EPA-650/2-75-066
July 1975
Environmental Protection Technology Series
HAZARDOUS
EMISSION CHARACTERIZATION
OF UTILITY BOILERS
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D. C. 20460
-------
f
EPA-650/2-75-066
HAZARDOUS
EMISSION CHARACTERIZATION
OF UTILITY BOILERS
U.S. FnvlrorfflarsU! Pwtodton
Region III Information Beseurw
Center (3PM52) -
Chataut She* '/
ic» ma r
C
by
Chatten Cowherd, Jr., Mark Marcus,
Christine M. Guenther, and James L. Spigarelli
Midwest Research Institute
425 Volker Boulevard
Kansas City , Missouri 64110
Contract No. 68-02-1324, Task 27
ROAP No. 21AUZ-002
Program Element No. 1AB015
EPA Project Officer: Ronald A. Venezia
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C. 20460
July 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-650/2-75-066
11
-------
PREFACE
This document was prepared for EFA/CSL under EPA Contract No.
68-02-1324, Task 27. The work reported herein was conducted in three
sections of the Physical Sciences Division of Midwest Research In-
stitute. The Program Coordinator was Dr. L. J. Shannon, Head, Environ-
mental Systems Section. Development of the overall sampling plan and
analysis of field and laboratory data were performed in this section
by Dr. Chatten Cowherd assisted by Ms. Christine Guenther. The field
sampling effort was carried out under the direction of Mr. Paul
Constant, Head, Environmental Measurements Section, with Mr. William
Maxwell serving as crew chief at the test site. The chemical analy-
sis of collected samples was directed by Dr. James Spigarelli, Head,
Analytical Chemistry Section, with assistance from Dr. Mark Marcus.
Approved for:
ID WEST RES
H. M. Hubbd
rd,
Physical Sciences Division
July 1975
111
-------
CONTENTS
Page
List of Figures vii
List of Tables ix
Sections
Summary. • 1
I Conclusions and Recommendations 3
II Introduction . . 6
Background 6
Program Objective 7
Related Studies 7
III Hazardous Pollutants from Fossil Fuel Combustion . 10
Hazardous Constituents in Utility Boiler Fuel. . 10
Hazardous Pollutant Impact Rating 14
IV Characteristics of Utility Boilers and Flue Gas
Environment. 21
Boiler Characteristics 21
Flue Gas Environment 32
V Field Testing Procedures 37
Test Facility Selection 37
Flue Gas (Mass-Rate) Sampling 38
Gross Coal and Ash Sampling 43
Inlet Air Sampling 46
Chemical Analysis of Collected Samples 46
Process Monitoring 46
-------
CONTENTS (Concluded)
Sections Page
VI Test Facility and Sampling Program 55
Test Facility 55
Sampling Program 60
VII Analytical Results and Quality Assurance 69
Analytical Results 69
Analytical Quality Assurance 90
VIII Calculated Test Results 94
Boiler Performance 94
Stack Gases and Inlet Air 100
Particle Size Distribution 100
Hazardous Pollutants 100
IX Discussion of Results 118
Mass Balance 118
Modifications to Sampling Train 121
Health Hazard Evaluation 123
Appendix A - Procedures for Handling, Preparation and
Analysis of Samples 127
Appendix B - Review of Chemical Analysis Methodology .... 138
Appendix C - Calculation of Boiler Steam Efficiency 154
Appendix D - Calculation Procedure for Mass Balance
Precision 162
Appendix E - Factors for Conversion to Metric Units 168
References 170
vi
-------
FIGURES
No. Page
1 Trace Element: Concentrations in Coal 13
2 Simplified Process Diagram for Steam-Electric
Generator 22
3 Prevalance of Coal-Firing Methods 23
4 Pulverized Coal-Firing Methods 24
5 Firing Method Vs Boiler Size 25
6 Trends in Utility Boiler Size 28
7 Utility Boiler Size Distribution 29
8 Utility Boiler Size Distribution 30
9 Utility Boiler Characteristics "Tree" 33
10 Control Devices Vs Boiler Size 34
11 Temperature History of Flue Gases 35
12 Flue Gas Sampling Train 41
13 Particle Size Sampling Train 44
14 Analyses for Flue Gas Sampling Train 49
15 Analyses for Particle Size Sampling Train 50
16 Dependence of Emissions on Load Factor 52
17 Heat Balance of Steam Generator 54
18 Widows Creek Unit 5 57
19 Simplified Diagram of Test Facility 58
20 Sampling Program - Run 2 62
21 Sampling Program - Run 3 63
22 Sampling Program - Run 4 64
23 Collector Inlet Duct Configuration 65
24 Stack Configuration 67
25 Variations in Coal Feed Rate - Run 2 97
26 Variations in Coal Feed Rate - Run 3 98
27 Variations in Coal Feed Rate - Run 4 99
28 Particle Size Distribution - Plates Only 105
vii
-------
FIGURES (Concluded)
No.
29 Particle Size Distribution - All Stages 106
30 Boiler Mass Imbalances 113
31 Dust Collector Mass Imbalances 114
32 Overall Mass Imbalances 115
33 Modified Flue Gas Sampling Train 122
34 Required Analyses for Mass-Rate Train 124
35 Sample Preparation and Analysis Diagram . 125
A-l Sample Preparation and Analysis Diagram 133
B-l General Procedure for Sample Treatment and Analysis. . . 140
C-l Steam Generating Unit Diagram 156
viii
-------
TABLES
No. Page
1 Origin of Coal Consumed by Electric Utilities in 1970. . 11
2 Trace Elements in U.S. Crude Oil 12
3 Trace Elements in Fossil Fuels Consumed by Electric
Utilities 15
4 Carcinogenic Polycyclic Organic Materials 16
5 Mass Balance Results—Cyclone-Fed Boiler 17
6 Volatility of Trace Elements in Coal 18
7 Hazardous Pollutant Impact Ranking 20
8 Size Categories for Steam-Electric Power Plants 27
9 Particulate Emission Control Equipment 31
10 Candidate Test Facilities 39
11 Statistical Model for Determination of Sampling
Frequency 45
12 Samples and Analyses 47
13 Chemical Analysis Methods 48
14 Operating Variables 51
15 Widows Creek Unit 5 Design Data 56
16 MRI Field Sampling Crew 61
17 Pollutant Concentration (ppm) in Coal, Ash, and Flue Gas
Streams 70
18 Comparison of Results with ORNL Data 76
19 Coal and Ash Properties 78
20 Particulate Mass (Grams) Collected in Flue Gas Sampling
Train 79
21 Pollutant Mass (Micrograms) Collected in Flue Gas
Sampling Train (Run 2, Dust Collector Inlet) 80
22 Pollutant Mass (Micrograms) Collected in Flue Gas
Sampling Train (Run 2, Dust Collector Outlet) 81
23 Pollutant Mass (Micrograms) Collected in Flue Gas Sam-
pling Train (Run 3, Dust Collector Inlet) 82
24 Pollutant Mass (Micrograms) Collected in Flue Gas Sam-
pling Train (Run 3, Dust Collector Outlet) 83
25 Pollutant Mass (Micrograms) Collected in Flue Gas Sam-
pling Train (Run 4, Dust Collector Inlet) 84
26 Pollutant Mass (Micrograms) Collected in Flue Gas Sam-
pling Train (Run 4, Dust Collector Outlet) 85
ix
-------
TABLES (Concluded)
No. Page
27 Particulate Mass (Grams) Collected in the Particle
Size Train 87
28 Pollutant Concentration (ppm) Versus Particle Size
Composite of Dust Collector Inlet Samples 88
29 Pollutant Concentration (ppm) Versus Particle Size
Composite of Dust Collector Outlet Samples 89
30 Quality Assurance Data 91
31 Boiler Conditions 95
32 Coal/Ash Mass Flow Rates—All Streams 96
33 Flue Gas Conditions 101
34 Stack Gas Composition 102
35 Inlet Air TSP Concentrations 103
36 Calculated Particle Size Distributions 104
37 Pollutant Mass Flow Rate (gra/min) in Coal, Ash, and
Flue Gas Streams 107
38 Average Mass Imbalances—All Runs 112
39 Uncontrolled Particulate Emission Factors 116
40 Pollutant Enrichment Ratios—Average, All Runs 117
41 Principal Sources of Inaccuracy and Imprecision in Mass
Balance 119
42 Mass Imbalance Versus Sampling Frequency 120
43 Potential Health Hazard Evaluation—Average All Runs. . 126
B-l Feasibility of Analytical Methods 142
B-2 Chemical Analysis Methods 144
B-3 Analysis Sensitivity of Atomic Absorption Spectrometry. 146
B-4 Required Sample Quantities for Analysis of Elemental
Pollutants by Flameless AAS 147
C-l ASME Test Form for Abbreviated Efficiency Test 157
C-2 ASME Test Form for Abbreviated Efficiency Test 158
D-l Definitions of Variances 163
-------
SUMMARY
This report presents the results of a field sampling program,
the most intensive single effort of its kind to date, aimed at the
quantification of potentially hazardous pollutants in the waste
streams of a representative coal-fired utility boiler. Fuel com-
bustion products that were identified as potentially hazardous air
pollutants included 22 trace elements, nitrates, sulfates, poly-
cyclic organic compounds, and polychlorinated biphenyls.
The test facility was a 125-MW boiler at the Tennessee Valley
Authority (TVA) Widows Creek steam electric generating station.
The boiler, fired with pulverized coal, was equipped with a mechani-
cal fly ash collector. Waste streams sampled included pulverized
coal, furnace bottom ash, superheater ash, collection ash and flue
gases at the inlet and outlet of the fly ash collection. An integrated
sampling train, especially designed and fabricated for this study,
was used to collect potentially hazardous particulates and vapors
from the boiler flue gases. Collected samples of coal, ash and flue
gases were quantitatively analyzed using standard instrumental
methods.
Quality assurance checks on the field and laboratory data in-
cluded complete mass balances on the sampled materials, analyses
of duplicate samples, determinations of recoveries from spiked sam-
ples, and analyses of certified samples of coal and fly ash from
the National Bureau of Standards (NBS).
Significant findings presented in this report include the fol-
lowing :
1. Acceptable mass balance was achieved for about half of
the elemental pollutants. The major causes of mass imbalance were:
(a) inefficient collection of vaporous metals in the flue gases; and
(b) possible analytical errors, particularly in the measurement of
trace constituents in coal. With few exceptions, the average mass
balance precision was within the expected tolerance (+ 25%).
-------
2. Trace metal enrichment was measured: (a) to a moderate
degree in cooler ash streams; and (b) to a high degree in the fine
particle portion of the fly ash.
Recommendations derived from this study focus on modifica-
tions to the methods for collecting and preparing samples. In gen-
eral, larger and more frequent samples of coal and bulk ash streams
are recommended to improve sample representativeness. The need is
emphasized for development of methodologies for estimating bulk
ash flows, to permit internal checks on the mass balances. Finally,
routine chemical analysis of NBS standard coal and fly ash will
improve quality assurance of the analytical methods.
-------
SECTION I
CONCLUSIONS AND RECOMMENDATIONS
The conclusions that are derived from this study fall into
two subject areas—test methodology and test results. These areas
will be discussed separately.
The following study conclusions pertain to the sampling and
analysis methods and the operation of the test boiler:
1. The boiler was operating near capacity and otherwise in
normal fashion during each of the sampling periods. The excess
combustion air averaged about 35% and there was an air infiltra-
tion rate of about 25% between the furnace and the stack. The
collection efficiency of the mechanical fly ash collector averaged
about 45%.
2. Wide variations in physical properties (e.g., color and
texture) of grab ash samples, particularly bottom ash and super-
heater ash, corroborated the large variations in the measured con-
centrations of trace elements. This suggests potential nonrepresenta-
tiveness of composite ash samples consisting of a small number of
grab samples.
3. The sampling train designed for collection of potentially
hazardous particulates and vapors from the flue gases functioned
satisfactorily for most pollutants, but did not efficiently trap
vaporous metals (antimony, arsenic, mercury and selenium) because
of a loss of oxidizing power of the impinger solution.
4. Analysis of particle size samples for trace metals was
limited by the small quantities of particulate collected on the
latter impaction stages and on the backup filter.
5. Analysis of duplicate samples indicated good precision
in the analytical techniques. However, the average percent re-
covery of spiked quantities spanned a wider range (80 to 115%).
Limited analysis of standard NBS coal and fly ash indicated that
determinations of trace metals in ash were more accurate than
determinations in coal.
3
-------
The conclusions which pertain to the calculated results of
the Widows Creek sampling study are as follows:
1. With few exceptions, measured mass balance precision was
within the expected tolerance of + 25%. Based on statistical analy-
sis of variance, the major cause of imprecision in mass balance
was nonrepresentative sampling; this source of imprecision can be
reduced by increasing the sampling frequency and volume sampled.
2. Mass imbalances on the boiler, dust collector and the com-
bination of the two were predominantly negative, indicating that
not all of the inlet mass flows could be accounted for in the cal-
culated outlet flows. Inefficient pollutant collection was the
major cause of the negative imbalance observed for the vaporous
elements (antimony, arsenic, mercury, selenium, and flourine). Mass
imbalance was consistently less for the dust collector than for
the boiler, reflecting the greater accuracy of pollutant measure-
ment in an ash matrix, and a higher degree of representativeness
in the samples collected from the flue gas stream.
3. There was a general tendency for progressive hazardous
pollutant enrichment of particulate matter, proceeding from coal
through the various ash streams to fly ash at the collector out-
let. Exceptions were antimony, barium, beryllium, manganese, tel-
lurium, titanium, and vanadium.
4. Fine particles were found to be enriched with most of the
trace metals, with the greatest degree of enrichment occurring for
beryllium, cadmium, copper, and zinc. Fine particle enrichment cor-
related with low removal efficiency by the fly ash collector, as
expected.
5. Comparison of flue gas concentrations of trace metals with
threshold limit values for industrial exposure indicates that only
beryllium is present in sufficient quantity to be of concern as
an air pollution health hazard.
As a result of the experience gained in the Widows Creek sam-
pling study, modifications to the methods for collecting and pre-
paring samples are recommended. A modified sampling train designed
to efficiently collect inorganic vapors along with particulates
and organic vapors is recommended for the collection of potentially
hazardous pollutants from the boiler flue gases. An alternate sam-
pling train developed by TRW is currently being used by KVB Engineer-
ing to sample hazardous emissions from industrial boilers. Neither
of these trains has been adequately field tested at this time.
4
-------
Because of the large short-term variations in the trace con-
stituent composition of coal and bulk ash streams, both sample size
and frequency should be increased to reduce the variance of mass
balance. The frequency of sampling corresponding to a desired tol-
erance may be determined by reworking the statistical model pre-
sented. Sample size must more closely conform with the specifications
of Method D2234 of the American Society for Testing and Materials.
In addition, methods should be developed for directly estimating
the flows of bulk ash streams.
In the Widows Creek study, it was assumed that soot buildup
on the boiler tubes had the same composition as superheater hopper
ash, and that soot plus superheater ash amounted to 20% of the in-
put coal ash. More data are needed to determine the rate of soot
buildup and samples should be obtained, if possible, for chemical
analysis.
A smaller digestion bomb should be used for treatment of sam-
ples from the Brink impactor. MRI has designed a bomb with a 5-ml
volume, one-tenth of the volume used in the Widows Creek study.
This device will facilitate treatment of small samples and transfer
of the 1-ml sample for analysis by atomic absorption spectrophoto-
metry.
Finally, it is recommended that NBS coal and fly ash standards
be analyzed routinely along with field samples collected in future
studies, to improve quality assurance of the analytical methods.
-------
SECTION II
INTRODUCTION
BACKGROUND
Recent studies of the hazardous air pollutant problem'' have
indicated that fossil fuel combustion in utility boilers is a ma-
jor contributing source.1-5/ This conclusion has been based pri-
marily upon the knowledge that coal and oil contain a variety of
potentially hazardous trace metals which may be discharged into
the atmosphere in the combustion waste gas stream. In addition,
the incomplete combustion of organic fuel is a source of poly-
nuclear compounds which are known carcinogens.£./ These conclu-
sion have been reinforced by limited measurements of hazardous
constituents in power plant emissions, primarily in fly ash.
The environmental impact of hazardous pollutants generated by
fossil fuel combustion in utility boilers is potentially much
greater than the relative mass emissions would suggest, because
some pollutants may be emitted in the form of vapors or fine parti-
culates which penetrate the customary particulate pollutant con-
trol devices as well as the conventional source sampling instru-
ments. The emissions of hazardous pollutants as fine particulates
(and vapors) intensify the potential adverse health effects for
two reasons: (a) pollutants in this form penetrate the natural
filters of the respiratory tract, and reach the air spaces of the
lung; and (b) fine particle emissions, largely formed by condensation
of volatile materials, are enriched in toxic elements in comparison
to the average composition of the earth's crust (a measure of the
acceptable metabolic tolerance level in humans).
Section 112 of the Clean Air Act defines hazardous pollutants
as those which may cause or contribute to an increase in
mortality or an increase in serious irreversible or in-
capacitating reversible illness.
-------
PROGRAM OBJECTIVE
EPA's Control Systems Laboratory has the responsibility to de-
velop technology to control hazardous emissions from utility boilers.
Necessarily, the control program must be based on field measurements
which adequately characterize and quantify the potentially hazard-
ous vapors and particulates in utility boiler waste gases. In the
absence of definitive data of this type, two primary objectives
were set for this investigative program:
1. The development of a comprehensive plan for measuring po-
tentially hazardous constituents in representative utility boiler
exhaust streams.
2. The field implementation of the test plan at a full-scale
utility boiler of representative design and operation.
The original intent of this program was to draw upon previ-
ously documented studies in the formulation of the plan for field
sampling of utility boilers and for chemical analysis of collected
samples. However, in the early stages of the program, the recogni-
tion of the inadequacies of currently available information (rela-
tive to the scope of the required testing effort) necessitated a
shift in emphasis from method application to method development.
Consequently, a major objective of the initial (field) ef-
fort to implement the test plan was the investigation of the re-
liability and accuracy of the sampling and analysis methods. TVA's
Widows Creek coal-fired steam-electric power plant was selected
for the test program.
RELATED STUDIES*
Little pertinent data are available from previous or current
multielement studies of trace materials in boiler waste gases.
