
25.00
30.30
Si02
40.15
40.50
55.60
56.80
55.00
55.10
59.00
61.00
53.70
52.00
52.00
49.20
55.29
53.10
22.00
50.45
50.70
42.70
47.70
51.00
43.30
51.40
49.90
52.40
55.10
53.00
56.75
52.40
TiOz
1.30
2.13
1.56
0.83
0.89
0.83
2.20
1.53
1.49
0.88
0.93
0.93
0.95
4.17
0.60
1.85
1.62
1.40
1.80
1.40
1.27
1.79
1.98
1.70
2.30
2.00
1.55
1.70
PzOs
0.34
0.54
0.14
0.17
0.17
0.18
0.19
0.16
1.06
0.18
0.21
0.24
0.60
0.90
0.39
0.23
0.28
0.34
0.52
0.33
0.36
0.32
0.32
0.40
0.36
0.17
0.28
0.51
S03*
1.70
1.83
0.32
0.17
0.32
0.24
1.05
0.77
0.79
0.26
0.36
0.46
0.58
1.11
4.80
0.70
0.57
0.69
1.30
0.96
0.64
0.37
0.42
0.29
0.23
0.36
1.12
0.45
LOI
0.33
0.20
0.74
0.60
0.65
0.60
0.45
0.50
1.49
0.45
0.40
0.45
0.35
0.44
0.41
7.70
1.00
5.50
2.10
l.BO
10.30
1.50
4.40
3.20
1.40
2.50
1.70
1.30
•Total Sulfur
12
-------
282
539
352
666
442 °C
827 °F
10"
lO'O
o
>-
K
>
*-
V>
I08
ASH
3
D 25
19
1.9
1.8
1.7 1.6 1.5
TEMPERATURE IOOO/T(°K)
1.4
Figure 1. Typical measured resistivity versus
reciprocal absolute temperature for
three ashes
13
-------
The experimental activation energy (6) is proportional to
the slope of the curves and was found to be between 1.0
and 1.1 eV for all the ashes. This suggests that a
similar conduction mechanism prevails for each ash.
An average value of 1.03 eV will be used later in the
report to establish the equation for predicting resistiv-
ity from ash chemistry and temperature. To develop
the relationship between resistivity and certain
chemical parameters, the resistivity at the reciprocal
temperature of 1000/T =1.6 was used. These data
(as measured) for all the ashes studied appear in
column two of Table 2 .
Intuitively, one would expect the volume conduction
process to be influenced by the amount of specimen
porosity. This was found to be true. When resistivity
was determined for several ashes fabricated to two
levels of porosity, the higher porosity produced greater
resistivity. Dalmon and Tidy7 also experienced this
relationship.
An empirical expression was developed to correct the
resistivity values acquired in this research to a
constant porosity. Using the initial approximation of
the curves relating resistivity to the concentrations
of lithium plus sodium and iron (shown in final form
later in the report) , the as-measured resistivity value
for each ash was normalized to a constant concentration
of lithium, sodium and iron. The normalized resistivity
value for each ash was then plotted against the porosity
of the respective ash specimen. The percent specimen
porosity was calculated from the measured bulk volume
and the determined helium pycnometer volume as follows:
%P = Vbulk - VHe Pyc x
Vbulk
The plot of normalized resistivity data versus porosity
was subjected to a linear regression analysis. The
following relationship was generated:
Log PC = log Pm + S (Pc-Pm) (5)
where
pc = resistivity corrected to porosity Pc
pm = resistivity measured at porosity Pm
S = Alog p/A%P = 0.04 .
14
-------
Table 2. RESISTIVITY, POROSITY, ATOMIC CONCENTRATIONS
Ash
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Resistivity
ohm- cm
As Measured at
1000/T(0K)=1.6
7.7x10'°
3. 4x10' °
1.5x10'°
8.5x10'
5.7x10'
4.5x10'
6.5x10'
1.4x10'
9.0x10*
1.0x10"
2.5x10'
4.1x10'
2.0x10'
5.3x10'
2.4x10'
1.9x10'
2.4x10'
3.0x10'
2.9x10'
1.5x10'
2.0x10'
6.0x10'
5.8x10'
1.0x10'
3.8x10'
1.9x10'
3.6x10'
5.8x10'
%
Porosity
47
42
36
39
37
33
39
39
46
34
32
35
40
39
50
42
34
44
36
33
31
40
40
42
42
40
37
36
Resistivity
ohm-cm at
1000/T(°K)=1.6
and
Porosity=40%
4.1x10'°
2.8x10'°
2.2x10'°
9.3x10'
7.5x10'
8.6x10'
7.1x10'
1.5x10'
5.2x10'
1.7x10'
5.2x10'
6.5x10'
2.0x10'
5.8x10'
9.6xl07
1.7x10'
4.2x10'
2.1x10'
3.8x10'
2.9x10'
4.6x10'
6.0x10'
5.8x10*
8.3x10'
3.2x10'
1.9x10'
4.8x10'
8.4x10'
Atomic
Concentration
Li+Na Fe
0.13
0.19
0.20
0.26
0.25
0.40
0.98
0.81
1.39
2.28
1.56
1.81
0.31
1.13
1.29
0.49
0.46
0.37
0.27
0.47
0.34
0.34
0.37
0.45
0.25
0.43
1.29
0.45
0.71
0.94
0.64
0.60
0.65
0.71
0.79
0.79
0.82
0.58
0.79
0.84
1.02
0.94
1.81
1.10
2.50
4.81
4.49
3.51
3.47
1.91
1.91
1.66
1.54
0.93
0.70
1.82
Resistivity
ohm-cm at
1000/T(°K)=1.6
Porosity =40%
Li+Na=0.4 a/o
_
-
-
-
-
-
-
-
-
-
-
-
-
-
8.0x10°
2.5x10'
5.4x10'
1.8x10°
1.9x10°
3.9x10'
3.4x10°
4.4x10'
5.0x10°
1.05x10'
1.3x10'
2.3x10'
4.2xl09
1.1x10'
15
-------
Although the positive slope of the normalized resistivity
vs percent porosity data was obvious, the twenty eight
data points presented considerable scatter. For confi-
dence limits of 95%, the slope, S, was 0.04 ± 0.02.