Until about 1970, particulate pollutant sampling and analysis
methods were not sufficiently refined to permit the accurate deter-
mination of trace constituents in fly ash. Furthermore, techniques
for sampling volatile trace materials in the vapor or fine aerosol
state are just now being developed.
This literature review was conducted in mid-1974; other per-
tinent field studies have been initiated more recently.
-------
One current program of interest is Oak Ridge National Labora-
tory's investigation of the disposition of trace elements from
coal combustion at the TVA Thomas A. Allen Steam Plant in Memphis,
T«»nn«»ss«»P-.3i7/ The test boiler, which drives a 290 MW generating
unit, is cyclone-fed with crushed coal. (Unfortunately, this fir-
ing method represents only 9% of the total generating capacity of
coal-fired utility boilers.) Mass balance data and the collection
efficiency of a new electrostatic precipitator have been deter-
mined for 34 elements including many trace constituents. Samples
of particulates collected with a cascade impactor are being analy-
zed to determine trace element concentration as a function of par-
ticle size.
In another study, by the University of Colorado, data have
been collected on the distribution of 16 particulate trace ele-
ments in the waste streams of a 180 MW generating unit fired with
pulverized coal, at the Valmont Power Plant near Boulder, Colorado.^/
Particulate emissions from the test boiler were controlled with a
mechanical collector followed by an electrostatic precipitator in
parallel with a wet scrubber. However, no data on particle size or
particulate collection efficiency were reported.
In other related work, the Edison Electric Institute is spon-
soring a study at the Battelle Memorial Institute to obtain a mass
balance for 14 critical elements around an experimental boiler
facility fired with either pulverized coal or residual oil. Battelle
has also conducted a recent project for EPA to determine the effect
of alternative sampling techniques on the amount and composition
of particulates measured in the effluent gases of oil- and coal-
fired combustion sources..27 In conjunction with this latter work,
approximate trace element analysis was performed on fly ash sam-
ples collected at the inlet and outlet of an ESP which controlled
emissions from a pulverized coal-fired boiler at the Edgewater Power
Plant in Loraine, Ohio.
Finally, the EPA has maintained interest in the fate of trace
elements in relation to various SOo removal demonstration projects.
For example, as subcontractor to the MITRE Corporation on the Cat-
Ox demonstration program, MRI collected coal ash samples and mea-
sured the concentrations of some 28 elements in the waste streams
of a pulverized coal-fired boiler with a mechanical collector.^/
The test boiler drove a 100 MW generator at Illinois Power Company's
Wood River, Illinois, plant.
-------
The remainder of this report is organized as follows:
Section Subject
HI
IV
VI
VII
VIII
Potentially Hazardous Pollutants Generated by
Utility Boilers
Typical Boiler Characteristics and Flue Gas
Environment
The Test Plan for Field Sampling and Chemi-
cal Analysis of Hazardous Emissions from
Representative Utility Boilers
The Widows Creek Test Boiler and Associated
Sampling Program
Analytical Results from the Widows Creek
Program
Test Results Calculated from Laboratory and
Field Data
-------
SECTION III
HAZARDOUS POLLUTANTS FROM FOSSIL FUEL COMBUSTION
The first major task in this program was to develop a list of
potentially hazardous pollutants originating from fossil fuel com-
bustion in utility boilers. To accomplish this, several lists of
hazardous pollutants from MRI's literature files were combined to
form a master list, which in turn was checked against information
on the trace elements in coal and fuel oil consumed by power plants.
HAZARDOUS CONSTITUENTS IN UTILITY BOILER FUEL
As the first step in the determination of representative con-
centrations of trace elements in utility boiler fuels, the origin
of coalli' and residual fuel oil—consumed by utility boilers
was determined. It was found that: (a) nearly all of the coal con-
sumed by power plants comes from two coal mining regions of the
country: the Appalachian (56%) and Interior Eastern (347.) regions
(Table 1) (the percentage of coal consumption originating from the
western region increased from 5.0% in 1970 to 6.7% in IQVZi^a/).
and (b) most of the domestic crude oil originates from three states--
Texas, Louisiana, and California (Table 2) (of the total amount of
crude oil processed in the United States, approximately 85% is
produced domestically with the balance being importedii').
Next, the most reliable information available on the poten-
tially hazardous elements in coal and oil—was analyzed to deter-
mine representative concentrations in the fuel from each producing
region. Figure 1 gives the extremes in average concentration of
trace elements for over 90% of the beds within each coal producing
region. It is important to note that for most trace elements, the
variations of concentration within a coal bed are frequently greater
than the differences between the averages for different beds.il/
The information on the origin of residual fuel oil and the hazard-
ous constituents in oil was much less complete than the information
on coal; however, the data for residual fuel oil are less criti-
cal because oil accounts for only 20% of the Btu output from utility
boilersJLft/
10
-------
ll/
Table 1. ORIGIN OF COAL CONSUMED BY ELECTRIC UTILITIES IN 1970
Consumption8-
(thousand short tons)
1 -
2 -
3 and
4 -
7 -
8 -
9 -
10 -
11 -
12 -
13 -
14 -
15 -
16 -
17 -
18 -
19 -
20 -
21 -
22 -
23 -
Origin of Coal
District
Pennsylvania, Maryland,
West Virginia
Pennsylvania
6 - West Virginia
Ohio
Virginia, West Virginia
Kentucky (east) ,
Tennessee (east), Virginia,
West Virginia
Kentucky (west)
Illinois
Indiana
Iowa
Alabama, Tennessee (west)
Arkansas
Kansas, Missouri, Oklahoma
Colorado
Colorado, New Mexico
New Mexico, Arizona
Wyoming
Utah
North Dakota, South Dakota
Montana
Washington, Alaska
Total
Region
A
A
A
A
A
A
IE
IE
IE
IW
A
IW
IW
SW
SW
SW
W
SW
N
N
N -f Alask
Electric
Utilities
31,089
8,804
36,809
41,893
1,192
62,009
47,844
50,745
15,956
812
11,296
—
6,981
522
2,357
6,525
6,405
1,005
4,870
a) 2>237
339,351
Total
46,647
39,581
50,053
55,699
37,128
161,022
53,360
67,660
22,641
882
20,511
1,000
7,625
593
6,602
6,498
7,215
4,586
5,916
2,773
597,992
a/ Total Electric Utility Consumption by Coal Source Region, in thousand
short tons.
Appalachian (A) 193,092
Interior - Eastern (IE) 114,545
Interior - Western (IW) 7,793
Western (W) 16,814
Southwestern (SW) (10,409)
Northern (N) 7.107
Total 339,351
11
-------
Table 2. TRACE ELEMENTS IN U.S. CRUDE OIL
137
(Parts Per Million)—
Element
Antimony
Arsenic
Barium
Manganese
Nickel
Tin
Vanadium
California
< 0.007
<0.007
<0.06
0.018
77
<0.6
48
Origin of Crude
Louisiana
0.05
0.05
0.09
0.027
4.4
0.5
1
Texas
<0.01
<0.12
<0.14
<0.05
3.3
<1.0
1.9
Weighted ,
• a/
Average-=-
< 0.024
<0.08
<0.11
<0.04
16
<0.8
9.0
/ 12/
-' From Petroleum Facts and Figures 1971—
State
California
Louisiana
Texas
1968 Production
(1.000 barrels)
375,496
817,426
1,133,380
12
-------
Arsenic
BPI> Ilium
Boion
Chromium
Cobalt
Copper
Fluoride
Lead
Mercury
Nickel
Tin
Vanadium
Zinc
0
1 ' '
k £LD * •*
1
• — r
i i i i i i 1 1
.1 i
. i i . . ...
. e $
5
A
X
.
A i 1
*•
X
a
i i i i i i i i
0 10
' 'x ' '
-------
Finally, representative concentrations of trace elements in
utility boiler fuel (Table 3) were derived by weighting the average
concentrations for each coal or oil producing region by the amount
of fuel from that region which is consumed by utility boilers. Also
shown in Table 3 are emission factors based on average heating values.
Although little information is available on hazardous organic
constituents in utility boiler emissions, there is strong reason to
suspect that polycyclic organic material (POM) and possibly poly-
chlorinated biphenyls (FOB) are formed during combustion of fossil
fuels and escape to the atmosphere prior to complete oxidation.
BenzoQaJpyrene, one of the key carcinogens present in the atmosphere,
has been specifically measured in power plant waste gases.i2jLiit±2'
Other POM with high carcinogenicity ratings are indicated in Table 4.
HAZARDOUS POLLUTANT IMPACT RATING
The relative potential environmental impact of hazardous pol-
lutants generated by utility boilers depends on three factors:
1. The amount of pollutant potentially liberated to the flue
gas stream per unit of heat input.
2. The volatility of the pollutant or its tendency to exist
in flue gases as a vapor or fine particulate.
3. The inherent toxicity of the pollutant.
Data on the volatility of trace constituents were obtained from
a recent study of the retention of elements during the oxidation of
coal at various temperatures.^' The boiling point of the pure ele-
ments was also taken as a measure of its volatility.Z/ Finally* for
the elements treated in the experimental studies cited above,Ai2-t§/
further indicators of high volatility were taken to be: (a) high
negative mass imbalance; (b) fine particle enrichment; or (c) high
relative mobilization to flue gases (Table 5).
Scientific judgment was used in developing the volatility clas-
sification from all of the applicable evidence; the results are
shown in Table 6. The temperature range of interest for the iden-
tification and control of hazardous pollutants is 200 to 800°F.
14
-------
Table 3. TRACE ELEMENTS IN FOSSIL FUELS CONSUMED BY ELECTRIC UTILITIES
Element
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Chlorine
Chromium
Cobalt
Copper
Fluorine
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Thallium
Tin
Titanium
Vanadium
Zinc
a/ Source:
b_/ Ratio of
c_/ Based on
d/ Based on
Concentration^/
(ppm)
5S/
32
5006-7
2.44
61
0.03
160
15.4
4.8
13.5
82
9.5
506-/
0.15
14.8
2.2
1£/ i
0.3^
0.9
385
26.4
12
Coal
Variance^/
5
6
2
2
1.5
3
3
3
3
>3
2
>5
Potential Pollutants in Fossil
highest to lowest
heating value of
heating value of
averages, for
11,200 Btu/lb
18,400 Btu/lb
Oil
Emission Factor Concentration— ' Emission Factor
(e/106 Btu)-7 (ppm) (e/106 Btu)d/
0.20 < 0.024 0.0059
1.3 < 0.08 0.002
20.2 < 0.11 0.003
0.099
2.47
0.001
6.48
0.624
0.194
0.547
3.32
0.38
2.02 < 0.04 0.001
0.0061
0.599 16 0.39
0.089
0.04
0.01
C.036 < 0.8 0.02
15.6
1.07 9 0.22
0.49
13/
Fuels.
coal regions or areas.
for coal as burned. 1ft'
for residual oil as burned. M(
£/ Estimated values based on data from various g«"r"a? 3»15,17/
-------
Table 4. CARCINOGENIC POLYCYCLIC ORGANIC MATERIALS
Compound Structure Carcinogenicity—
1. 7,12 Dimethylbenz[a]-
anthracene
2. Dibenz[a,h]anthracene
3. BenzoCcJphenanthrene
4. 3 Methylcholanthrene f^\ +4
5. Benzo[a_Jpyrene
6. Dibenzo[a_,h_jpyrene
7. Dibenzo[a,i]pyrene
8. Dibenzo[c,g]carbazole H +3
at Taken from Ref. 6; compounds of highest carcinogenicity, (b > 3 > 2...)
from a list of over 50 compounds rated from (-) to (+ 4).
16
-------
Table 5. MASS BALANCE RESULTS--CYCLONE-FED BOILER^'
Element
Arsenic
Barium
Beryllium
Boron
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Titanium
Vanadium
Zinc
a/ NAA =
SSMS =
AAS =
Analysis
Method^/
NAA
SSMS
NAA
SSMS
SSMS
SSMS
SSMS
NAA
SSMS
NAA
SSMS
SSMS
SSMS
NAA
SSMS
AAS
SSMS
NAA
NAA
SSMS
NAA
SSMS
SSMS
Mass Flow (em/rain)
Coal
4.
6.
99
130
< 6.
250
0.
26
37
4.
9
63
< 25
67
130
0.
<; 130
4.
890
880
26
37
110
Neutron Activation Analysis
Spark Source Mass
Atomic Absorption
Slag Tank
7 0.05
2 0.22
66
33
3 < 1.1
33
63 0.22
20
< 22
1 2.1
4.4
22
0.33
46
110
079 0.0099
55
0 1.5
330
220
14
11
11
(±107.)
% Mobilized to
Flue Gases
99
96
33
75
-83
87
65
23
> 40
49
51
65
< 99
31
15
87
s 58
62
63
75
46
70
90
Spectroscopy (±507.)
Spectrosc
opy (±57.)
17
-------
Table 6. VOLATILITY OF TRACE ELEMENTS IN COAL
< 300°F
Volatility Index and Temperaturea»W
300-850°F
850-1300°F
> 1300°F
mercury
fluorine
thallium
antimony
selenium zinc0-' copper
arsenic , cobalt
c/
barium-'
lead
chlorine
manganese
tellurium ,.
d/
nickel-
chromium-'
e/
cadmium^'
beryllium
boron
titanium
vanadium
tin
a/ Entries above dashed lines are from Occurrence and Distribution of
Potentially Volatile Trace Elements in Coal.jj*/
b_/ Temperature ranges within which volatilization of an element occurs.
cl Preferentially concentrated in fine particles of fly ash (pulverized firing)
d/ Concentrated in crust of moderate-sized particles of fly ash (pulverized
firing).!/
_e/ Large percentage to bottom ash (pulverized firing).—'
18
-------
The relative toxicity index of each constituent was based on
the threshold limit values determined by the American Conference
of Governmental Industrial Hygienistsi2' and on the review article
by Dr. H. A. Schroeder.l?/
Table 7 presents the impact ranking based on concentration,
volatility, and toxicity of each potentially hazardous element. In
the composite ranking, toxicity has been given more weight than
the other two factors. Note that boron was eliminated from further
consideration because of its low impact ranking coupled with in-
herent difficulties in chemical analysis procedures. It was judged
that the added costs of boron analysis were not justified by the
relative significance information to be gained.
The remainder of this document focuses on coal-fired boilers
which account for 807. of the Btu output of utility boilers.ifL/
19
-------
Table 7. HAZARDOUS POLLUTANT IMPACT RANKING
fo
O
Concentration in Fuel
Element
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Chlorine
Chromium
Cobalt •
Copper
Fluorine
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Thallium
Tin
Titanium
Vanadium
Zinc
g/106 Btu£/
0.16
1.04
16.2
0.079
1.98
0.0008
5.2
0.50
0.15
0.44
2.66
0.30
1.62
0.005
0.56
0.071
0.032
0.0096
0.033
12.5
0.90
0.39
a/ Concentration factors for coal
b/ 1 = > 10
£/ See Table
g/106 Btu, 2 = 1-10,
Index*
3
2
1
4
2
5
2
3
3
3
2
3
2
5
3
4
4
5
4
1
3
3
Volatility
I/ IndexS/
1-2
2
3
5
5
4
2
4
4-5
4
1
4
4
1
4
2
2-3
1
4
4-5
4
3
Toxicity Composite Impact
Index!/ Index!/
3
3
4l/
1
5
2
4
3
2-3l/
3
3-4
2
4
z£/
3
2
2
2
4
5
af/
3-4
40
36
48
20
250
80
64
108
84
108
24
48
128
20
108
32
40
20
256
112
108
110
Ranking
7
6
9
1
21
12
11
14
13
14
4
9
20
1
14
5
7
1
22
19
14
18
and oil (Table 3) weighted 4 to 1, respectively.
3 = 0.
1-1, 4 = 0.01-0.1, 5 =
< 0.01.
6 for index categories.
d/ 1 = Extremely toxic, threshold
" 3 = Very
£/ Composite
toxic, 0.5 * TLV < 2;
limit
value (TLV) < 0.1 mg/m
4 = Toxic, 2 £ TLV < 15; 5 =
= (Toxicity)2 x Volatility
x Concentration.
3; 2 = Highly toxic
Mildly toxic, TLV ;
, 0.1 s
* 15.
s TLV < 0.5;
_f/ Value adjusted from information in Reference 17.
-------
SECTION IV
CHARACTERISTICS OF UTILITY BOILERS AND FLUE GAS ENVIRONMENT
The typical coal-fired steam-electric plant is comprised of
a boiler, generator, condenser, fuel handling equipment, dust col-
lection and disposal equipment, water handling and treatment fa-
cilities, and heat recovery systems such as economizers and air
heaters. A simplified process schematic is given in Figure 2.
BOILER CHARACTERISTICS
The major design characteristics which differentiate utility
boilers are: (a) fossil fuel type; (b) method of firing; (c) fur-
nace temperature; (d) gross generating capacity; and (e) air pol-
lution control equipment. Each of these factors will be discussed
below.
Method of Firing
The mechanical firing methods for coal-fired boilers in pres-
ent day usage include pulverized, cyclone (utilizing a coarser mix
of pulverized coal), and stoker types, with pulverized firing com-
prising nearly 90% of the total. Figure 3 shows the relative prev-
alence of firing methods, as derived from a recent (9 January 1974)
National Emissions Data System (NEDS) listing.—/ Boilers fired
with pulverized coal are classified with respect to the firing posi-
tion of the burners as shown in Figure 4.±i' Figure 5 shows the prev-
alence of pulverized-coal firing method as a function of boiler size,
based on recent data provided by the Edison Electric Institute (EEI)._'
21
-------
STACK
AIR
FD FAN
i n CAM
A.R CONTROL
CATCH I
A*
A*
FUEL
STEAM
AIR PREHEATER
*
BOILER
M EXHAUST GASES
FEEOWATER
EXHAUST (
|
ECONOMIZER
WORK >l
# OR
-------
COAL-FIRED UTILITY BOILERS
10
Control
Equipment:
None 124
G 49
C 263
WS 12
ESP 330
Other 3
G-C 2
G-ESP 40
C-ESP 98
C-C 7
ESP-ESP 8
G-Mist 1
G = Gravitational Collector
C = Cyclone
WS = Wet Scrubber
ESP= Electrostatic Precipitator
Mist = Mist Eliminator
Other = Process Change or
Unknown
13.4%
Pulverized
Wet Bottom
(222 Units)
9.5%
Cyclone
(81 Units)
73.8%
Pulverized
Dry Bottom
(937 Units)
3.3%
Stoker
(263 Units)
GENERATING CAPACITY
Figure 3. Prevalence of coal-firing
-------
Primary Ai
and Coal
Primary air
and Coal
Fan tail Multiple Intertube
(a) Vertical Firing
Primary Air •
and Coal
Plan View of Furnace
(b) Tangential Firing
Primary Air Primary Air
and Coal and Coal
I
Secondary Air Secondary Air <
Multiple Intertube Circular
(c) Horizontal Firing
Secondary Air
Primary Air
and Coal
Cyclone'
Primary Air
and Cool
(d) Cyclone Firing
(e) Opposed-Inclined Firing
Figure 4. Pulverized coal-firing methods.