Since the porosity correction factor is somewhat
uncertain and since the average porosity for all
resistivity specimens was about 40%, the as-measured
resistivity data were corrected to a porosity level of
40% for subsequent use. The porosity values and the
resistivity data corrected to 40% appear in columns 3
and 4 respectively in Table 2.
The chemical concentrations used to show the relationship
between resistivity and ash chemistry are given in atomic
percent in columns 5 and 6 of Table 2. These values were
arrived at in the following manner. The weight percent
chemical analyses shown in Table 1 were normalized to
total 100% excluding the loss on ignition. The weight
percentages were then converted to molecular percentages
from which the atomic concentration of lithium, sodium,
and iron were calculated.
In summary. Table 2 shows for each of the twenty eight
ashes studied: the as-measured resistivity at 1000/T(°K) =
1.6, the porosity of the resistivity specimen, the
resistivity corrected to 40% porosity, and the atomic
percentages of lithium plus sodium and iron in the
specimen. The last column in Table 2 will be referred
to later.
RESISTIVITY-ASH CHEMISTRY CORRELATION
In reference 2 it was proposed that the volume conduction
mechanism is ionic and that the alkali metals, principally
sodium, are the charge carriers. This was substantiated
by direct evidence of the proportionality between mass
transfer and the quantity of electricity conducted and by
the chemical identification of the migrating elements.
For fly ashes of limited compositional range, it was
shown that resistivity was inversely proportional to the
concentration of lithium and sodium. The above points as
well as other aspects of the conduction process were
discussed in detail in the subject reference and will not
be reiterated.
16
-------
The research reported herein was designed to verify the
initial results, expand the scope to include a greater
number of ashes so that a wide range of ash composition
might be evaluated, and develop an expression capable
of a reasonable prediction of volume resistivity based
on ash chemistry, temperature, and ash layer porosity.
To satisfy the objectives, 28 ashes were acquired. The
ashes were produced mainly from bituminous and sub-
bituminous coals but also from lignite, anthracite and
coal prepared for metallurgical use. It is believed
that the ash compositional ranges are sufficiently broad
for the effort. The resistivity data were presented
in Table 2.
These data were initially examined by testing the
inverse proportionality between resistivity and the
combined concentrations of lithium and sodium. This
was done by plotting on a log-log scale the resistivity
of the 28 ashes taken at 1000/T(°K) =1.6 (^350°C or
665°F) and corrected to 40% porosity versus the atomic
percentage of lithium plus sodium. Figure 3 illustrates
the result. The ash designation numbers that are
underlined are Eastern ashes while the remainder are
Western. Viewing the data points showing the spread in
resistivity at 0.4 atomic percent lithium plus sodium
or the spread in concentration at a resistivity of
5x108 ohm-cm, one would doubt that a correlation exists.
It was obvious that some factor in addition to the
lithium-sodium concentration was affecting the results.
To verify the earlier results from reference 2, a linear
regression analysis was made between resistivity and
lithium-sodium concentration for those Western ashes
containing <1.0 atomic percent iron. This level of iron
was chosen from the earlier work. The curve constructed
in Figure 3 with a slope of -1.84 is the result of the
regression analysis. The coefficient of correlation
between the calculated curve and the data points used
was 0.98, and the slope was within 5% of that previously
determined with similar ashes.
With the correlation between resistivity and lithium-sodium
concentration verified for a select group of ashes, the
problem became one of explaining the observation that
about half the ashes examined did not conform. Although
the concentrations of other elements were considered,
inspection of the resistivity and chemical analysis data
suggested that the iron concentration affected resistivity.
17
-------
108
ATOMIC PERCENT LITHIUM PLUS SODIUM
10.0
Figure 3. Resistivity versus atomic percentage lithium
plus sodium at 1000/T(°K) =1.6 and 40% porosity
18
-------
The effect of iron was evaluated by normalizing the
resistivity values to a constant amount of lithium and
sodium for all Eastern ashes and the Western ash having
an iron concentration >1.0 atomic percent. Resistivity
values normalized to 0.4 atomic percent sodium plus
lithium were obtained by drawing a line through the
specific data point parallel to the constructed curve
in Figure 3 and reading the resistivity at the point
where this line intersected the 0.4% ordinate.
Resistivity data acquired in this manner are tabulated
in the last column in Table 2.
Figure 4 shows the resistivity data normalized to a
constant lithium-sodium concentration plotted against
the iron concentration on a log-log scale. The data
point marked (*) represents the average resistivity
and iron concentration for the Western ashes having
<1.0 atomic percent iron. Again, the Eastern ash
numbers are underlined and the Western is not.
A linear regression analysis of all data points in
Figure 4 with the exception of numbers 22 and 23
produced the curve shown. This curve has a slope of
-1.65 and a coefficient of correlation with the data
points used of 0.99. No definite explanation can be
given for the two points that lie significantly below
the constructed curve. It should be noted that there
are many opportunities to introduce errors: 1) during
the selection of ash samples and the fabrication of
specimens, 2) in the measurement of weight percent
chemical analyses, resistivity and porosity, and 3)
during the subsequent manipulation of these data. On
the other hand, it is conceivable that these points
would become part of a band of data if a very large
number of ashes were examined.
From Figure 4 it becomes apparent that the volume
resistivity of fly ash is also inversely proportional
to the iron concentration. The slopes of the curves in
Figures 3 and 4 suggest that resistivity is only slightly
less sensitive to iron concentration than it is to
lithium-sodium concentration. Figure 4 empirically
explains the scatter of data shown in Figure 3.
19
-------
10'0
27
X
AS
xo
i- o
OTOC
o
UJ Q.
K8S
O
10
8
0.5
Figure 4.
1.0 2.0
ATOMIC PERCENT IRON
5.0
10.0
Resistivity versus atomic percentage iron
at 1000/T(°K) =1.6, 40% porosity and
0.4 a/o lithium plus sodium
20
-------
In summarizing the resistivity data, one finds that the
28 ashes have the same resistivity value plus or minus
25% when temperature, porosity, and concentrations of
iron, lithium and sodium are constants. Inspection of
the resistivity and chemical analysis data does not
suggest an additional factor to justify the variation
of ±25% in resistivity. It is possible that this
variation is due to the unknown physical and chemical
inequities associated with the inherent character of
the material being studied and possible inadvertent
experimental error.
Previously it has been stated that the role of lithium
and sodium in the volume conduction of fly ash was one
of suppling mobile charge carriers to the glassy
portion of the ash. One therefore would expect the
resistivity to be inversely proportional to the concen-
tration of these elements. Now it has been observed
that iron concentration has a similar relationship to
resistivity. Both from the fundamental viewpoint as well
as the pragmatic with the respect to ash conditioning,
it was desirable to explore the role of iron.
ROLE OF IRON
Background
Several investigators have mentioned the effect of iron
with regard to fly ash resistivity. Shale e_t als
experienced little difference in resistivity between
two ashes having significantly different levels of iron.
However, in this comparison, the ash having the lower iron
content contained about twice as much Na20 as the other
ash. This coupled with the absence of information
regarding lithium content and specimen porosity makes
comment impossible.
A regression analysis9 of ash resistivity versus iron
concentration for 38 low-sulfur Western coals, suggested
that resistivity increases with increased iron concentra-
tion. The coefficient of correlation for these data was
too low to consider the data meaningful. Without normaliza-
tion of data for other parameters of influence, the effect
of iron can be masked.
21
-------
Dalmon and Raask10 observed a decrease in resistivity of
accessory coal minerals that contained iron and sodium
and were subjected to fusion at high temperatures. These
authors also found that resistivity could be significantly
lowered by coating the particles with an iron salt
followed by oxidation at an elevated temperature.
Utilizing synthetic coal slags, Frederikse and Hosier11
showed a large decrease in resistivity with increase in
iron content at a given elevated temperature and oxygen
partial pressure. They attribute the lower resistivity
of these slags to electronic conduction.
In general, iron content is held to a minimum in commercial
glasses and other ceramics designed to have electrical
insulative properties since this element decreases
resistivity. Little work has been done with reference
to small quantities of iron in glass since the element
also has a great effect on optical character. Morey12
mentions some early research that showed an initial
increase followed by a dramatic decrease in surface
resistivity as the iron content was increased in a simple
sodium silicate glass. Considerable literature13'1"'
is available showing the effect of iron in producing
semiconducting glasses. The low resistivity is associated
with electronic conduction. These glasses however,
are oxide glasses containing no silica, chalcogenide
glasses or inverted silicate glasses that are unlike the
glass one would probably find in fly ash.
From the foregoing, it would seem there is mixed evidence
regarding the effect of iron on fly ash resistivity.
With respect to silicate glasses, the decrease in
resistivity with increase in or introduction of iron
is associated with electronic conduction in invert
glasses.