24
21/
-------
ro
Ui
PULVERIZED COAL-FIRED (DRY BOTTOM) UTILITY BOILERS
FRONT OR BACK
100
200
300
400 500 600
BOILER SIZE (MEGAWATTS)
700
800
900
Vertical and Front or Back not in larger boilers
Opposed not in smaller boilers
1000
Figure 5. Firing method vs boiler size.——•
221
-------
For the most part, the burner configuration in units fired with
pulverzied coal does not in itself have a major effect on emission
characteristics.22/ However, based on Limited data which show that
the amount of unburned combustibles in the coal ash is affected by
method of firing,15*16*23^ it is likely that the firing method may
have a substantial effect on the generation of POM's. The control-
ling factors appear to be furnace temperature distribution and
residence time.
Furnace Temperature
Pulverized coal-fired boilers are also classified as either
wet-bottom (slag-tap) or dry-bottom, depending on the furnace tem-
perature relative to the ash fusion temperature. All cyclone-type
furnaces are wet-bottom.
Gross Generating Capacity
Utility boilers may be categorized by size as shown in Table
8. The trends in the average size of fossil-fueled steam-electric
generators in this country are depicted in Figure 6.1ft/ In 1968,
the average size steam-electric unit was 66 MW, but the average
size of utility boilers under construction in 1970 was 460 MW. (Since
1950 it has been common practice to have one boiler per turbine gen-
erator. 2ft/)
From recent inventories of utility boilers, the size dis-
tribution of the "current" population of pulverized coal-fired
(dry-bottom) utility boilers has been derived. Figure 7 shows the
distribution for boilers larger than 50 MW and Figure 8 for all
utility boilers, based on a recent (9 January 1974) NEDS inven-
tory.12/ From these figures it is evident that the boiler popula-
tion closely follows a log-normal size distribution. Also shown
in Figure 7 is a size distribution for pulverized coal-fired util-
ity boilers derived from the EEI inventory.ll/
Air Pollution Control Equipment
The most common types of particulate emission control equip-
ment for coal-fired utility boilers are centrifugal collectors and
electrostatic precipitators. Table 9 gives the distribution of con-
trol equipment currently in use on coal-fired utility boilers,
based on a recent listing from the National Emissions Data System.—/
26
-------
Table 8. SIZE CATEGORIES FOR STEAM-ELECTRIC POWER PLANTS
Btu per hour
Boiler horsepower
Pounds of steam per hour
Megawatts
Large
& 5 x 108
;> 15,000
^ 500,000
* 50
Intermediate
3 x 10s - 5 x 108
10-15,000
350-500,000
< 50 MW
Small
< 3 x 105
< 10
< 330
Not
applicable
27
-------
1000
I
Z
o 100
N
in
LU
O
FOSSIL-FUELED
UTILITY BOILERS
A New Units
OAII Umts
NEDS (Cool-fired) ' •
20
10
I
I
I
I
I
1930 1940 1950
1960
YEAR
1970 1980
1990
247
Figure 6. Trends in utility boiler size.
28
-------
% OF BOILERS GREATER THAN STATED SIZE
«M m ni
8
in
•i
O
Ul
M
./-)
Q
U I
_J
o
m
o
10
tl « w M >fi in »n in 10 ?n in
I m oi «i e« 801
rTrnTrmi
i-Eioi^oocj
PULVERIZED COAL-FIRED
(DRY BOTTOM)
UTILITY BOILERS,
£50 MW ONLY
MOST PROBABLE AVERAGE
SIZE (MW) SIZE (MW)
20/
<
L! !
11 t 10 ;U JO 4J >U bu /" IU 1MJ 15 *M
% OF BOILERS LESS THAN STATED SIZE
-------
% OF BOILERS GREATER THAN STATED SIZE
M« Ht •>•
1000
1
CD
N
to
100
10
! : |
jj
— • * •
- j ;
i !|
j ;
— :
; : :
h
; . i
• i
U
'
.
.
ilLL
PULVERIZED
COAL
-r
5f
:IRED
(DRY BOTTOM)
UTILITY
Most probable
Averaqe size
j
•
• •
i
-i *
. i
~ t
"\ —
•
•
— ^T
TT~H
': :
3 :
.: .
— i .
'
Itf
t •
1 i
n
i-
'':•
~ ~
....
-,..
.'.I:
.( j
;>:
:\:'
sii
"',
.
-;:
>
.
. t..
i
-',•
•fh;
JH:
: I :•
!:.
. : : .
...
. /
7?
/
Mi
' r
• i
— . . ,
.
J_
BOILERS
size = 62 MW
= 120 MW
__-
• ,
-
-;
/
/
1 i
* . ' •
:
'
.1 --
-
/
..
- . i
:T"
J
.}/
r-' •
.'.'.'
.
t
/
f
rt
::
• ' :
,li:
y
*
: .
• t
--•
-:.
jfjl
^
/
..1:
. . •
; . :
•-f
:'l
: •'
:
• : ;
l.i.
;i!l
J
7
f
•:-:
••
: ; : :
:::.
Hi)
|
'••••i
t
r
S
m
r
• ••
•
---
ilii
hi
i
(
;•
"i'
1
'-
.". '
•
"
;—
.:!
[
• : : .
i ...
ijlj
. . • ,
• •• i
i ; ; i
: •
j
3
i
1- •
4-f
•
i
: 1
:
1 1
- i
It .
.
.,.
Y
i '
• h
..
y
1 j
; :
w
rr- '-
£4
~r — '
• ' '. '.
ijj
: ;
1
• :
'
i
.
.
1
: i
! -
B
1 !
H
:
--^-
j ;
d i
! - _
I
77l~
, ;
i : J
fnr
H
.
;T ~
-i_, —
I
. i
«ni on •! e; o» i 2
M u MI M
••'i
CM
a
o
," i
v
-
u
0
.
.
.1:
9
BO
% OF BOILERS LESS THAN STATED SIZE
30
-------
Table 9. PARTICULATE EMISSION CONTROL EQUIPMENT
(Coal-Fired Utility Boilers®/)
Number of Emission Sources, by
Boiler Type
(Firing)
Pulverized -
Dry Bottom
Pulverized -
Wet Bottom
Cyclone
Stoker
TOTAL
No
Control
108
40
15
28
191
Gravitational
Collector
83
3
1
15
102
Centrifugal
Collector
355
76
15
77
523
Control Type
Electrostatic
Precipitator
330
87
48
1_
466
Wet
Scrubber
10
4
3
0
17
Total
886
210
82
121
1,299
a/ Boiler size > 100 million Btu/hr (10 MW).
Source: NEDS listing, dated 8/28/73.^
-------
Figure 9 is a boiler characteristics "tree" which was devel-
oped from a NEDS utility boiler inventory (9 January 1974) ,12/ It
divides the current coal-fired boiler population into firing types
and gives detailed information on the pulverized dry-bottom seg-
ment of the population. For each specified size range of pulverized
dry-bottom boilers, the figure specifies:
1. The percentage of the total generating capacity of coal-
fired utility boilers.
2. The number of boilers (indicated on the outer branches).
3. The distribution of particulate emission control devices.
Figure 10 presents the distribution of control devices as a
function of boiler size.
The emerging technologies for the control of gaseous emissions
from utility boilers should also be considered as potential sources
of hazardous emissions. Flue gas desulfurization systems are in-
creasing in number throughout the electric utility industry and
will soon have a substantial effect on utility boiler emissions.
In addition to fuel characteristics, the most important
process parameters in utility boiler operation are: (a) load fac-
factors (percentage of rated boiler capacity at any point in time);
(b) excess air; and (c) soot blowing cycle. For a given boiler,
the furnace temperature is determined by the first two of these
parameters.
FLUE GAS ENVIRONMENT
As flue gases travel from the point of generation to the
point of atmospheric discharge, two important changes take place:
(a) the temperature drops from near 3000°F to about 300°F and
(b) most of the large particles of fly ash are removed by a par-
ticulate collection device.
The temperature history of the flue gases (Figure 11) has a
major effect on the physical and chemical form of potentially
hazardous flue gas constituents.
32
-------
None
G.
C
WS
ESP
G-C
G-ESP
C-ESP
C-C
ESP-ESP
1
2
14
1
29
2
10
12
1
1
None 4
G 1
C 20
WS 2
ESP 14
G-ESP 2
C 1
ESP 15
C-ESP 5
G-Mist 1
22
G 1
C 6
WS 1
ESP 14
G-ESP 1
C-ESP 7
r r
WS 2
ESP 15
G-ESP 1
C-ESP 6
24
/
C 1
WS 3
ESP 9
G-ESP 2
ESP-ESP 3
/18 C
WS
M
' ESP
G-ESP
1
1
10
3
C
ESP
1
18
.'
ESP
Other
5
2
Pulverized Wet-Bottom
Stoker and Cyclone
GENERATING CAPACITY OF
COAL-FIRED UTILITY BOILERS
G - Gravitational Collector
C - Cyclone
WS - Wet Scrubber
ESP - Electrostatic Precipitator
Mist - Mist Eliminator
Other - Process Change or Unknown
Figure 9. Utility boiler characteristics "tree.1
-------
PULVERIZED COAL-FIRED (DRY BOTTOM) UTILITY BOILERS
1.0
0.8
-
CO
<
CD
O 0.6
LU
u
UJ
O
o;
t—
z
O
u
0.4
0.2
OTHER PRIMARY
SECONDAR
100
200
300
400 500 600
BOILER SIZE (MEGAWATTS)
700
800
900
1000
Figure 10. Control devices vs boiler size.±H
20/
-------
2800
2400
2000
1600
I
3
§
1200
•800
400
Typical Coal-Fired Boiler
Burner-
Furnace
Boiler/Economizer
/Air- heater
Precipitator
•^^•^•^^-^^
Stack
o
Figure 11. Temperature history of flue gases.
-------
In the practical temperature range of interest (200 to 800°F),
hazardous pollutants may exist either in the vaporous state or
as one of the following categories of particulates: (a) solid
or molten particles, largely inorganic oxides; (b) nucleated
material from supersaturated vapors (partial pressure exceeding
vapor pressure of condensed phase) in the gradually cooling flue
gases; (c) material added to solid or liquid particles by physi-
cal or chemical sorption; and (d) condensible material that
originates through chemical reaction. The volatility of an in-
dividual constituent is a major factor in determining its physi-
cal form. In most cases, the more volatile the element, the more
likely it will escape as an uncondensed vapor, or as a fine par-
ticulate. Most trace metals are released as oxides although some
heavy metals such as mercury are so electropositive that they
may appear as the free element.
36
-------
SECTION V
FIELD TESTING PROCEDURES
This section describes (a) the selection of the test boiler,
(b) sampling and analysis methods used to measure hazardous emis-
sions and (c) process monitoring procedures. The testing program
was designed to yield essential information on the following as-
pects of individual hazardous pollutants:
1. Mass emission rate.
2. Physical state (including particle size) versus loca-
tion in flue gas stream.
3. Efficacy of removal by commonly used particulate emis-
sion control devices.
4. Distribution among process waste streams (mass balance).
5. Effect of process variations.
TEST FACILITY SELECTION
The following representativeness criteria (as developed in
Section III) were used in the selection of the test boiler:
1. Pulverized coal-firing, dry-bottom (representing 73.8%
of coal-fired utility boiler generating capacity).
2. Representative megawatt size, i.e., most probable size
(~ 122 MW) of pulverized coal-fired utility boilers larger than
50 MW (Figure 7).
3. Available coal from the Appalachian or Interior Eastern
source region (accounting for 91% of utility boiler coal consump-
tion).
37
-------
4. Common particulate emission control device (Figures 9
and 10).
In addition, it was considered highly desirable that exten-
sive data from previous tests be available so that specific sam-
pling requirements could be determined without the necessity of
preliminary testing. Finally, the cooperation of plant personnel
was considered essential to the success of the program.
Table 10 lists utility boilers in the high density size
range (100 to 150 MW) which generally were found to meet the
representativeness criteria; however, in some cases preliminary
test have may have been lacking and/or utility personnel may not
have been favorably disposed to the proposed testing. A 125-MW
boiler at TVA's Widows Creek steam-electric generating station
was judged to best comply with the selection criteria.
FLUE GAS (MASS-RATE) SAMPLING
The major elements of a flue gas sampling methodology con-
sist of the sampling train design, the selection of test loca-
tions, and the detailed sampling procedures. Each of these ele-
ments will be discussed below, particularly with regard to special
problems which were anticipated in the sampling of hazardous pol-
lutants. A number of comprehensive treatises on sampling meth-
QdolQgy26a27/ have been written which review general requirements
and describe the reference procedures which have been developed.
Sampling Locations
Two general flue gas sampling locations were dictated by the
test objectives: (a) control device inlet (400°F); and (b) con-
trol device outlet (300°F). The following criteria were used in
selecting specific locations:
1. The flow within the duct should be relatively free of
mechanical disturbances (indicated by large-scale velocity fluctu-
ations).
2. Adequate space for sampling equipment must be available
outside the duct; that is, there should be no physical obstructions
within a distance of at least one duct diameter.
3. Standards of safety for the test crew must be met.
38
-------
Table 10. CANDIDATE TEST FACILITIES (100-150 MW)£/
\o
Utility
Cincinnati Gas & Electric
Dayton Power & Light
Detroit Edison
Duke Power
Georgia Power
Gulf Power
Illinois Power
Kansas Power & Light
Northern Indiana P. S.
Pennsylvania Power
Potomac Electric
TVA
Location
Ohio
Ohio
Michigan
So. Carolina
Georgia
Florida
Illlnola
Kansas
Indiana
Pennsylvania
Washington, D.C.
Alabama
Previously
Plant OAF CSL
W.C. Beckjord
Miami Fort
F.M. Talc
St. Clalr
Lee X
Hammond
Yates
Lansing Smith
Wood River X X
Lawrence X X
Mitchell X
Martins Creek
X
Widows Creek X
Tested
Cooper- Boiler Site
atlve*-' fMWI
X 105, 136
X 143
X 124
146(3)
132
X 104(3)
X 116(3). 134(1)
130, 141
X 110
125
115
X 138(2)
139
125^
NEDS DATA —
Control Eauloment Coal Ash
ESP, C-ESP
C-ESP
C-ESP
C-ESP
C
C-ESP
G-ESP
ESP
C-ESP
WS
ESP
ESP's
ESP
*f
Eff (H l
95. 96
96
97.5
98.1
85
85-98
98
97
99.5
99
98
99.8
96
Content (I)
16.9
13.5
13.1
12.9
6.8
10.3
10.7
-
10.6
12.8
9.8
13.3
9.5
Coal Sulfur
Content (1)
2.8
3.7
1.6
3.0
1.0
2.46
2.31
~
2.9
3.8
3.07
3.08
1.0
Firing Deaulf Coal
Type^7 System^ Source^
A, IE
A. IE
T A
F A
A,
IE,
IE,
IE.
X IE
X IW
T or F 1976 IE
F A
T A
A
IE
A
A
A
a/ Pulverized coal-fired utility boilers.
b/ Known to be cooperative by Control Systems Laboratory (CSL).
c/ From Edison Electric Institute llstlng:2i/ T - tangent; F - front-fired.
if From Ale/Water Pollution geport.
e/ From Keystone Coal Industry Manual: A - Appalachian; IE - Interior Eastern, IW - Interior Western.
fj Hot Included In NEDS listing.
-------
Sampling Procedures
Fending the results of current research on the reliability
of pollutant source sampling methods as a function of parent gas
conditions, it was decided that the flue gas sampling procedures
should be patterned after the widely used EPA Standard Methods.^/
This includes the following determinations:
1. Sampling point distribution by Method 1.
2. Preliminary velocity profile by Method 2.
3. Dry carrier gas composition by Method 3.
4. Preliminary moisture content by Method 4.
5. Hazardous pollutant emissions (particulate and vaporous)
by Method 5 (modified).
6. SOp emissions by Method 6.
7. NOX emissions by Method 7.
Hazardous Pollutant Sampling Train
A single integrated sampling train, modeled after the Method
5 train, was developed for measurement of mass flow rate of hazard-
ous pollutants. In the absence of the availability of adequately
developed sampling trains for hazardous pollutants, an integrated
train offered the advantage of physical manageability in typically
confined sampling locations. The mass rate train (diagrammed in
Figure 12) was designed for efficient collection of a variety of
potentially hazardous compounds: trace elements, organics, minor
cationic elements, sulfates and nitrates. Collection devices were
positioned to minimize interpollutant interferences during subse-
quent chemical analysis.
Specifically, the sampling train consisted of:
1. A Teflon-lined flexible probe, 12 ft long and heated to
stack temperature.
40
-------
Stock Temperature
Stack
Sample Box
Sample
Intake
L
J L. _JI
Figure 12. Flue gas sampling train.
-------
2. A quartz-fiber filter, heated to stack temperature (for
collection of particulates). A glass cyclone, shown in the diagram
(Figure 12), is generally used but was not for this study.
3. Eight impingers in an ice-water bath for collection of
condensibles and vaporous species: first two impingers containing
water (for removal of condensed water, POM's and PCB's); third im-
pinger, dry (for removal of carry-over from second impinger);
fourth impinger, 10% sodium carbonate (for removal of selenium and
sulfur dioxide); fifth and sixth impingers, 10% sulfuric acid, 3%
permanganate (for removal of inorganic vapors); seventh impinger,
dry (for removal of carry-over from sixth impinger); and eighth
impinger, silica gel (for removal of residual water vapor).
A. Quartz-fiber filter at approximately 80°F between third
and fourth impingers for collection of condensed particulate.