Observations And Experimentation
Several hypotheses were advanced that could explain the
relationship illustrated in Figure 4. First, higher iron
concentration reduces the fusion temperature of an ash
and therefore, for a sintered-disc type resistivity
specimen, could promote enhanced particle to particle
bonding. Second, with sufficient concentration and in a
particular state of oxidation, the iron could add an
electronic component to the conduction process thereby
22
-------
lowering resistivity. Third, the iron may affect the
amorphous phase of the fly ash so that the effective
concentration, the mobility, or the type of alkali metal
serving as a charge carrier was affected. The following
observations and experiments are discussed in an attempt
to clarify the role of iron in the volume conduction
process for fly ash.
Transference Experiments -
Transference experiments were conducted on four ashes
(Nos. 19, 20, 22 and 26) to evaluate certain facets of
the aforementioned hypotheses. These ashes contained
a relatively uniform concentration of alkali metals
and represented the total range of iron concentration
encountered in this work. The experimental conditions
and the character of the materials investigated precludes
the extraction of unequivocal data. These limitations
were discussed in reference 2. The data are used only
for the qualitative understanding of the conduction
process.
Gravimetric data for the transference experiments are
shown in Figure 5. In this figure, the mass transferred
out of the ash disc adjacent to the positive electrode
toward the negative electrode is plotted against the
quantity of electricity passed during the experiment.
The lines labeled with the names of alkali metals
represent Faraday's Law. For example, if conduction
were entirely ionic and potassium were the only charge
carrier, one would expect a weight loss of about 50 mg for
the passage of 120 coulombs of electricity.
The open circles represent the experimental data points.
The data for ashes 3 and 10 were acquired at an earlier
date for ashes of the lowest iron level. Figure 5
strongly suggests that the electricity passed is accountable
for by a mass transfer and that sodium is the principal
charge carrier. The only other way in which the data points
could occur near the sodium line would be due to some
fortuitous situation whereby charge carriers both heavier
and lighter than sodium ions participated to yield an
average weight change equivalent to that of sodium.
The above is not meant to imply that sodium was the exclusive
carrier. The migration of lithium was obvious; however,
quantitatively this represents a small contribution. Also,
the migration of potassium can be detected with the support
of microprobe analysis; however, it too amounts to a small
23
-------
50
CO
(9
CO
O
CD
UJ
40
UJ
Q
g 30
u
Id
_J
UJ
Ul
> 20
H
CO
O
0.
10
30 60 90 120
ELECTRICITY PASSED In COULOMBS
ISO
Figure 5. Gravimetric data for transference experiments
24
-------
quantitative contribution. Since the amount of potassium
migrating is extremely small in comparison to the relatively
large concentration of this element in the ash, it is
excluded from resistivity-ash chemistry correlations.
In Table 3, the results of the chemical analyses for the
transference experiments are given. The data show a
trend similar to that experienced previously for trans-
ference tests on low iron specimens and that these data
compliment or support the gravimetric data expressed in
Figure 5. For ashes containing from 5 to 22% weight
percent iron, only the migration of sodium and lithium
from the positive to the negative electrode can be
unquestionably observed. The small variations in
potassium and iron concentrations in Table 3 are thought
to be within the data error for the technique of analysis
and the selection of random samples. From this informa-
tion, it was indicated that the iron does not act as an
ionic carrier, and its presence in increased amounts does
not induce an increased participation by potassium.
In carefully examining the data in Table 3, it was observed
that the percentage of lithium and sodium that had migrated,
relative to the amounts initially present, increased with
increasing iron concentration. An empirical parameter
was devised to demonstrate this point. For a constant
amount of electricity passed in each test, the percent
increase in sodium and lithium content at the negative
electrode over that contained initially by the ash was
computed as "relative effectiveness". When this parameter
was plotted against the iron concentrations of the four
ashes studied, the result shown in Figure 6 was obtained.
The increase in the relative effectiveness parameter with
increasing iron concentration suggests that the role of
iron is indirect in that it seemingly enhances the
participation of lithium and sodium in the conduction
process.
Although the small deviations in experimental activation
energies would not suggest it, it is possible that the
iron concentration affects the mobility of the alkali
metal ions. However, it is quite possible that
the iron concentration could influence the number of
mobile charge carriers. Although the total concentration
of a particular ion species is used to graphically display
25
-------
TABLE 3, TRANSFERENCE EXPERIMENTS,CHEMICAL ANALYSES OF
SPECIMENS IN WEIGHT PERCENT
DISC CONTIGUOUS BASELINE
ASH OXIDE TO POSITIVE ELECTRODE COMPOSITION
DISC CONTIGUOUS
TO NEGATIVE
ELECTRODE
19
20
22
26
Li20
NA20
K20
FE203
Li20
NA20
K20
Li20
NA20
K20
FE203
0,013
2,9
21,1
Li20
NA20
K20
FE203
0,03
0,29
3,8
10,0
0,030
0,40
3,9
4,8
0,019
0,29
3,1
21,6
0,024
0,45
2,9
16,8
0,04
0,39
4,1
10,2
0,04
0,48
3,9
4,9
0,027
0,53
3,2
21,0
0,041
0,80
2,9
16,6
0,05
0,48
4,0
10,2
0,049
0,56
4,0
4,9
26
-------
1.0
CO
CO
LJ
z
111
>
I-
o
UJ
u.
u.