5. Pyrex tube packed with Tenax-GC®for collection of organic
vapors (polymeric material used as packing for chromatographic
columns).
6. Gas meters and pump as prescribed by EPA Method 5.
The individual components of the specified train were com-
mercially available; however, some modifications were necessary
to properly assemble the components.
The procedures used for removing collected samples from the
sampling train are detailed in Appendix A.
Particle Size Analysis
Cascade impactors were selected over other in-stack devices
for particle sizing because of the following operational advantages:
1. Measurement is effected in the size range of interest for
air pollution quantification and control.
2. Aerodynamic impaction is useful over a wide range of particle
sizes and concentrations.
3. Aerodynamic size distribution relates directly to environ-
mental effects and controllability.
42
-------
4. Measurements are not affected by optical or electrical
properties of particles.
5. Overall efficiency and fractional efficiency of parti-
culate emission control devices can be measured at the same time.
Brink in-stack cascade impactors were used for particle size
analyses at the two flue gas sampling locations* The sizing train
(Figure 13) was the basic configuration suggested by the manufac-
turer of the Brink impactor. The impactors were fitted with tared
aluminum foil liners for collection cups.
GROSS COAL AND ASH SAMPLING
To determine the required number of coal and ash samples for
accomplishment of test objectives, a simple boiler stream flow
model was postulated. The model boiler had one input (coal) stream
and three output (ash) streams. True average mass flows of hazard-
ous hazardous pollutant were designated for each stream (Table 11)
and the ratio of concentration extremes (i.e., maximum/ minimum)
was assumed to be a factor of two for each stream during a given
test.
Statistical analysis techniques were applied to determine the
number of samples required to estimate the true average mass flow
rate of the hazardous pollutant, corresponding to a given tolerance
(error), d , and confidence level. The analysis was based on the
following assumptions:
1. Measurement and stochastic errors are normally distributed
with standard deviations indicated in Table 11.
2. Errors are mutually independent.
The results of the analysis are presented in Table 11.
To achieve acceptable measurement accuracy (i.e., tolerance
d = 25% at the 95% confidence level), the model indicated that at
least three samples of each coal and ash stream (e.g., furnace
bottom ash and fly ash removed by control equipment) were required
for each run. Grab samples were sealed in heavy polyethylene bags.
43
-------
Cyclone
Probe Tip
Brink
Impactor
Stages
Filter
I I I I I QJ
Pump
Manometer
Figure 13. Particle size sampling train.
44
-------
Table 11. STATISTICAL MODEL FOR DETERMINATION OF SAMPLING FREQUENCY
Stream
Input:
Coal
.p. Output :
in
Bottom Ash
Collector Ash
Fly Ash
True
Mass
Flow
100
20
50
30
Stream
Flow
2.5
10.0
10.0
2.5
Relative Standard
Pollutant
Concentration
10.0
10.0
10.0
10.0
Deviation (%)
Pollutant Stochastic
Flow Error
10.3 16.7
11.4 16.7
11.4 16.7
10.3 16.7
Required
90% Confidence
d=107. d=20%
11 3
12 3
12 3
11 3
Number of
Level
d=25%
2
2
2
2
Samples
Per Run
95% Confidence
d=107.
15
16
16
15
d=20%
4
4
4
4
Level
d=25%
3
3
3
3
Note: d = Desired tolerance or an estimate of the range of mass flow rate as a percentage of the average mass flow rate.
-------
INLET AIR SAMPLING
09 /
During each run, a conventional high volume filtration unit—'
was used to sample inlet air near the intake to the forced draft
fan.
CHEMICAL ANALYSIS OF COLLECTED SAMPLES
This section lists the methods selected for chemical analysis
of collected samples. Table 12 presents a summary of the types of
samples obtained, the analyses performed on each sample type and
the chemical analysis methods. The potentially hazardous pollu-
tants selected for study and the methods that were for chemical
analysis of each pollutant are given in Table 13* Figures 14 and
15 indicate the specific analyses performed on discreet samples
taken from various components of the hazardous pollutant sampling
train and the particle sizing train, respectively.
The procedures used for sample handling, preparation and
analysis are described in Appendix A. A review of chemical analy-
sis methods which are applicable to the determination of hazard-
ous constituents in utility boiler flue gases, is presented in
Appendix B.
PROCESS MONITORING
The key process operating parameters which affect flue gas
temperature and composition (and, presumably, hazardous pollutant
characteristics) are listed in Table 14. The test plan required
that these quantities be precisely determined and held constant
(within normal control limits) during each run. Excess air was
measured at the flue gas sampling locations and the other param-
eters were derived from boiler and generator gauge readings.
For each test, the load factor was maintained near 100%.
Previous MRI test data (Figure 16) indicate that particulate
emissions are linearly dependent on load factor for loads above
75% of capacity.
Also during each run, boiler heat input was measured to pro-
vide the basis for hazardous pollutant emission factors and to
check compliance with federal performance standards for utility
boilers. Heat input was determined from fuel flowmeters and con-
firmed (a) by fuel analysis data (see Gross Coal and Ash Sampling
and Analysis Section) and measured flue gas characteristics and
(b) by a heat and mass balance over the steam generation system.
46
-------
Table 12. SAMPLES AND ANALYSES
Sample type
Coal
Ash (bottom ash,
superheater ash,
dust collector ash)
Flue gas
Particulates and
vapors
Particle size
Criteria gases
Carrier gases
Inlet air
Analyses
Combustion properties
Composition
Hazardous pollutants
Combustion properties
Hazardous pollutants
Hazardous pollutants
Total particulates
Hazardous pollutants
Total particulates
Sulfur dioxide
Nitrogen oxides
Oxygen
Carbon dioxide
Carbon monoxide
Nitrogen
Total suspended par-
ticulates
Analytical methods
Proximate analysis
a/
a/
Ultimate analysis-
Table 13
Proximate analysis
Table 13
Table 13
Gravimetry
Table 13
Gravimetry
EPA Method 6
EPA Method 7
Orsat (EPA Method 3)
Gravimetry
a./ American Society for Testing and Materials (ASTM) Method D271-70.
47
-------
Table 13. CHEMICAL ANALYSIS METHODS
Pollutant Methods of analysis-
Trace elements (cations)
Antimony
Arsenic
Barium 1
Beryllium 1
Cadmium 2
Chromium 1
Cobalt 1
Copper 1
Lead 2
Manganese 1
Mercury 2£/
Nickel 1
Selenium 1-'
Tellurium 1
Tin 2
Titanium 1
Vanadium 1
Zinc 1
Minor elements (cations)
Calcium 1
Iron 1
Anions
Chloride 3
Fluoride 4
Nitrate 7
Sulfate 8
Organics
POM 6
PCB 5
a/ The methods of analysis are as follows:
(1) Atomic Absorption Spectrometry (AAS),
conventional flame methods;
(2) AAS, micro flameless methods;
(3) AgN03 titration, electrochemical (EC) detection;
(4) EC, fluorine selective electrode;
(5) Gas chromatography (GC), electron capture
detection;
(6) GC, flame ionization detection;
(7) Spectrophotometric, phenol disulfonic
acid complex; and
(8) Barium perchlorate titration.
b_/ AAS, hydride generation methods.
c_/ AAS, cold vapor method.
48
-------
Mass Rate Train:
*\LJ •\£>j
Probe Heated
Quartz
Filter
H2O
H20
•_r e 1 _
Back-up
Quartz
Filter
fcifti
Tenax-GC
Plug
No
Acid
KMnO4
Acid
KMnO4
\D
""^^.^^ Sample
Analysis ^"•*-.*^>^
Elemental
Sb, As, Hg, Se
Chloride, Fluoride
Sulfate, Nitrate
DflM DfB
r\JN\, rv^D
Parti culates
123456789
• •
• • • • •
• • •
* «
• • • •
Figure 14. Analyses for flue gas sampling train.
-------
Sizing Train
Cyclone & Individual Back-up
Probe Tip Impactor Filter
Rinse Stages
^
Analysis
Metals
Sample
Particulates
1 2 3
• • •
• • •
Figure 15. Analyses for particle size sampling train.
50
-------
Table 14. OPERATING VARIABLES
Variable
Load Factor^'
Excess Air
Soot Bloving
Practical Range
50 to 100%
10 to 35%
Time between soot
blowing operations
Primary Effects on Flue Gas
. Flue gas flow rate
. Flue gas temperature
profile
. Flue gas composition
. Combustion efficiency
„ Flue gas composition
. Combustion efficiency
. Flue gas flow rate
. Flue gas composition
(particulate loading)
a/ Percentage of rated boiler capacity at any point in time.
51
-------
in
10
10,000 r
9,000 -
8,000 -
7,000 -
6,000 -
I/I
I/I
u
5,000 -
4.000 -
3,000 -
2,000 -
WOOD RIVER POWER PLANT
(100 MW UNIT)
Air Preheater
o Stack
10
40 50 60
LOAD FACTOR (%)
Figure 16. Dependence of emissions on load factor.
400,000
- 300,000
u
<
- 200,000
o
- 100,000
-------
For the heat and mass balances, the mass flow rates of major
input/output process streams (Figure 2) were measured during each
run. Data for these determinations included: (a) continuous read-
ings from boiler control room meters, recorded continuously or
periodically; (b) scale dump counts; and (c) flue gas measurements.
Boiler steam efficiency (Figure 17) was determined for each
run using the abbreviated heat loss method specified in American
Society of Mechanical Engineers (ASME) Performance Test Code 4.1
(1964). The calculation method is presented in Appendix C.
53
-------
HEAT IN FUEL (H,) (CHEMICAL)
PTC 4.1 - 1964
INPUT
ENVELOPE.
HEAT IN ENTERING AIR
HEAT IN ATOMIZING STEAM
SENSIBLE HEAT IN FUEL
PULVERIZER OR CRUSHER POWER
BOILER CIRCULATING PUMP POWER
PRIMARY AIR FAN POWER
RECIRCULATING GAS FAN POWER
• A
HEAT SUPPLIED BY MOISTURE
IN ENTERING AIR
B
w HEAT IN COOLING WATER
CREDITS (B>
BOUNDARY
C^
C7
- HEAT IN PRIMARY STEAM
- HEAT IN DESUPERHEATER WATER AND CIRCULATING PUMP INJECTION WATER
* HEAT IN FEEDWATER
•*• HEAT IN SLOWDOWN AND CIRCULATING PUMP LEAK-OFF WATER
•*- HEAT IN STEAM FOR MISCELLANFOUS USES
-*• HEAT IN REHEAT STEAM OUT
— HEAT IN OESUPERHEATER WATER
—* HEAT IN REHEAT STEAM IN
LOSSES (L)
UNBURNED CARBON IN REFUSE
HEAT IN DRY GAS
-„.
MOISTURE IN FUEL
MOISTURE FROM BURNING HYDROGEN
MOISTURE IN AIR
HEAT IN ATOMIZING STEAM
CARBON MONOXIDE
UNBURNED HYDROGEN
-UHC UNBURNED HYDROCARBONS
RADIATION AND CONVECTION
RADIATION TO ASH PIT. SENSIBLE HEAT IN
SLAG A LATENT HEAT OF FUSION OF SLAG
SENSIBLE HEAT IN FLUE DUST
HEAT IN PULVERIZER REJECTS
HEAT IN COOLING WATER
SOOT BLOWING
OUTPUT = INPUT - LOSSES
DEFINITION: EFFICIENCY (PERCENT) = (ft) =
» 10<> =
" '°°
HEAT BALANCE: H, + B = OUTPUT + L OR ^ (%) = 1 - ^-j^
100
Figure 17. Heat balance of steam generator.
54
-------
SECTION VI
TEST FACILITY AND SAMPLING PROGRAM
TEST FACILITY
Unit 5 at the Tennessee Valley Authority Widows Creek steam
electric power plant was selected for the testing program. The
Widows Creek six-unit installation is located on the west shore
of Guntersville Lake (Tennessee River), 5 miles southwest of
Bridgeport, Alabama. Unit 5 was put into commercial operation in
May 1954.
The major design features of Widows Creek Unit 5 are listed
in Table 15. An elevation view of this boiler is given in Figure
18.
Widows Creek Unit 5 is representative of the current popula-
tion of coal-fired utility boilers and meets the following criteria
established for the test facility:
. Pulverized coal firing; dry-bottom furnace.
125 MW generating capacity.
. Utilizes coal from Appalachian sources.
. Equipped with common particulate emission control device,
i.e., mechanical fly ash collector.
A simplified design drawing for Unit 5 is shown in Figure 19.
Pulverized coal is pneumatically conveyed from four pulverizers to
16 forced-draft burners, eight in each half of the divided furnace.
Combustion gases leaving either side of the boiler pass vertically
through a Ljungstrom-type air preheater and then turn abruptly into
the mechanical fly ash collectors.
55
-------
Table 15. WIDOWS CREEK UNIT 5 DESIGN DATA
Power capability; 125 MW
Boiler type; Single pass, divided furnace, water wall, dry bottom,
radiant, reheat boiler
Boiler efficiency; 88.45% at 112.5 MW
Firing method; Horizontal firing of pulverized coal
Fly ash collector; Two four-bank multiclone units
Bottom ash system; V-type furnace bottom hoppers with water sluice
56
-------
f"1 SECONDARY
1-, j SUPERHEATER
PULVERIZERS 6AS RECIRCULATINO
FAN •-
Figure 18. Widows Creek Unit 5.
57
-------
WIDOWS CREEK
STEAM PLANT
UNIT 5
DUST COLLECTOR
COAL
SCALES
PRIMARY
UPERHEATER
REHEATER
SECONDARY
SUPERHEATER
INDUCED
DRAFT FAN
SAMPLING LOCATIONS
1 Inlet Air
2 Pulverized Cool
3 Bottom Ash
4 Superheater Hopper Ash
5 Flue Gas - Collector Inlet
6 Collector Hopper Ash
7 Stack Gas
PULVERIZERS
Figure 19. Simplified diagram of test facility.
58
-------
Flue gases from each collector are drawn down to the induced
draft fan and then are discharged through breeching to the 150-
ft stack and finally into the atmosphere. Figure 19 indicates the
locations for sampling inlet air, coal, flue gas and ash.
This unit presented a number of sampling problems which are
typical of utility boilers. These are listed below in order of
sampling location (Figure 19):
. Station 3; Bottom ash (in sluice water) could be sampled
from the ash hopper, but it was not possible to estimate
directly the total mass flow rate of bottom ash.
. Stations 4 and 6; Fly ash could be sampled from the super-
heater and mechanical collector hoppers, but it was not
possible to estimate directly the total mass flow rate of
these ash streams.
. Station 5; The number of equivalent duct diameters be-
tween the Ljungstrom air heater and the bend to the dust
collectors was insufficient to assure a sampling location
with rectilinear flow and without significant turbulent
fluctuations. Also, at this station, there was insufficient
space between the accessible sides of the ducting and the
dust collector hoppers (and bracing) to permit use of a
rigid probe.
. Station 7; The ducting between the dust collector outlets
and the inlet to the induced draft fan followed a tortuous
flow path, which was judged to be unacceptable for sam-
pling. This made it necessary to sample on the stack, which,
prior to this test program, was not equipped with sampling
ports or platforms.
The most serious of these sampling problems were those produced
by the flow obstructions upstream of Station 5. However, the flow
disturbances caused by the fins of the air heater and the transition
from half-circular to rectangular cross section at the air heater
outlet, were judged to be much less disruptive than those caused
by other types of duct configurations. The potential impact of flow
disturbances on representative sampling at Station 5 was minimized
by increasing the number of sampling points. In addition, to overcome
59
-------
external space limitations, a flexible, heated probe was designed
and fabricated for use at this station.
At the stack (Station 7) the number of sampling points was
also increased so that distance of the sampling cross section
above the diameter reduction (at the base of the stack) could be
held to a minimum (approximately two stack diameters). A painters
platform hung from the top of the stack was used to support men
and equipment at this station.
The sampling problems encountered at Widows Creek were judged
to be about average for a utility boiler. On the positive side,
Station 5 was conveniently located on the roof of the plant, and
there were easily accessible ports for coal and ash sampling.
SAMPLING PROGRAM
Table 16 gives the composition of the MRI field crew who
carried out the sampling program at Widows Creek. Because of the
split-duct design of the flue gas flow system, duplicate sampling
teams were required at the two collector inlet sampling locations
(Stations 5A and 5B).
The Widows Creek sampling program was conducted during the
period 15-24 August of 1974. Figures 20 through 22 summarize the
main sampling program (Runs 2 through 4). For each sampling sta-
tion, the times and duration of sampling and the number of sam-
ples are specified.
Flue Gas Sampling
Hazardous Pollutants - Flue gas sampling to determine mass flow
rate of hazardous pollutants (particulates and vapors) and particle
size distribution was performed at the inlet and outlet to the
dust collector (Stations 5 and 7, respectively). In addition, at
the outlet of the dust collector, flue gases were sampled for
sulfur dioxide, nitrogen oxides and bulk carrier gas composition.
Figure 23 shows the ductwork configuration and the distribution
of sampling points at either of the dust collector inlet sampling
stations. During a run, each of the 48 points was sampled for 4
min. Both inlet stations were sampled simultaneously with separate
sampling trains.
60
-------
Table 16. MRI FIELD SAMPLING CREW
Sampling Number of
Responsibility station personnel
Crew chief - 1
Flue gas sampling
Mass train console 5A 1
5B 1
7 1
Mass train probe 5A 1
5B 1
7 1
Size train 5A
5B
7 1
Gas trains 7 1-
Coal/ash samplinga/ 2, 4, 6 2
Inlet air sampling 1 1-
Process monitoring 1
Field laboratory
Mass train 2
Size train _1
Total 16
a_/ Bottom ash samples were collected by Widows Creek plant personnel.