Ul
UJ
>
UJ
oc
0.8
0.6
0.4
0.2
LITHIUM
SODIUM
20s,
19
0.0
ATOMIC PERCENT IRON
Figure 6. Relative effectiveness of lithium and sodium
as charge carriers as a function of iron
concentration
27
-------
data, it is highly probable that not every ion of the
given type is free to migrate. It can be suggested that
the role of iron is to alter the amorphous ash structure
allowing the participation of a greater percentage of the
total available alkali metal ions or to enhance their mobility.
Miscellaneous Observations And Ancillary Experiments -
Although the transference experiments suggest that the role
of iron is not related to an electronic contribution to
the conduction process, this point was additionally
considered. In the course of this research, as well as
in previous efforts, it has been observed that within the
estimated experimental error the experimental activation
energy for conduction has varied little among ashes.
If the high iron ashes imparted an electronic contribu-
tion to the total conduction process, one might expect
a change in activation energy. Also, it has been noted
that the current decreases with time for a given voltage
potential using blocking electrodes. If there were an
electronic conduction mechanism, the polarization effect
should not be apparent.
Two additional experiments were run regarding the potential
electronic contribution of iron. In one case, ash 22
was repeatedly passed through a magnetic separator prior
to fabricating a resistivity specimen. The iron concen-
tration was reduced by 50% without significantly altering
the resistivity. It is believed that the iron removed
was that fraction which most likely would have provided
an electronic contribution, while that which remained
was principally involved in the glassy portion of the
ash. The observation that the removal of the magnetic
fraction did not increase the resistivity indicates the
absence of electronic conduction. Solid electrolyte
experiments16 were conducted on sintered specimens of
ashes 10 and 19 using a calcium stabilized zirconia as
a standard. Both ashes produced sufficient emf so that
the consideration of an electronic contribution to the
overall conduction process for the ashes could be dismissed.
Intuitively, one might suggest that ashes containing high iron
concentrations develop better particle to particle contacts
thereby lowering the resistivity of sintered disc specimens.
The following observations are submitted to clarify this
thought:
28
-------
a) The previously described experiment
involving the magnetic removal of iron
prior to specimen fabrication did not
alter the resistivity.
b) If superior particle contact played
an important role, one would neither
anticipate the excellent correlation
between resistivity and ash chemistry
at all levels of iron nor the inter-
pretation available from the trans-
ference experiments.
c) Superior particle contact should be
reflected by variations in experimental
activation energies for conduction.
Little difference in activation energy
was noted among the ashes utilized.
When low, medium and high iron ashes
were intentionally sintered to produce
specimen shrinkage and increase particle
bonding, the resistivity and experi-
mental activation energies were
dramatically reduced.
d) Two ashes having high and low iron
concentrations were used to measure
resistivity in the temperature range
of 200 to 450°C utilizing the ASME,
PTC 28 apparatus. These tests
employing loose, as-received ash
produced resistivity values almost
identical to their respective sintered
disc counterparts when the sintered
disc data were corrected to the same
porosities as the PTC 28 specimens.
The preceding observations indicate the sinterability of
the various ashes does not bias the resistivity data
when specimens are prepared at a high level of porosity
without specimen shrinkage.
• 29
-------
Summary
An unequivocal interpretation of the role of iron in the fly
ash conduction process is not available. However, it is
reasonably certain that the pronounced influence of the
iron concentration shown in Figure 4 is neither related to
an electronic contribution to the conduction process nor is
it an aberration due to the particle to particle bonding
in the sintered disc resistivity specimen. The evidence
obtained indicates that the presence of increased iron
concentration does not alter the previously identified
mechanism for volume conduction, i.e., the ionic migration
of the alkali metals. It is conceivable that the iron
concentration associated with the principal fraction of
the ash, the glassy phase, alters the structure of the
glass so that either a greater number of the total alkali
metal ions present are capable of migration or the ions move
at an increased velocity. Although no direct proof of this
rationalization is available, the concept is compatible
with: 1) only obvious evidence of lithium and sodium
migration has been found, 2) no electronic component to
the conduction process has been shown, and 3) very little
variation in conduction activation has been observed.
30
-------
SECTION VII
PREDICTION OF VOLUME RESISTIVITY
GENERAL
One of the primary goals of this research was to obtain
sufficient understanding of the volume conduction process
so that an attempt could be made to predict volume
resistivity from the chemical analysis of an ash. It
has been observed that the volume resistivity of an
ash is a function of the lithium-sodium concentration,
the iron concentration, the temperature and the porosity
of the ash layer. Since only a small unknown fraction
of the total potassium present participates as carriers,
it is impossible to include this element. When the
resistivities for 28 ashes were normalized so the four
parameters above were constant, the total range of
resistivity was reduced from about 3 orders of magnitude
to about one-half order of'magnitude. Inspection of the
data did not suggest additional approaches with which the
data might be interpreted. The resultant half order of
magnitude range of resistivities could be related to
accumulated errors, an undetected chemical effect, the
percentage of glassy phase in the individual ashes, etc.
In this prediction of volume resistivity the effect of
water vapor, SOs, etc., has not been considered. It is
believed that these agents are not of influence at
temperatures above 250°C where volume conduction is most
important. Also, the type of specimens used in this
research precluded the consideration of the influence of
unburned carbon. However for ashes containing less than
8% unburned carbon, Shale8 has indicated little effect
on resistivity for this constituent. Almost all the ashes
used in the current work in the as-received condition
contained less than 5% unburned carbon.
CALCULATION OF RESISTIVITY
Effect Of Chemical Concentrations
The effect of lithium plus sodium on resistivity at a
given temperature and porosity was shown in Figure 3.
The effect of iron on resistivity at a given temperature,
31
-------
porosity and lithium-sodium concentration was given in
Figure 4 . Both sets of data correlated well with a linear
interpretation of a log-log plot. Since Figure 4 represents
the variation of resistivity with iron for a constant
level of sodium plus lithium, the equation for this line
can be substituted for the value of resistivity in
Figure 3 at the 0.4 atomic percent lithium-sodium ordinate
and an equation developed yielding resistivity in terms
of both concentration variables.
The general expression relating resistivity to iron con-
centration in Figure 4 is:
log P(pe) = Io9 Po(Fe) ~ b(log
With a known line slope (b) and an assumed value for p
which defines the iron concentration in atomic percent
(a/o) , one can calculate log Po(Fe)» tne intercept.
For:
p = 109 (assumed)
log p = 9.0000
b = -1.65 (from Figure 4)
a/o Fe = 1.75 a/o (from Figure 4 at p = 10 9)
log a/o Fe = 0.2430
substituted into equation (6), log Po(Fe) is calculated
to be 9.4010, and equation (6) can be rewritten as:
log p(Fe) = 9.4010 - 1.65 (log a/o Fe) (7)
Equation (7) is the specific expression for the line in
Figure 4 and also defines the resistivity values in Figure
3 along the 0.4 a/o lithium-sodium ordinate for all iron
concentrations.
32
-------
The general expression relating resistivity to lithium-
sodium concentration in Figure 3 is
log P(Li-Na) = Io9 Po(Li-Na) ~ b(log a/o Li-Na)
(8)
Using the value of log P(Fe) defined in equation (7) at a
lithium-sodium concentration of 0.4 a/o and the known value
of (b) taken from Figure 3, one can calculate log Po
for equation (8) in terms of the iron concentration.
For:
P(Li-Na)
b =
a/o Li-Na
log a/o Li-Na
9.4010-1.65 (log a/o Fe)
(equation 7)
-1.84 (from figure 3)
0.4 a/o
-0.3979
substituted into equation (8), log Po(Li-Na) is calculated to
be 8.6689 - 1.65 (log a/o Fe), and equation (8) can be
rewritten as
log P(Li-Na,Fe) = [8.6689-1.65(log a/o Fe)]-1.84 (log a/o Li-Na)
(9)
Equation (9) expresses resistivity in terms of the two
critical concentration parameters obtainable from the
chemical analysis of fly ash. The value of resistivity
calculated from equation (9) is restricted to an ash layer
porosity of 40% and a temperature of 625°K.
Effect of Temperature
In Figure 2, the linearity of the plot of log p versus
1/T indicates the resistivity-temperature relationship can
be expressed with an Arrhenius equation. In logarithmic
form,
log p = log PO + [(8/k) log e] (1/T)
(3)
33
-------
Using the value of log p at 625°K and 40% porosity given
in equation (9), one can calculate log p? in equation (3)
and arrive at an expression for resistivity in terms of
temperature and chemical concentrations for an ash layer
of 40% porosity.
Given:
log p = 8.6689-1.65(log a/o iron)-1.84(log a/o
lithium-sodium), from equation (9)
0 = 1.03 eV, average experimental
activation energy
eV = 1.602x10-12 erg
k = 1.380x10"16 erg "K"1, Boltzmann's constant
log e = 0.4343
[(6/k) log e] = 5.193xl03 °K, calculated from above
[(6/k) log e] (1/T) = 8.3088, for T = 625°K
Substituting in equation (3) for T = 625°K,
log PO = 8.6689-1.65 (log a/o Fe)-1.84(log a/o Li+Na)-8.3088
Rewriting equation (3) incorporating this value of log p0
yields an expression for log p for all applicable values
of temperature and determined values of chemical concen-
trations at a porosity of 40%,
log p,F Ti_N „,, = [8.6689-1.65(log a/o Fe)-1.84(log a/o
(Fe,Li Na,i) Li+Na) _ 8.3088] + 5.193xl03 °K x
(1/T) °K~1 (10)
Effect Of Porosity
Equation (10) defines the resistivity in terms of tempera-
ture, iron concentration, and lithium-sodium concentration
for one value of ash porosity, 40%. Previously in this
report, an empirical relationship was given relating volume
34
-------
resistivity to percent porosity, equation (5). This
equation can be used to determine log p /pe Li._Na T)
equation (10), for porosities other than 46%.