61
-------
RUN 2
Sampling Time
ro
Sampling Location
1. F.D. Fan Intake
2. Pulverizer Lines (4)
3. Furnace Hopper (2)
4. Superheater Hopper (4)
5. Collector Inlet A
Collector Inlet B
6. Dust Collector
Hopper (8)
7. Collector Outlet
Sample Type
• Continuous Sample
X Grab Sample
O Grab Sample (Estimated Time)
1000 1100 1200 1300 1400
1500
Inlet Air
Pulverized Coal
Bottom Ash
Superheater Ash
Flue Gas
Hazardous
Pollutants
Particle Size
Flue Gas
Hazardous
Pollutants
Particle Size
Collector Ash
Stack Gas
Hazardous
Pollutants
Particle Size
S02
NOX
Carrier Gases
*
t
§•
M
•M
•1
X
X
X
• ^^m •
m
• ^^m •
-
X
••• i
(
i
X
X
X
X
• ^^H
• ^Ml
X
M^M
)
B
X
(
IM
•a
M
••
X
X
)
X
1 H^M •
1 ^HM •
X
^^^•M
•
X
m ^^
M mmm
B^^
X
315
-
-
192
7
192
7
-
192
8
60
-
-
Sampling
Duration No. of
(min.) Samples
1
12
6
10
1
1/2-
1
1/2
24
1
1
4
2
Figure 20. Sampling program - Run 2
-------
RUN 3
Sampling Time
a*
u>
Sampling Location
1. F.D. Fan Intake
2. Pulverizer Lines (4)
3. Furnace Hopper (2)
4. Superheater Hopper (4)
5. Collector Inlet A
Collector Inlet B
6. Dust Collector
Hopper(8)
7. Collector Outlet
Sample Type
1000 1100 1200 1300 1400 1500
Inlet Air
Pulverized Coal
Bottom Ash
Superheater Ash
Flue Gas
Hazardous
Pollutants
Particle Size
Flue Gas
Hazardous
Pollutants
Particle Size
Collector Ash
Stack Gas
Hazardous
Pollutants
Particle Size
S02
NOX
Carrier Gases
(
•
mf
•H
< X
>
X
._-
X
^•i ••
X
X
•—
X
«^—
•M
X X
mmmm mm
•
mfmm^m
(
^
X
X
X
• mmmm mmt
X
mmmmm
i
^_
X
X
1
•
X
345
-
-
192
7
192
7
-
192
8
60
-
-
mm Continuous Sample
O Grab Sample
X Grab Sample (Estimated Time)
Sampling
Duration No. of
(min.) Samples
1
12
4
9
1
1/2'
1
1/2.
24
1
1
4
2
Figure 21. Sampling program - Run 3,
-------
RUN 4
Sampling Time
^ Continuous Sample
X Grab Sample
O Grab Sample (Estimated Time)
Sampling
Sampling Location Sample Type
1. F.D. Fan Intake Inlet Air
2. Pulverizer Lines (4) Pulverized Coal
3. Furnace Hopper (2) Bottom Ash
4. Superheater Hopper (4) Superheater Ash
5. Collector Inlet A Flue Gas
Hazardous
Pollutants
Particle Size
Collector Inlet B Flue Gas
Hazardous
Pollutants
Particle Size
6. Dust Collector Collector Ash
Hopper (8)
7. Collector Outlet Stack Gas
Hazardous
Pollutants
Particle Size
SO2
NOX
Carrier Gases
0800 09
]
X
MM
MM
MM
•
X
00 1000 1100 1200 1300
E 3
X
MM MM 1
•
MB MMI
-
X
MB MM
<
• M
X
X
X
Ml M
Mi Ml
X
MM ••
>
• 1
X X
X
X
X
X
• MM Ml
• MM Ml
X
MM. M
Ml
MM
1 MM
MM*
•
uurarion mo. or
(min.) Sample:
240 1
£fW 1
12
4
9
192 1
7 1/2-
192 1
7 1/2-
24
192 1
8 1
60 1
4
2
Figure 22. Sampling program - Run 4.
-------
'
9'-
0"
'
1
~7'-
.
~201-6"ID
•8" ID
Roof
/
X
.
t~8"~
/^
I
76 875" •-- • J
4& 125" _.-u, .J
~~| 15. 375" -M)
•
'
'
!!!!!!!,' I
i ' I 1
—-»• — u u y u uu u ITTTT
15.375"— »j U- == = = = =
^^ ' i^s co n K o> c*
•o CN o -o >o r"
>/ Right Duct *>' oo n g
Cyclone
•*• Air Flow Cyclone
r t
1 Sampling ~ 8'-0"
Port 1 Roof
i
~12'-0"
i
Not to Scale
Air Heater
Figure 23. Collector inlet duct configuration.
-------
Figure 24 shows the ductwork configuration and the sampling
traverses for the collector outlet sampling location, i.e., the
stack. Twenty-four sampling points were distributed along each
traverse with spacing as specified by EPA Method 1. Each of the
48 points was sampled for 4 rain.
During each run, the inlet and outlet locations were sampled
nearly simultaneously (Figures 20 through 22). The duration of
sampling and the dry gas volume sampled at each station exceeded
the minimums (i.e., 2 hr and 60 cu ft, respectively) specified
for the performance testing of utility boilers.,28/
Particle Sizing - During each run, two particle size samples were
obtained, one each at the collector inlet and outlet sampling sta-
tions. Each of the two inlet ducts was sampled (with the same
train) for a period of 7 rain to obtain a single composite sample;
samples were drawn from one point in each duct, about 24 in. from
the interior wall* The stack was sampled for 8 min at one point
about 24 in. from the interior wall. Prior to each sizing test,
the Brink unit was heated in the stack for a period of 30 min to
1 hr to establish thermal equilibrium.
SC>2, NOX, and Carrier Gases - During each run, gas sampling was
distributed over the time required for particulate sampling.
A 1-hr SO2 sample was obtained -by sampling for three separate
periods of at least 15 min. Also, four grab samples of NOX and
two grab samples of dry carrier gases were obtained. Samples were
drawn from points in the flow about 24 in. from the interior wall
of the stack.
Coal/Ash Sampling
Based on statistical analysis of a boiler stream flow model
(Section IV), it was determined that at least three representative
samples of pulverized coal and each ash stream should be taken dur-
ing each run in order to achieve acceptable measurement accuracy
(i.e., tolerance of 25% at the 957. confidence level). A represen-
tative grab sample of coal or ash consisted of portions removed
from each segment of a divided flow stream, i.e., from the four
pulverized coal lines, two bottom ash hoppers, two superheater
hoppers, and eight dust collector hoppers.
66
-------
-0" Dia.
Not to Scale
Inlet Breeching
Figure 24. Stack configuration.
67
-------
Inlet Air Sampling
During each run, inlet air was sampled near the intake to
the forced draft fan.
68
-------
SECTION VII
ANALYTICAL RESULTS AND QUALITY ASSURANCE
This section presents the results of the quantitative analysis
of samples obtained in the Widows Creek sampling program. Reporting
of results, i.e., mass or concentration of constituents found, is
done without direct reference to sampling parameters. However, in-
ferences are made about the collection efficiency of the sampling
train. Problems in carrying out the analytical techniques are also
discussed along with corrective measures that were taken. Finally,
the results of experiments to determine the quality (precision and
accuracy) of the analytical results are presented.
ANALYTICAL RESULTS
The analytical results obtained for the various samples col-
lected at Widows Creek are discussed in the following order: (a)
coal and ash samples; and (b) flue gas samples.
Coal and Ash Samples
Table 17 lists the results obtained for the hazardous pol-
lutant analyses of coal, bottom ash, superheater ash, inlet fly
ash, dust collector ash, and outlet fly ash. The inlet and outlet
fly ash results are discussed in more detail in the next section.
"Less than (<)" numbers indicated in this table are based on the
concentrations required to produce a signal that is twice the
noise level.
The values reported for each of the hazardous pollutants ap-
pear to have internal consistency. The relative magnitude of the
value for coal tends to be reflected in the corresponding ash
values. Also, values for a given pollutant are consistent over the
three runs.
69
-------
Table 17. POLLUTANT CONCENTRATION (ppm)£/ IN COAL, ASH, AND FLUE GAS STREAMS
-o
o
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Run
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4.
Avg.
2
3
4
Avg.
Coal
< 1.3
0.7
< 1.2
< 1.1
13.0
13.5
16.3
14.3
< 167
< 173
< 165
< 168
1.6
1.5
1.4
1.5
1.25
0.31
1.40
0.99
24
24
23
24
Bottom
ash
1.3
1.4
1.3
1.3
5.6
7.6
4.2
5.8
905
844
444
731
8.0
7.4
6.5
7.3
0.50
2.01
0.74
1.08
125
132
116
124
Superheater
ash
0.31
1.3
1.1
0.90
12.1
3.2
5.7
7.0
1,119
592
715
809
6.4
5.6
6.2
6.1
1.46
1.35
1.98
1.60
109
105
130
115
Inlet
fly ash
.
0.55
< 1.0
< 1.3
< 0.95
8.2
8.1
8.8
8.4
1,054
604
986
881
8.2
7.7
8.2
8.0
4.42
4.18
10.73
6.44
296
168
153
206
Dust
collector ash
y
1.4
0.32
< 0.9
b/
W
b/
w
1,213
916
1,367
1,165
7.2
6.8
9.7
7.9
2.89
1.14
2.00
2.01
133
191
128
151
Outlet
fly ash
1.54
1.36
1.5
1.5
7.4
5.5
12.0
8.3
1,028
1,262
931
1,074
10.0
8.5
9.5
9.3
6.29
3.88
14.09
8.09
316
170
174
220
-------
Table 17. (Continued)
Pollutant
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Run
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
Coal
1.84
1.51
0.99
1.45
10
12
9
10
3.68
2.26
5.23
3.72
24
30
51
35
1.88
1.91
1.93
1.91
18
16
12
15
Bottom
ash
5.74
1.95
3.20
3.63
51
48
45
48
6.94
12.3
5.07
8.10
125
377
184
229
< 0.541
< 0.489
< 0.502
< 0.51
45
84
58
62
Superheater
ash
5.86
4.45
3.93
4.75
54
46
45
48
11.9
10.4
9.41
10.6
217
265
326
269
< 0.58
6.90
46.4
< 18.0
108
94
102
101
Inlet
Oust
fly ash ' collector ash
7.09
6.49
4.36
5.98
75
65
64
68
21.8
26.1
48.2
32.0
153
222
371
249
16.7
23.8
18.3
20.0
178
128
97
134
10.5
6.87
6.41
7.93
59
74
39
57 .
21.7
11.6
11.5
14.9
169
287
268
241
< 1.21
< 1.17
< 1.17
< 1.18
88
98
60
82
Outlet
fly ash
3.61
2.78
4.68
3.69
81
70
72
74
18.7
29.9
61.2
36.6
164
154
285
201
23.3
2.2
25.4
' 17.0
206
86
86
126
-------
Table 17. (Continued)
Pollutant
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Run
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
Coal
< 6.1
< 6.1
6.0
< 6.1
< 30
< 30
< 30
< 30
1.73
1.70
1.65
1.69
1,090
833
758
895
61
61
64
62
36
17
111
55
Bottom
ash
< 5.5
< 5.9
< 5.4
< 5.6
62
41
26
43
2.83
1.81
1.45
2.03
6,900
5,520
6,010
6,150
272
419
369
353
68
275
107
150
Superheater
ash
< 6.2
< 5.7
< 5.5
< 5.8
30
< 27
< 27
< 28
2.11
1.59
2.08
1.93
5,430
5,480
5,200
5,370
229
215
342
262
133
110
186
143
Inlet
Dust
fly ash collector ash
27.9
24.1
27.5
26.5
31
< 30
< 30
< 30
3.04
3.31
2.07
2.81
6,420
6,990
5,930
6,450
308
238
478
341
163
201
691
352
b/
b/
< 12.5
< 12.5
30
31
28
30
1.74
3.48
3.45
2.89
5,150
6,260
3,940
5,120
275
190
240
235
154
131
164
150
Outlet
fly ash
< 18.9
< 13.1
18.2
< 16.7
< 35
35
29
< 33
1.69
2.38
1.69
1.92
6,840
7,410
6,400
6,890
359
262
623
415
212
150
736
366
-------
Table 17. (Continued)
Pollutant
Minor elements (cations)
Calcium
Iron
Sulfur
Anions
Chloride
Fluoride
Nitrate
Run
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
Coal
11,200
11,300
13,500
12,000
19,500
23,600
26,600
23,200
26,000
38,500
39,500
34,700
396
43
18
152
135
124
104
121
Bottom
ash
9,500
49,300
45,500
34,800
120,000
290,400
288,400
212,900
900
2,200
1,700
1,600
92
91
76.5
87
9.0
10.6
12.3
10.6
15.5
17.7
14.7
16.0
Superheater
ash
24,100
35,000
49,000
36,000
190,500
258,700
313,900
254,400
4,000
4,200
4,800
4,330
27.5
37
88
51
43.0
42.8
40.7
42.2
34.0
28.6
18.9
27.2
Inlet
fly ash
9,200
18,700
40,900
22,900
95,500
156,600
172,600
142,600
3,400
808
2,260
2,160
796
564
512
624
178
307
57.6
181
Dust
collector ash
10,500
18,900
32,200
20,500
84,600
188,700
116,500
129,900
11,000
3,000
3,950
5,980
62
31
12
35
45.5
22.5
20.6
29.5
42.1
27.5
33.2
34.3
Outlet
fly ash
7,100
17,200
34,700
19,700
84,200
131,000
124,000
113,000
1,335
125
332
597
830
559
624
671
103
88.8
64.9
85.6
-------
Table 17. (Concluded)
Pollutant
Sulfate
Organic s—
POM (1)
POM (2)
POM (3)
PCB's (all)
a/ Parts per million
b/ No sample left.
c/ POM compounds :
Run
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
by weight.
Coal
2.5
9.6
2.1
4.7
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Bottom
ash
116
1,090
818
675
0.2
ND
ND
< 0.2
0.2
ND
ND
< 0.2
0.2
ND
ND
< 0.2
0.04
ND
0.02
0.02
Superheater
ash
7,130
6,580
7,430
7,050
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.08
0.06
0.12
0.09
Inlet Dust
fly ash collector ash
5,570 2,110
7,000 2,520
8,400 3,510
6,990 2,710
0.2
0.2
0.2
0.2
ND
ND
ND
ND
ND
ND
ND
ND
0.16
0.04
0.02
0.07
Outlet
fly ash
3,970
4,310
8,020
5,430
(1) 7,12-Dimethylbenz[a]anthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
Note: ND = None detected.
-------
Another check of the validity of the measured concentrations
can be made by comparing the values found in this study with those
found by Oak Ridge National Laboratory (ORNL) in a similar study.^2/
ORNL data are the most comparable that we could find based on the
source of coal used in both programs and metals analyzed. ORNL
performed all of their analyses by neutron activation, an analysis
procedure totally independent from the atomic absorption procedures
used in this study.
Table 18 gives a comparison of MRI average values with ORNL
values for coal, bottom ash, and inlet and outlet fly ash. Because
ORNL tested a cyclone-fed boiler with an electrostatic precipitator,
systematic differences may be expected when the ORNL values for
outlet fly ash and bottom ash are compared to those reported by
MRI. For example, because the control device at Widows Creek was
relatively inefficient, metal concentrations in the inlet and out-
let fly ash are similar; on the other hand, the more efficient
precipitator used in the ORNL study resulted in higher concentrations
in the outlet fly ash which was composed of finer particles that
were more enriched in trace metals.
The results of the proximate analyses of the coal and ash
samples, which were performed by the Industrial Testing Laboratory
of Kansas City, are reported in Table 19* One blind duplicate sample
of each sample type (i.e., Run 3 coal, Run 3 bottom ash, Run 2
superheater ash, and Run 4 dust collector ash) was submitted for
analysis as a precision check. The results obtained for the check
samples agreed closely with the duplicates except for the heat of
combustion of dust collector ash samples from Run 4. Bottom ash
analysis gave negative results for fixed carbon, which is deter-
mined by differences; the high iron content caused these samples
to gain weight during ignition due to the formation of iron oxides.
Flue Gas Samples
Mass-Rate Sampling Train - The quantities of total particulates
and hazardous pollutants which were found in the flue gas sampling
trains operated at the inlet and outlet of the dust collector are
given in Tables 20 through 26. Some of the results from hazardous
pollutants are suspect because of problems encountered in operating
the sampling trains as discussed below.
75
-------
Table 18. COMPARISON OF RESULTS WITH ORNL— DATA
Pollutant concentrations (ppm)
Trace metal
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
Coal
< 1.1
0.5
14.3
14
< 168
59
1.5
a/
0.99
0.46
24
20
1.45
a/
10
a/
3.7
a/
35
34
1.91
a/
15
a/
< 6.1
a/
Bottom
ash
1.3
0.04
5.8
10
731
440
7.3
£/
1.08
1.1
124
150
3.63
a/
48
a/
8.1
a/
229
300
< 0.51
a/
62
£/
< 5.6
a/
Inlet
fly ash
< 0.95
12
8.4
120
881
450
8.0
a/
6.44
8.0
206
310
5.98
a/
68
a/
32.0
£/
249
290
20
a/
134
a/
26.5
a/
Outlet
fly ash
1.5
440
8.9
440
1,074
750
9.3
a/
8.09
51
220
900
3.69
a/
74
a/
36.6
a/
201
430
17
a/
126
a/
< 16.7
£/
76
-------
Table 18. (Concluded)
Pollutant concentrations (ppm)
Trace Metal
Tellurium
Tin
Titanium
Vanadium
Zinc
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
MRI
ORNL
Coal
< 30
£/
1.69
«/
895
510
62
28
55
46
Bottom
ash
43
£/
2.03
a/
6,147
4,100
353
260
150
100
Inlet
fly ash
< 30
£/
2.81
•/
6,448
6,080
341
440
352
740
Outlet
fly ash
< 33
a/
1.92
£/
6,887
10,000
415
1,180
366
5,900
a/ Not determined.
77
-------
Table 19. COAL AND ASH PROPERTIES
-J
00
Percent by weight, as
Stream
Coal
Bottom ash
Superheater
Ash
Dust
Collector
Ash
Run
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
A.V&.