COMMENTS
The foregoing expressions represent an initial attempt
to predict volume resistivity as a function of ash
chemistry, temperature and porosity. A particular type
of test specimen was used and twenty-eight fly ashes
were examined to acquire the necessary data. The
critical parameters controlling volume resistivity have
been reasonably well documented with respect to the
magnitude and mechanism of the effect.
Predicted resistivity values have agreed well with experi-
mentally determined values for ashes coming into this
laboratory. As-measured resistivity data have been
plotted against predicted values for the temperature
parameter of 1000/T(°K) =1.6. In predicting the
resistivity, the measured porosity of the individual
ash specimens was utilized. The good correlation
between predicted and measured volume resistivity data
can be seen in Figure 7. The individual data points of
maximum departure from the curve are less than one
half order of magnitude removed.
Whether predicted volume resistivity values will duplicate
or at least correlate with field measurements or labora-
tory measurements utilizing other experimental techniques
remains to be shown. Many factors can contribute to
variations in resistivity values resulting from the
different techniques, locations and conditions of
measurement. No opportunity to evaluate predicted
resistivity against high temperature field measurements
has been available. In this laboratory, surface
resistivity has been determined for a number of fly ashes
up to a temperature of 250°C using the ASTM, PTC 28
resistivity apparatus. It has been observed that in most
cases the predicted volume resistivity data serves well
in extending the surface resistivity data to higher
temperatures. This evaluation of the method for resis-
tivity prediction is being continued.
35
-------
10
II
^ 10
10
to
o
<5 I09
108
I40 O" 0022
°/28 23
OI2 XL
18.
HO
O
10
I08
109
RESISTIVITY OHM-CM
PREDICTED AT IOOO/T(°K)= 1.6
10"
Figure 7. Correlation between as determined and
predicted resistivities
36
-------
SECTION VIII
SUMMARY
From characterizations of ashes for this research and
previous work, it was estimated that an average ash is
made up of 0-10% unburned carbon, 5-15% crystalline
compounds, and 75-95% amorphous material. The amorphous
material appears to be a heterogeneous glass that forms
a continuous matrix in an ash layer. It probably forms
by the volatilization and/or fusion of auxiliary minerals
present in the coal at the peak boiler temperature and
is retained as a glass due to rapid cooling. The particles
are primarily spherical in shape and vary in diameter from
100 microns to submicron size. The ashes have a helium
pycnometer density of 2.0 to 3.0 grams/cc. When placed
into an ASTM, PTC 28 electrode, a porosity of 50 to 75%
usually occurs.
The writer previously suggested that volume conduction
takes place due to an ionic mechanism in which charge
carriers migrate through the continuous matrix of glassy
particles. It was shown in this earlier work and
confirmed in this research that the principal charge
carriers are the alkali metal ions, mainly sodium. It
has been observed that the amount of electricity
conducted is proportional to a mass transfer, and the
migration of sodium and lithium has been chemically
and visually demonstrated. Graphically it has been
illustrated that the volume resistivity of fly ash is
inversely proportional to the concentration of these
elements for ashes of limited compositional range.
The objective of developing an empirical expression with
which resistivity could be predicted from chemical
analyses of ash required a broad spectrum of ash compo-
sition to be evaluated. It was observed that resistivity
was disproportionately low for a given lithium-sodium
concentration when ashes having high iron content were
examined. A good linear correlation was established
between resistivity and the atomic percentage of iron
present in the ash for a specific level of lithium and
sodium. From this information, an expression was developed
to predict volume resistivity as a function of the
37
-------
combined lithium and sodium concentrations and the iron
concentration for a given temperature and ash layer
porosity in the temperature range where volume conduc-
tion occurs. Additional expressions were given to account
for changes in resistivity with temperature and porosity.
These predictive expressions are considered initial
approximations that are in need of substantiation with
field tests and laboratory experiments utilizing other
techniques.
An effort was made to determine the mechanism by which the
iron concentration influenced resistivity. Unequivocal
proof was not attained. Several items of circumstantial
evidence indicate that the data were not distorted by
superior bonding of particulates when preparing sintered-
disc resistivity specimens containing above minimum iron
concentration. No observations and no specific experi-
ments produced data to indicate an electronic contribution
to the conduction process due to the presence of elevated
iron content. Transference experiments showed that at
all levels of iron only sodium and lithium had significantly
migrated. As the iron content among ashes increased, a
disproportionate percentage of lithium and sodium was
transferred based on the initial concentrations of these
elements in their respective ashes for a constant quantity
of electricity passed. This observation suggests that
the iron present in the glassy phase alters the structure
of the glass to permit the migration of a greater percentage
of the total number of alkali metal ions available.
Since the experimental activation energy did not vary
significantly among all experiments with respect to the
estimated error, it was considered not likely that the
iron affected the carrier mobility.
It is generally noted in glass technology that compositional
changes altering structure have a pronounced effect on
properties. It is conceivable that this phenomenon occurs
in fly ash. One might visualize the structural modifica-
tion being a physical or chemical change that is particularly
sensitive to an electric field.