Moisture
1.25
(1.59
U.65
1.38
1.42
0.09
(0.01
10.13
0.08
0.08
(0.13
10.14
0.19
0.29
0.21
0.11
0.07
(0.10
10.03
0.08
Ash
14.95
(14.72
U4.71
17.87
15.84
a/
*/
93.79
-
(94.41
194.36
95.51
87.76
92.55
90.29
93.21
(89.05
(91.76
91.30
Volatile
matter
32.38
(36.12
\36.07
34.35
34.28
1.01
(1.37
12.28
1.52
1.45
(2.82
12.47
3.40
4.76
3.60
2.15
2.52
(3.08
13.31
2.62
received
Fixed
carbon
51.42
(47.57
147.57
46.40
48.46
a/
*/
4.61
-
(2.64
13.03
0.90
7.19
3.64
6.45
4.20
(7.77
15.90
5.83
Sulfur
2.60
(3.84
13.87
3.95
3.47
0.09
(0.23
10.21
0.17
0.16
(0.42
10.37
0.42
0.48
0.43
1.10
0.30
(0.43
10.36
0.60
Heat of
combustion
(Btu/lb)
12,508
(11,926
112,041
11,541
12,011
137
(134
1 130
172
147
(657
1614
484
660
593
659
852
(1,528
I 105
776
a/ Ash gained weight due to high iron content.
-------
Table 20. PARTICUIATE MASS (GRAMS) COLLECTED IN FLUE GAS SAMPLING TRAIN
vO
Net weight (g)
Probe
Heated
Water
Imping
Sample
quartz filter
impingers
er rinse
Run
Inlet
•^^^^••••^
18.7677
12.3005
0.1787
0.0018
0.0360
2
Outlet
5
8
0
0
0
.8747
.3259
.0816
.0003
.0314
Run
Inlet
^^taHB^^^KW
13.2549
12.0023
0.4112
0.0037
0.0297
3
Outlet
3.3076
9.0423
0.0624
0.0008
0.0175
Run
Inlet
19
13
0
0
0
.7028
.1129
.2279
.0022
.0595
4
Outlet
4.5901
11.7779
0.1011
0.0015
0.0296
Total
31.2847 14.3139 25.7018 12.4306 33.1053 16.5002
-------
Table 21. POLLUTANT MASS (M1CROGRAMS) COLLECTED IN FLUE GAS SAMPLING TRAIN
(Run 2, Dust Collector Inlet)
Pollutant
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
CO Tin
O Titanium
Vanadium
Zinc
Calcium
Iron
Anlons
Chloride
Fluoride
Nitrate
Sulfate
Organic s-^
POM (1)
POM (2)
POM (3)
PCB's (all)
Probe
and
cyclone
is)
7.13
128
19,820
143
73.6
7,390
133
1,480
541
3,230
178
4,620
SOS
582
35.1
118,000
5,780
3,700
in)
191,430
1,850,500
100.800
16,900
4,570
114,900
ND
ND
186
ND
Water First
Water Implnger Back-up Tenax acid
Filter Implngers rinse filter plus Implnper-'
9.84 0.78
128 2.4
12,920
112
63.8
1,810
861
138
1,510
341 2.18
910
363 < 0.50
381
59.5
81,400
3,780
1,353
93,480
1,115,700
4,690
7,850
947
58,020
20 ND ND ND ND
20 ND ND ND ND
39 ND ND ND ND
ND ND 6.2 ND 5,700
Second
acid H202
iaplngerfe/ rinse-'
0.13 0.44
2.0 47.4
1.22 0.96
0.93 1.1
Note: ND - None Detected
a/ POM compounds:
(1) 7,12-DlmethylbenzfaJanthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
b/ Vaporous elements
-------
Table 22. POLLUTANT MASS (MICROGRAMS) COLLECTED IN FLUE CAS SAMPLING TRAIN
(Run 2, Dust Collector Outlet)
00
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Antons
Chloride
Fluoride
Nitrate
Sulfate
Organic a—'
POM (1)
POM (2)
POM (3)
PCB's (all)
Probe
and
cyclone
5. 35
31.7
5,980
52.3
7.81
3,125
21.4
517
28.8
1,350
327
2.340
36.4
247
10.1
40,450
1,770
1,840
51,100
511,000
17,920
5.440
792
28,560
ND
ND
ND
ND
Filter
16.6
73.3
8,630
89.9
81.6
1,370
29.9
641
236
982
< 4.7
591
232
< 241
14.0
56,840
3.330
1,170
49,960
686,000
1,082
6,360
674
27,970
12
12
12
ND
Water First Second
Water impinger Back-up Tenax acid acid H202
Imp inner s rinse filter plug Impinaerk/ implnger^/ rinse-/
0.20 0.16 0.42
2.2 1.5 3.3
1.08 0.62 0.57
1.7 0.89 1.1
ND ND ND ND
ND ND ND ND
ND 1 ND ND
0.1 ND 0.2 13.2
Note: ND • None Detected
a/ POM compounds:
(1) 7,12-Dlmethylbenz[aJanthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
b/ Vaporous elements
-------
Table 23. POLLUTANT MASS (MICROGRAMS) COLLECTED IN FLUE GAS SAMPLING TRAIN
(Run 3, Dust Collector Inlet)
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
00 Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Anlons
Chloride
Fluoride
Nitrate
Sulfate
Organ ica^f
POM (1)
POM (2)
POM (3)
PCB's (all)
Probe
and
cyclone
< 13.3
96.8
7,300
92.8
54.3
2,480
101
862
350
3,300
297
2,150
425
411
35.4
92,400
3,300
2,640
267,700
2,244,000
20,100
10,700
6,170
124,700
ND
ND
105
ND
Water
Filter implngers
8.88
107
7,960
101
51.2
1,760
62.8
780
310
2,300
304
1,090
182
< 336
48.5
84,200
2,710
2,450
204,000
1,712,000
216
3,490
1,560
52,000
ND ND
ND ND
23 ND
ND ND
Water First Second
Imp Inge r Back-up Tenax acid acid H2Q2
rinse filter plug ImplngerV implnger^/ rinse^/
0.38 0.36 0.70
68.5 3.6 9.6
3.4 1.43 0.75
1.6 1.5 2.6
ND ND ND
ND ND 11.4
ND NO ND
8.5 0.4 12
Note: ND " None Detected
a/ POM compounds:
(1) 7,12-Dimethylbenz[a]anthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
b/ Vaporous elements
-------
Table 24. POLLUTANT MASS (MICROGRAMS) COLLECTED IN FLUE GAS SAMPLING TRAIN
(Run 3, Dust Collector Outlet)
00
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Antons
Chloride
Fluoride
Nitrate
Sulfate
Organic s—
POM (1)
POM (2)
POM (3)
PCS' 8 (all)
Probe
and
cyclone
4.63
23.8
3.840
27.1
18.0
774
9.20
265
123
691
21.5
409
< 20.8
162
5.82
24,390
814
853
54,600
447,000
456
2,600
400
21,940
ND
ND
ND
ND
Filter
12.1
44.3
11,750
77.8
29.9
1,320
606
246
1.211
< 5.3
660
141
271
2.360
67,180
2,430
994
158.240
1,167,000
1,085
4,300
697
31,240
15
30
45
ND
Water First Second
Water impinger Back-up Tenax acid acid ^QZ
impingers rinse filter plug impinger"^ impinger—' rinsek/
0.13 0.27 0.67
2.9 3.3 3.7
0.41 0.99 0.26
1.3 0.87 2.0
ND ND ND ND
ND ND ND ND
ND ND ND ND
0.1 7.0 0.2 0.5
Note: ND = None Detected
a/ POM compounds
(1) 7,12-Dimethylbenz[aJanthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
b/ Vaporous elements
-------
Table 25. POLLUTANT MASS (MICROGRAMS) COLLECTED IN FLUE CAS SAMPLING TRAIN
(Run 4, Dust Collector Inlet)
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Anions
Chloride
Fluoride
Nitrate
Sulfate
Organic s—
POM (1)
POM (2)
POM (3)
PCB's (all)
Probe
and
eye lone
< 25.6
174
19,200
146
142
3,170
96.7
1,140
786
7,090
593
1,950
439
611
34.9
114,800
8,310
11,000
782 .200
3,420,000
55,400
11,800
1.340
186,300
39
79
39
ND
Water First
Water Implnger Back-up Tenax acid
Filter implngers rinse filter plug implnger^/
15.7 0.09
138 <2.96
13,200
123
210
1,850
46.4
970
797
5,075
< 7.34 1.42
1,230
463 0.90
< 380
32.9
79,900
7,380
11.700
559,900
1,718,000
18,750
4.980
548
89,300
ND ND ND ND ND
ND ND ND ND ND
18 ND 16 ND ND
ND ND 7.9 0.3 ND
Second
acid HjOj
ImplngerS/ rinse-7
0.02 0.27
5.1 6.0
1.40 0.72
< 1.03 2.7
Note: ND ° None Detected
a/ POM compounds:
~ (1) 7,12-Dlmethylbenz[aJanthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
b/ Vaporous elements
-------
Table 26. POLLUTANT MASS (MICROGRAMS) COLLECTED IN FLUE GAS SAMPLING TRAIN
(Run 4, Dust Collector Outlet)
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
__ Selenium
oo
Ui Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Anions
Chloride
Fluoride
Nitrate
Sulfate
Organic sS/
POM (1)
POM (2)
POM (3)
PCB's (all)
Probe
and
cyclone
6.89
33.0
3,140
41.3
56.2
1,060
10.5
349
288
1,500
10.7
555
83.5
£/
~7.94
30,570
2,220
3,390
149,200
633,900
464
3,270
425
30,830
45
ND
28
ND
Filter
c/
~163
12,080
114
174
1.790
66.2
824
714
3,160
405
860
c/
~342
19.7
74,220
7,970
8,660
418,100
1,401,600
4,970
6,950
638
100,500
33
66
ND
ND
Water First Second
Water Implnger Back-up Tenax acid acid H202
implnfters rinse filter plug Impinge r^/ implneer^/ rlnset/
0.23 0.28 0.06
1.5 0.9 3.0
.
0.45 0.43 0.79
1.5 < 0.51 < 0.51
ND ND ND ND
ND ND ND ND
ND ND ND ND
0.1 ND 0.3 ND
Note: ND • None Detected
a/ POM compounds:
(1)7,12-Dimethylbenz[aJanthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
b/ Vaporous elements
cV No sample left.
-------
The water impingers, containing 100 ml of water, could not
be maintained at a temperature below 20°C. Recent studies by
MRI—' have shown that these conditions are not satisfactory for
efficient condensation of POM and PCB. However, it should be
noted from this work that the bulk of the POM is found on the ash
collected in the probe and heated filter. The association of POM
primarily with ash has been observed in subsequent source testing
as reported for Winnetka Organic Test 2.2I/
The results from additional source testing-Li/ using a Tenax-
GC® plug indicate that it is an efficient trap for PCB vapor and
the data from Widows Creek should be reliable even though most of
the PCB was not condensed in the water impingers.
The acid permanganate impingers did not hold oxidizing strength
through any of the three runs on the inlet and outlet, indicating
elemental vapor penetration through these impingers. Only four
metals (arsenic, antimony, selenium, and mercury) were analyzed in
the acid permanganate impingers. In subsequent source tests,.!!/ we
have found that acid permanganate will not trap mercury efficiently
at isokinetic sampling rates even if maximum oxidizing strength is
maintained. We have also found in tests following the Widows Creek
sampling that selenium and antimony are trapped in sodium carbonate
solution (not analyzed in this study).
Size Train - The Brink particle size train samples were analyzed
for metals from the composite inlet and the composite outlet sam-
ples. Because of the low quantities of particulate collected (Table
27) only 15 of the 21 metals were analyzed, i.e., only those ele-
ments that can be analyzed by the carbon rod technique. Elemental
concentrations on each collection stage are given in Tables 28 and
29. (Brink outlet Stage 2 was contaminated during preparation for
hydrofluoric acid digestion.)
The results for the filter are subject to large errors due to
high filter to ash weight ratios. Three combined filter-ash sam-
ples were digested together because of the difficulty in accurately
transferring submilligram quantities of ash from Tissuquartz filters.
Therefore, relatively high background values had to be subtracted
from elements detected.
86
-------
Table 27. PARTICULATE MASS (GRAMS) COLLECTED
IN THE PARTICLE SIZE TRAIN
Weight (%)
Inlet weights (g)
Stage
Cyclone
1
2
3
4
5
Back-up
filter
Total
Run 2
0.09010
0.01200
0.01187
0.00134
0.00023
0.00013
0.00150
Run 3
0.08628
0.01394
0.00561
0.00110
0.00034
0.00009
0.00100
Run 4
0.09790
0.01821
0.00766
0.00224
0.00059
0.00016
0.00140
Total
0.27428
0.04415
0.02514
0.00468
0.00116
0.00038
0.00390
0.35369
All
stages
77.55
12.48
7.11
1.32
0.33
0.11
1.10
Plates
only
58.47
33.29
6.20
1.54
0.50
Outlet weights (g)
Cyclone
1
2
3
4
5
Back-up
filter
Total
0.05620
0.00971
0.00253
0.00064
0.00052
0.00027
0.00120
0.05296
0.00748
0.00194
0.00044
0.00013
0.00012
0.00160
0.04302
0.00643
0.00401
0.00086
0.00052
0.00004
0.00040
0.15218
0.02362
0.00848
0.00194
0.00117
0.00043
0.00320
0.19102
79.67
12.37
4.44
1.02
0.61
0.22
1.67
66.26
23.79
5.44
3.29
1.21
87
-------
Table 28. POLLUTANT CONCENTRATION (ppm)-/ VERSUS PARTICLE SIZE
COMPOSITE OF DUST COLLECTOR INLET SAMPLES
oo
CO
Pollutant
Trace elements (cations)
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Nickel
Tellurium
Thallium
Tin
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Cyclone
6.6
58
381
29
75
13.4
178
450
3.5
1.2
4.0
422
426
23,000
134,000
Stage 1
(> 3.96
nm)
5.2
46
251
26
147
14.5
149
460
2.3
1.2
9.3
475
791
20,000
145,000
Stage 2
(2.35 to
3.96 um)
8.7
164
458
30
261
15.5
274
840
2.3
1.4
7.6
818
1,110
35,000
155,000
Stage 3
(1.61 to
2.35 um)
4.7
92
1,080
28
152
29.5
189
690
3.1
2.1
4.3
592
781
20,000
295,000
Stage 4
(0.87 to
1.61 um)
5.4
135
3,080
58
564
12.1
569
2,460
6.6
4.7
10.3
741
2,670
80,000
121,000
Stage 5
(0.56 to
0.87 um)
36.1
447
4,510
75
2,660
18.4
654
689
17.4
28
52
1,550
4,620
81,000
184,000
Back-up
filter
97
571
1,740
66
3,380
10.2
692
1,380
41
55
32.8
1,970
4,100
133,000
102,000
a/ Parts per million by weight.
-------
00
\D
Table 29. POLLUTANT CONCENTRATION (ppm)-/ VERSUS PARTICLE SIZE
COMPOSITE OF DUST COLLECTOR OUTLET SAMPLES
Pollutant
Trace elements (cations)
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Nickel
Tellurium
Thallium
Tin
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Cyclone
8.5
36
284
46
183
12.6
154
401
1.8
1.8
3.0
278
876
28,000
126,000
Stage 1 Stage 2^
(> 2.87 (1.69 to
urn) 2.87 urn)
12.6
80
347
55
177
10.5
199
414
2.5
1.3
5.1
630
1,780
35,000
105,000
Stage 3
(1.16 to
1.69 urn)
8.3
165
577
48
212
9.3
288
319
6.7
2.9
13.4
500
2,260
47,000
93,000
Stage 4
(0.61 to
1.16 pm)
3.5
435
709
67
264
6.8
376
478
11.2
4.5
14.5
680
1,880
100,000
68,000
Stage 5
(0.39 to
0.61 urn)
86
429
1,790
218
4,500
18.5
1,560
512
22.1
9.3
69.7
1,660
4,640
169,000
185,000
Back-up
filter
88
312
2,030
65
4,120
6.2
656
437
22.8
30.9
9.4
1,970
3,130
159,000
62,000
a/ Parts per million by weight.
b/ Sample contaminated during digestion.
-------
Samples from Brink Stages 4 and 5 were less than 2 mg after
composite samples were made. The samples had to be digested and
brought to known volume before analysis. The smallest volume was
1 ml which resulted in a large dilution factor for these samples.
Weighing error, in combination with the high dilution factor,
made accurate analysis of these samples impossible.
ANALYTICAL QUALITY ASSURANCE
Table 30 presents the results of the precision, recovery and
accuracy determinations made during the analysis of the Widows
Creek samples.
Precision
Duplicate analyses were performed on coal, bottom ash, super-
heater ash, and dust collector ash from each of the three runs.
These duplicate samples were taken through the entire digestion
and analysis procedures. Precision values are reported as a pooled
relative standard deviation (PRSD) for each analysis because of
the small number of analyses for any given sample and the rela-
tively large number of duplicate analyses performed for any given
element.
The standard deviation for the duplicate samples was cal-
culated by:
a.
0.889
The pooled relative standard deviation was then calculated
by the following equation:
N 2
PRSD = / £ RSDi
i=l
N
90
-------
Table 30. QUALITY ASSURANCE DATA
vO
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Tltantlum
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Anlons
Chloride
Fluoride
Nitrate
Sulfate
/
Ontanlcs£'
POM (1)
POM (2)
POM (3)
FCB's (all)
a_/ Spike too large
b_/ Spike too small.
c/ Not certified by the
d/ Nonflane AA method.
e_/ POM compounds:
Pooled relative standard
deviation of duplicate
analyses A)
12.5
12.2
11.9
2.0
8.6
7.8
9.8
S.8
11.8
2.8
2.9
5.6
15.7
12.8
11.2
5.2
3.3
8.9
4.6
6.4
6.3
7.1
National Bureau of Standards (NBS).
Average percent
recovery from
fortified samples
82
it
85
83
94
99
1'
88
«/
90
90
98
80
b/
k/
90
y
93
94
US
70
103
Percent
certified
Coal
£/
it
£/
153
£/
120
£/
33
166
85
i/
106
if
£/
100
£/
110
81
£/
c/
ll
c/
£/
c/
£/
sJ
c/
of NBS
values
Ply ash
<•/
it
£/
108
i>
90
£/
94
111
59
if
118
it
£/
£/
£/
94
81
£/
£/
£/
£/
£/
£/
£/
£/
£/
£/
(1) 7,L2-Dlmethylbenz[a3anthracene
(2) 3,4-Benzopyrene
(3) 3-Methylcholanthrene
-------
where N = 8 (two runs and four sample types, i.e., coal, bottom
ash, superheated ash, and dust collector ash), and
RSD = [ - ] 100
•©
The factor 0.889 is a statistically more valid number to use than
the usual factor n-1 when there are only two numbers used to cal-
culate a a .
At least two-thirds of the elements determined had PRSD less
than 10% (Table 30). The maximum PRSD was 15.77. (for selenium). In
terms of conventional relative standard deviations, all elemental
analyses were below 7%.
Recovery from Fortified Samples
One sample of each solid type was fortified with all 21 metals
of interest before the digestion. The fortification level for each
metal was estimated from previous reports for that type of sample.
The added metal was in the range of 50 to 100% of that expected in
the sample. Calculated values for percent recovery averaged for
all types of samples are given in Table 30. In five cases the spike
was either too large or small to determine recovery.
The samples that were extracted and analyzed for PCB and POM
were spiked with Aroclor-1260 and benzoQaJpyrene before extrac-
tion. The results for Aroclor-1260 recovery are quite good. The
results for benzoQaJpyrene are low (70%), which indicates a
problem with the extraction procedure. It has been found that
benzo[ajpyrene is unstable during the Soxhlet extraction proce-
dure. Decomposition during extraction would account for the low
recovery.
After completion of the analysis of the Widows Creek samples,
National Bureau of Standards (NBS) certified Coal No. 1632 and Fly
Ash No. 1633 became available (mid-January 1975). Duplicate sam-
ples of each material were digested in a manner described in Sec-
tion IV* These samples were analyzed by flame atomic absorption
for the elements certified by NBS. The analyzed percent of the
NBS values are also indicated in Table 30. The NBS certification
92
-------
requires a minimum of 250 mg/sample for the certified values to
be valid. We could digest only 100 mg of either of the two mate-
rials, thus introducing some error in the comparison of our analyses
with the certified values.
It was not possible to check certified values for those ele-
ments run by carbon rod atomization technique because the instrument
used for Widows Creek samples (Varian AA-5) was modified after
completion of the project and before receiving the NBS certified
samples. The modification included conversion to a Model AA-6 with
different photomultiplier, amplifier, and a background corrector.
Any analysis of sample by carbon rod with this instrument con-
figuration would not yield valid numbers when compared to those
results obtained with the actual Widows Creek samples. The same
problem exists with mercury cold vapor and arsenic, selenium, and
antimony hydrides, since the Widows Creek samples were analyzed on
the instrument prior to modification.
93
-------
SECTION VIII
CALCULATED TEST RESULTS
This section of the report presents the results calculated
from field sampling data and process monitoring data obtained dur-
ing the Widows Creek test program and from the analytical results
presented in the previous section.
BOILER PERFORMANCE
Table 31 summarizes boiler conditions during the Widows Creek
sampling program. The boiler was operating near capacity for each
of the three runs. Generator load was recorded periodically from
the appropriate meter in the boiler control room. The values for
the other quantities shown all fall within the expected ranges.
Table 32 gives the coal feed rate and the mass flow rates of
the various ash streams for each of the runs. As indicated, ap-
proximately half of the ash produced by coal combustion, i.e.,
input ash, settled either in the furnace bottom hopper or the
superheater ash hopper or was accumulated on boiler tubes the
remaining half was mobilized to the flue gas stream. For purpose
of this study it was estimated that 607. of the settled ash was
bottom ash; this estimate was based on (a) reported fly ash emis-
sion facors for utility boilers,^!/ (b) reported data on rates of
soot buildup.10/ and (c) the analytical results presented below.
As indicated in Table 32, the collection efficiency of the
mechanical fly ash collector was approximately 45%.
Figures 25, 26, and 27 show the variations in the coal feed
rate for each of the three runs. In each case the coal feed rate
was constant to within plus or minus 4%. Coal feed rate was deter-
mined from periodic scale dump counts.
94
-------
Table 31. BOILER CONDITIONS
Run
2
3
4
Avg.
Capacity
factor
(W
96
99
98
98
MW
Load
120
124
123
122
Heat
input
(106 Btu/hr)
1,330
1,290
1,330
1,320
Heat
rate
(Btu/kw-hr)
1.11 x 104
1.04 x 104
1.09 x 104
1.08 x 104
Flue gas
generation
(dscfm/Mw)
2,070
1,920
2,100
2,030
Excess
air
(%)
«s/
32^
36
37
a/ Approximate value.
95
-------
Table 32. COAL/ASH MASS FLOW RATES--ALL STREAMS
Coal feed rate
Run
2
3
4
Avg.
(tons/hr)
53.1
53.8
57.6
54.8
(kg/min)
803
813
871
829
Input
ash
(kg/min)
120
120
156
132
Settled
ash^
(ke/min)
56.3 (47)£/
64.2 (53)
83.9 (54)
68.1 (52)
Inlet
fly ash
(kg/min)
63.7 (53)
55.8 (47)
72.1 (46)
63.9 (48)
Collected
fly ash^
(kg/min)
28.8 (24)
26.5 (22)
31.9 (20)
29.1 (22)
Outlet Collection
fly ash efficiency
(kg/min) (%)
34.9 (29)
29.3 (25)
40.2 (26)
34.8 (26)
45
47
44
45
a_/ Determined by difference: settled ash = input ash - inlet fly ash.
b/ Determined by difference: collected fly ash = inlet fly ash - outlet fly ash.
c/ Values in parenthesis represent percentage of the coal feed rate.
-------
RUN: 2
DATE: August 21, 1974
1600
Figure 25. Variations in coal feed rate - Run 2.
-------
vO
oo
60
59
? 58
»
O 5*
Wl
53
52|-
51
50L-
0800
RUN: 3
DATE: August 22, 1974
•Test Period
0900
1000
1100
1200
TIME
1300
1400
1500
1600
Figure 26. Variations in coal feed rate - Run 3.
-------
vO
vO
60
59
T 58
56
I 55
IS)
u
O
53
52
51
50
0800
RUN: 4
DATE: August 23, 1974
Average
-Test Period
I
0900
1000
1100
1200
TIME
1300
1400
1500
1600
Figure 27. Variations in coal feed rate - Run 4.
-------
Table 33 summarizes the boiler flue gas conditions. As indi-
cated, there was a measured air infiltration (leakage) rate of
about 207. between the dust collector inlet sampling locations and
the stack. The measurement of excess air obtained upstream of the
air heater during Run 4 indicated an additional leakage of about
7% in the air heater.
Boiler steam efficiency was determined for Run 4 using the
abbreviated heat loss method specified in ASME Performance Test
Code 4.1 (1964). The calculated value of 88.1% is close to the
design value for Unit 5. The calculation scheme is presented in
detail in Appendix C.
STACK GASES AND INLET AIR
Table 34 gives the breakdown of major components in the stack
gases as determined for each run. Also given are the concentrations
of sulfur dioxide, nitrogen oxides, and total particulate matter
in the stack gases.
Table 35 gives the concentration of total suspended parti-
culates in the air which was drawn into the combustion system. The
grain loading in the inlet air was about 0.1% of the grain load-
ing in the stack gases.
PARTICLE SIZE DISTRIBUTION
Table 36 lists the data used for the calculation of particle
size distributions. Included are effective cutoff diameters for
each stage of the Brink impactor. These values were calculated from
particle size sampling parameters and impactor design data for each
stage.
The particle size distributions are plotted in Figures 28 and
29.
HAZARDOUS POLLUTANTS
Table 37 gives the calculated flow rates of hazardous pol-
lutants in each of the process flow streams. Also shown is the
efficiency of removal of each pollutant from the flue gases by
the mechanical dust collector.
100
-------
Table 33. FLUE GAS CONDITIONS
Run
2
3
4
Location
Inlet A
Inlet B
Outlet
Inlet A
Inlet B
Outlet
Inlet A
Inlet B
Outlet
Flue gas
temperature
(°F)
376
387
367
364
362
339
367
355
353
Static
pressure
(in. He)
-0.55
-0.55
-0.05
-0.55
-0.55
-0.05
-0.55
-0.55
-0.05
Flow
rate
(dscfm)
126,000
141,000
331,000
114,000
141,000
306,000
130,000
147,000
313,000
Excess
Leakage air
(dscfm) (%)
} *
64,000 92
r
51,000 73
\ 52
J
36,000 74
Particulate
loading
(Er/acf)
) 2.05
0.984
^1.95
0.916
12.31
j
1.19
-------
Table 34. STACK GAS COMPOSITION
Units
Run
C02
02
CO
N2
Moisture
S0
as N02
Farticulate
loading
% by volume
% by volume
% by volume
% by volume
% by volume
ppm by volume
ppm by volume
gr/acf
8.5
9.8
0.2
76.3
5.2
1,370
303
9.2
8.5
0.2
75.6"
6.5
2,350
271
0.984
0.916
9.4
8.4
0.1
74.9
7.2
1,950
380^'
1.19
a/ Estimated value.
102
-------
Table 35. INLET AIR TSP CONCENTRATIONS
Total
Vol. of air participate TSP
sampled collected concentration
Run (m3) (ing) ytg/m3 gr/acf
2 322 719.7 2,235 9.77 x 10"4
3 337 617.8 1,831 8.00 x 10~4
4 245 498.5 2,032 8.88 x 10~4
Avg. 2,033 8.88 x 10"4
103
-------
Table 36. CALCULATED PARTICLE SIZE DISTRIBUTIONS
ECD
Inlet
Outlet
Cumulative weight %
Stage (um) All stages
Cyclone 7.5
1 3.96
2 2.35
3 1.61
4 0.87
5 0.56
77.55
90.03
97.14
98.46
98.79
98.90
ECD
Cumulative weight %
Plates only (um) All stages
58.47
91.76
97.96
99.50
100.00
7.5
2.87
1.69
1.16
0.61
0.39
79.67
92.04
96.48
97.50
98.11
98.33
Plates only
66.27
90.06
95.50
98.79
100.00
Filter
100.00
100.00
Note: ECD = Effective Cutoff Diameter, in microns.
104
-------
10.0
WEIGHT % GREATER THAN STATED SIZE
« « 90 10 7(1 tO SO in ^ ? ) 0 * 0 7 0 1 0 04 BOI,
P
u
^
5
UJ
_l
y
<
DL
WEIGHT % LESS THAN STATED SIZE
Figure 28. Particle size distribution - plates only.
105
-------
WEIGHT % GREATER THAN STATED SIZE
10.0
10 70 60 JO 40 30 10 10 5
1 \ O.S 02 0 I 0 W 0 01
c
2
u
Of.
L!_J
!—
LU
<
Q
ILj
dm O.OS 0.1 0? Ob 1 I 5 tO 20 30 40 SO 60 70
90 9S 98
WEIGHT % LESS THAN STATED SIZE
Figure 29. Particle size distribution - all stages,
106
-------
Taule 37.
POLLUTANT MASS FLOW RATE (gm/mln) IN COAL, ASH, AND
FLUE CAS STREAJS
Follucanc
Run
Coal
Pollutant
Bottom Superheater
aah ash
mass flow rate (Km/mln)
Inlet Inlet Duet
fly ash vapor collector ash8-^
Trace elenentfl (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
2
3
4
Avg.
2
3
4
Avg
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
< 1.04
0.569
< l.OS
< 0.89
10.4
11.0
14.2
11.9
< 134
< 141
< 144
< 140
1.28
1.22
1.22
1.24
1.00
0.2S2
1.22
0.824
19.3
19.5
20.0
19 6
1.48
1.23
0.862
1.19
0.044
0.054
0.065
0 054
0 189
0.293
0.211
0.231
30.6
32.5
22.4
28.5
0.270
0.285
0.327
0.294
0.017
0.077
0.037
0.044
4.22
5.08
5.84
5.05
0.194
0.075
0.161
0.143
0.007
0.033
0.037
0.026
0.272
0.082
0.191
0.182
25.2
15.2
24.0
21.5
0.144
0.144
0.208
0.165
0.033
0.035
0.066
0 045
2.45
3.03
4.36
3.28
0.132
0.144
0.132
0.126
0.0350 0.0027
< 0.0558 0.0031
< 0.0937 0.0008
< 0.062 0.0022
0.522 0.106
0.452 0.177
0.634 < 0.031
0.536 < 0.105
66
34
71
57
0.522
0.430
0.591
0.514
0.282
0.233
0.744
0.420
18.9
9.37
11.0
13.1
0.452
0.362
0.314
0.376
-
0 037
0.010
0.024
_
.
.
-
34.9
24.3
43.6
34 3
0.207
0.180
0.309
0.232
0.083
0.030
0.064
0.059
3.83
5.06
4.08
4.32
0.302
0.182
0.204
0 229
Oust
Overall
mass
Outlet Outlet collector bi Imbalance
fly aah vapor efficiency (X)~ (I)
0.0537 0.0019
0.0398 0.0025
0.0603 0.0014
0.0513 0.0019
0.258 0.0171
0.161 0.0236
0.482 0.0132
0.300 0.0180
36
37
37
37
0.349
0.249
0.382
0.327
0 219
0.144
0 566
0.300
11 0
4.98
6.99
7.66
0.126
0.081
0.188
0.132
-
~ 29
- 36
~ 17
51
64
24
44
45
0
48
35
33
42
35
36
22
51
27
29
42
47
36
42
72
78
40
65
-80
-71
-67
-72
-93
-95
-94
-94
89
55
76
73
-24
-30
0
-18
-65
2
-40
-46
11
-7
6
4
-49
-63
-21
-47
-------
Table 37. (Continued)
O
00
Pollutant mass flow rate (am/rain)
Bottom Superheater
Pollutant
Copper
lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Run
2
3
-4
Avg.
2
3
4
Avg.
2
3
4
AVg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
It
Avg.
Coal
8.03
9.76
7.84
8.S4
2.96
1.84
4.55
3.12
19.3
24.4
44.4
29.4
1.51
1.55
1.68
1.58
14.5
13.0
10.5
12.7
< 4.90
<4.90
5.23
< 5.01
< 24.1
< 24.4
< 26.1
< 24.9
ash
1.72
1.85
2.27
1.95
0.234
0.412
0.255
0.300
4.22
14.5
9.26
9.33
< 0.018
< 0.019
< 0.028
< 0.022
1.52
3.24
2.92
2.56
< 0.186
< 0.23
< 0.272
< 0.229
2.09
1.58
1.31
1.66
ash
1.22
1.18
1.51
1.30
0.268
0.270
0.316
0.285
4.89
6.81
10.9
7.53
< 0.013
0.177
1.56
< 0.58
2.43
2.41
3.42
2.75
< 0.140
< 0.146
< 0.185
< 0.157
0.676
< 0.69
< 0.906
< 0.757
Inlet Inlet
Dust
fly ash vapor collector ashS/
4.78
3.63
4.61
4.34
1.40
1.46
3.48
2.11
9.75
12.4
26.7
16.3
1.06 0.0089
1.33 0.0117
1.32 0.0077
1.24 0.0094
11.3
7.14
6.99
8.48
1.78 < 0.0052
1.34 0.0124
1.98 < 0.0101
1.70 < 0.0092
1.97
< 1.67
< 2.16
< 1.93
1.70
1.96
1.24
1.63
0.625
0.307
0.367
0.433
4.87
7.61
8.55
7.01
< 0.035
< 0.031
< 0.037
< 0.034
2.53
2.60
1.91
2.35
.
-
< 0.399
< 0.399
0.864
0.82
0.893
0.859
Outlet Outlet
Dust
collector
Overall
mass
. . Imbalance
fly ash vapor efficiency (X)2' (TL)
2.83
2.05
2.89
2.59
0.653
0.876
2.46
1.33
5.72
4.51
11.5
7.24
0.813 0.0055
0.064 0.0040
1.02 0.0041
0.63 0.0045
7.19
2.52
3.46
4.39
0.660 0.0090
< 0.384 0.0099
0.732 < 0.0062
< 0.592 < 0.0084
< 1.22
1.03
1.17
< 1.14
41
44
37
40
53
40
29
37
41
64
57
56
23
95
23
49
36
65
51
48
> 63
> 71
63
— 65
38
38
46
41
-7
-28
1
•12
-40
1
-25
-25
2
48
-9
6
-44
-83
56
-6J£/
-6
-17
12
-5
-66
-84
-78
-76
-65
-69
-71
-68
-------
Table 37. (Continued)
O
vO
Pollutant Run
Tin
Titanium
Vanadium
Zinc
Minor eleme
Calcium
Iron
Sulfur
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
nts (cations)
2
3
4
Avg.
2
3
4
Avg.
2
3
4
Avg.
Coal
1.39
1.38
1.44
1.40
878
677
660
738
49.0
49.6
55.7
51.4
28.9
13.8
96.7
46.5
8,990
9.190
11,800
9.990
15,700
19,200
23,200
19,400
20,900
31,300
34,400
28,900
Bottom
ash
0.096
0.070
0.073
0.080
233
213
303
250
9.19
16.1
18.6
14.6
2.29
10.6
5.39
6.09
321
1,900
2,290
1,500
4,050
11,200
11,500
8,920
50.7
159
143
118
Pol lutai
Superheater
ash
0.048
0.041
0.070
0.053
122
141
174
146
5.16
5.52
11.5
7.39
3.00
2.82
6.24
4.02
543
900
1,640
1,030
4,290
6,640
10,500
7,140
225
303
403
310
it maaa flow rate (
fDK/min)
Inlet Inlet Dust
Outlet Outlet
Dust
collector
Overall
mass
... imbalance
fly ash vapor collector ash"/ flJEJUh vapor efficiency nf (I)
0.194
0.185
0.149
0.176
409
390
428
409
19.6
13.3
34.5
22.5
10.4
11.2
49.8
23.8
586
1,040
2,280
1,300
6,080
8,740
12,400
9,070
.
-
-
-
0.050
0.092
0.110
0.084
148
166
126
147
7.92
5.04
7.66
6.87
4.44
3.47
5.23
4.38
302
501
1,030
611
2,440
5,000
3,720
3,720
317
79.5
128
175
0.059
0.070
0.068
0.066
239
217
257
238
12.5
7.68
25.0
15.1
7.40
4.39
29.6
13.8
247
504
1.400
717
2,940
3,840
4.990
3,920
18,000
29,100
24,900
24,000
70
62
54
62
42
44
40
42
36
42
28
33
29
61
70
42
58
52
39
45
52
56
60
57
m
_
-
-82
-80
-78
-80
-15
9
30
6
-29
-31
13
-14
-41
54
-52
-39
-84
-59
-46
-61
-12
39
33
22
-11
-5
-26
-15
-------
Table 37. (Concluded)
Pollutant Run
Aoiona
Chloride 2
3
4
Avg.