38
-------
SECTION IX
REFERENCES
1. R. E. Bickelhaupt and G. B. Nichols, "Investigation of
the Volume Electrical Resistivity of Fly Ash from the
Sundance and Wabamun Power Stations", Final Report
A1231-2865-I from Southern Research Institute to
Calgary Power, Ltd., Calgary, Alberta, Canada,
June 30, 1972.
2. R. E. Bickelhaupt, "Electrical Volume Conduction in
Fly Ash", APCA Journal 24 (3) 251-255 (1974).
3. "Tentative Method of Test for D-C Volume Resistivity
of Glass", Designation C657-70T, Part 13, American
Society for Testing Materials, Philadelphia, April 1972.
4. W. D. Kingery, Property Measurements at High Temperatures,
Chapter 13, John Wiley & Sons, Inc., New York, 1959.
5. B. D. Cullity, Elements of X-ray Diffraction, Chapter
14, pp 396-398, Addison-Wesley Publishing Company, Inc.,
Reading, Massachusetts, 1959.
6. "Studies into the Effects of Coal Ash Composition on
Electrostatic Precipitation Performance", progress
report covering period of September 1971 to May 1972,
prepared for Environmental Protection Agency,
Contract 02-513-4. University of Belgrade, Belgrade,
Yugoslavia, May 1972.
7. J. Dalmon and D. Tidy, "A Comparison of Chemical
Additives as Aids to the Electrostatic Precipitation
of Fly Ash", Atmospheric Environment 6 (10) 721-734
(1972) .
8. C. C. Shale, J. H. Holden, and G. E. Fasching,
"Electrical Resistivity of Fly Ash at Temperatures
to 1500°F", R17041, Bureau of Mines, U.S. Department
of the Interior, 1968.
9. S. J. Selle, P. H. Tufte, and G. H. Gronhovd, "A Study
of the Electrical Resistivity of Fly Ashes from Low-
sulfur Western Coals using Various Methods", Paper
72-107 presented at the 65th Annual Meeting of the
Air Pollution Control Association, Miami Beach,
Florida (1972).
39
-------
10. J. Dalmon and E. Raask, "Resistivity of Particulate
Coal Minerals", J. Inst. Fuel 46 (4) 201-205 (1972).
11. H. P. R. Frederikse and W. R. Hosier, "Electrical
Conductivity of Coal Slag", J. Am. Cer. S. 56 (8)
418-419 (1973).
12. G. W. Morey, The Properties of Glass, Chapter 20,
Reinhold Publishing Corporation, New York (1938).
13. H. J. L. Trap and J. M. Stevels, "Ionic and Electronic
Conductivity of Some New Types of Glass-like Materials",
Physics and Chemistry of Glasses 4_ (5) 193-205 (1963) .
14. Robert H. Doremus, Glass Sciences, Chapter 10,
John Wiley & Sons, New York (1973).
15. J. D. Mackenzie, "Semiconducting Oxide Glasses:
General Principle for Preparation", J. Am. Cer. S. 4_7
(5) 211-214 (1964).
16. Kalevi Kiukkola and Carl Wagner, "Measurements on
Galvanic Cells Involving Solid Electrolytes",
J. Electrochemical Soc. 104 (6) 379-387 (1957).
40
-------
TECHNICAL REPORT DATA
(Please read InUmcnoiis on the reverse before completing)
1 REPORT NO.
EPA-65Q/2-74-074
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Influence of Fly Ash Compositional Factors on
Electrical Volume Resistivity
S. REPORT DATE
1974
July
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R. E. Bickelhaupt
8. PERFORMING ORGANIZATION REPORT NO.
SORI-EAS-74-247
2938-F
9. PERFORMING OR6ANIZATION 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
68-02-0284
12. SPONSORING AGENCY NAME AND ADDRESS
EPA5 Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; Through July 1974
14. SPONSORING AGENCY CODE
IS. SUPPLEMENTARY NOTES
16. ABSTRACT
The report gives results of a study during which 28 fly ash samples --
representing a broad spectrum of ash compositions produced by burning coal in
commercial power station boilers—were characterized, chemically analyzed, and
fabricated into sintered-disc resistivity specimens. Resistivity and transference
experiments were performed in the temperature range in which volume conduction
prevails. Results confirmed conclusions of an earlier investigation: the volume
conduction mechanism for fly ash is ionic; and the charge carriers are the alkali
metal ions, principally sodium. Increasing iron concentration caused a resistivity
decrease for a given level of sodium and lithium. No evidence of biased data or
electronic conduction was found. It was rationalized that, in a manner analagous to
that for glass, the iron affected the structure of the predominant glassy phase of the
ash, thereby inducing the participation of a greater percentage of the available alkali
metal carrier ions. From these data, empirical equations were developed to predict
the volume resistivity of fly ash as a function of ash chemistry, temperature,
and porosity.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI 1 icId/Group
Air Pollution
Fly Ash
Chemical Analysis
Coal
Combustion Products
Boilers
Air Pollution Control
Stationary Sources
Characterization
Resistivity
Volume Conduction
13B
21B
07D
21D
ISA
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21 NO OF PAGES
49
20. SECURITY CLASS (Thispage)
Unclassified
22 PRICE
EPA Form 2220-1 (9-73)
41
------- |