Fluoride 2
3
4
Coal
318
35.0
15.7
123
108
101
90.6
99.9
Pollutant
Bottom Superheater
ash ash
3.11
3.51
3.85
3.49
0.30
0.408
0.619
0.442
0.619
0.95
2.95
1.51
0.968
1.10
1.37
1.15
mass flow rate (em/mln)
Inlet Inlet Dust
fly ash vapor collector ash—
217 1.79
45.1
163
142
50.7
31.5
28.6
36.9
0.822
0.383
0.998
1.31
0.596
0.657
0.854
Dust
Overall
mass
Outlet Outlet collector b/ Imbalance
fly ash vapor efficiency (TL) ~ OL)
46.6 79 -84
3.66
13.3
21.2
29.0
16.4
25.1
23.5
92
92
85
43
48
12
36
-74
30
-78
-71
-82
-69
-74
»/ Superheater ash Includes soot build-up.
b/ Average collection efficiency based on average inlet and average outlet values.
£/ Based on two runs only.
-------
Table 38 lists the mass imbalances (averaged for the three
runs) for each of the pollutants. Mass imbalance is a measure of
the degree to which the output pollutant mass flow rates match
the corresponding input values. This may be mathematically ex-
pressed as follows:
Mass imbalance (%) = ~ inPUt x 100
input
The average mass imbalance for the three runs was calculated
based on average inlet and average outlet values; all "less than"
(<) values were considered as one-half the value.
As shown in Table 38, nearly all of the mass imbalances values
are negative, indicating that there was less mass flow measured in
the various output streams than in the input stream. This would be
expected, for example, for those trace elements which are highly
concentrated in the vaporous state, and inefficiently sampled in
the output streams. Values for mass imbalance around the dust col-
lector were consistently smaller in magnitude than corresponding
values around the boiler. The overall mass imbalances are heavily
weighted by the respective boiler imbalances.
Figures 30, 31, and 32 illustrate the range of mass imbalance
values obtained for the three runs. Again the ranges are smallest
for mass balance around the dust collector.
Table 39 gives the uncontrolled emissions factors for the ele-
mental pollutants in particulate form. These factors represent the
degree to which each element present in coal is converted to fly
ash in the flue gas stream. These emission factors do not account
for significant amounts of vapors. From another MRI study33/ ap-
proximately 90 to 100% of arsenic, mercury, and selenium should
exist in the vaporous state.
Table 40 gives the progressive enrichment ratios for each ash
stream scaled against the ratio of the hypothetical concentration
in coal ash divided by the concentration of the element in coal.
The concentration in coal ash is the value that would occur assum-
ing that all of the element were retained in the ash after combus-
tion.
Tables 28 and 29, presented earlier, exhibit a strong tendency
for trace metal enrichment with decreasing particle size.
Ill
-------
Table 38. AVERAGE MASS IMBALANCES—ALL RUNS
-
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Sulfur
Anions
Chloride
Fluoride
Boiler
-83
-90
53
-21
'38
9
-46
-11
-13
13
-15*'
9
-44
-72
-78
9
-13
-27
-62
29
-
20
-61
Mass imbalance (%)
Dust
collector
78
-50
24
9
-15
-8
-4
-3
-17
-12
-62l/
-21
-65
39
-15
-6
-3
-24
2
-16
-
-84
-34
Overall
-72
-94
73
-18
-46
4
-47
-12
-25
6
-63£/
- 5
-76
-68
-80
6
-14
-39
-61
22
-15
-78
-74
£/ Based on Runs 2 and 3 only.
112
-------
u>
Sb
As
Ba
Be
Cd
Cr
Co
Cu
Pb
Mn
£ "•
< Ni
3 Se
0 Te
a.
Sn
Ti
V
Zn
Ca
Fe
cr
F-
. *lSbo
- **
«. lfc o
11 • 1 °
1 o Cd
1 °
A JC° 0
Cu 1 .
| H.
* 1°
• 1 A
' \ *
0 1 A5® .
01 lTe .
- S^'
1 V
^ ^1
0 1
° 1
Ca i
. IF|- . "*
03— f— •
1 1 1 1 1 1 1 1 1
^l 0
a | o
_ |Cr lmhnlnnr«><:
0 | •
• Run 2
* Run 3
^— ^_J"1 H J
0 0 Run 4
0- |M" £L Probable nutlier
' H9
A . Ni 0
A1' 0
Zn
Fe 1 A o
_cr A
i i i i i i i i i
-100
-80
-60
-40
-20 0
IMBALANCE (%)
20
40
60
80
Figure 30. Boiler mass imbalances.
-------
Sb
Ac
Rn
Be
Cgt
Cr
Co
Cu
Pb
Mn
Z Ha
Z Hg
h- Mi
"
2
Sn
Ti
V
Zn
Co
To
re
G-
F-
A . 1* 0
a • | 0
A dCd
q
• lCr
1
A -i
* T
o •!
^Pb t '
0 1 Mn A
H9 1
A dN! .
A q .
1 Se o
1 °
. 1 A5"
H£.
H^
a 1 ?n.
0 1
^£2
o |Fe.
" 1
» |CI".
A . I Fl' 0
"• * 1 °
1 1 1 1 1 1 1 1 I
Sbl a
1 A
. 0 1 ta A
-I- n
1 °
)__A Dust Collector
V*n Mass Imbalances
Cu ' Run 2
^ A Run 3
0 Run 4
u | Average
,Te
1 * *
—»
1 0
i °
1 1 1 1 1 1 1 I
-100
-80
-60
-40
20
-20 0
IMBALANCE (%)
Figure 31. Dust collector mass imbalances.
40
60
80
-------
Sh
As
n*.
DO
Be
CA
Co
Cu
Pk
A/n
Ha
ng
t M;
Z INI
^ Se
d Ie
2 Sn
T;
1 1
7n
Co
a_
F —
- Sb|A 0
A . 1 Be C
0 •- | t
1 o Cd
• | 0
A .1 Co
' , o.| -
*. - A ^
• 9
• . Hg
A JNi
"
A ol Se- •
'.Te
• Sn n
M*>
A. 1 V
A* |
• C° 1 A 0
0 S| r A
0 |
. 1 A Cl~
A Fl"| .0
1 1 1 1 1 1 1 1 1
Ba I o
a jo •
Overall Mass
Imbalances
Cr • Run 2
A Run 3
O Run 4
O | Average
& Probable Outlier
• Mn fl
1
Ti 1 A o
Zn «
1=6 1 0 A
—I"!
1 1 1 1 1 1 1 1 1
-100
-80
-60
-40
-20 0
IMBALANCE (%)
20
40
60
80
Figure 32. Overall mass imbalances.
-------
Table 39. UNCONTROLLED PARTICIPATE EMISSION FACTORS
Particulate emission factor (%)
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Anions
Chloride
Fluoride
Run 2
> 3.37
5.02
> 49.3
40.8
28.2
97.9
30.5
59.5
47.3
50.5
70.2
77.9
> 36
> 8.2
14.0
46.6
40.0
36.0
6.52
38.7
68.2
46.9
Run 3
< 0.81
4.11
> 24.1
35.2
92.5
48.1
29.4
37.2
79.3
50.8
85.8
54.9
> 27
~ 6.8
13.4
57.6
26.8
81.1
11.3
45.5
129
31.2
Run 4
< 8.92
4.46
> 49.3
48.4
61.0
55.0
36.4
58.8
76.5
60.1
78.6
66.6
37.9
~ 8.3
10.3
64.8
61.9
51.5
19.3
53.4
1,040
31.6
Average^
~ 7.0
4.5
> 40.7
41.5
51.0
66.8
31.6
50.8
67.6
55.4
78.5
66.8
> 34
~ 7.8
12.6
55.4
43.8
51.2
13.0
46.8
115
36.9
a/ Based on average inlet fly ash
average
Note: Emission factor =
coal
inlet fly ash
v inn
coal
116
-------
Table 40. POLLUTANT ENRICHMENT RATIOS—AVERAGE, ALL RUNS
Pollutant
Trace elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements (cations)
Calcium
Iron
Sulfur
Anions
Chloride
Fluoride
Nitrate
Sulfate
a/ Coal ash = ppm
Coal ash-/
< 6.7
90
< 1,072
9.58
6.1
151
9.38
66.3
23.1
217
12.1
98.7
< 38.6
191
10.8
5,740
394
326
75,700
146,600
218,900
1,010
776
58.2
27,650
in coal
Bottom ash
coal ash
0.21
0.064
> 0.68
0.76
0.18
0.82
0.39
0.72
0.35
1.1
0.04
0.63
0.15
> 0.23
0.19
1.1
0.90
0.46
0.46
1.5
0.007
0.086
0.014
0.27
0.024
Superheater
ash
coal ash
0.14
0.078
> 0.75
0.64
0.26
0.76
0.51
0.72
0.46
1.2
1.5
1.0
0.15
0.15
0.18
0.93
0.66
0.44
0.48
1.7
0.020
0.050
0.054
0.47
0.25
Inlet fly
ash
coal ash
< 0.15
0.093
> 0.82
0.83
1.1
1.4
0.64
1.0
1.4
1.1
1.7
1.4
> 0.69
0.16
0.26
1.1
0.87
1.1
0.30
0.97
-
2.1
0.80
3.1
0.25
Dust collector
ash
coal ash
< 0.143
-
> 1.1
0.82
0.33
1.0
0.85
0.86
0.65
1.1
< 0.10
0.83
0.32
> 0.16
0.27
0.89
0.60
0.46
0.27
0.89
0.027
0.035
0.038
0.59
0.10
Outlet fly
ash
coal ash
0.24
0.092
> 1.0
0.97
1.3
1.5
0.39
1.1
1.6
0.93
1.4
1.3
0.43
0.17
0.18
1.2
1.1
1.1
0.26
0.77
-
0.59
0.86
1.5
0.20
fraction ash content in coal
-------
SECTION IX
DISCUSSION OF RESULTS
This section presents more extensive discussion of key re-
sults from the Widows Creek test program.
MASS BALANCE
The primary indicator of the reliability of test results ob-
tained in this study is overall mass balance. The average and the
range of mass imbalance for each pollutant are given in Table 41
(second column).
With the exception of cadmium and manganese, the precision of
measured mass balance is within the expected tolerance (Table 42)
based on the statistical model described earlier (Table 11). Addi-
tional statistical analysis (using techniques described in Ap-
pendix D) indicates that the primary sources of mass balance im-
precision are analytical imprecision and nonrepresentative sampl-
ing.
As indicated in the last two columns of Table 41, nonrepre-
sentative sampling accounts for most of the mass balance impreci-
sion observed. This is apparently due to the nonuniformities of
elemental concentrations in the bulk ash streams. Physical evi-
dence to support this is the wide range of color and texture ob-
served for the bottom ash and superheater ash samples collected
at different times.
Of greater importance is the average degree of mass imbalance
(i.e., inaccuracy in mass balance) for each pollutant. The two
major causes of statistically significant mass imbalances (exceed-
ing I + 25% I ) are inefficient collection of vaporous elements
(resulting in highly negative imbalance) and analytical error.
118
-------
Table 41. PRINCIPAL SOURCES OF INACCURACY AND IMPRECISION IN MASS BALANCE
Pollutant
Trace Elements (cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor Elements (cations)
Calcium
Iron
Sulfur
Anions
Chloride
Fluoride
Overall Mass
Imba lance
(%)
-72 + 6
-94 + 1
+ 73+17
-18+15
- 46 ± 34
4+9
- 47 + 21
- 12 ± 15
-25+21
6 ± 29
- 63 ± 20V
- 5 ± 16
-76 + 9
-68 + 3
-80+2
6+22
- 14+22
- 47 + &
-61+19
22 + 25
-15+10
- 79 ± 5^X
-74+6
Sources of
Inaccuracy^' Sources of
Inefficient (percentage
Pollutant Analytical Analytical
Collection Error Imprecision
X 3.0
X
X£/ 40.6
2.8
X 1.7
24.0
X 4.1
15.0
4.4
3.3
X 0.3
24.3
X 1.5
X£/ 12.8
VC/ Q Q
A^- O*O
7.1
1.9
X 0.3
X 1.4
6.7
X
X
Imprecision
contribution)
Nonrepresentative
Sampling
97.0
59.4
97.2
98.3
76.0
95.9
85.0
95.6
96.7
99.7
75.7
98.5
87.2
91.2
92.9
98.1
99.7
98.6
93.3
£/ Average overall mass imbalance exceeding ± 307.,.
b_/ Based on two runs only.
c/ Near the limit of detection of the analytical method.
-------
Table 42. MASS IMBALANCE VERSUS SAMPLING FREQUENCY
Tolerance
d(7.)-'
5
10
15
20
25
30
40
50
907. Confidence
Level
59
15
7
4
3
2
1
1
Number of samples
per run (all streams)
957. Confidence
Level
83
21
10
6
4
3
2
1
997. Confidence
Level
143
36
16
9
6
4
3
2
a/ Expected range of mass imbalance for nonvolatile elements.
120
-------
Inefficient pollutant collection accounts for the highly negative
imbalance observed for antimony, arsenic, mercury, selenium,
and fluorine. For the other six elements for which mass imbalance
was significant, analytical errors are thought to be important,
although the data on analytical accuracy (Table 30) are not ex-
tensive enough to quantitate this dependence. For cobalt, tin, zinc,
and chlorine, inconsistencies in stream enrichment (Table 40) sup-
port the probability of systematic analytical errors. Measured
levels of tellurium and tin were near the detection limits.
MODIFICATIONS TO SAMPLING TRAIN
To improve the efficiency of collection of vaporous elements
from the flue gases, modifications to the sampling train are recom-
mended. The modified sampling train (Figure 33) has the following
basic features:
1. Pyrex-lined probe (or stainless if longer than 8 ft)
heated to stack temperature.
2. Quartz fiber filter, heated to stack temperature, for
collection of particulates; filter preceded by stainless cyclone
if high grain loading.
3. Eight impingers in an ice-water bath for collection of
condensibles and inorganic vaporous species: first two impingers,
saturated sodium carbonate (for removal of condensed moisture,
sulfur dioxide, condensed organics, selenium and antimony); third
impinger, dry (for removal of carry-over from second impinger);
fourth and fifth impingers, acid dichromate (for removal of in-
organic vapors); sixth impinger, acid permanganate (for removal
of inorganic vapors); seventh impinger, dry (for removal of carry-
over from sixth impinger); and eighth impinger, silica gel (for
removal of residual water vapoor).
4. Quartz filter (optional) at 50 to 60°F between third and fourth
impingers for collection of condensed particulate.
5. Tenax-GC®plug for collection of organic vapors.
6. Gas meters and pump as prescribed by EPA Method 5.
121
-------
Stack Temperature
Stack
Sample Box
Filter
Sample
Intake
Ice Bath
L.
i
J L. Jl
fO
To Meters
and Pump
Figure 33. Modified flue gas sampling train.
-------
The analyses to be performed on each component of the modified
sampling train are indicated in Figure 34. Procedures to be used
for removing collected samples from the train are identical to those
described in Section V, with the exception that nitric acid-potas-
sium dichromate is suggested for rinsing the acid-solution im-
pingers. A schematic of sample treatment and analysis steps is shown
in Figure 35. The recommended analytical techniques are those used
in the Widows Creek Study.
HEALTH HAZARD EVALUATION
Table 43 presents an approximate evaluation of the potential
health hazard associated with the ground-level concentrations of
pollutants resulting from the emissions of toxic elements in boiler
flue gases. The "worst case" concentrations are based on the as-
sumption that all of the element in coal is emitted with the stack
gases. These concentrations are more realistic than the measured
concentrations for the vaporous elements.
The health hazard assessment is made by comparing the maximum
ground-level concentration with the corresponding threshold limit
value (TLV)..33_/ A factor of 10^ is used to account for stack gas
dilution between the point of emission and the maximum ground-level
receptor point. Because the threshold limit value specifies the con-
centration to which a person may be safely exposed during a normal
40-hr work week, this value should be reduced by at least a fac-
tor of 10 if it is to be applied to continuous exposure situations.
With this taken into account, only beryllium concentration is near
the level of potential concern.
123
-------
Mass Rate Train:
Probe &
Cyclone
Heated
Quartz
Filter
Back-up Tenax-GC
Quartz Plug
Filter
Acid
Acid
Acid
KMnO4
N>
Sample
Analysis
Elemental
POM, PCB
Sulfates, Nitrates
Se, As, Sb, Hg
Particulates
678
Figure 34. Required analyses for mass-rate train.
-------
N3
Acid Dichromate
Back-up Filter (Optional)
Tenax Plug
HF
Digestion
NaOH
Digestion
H2O
Extraction
Benzene
Extraction
& Column
Chromatography
Reduce with
SnCI2
Reduce with
1
Elemental Analysis:
AAS
Hg Analysis:
Cold Vapor AAS
As, Sb, Se Analysis:
AAS of Hydrides
Cl~ Analysis:
Chloride meter
F~ Analysis:
Specific Ion Electrode
SO42~ Analysis:
Barium-Thorin
Analysis:
Phenoldisulfonic Acid
PCB Analysis:
Electron Capture GC
POM Analysis:
Electron Capture GC
Figure 35. Sample preparation and analysis diagram.
-------
Table 43. POTENTIAL HEALTH HAZARD EVALUATION--AVERAGE ALL RUNS
Pollutant
Trace elements
(cations)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Tin
Titanium
Vanadium
Zinc
Minor elements
(cations)
Calcium
Iron
Anions
Chloride
Fluoride
Nitrate
Sulfate
Concentration
in coal
(ppm)
<1.1
14.3
< 168
1.5
0.99
24
1.45
10
3.72
35
1.91
15
< 6.1
< 30
1.69
895
62
55
12,000
23,200
152
121
9.1
-
Concentration in flue gases
(mg/m3)
Dust collector
inlet
< 0.00464
0.0404
4.26
0.0387
0.032
0 . 983
0.028
0.327
0.158
1.22
0.0938
0.640
0.128
< 0.146
0.013
30.9
1.68
1.77
114
686
10.5
2.99
0.837
33.9
Dust collector
outlet
0.00347
0.0203
2.50
0.0221
0.020
0.514
0.009
0.175
0.091
0.490
0.0421
0.293
< 0.0399
< 0.0772
0.0045
16.2
1.02
0.934
49
268
1.39
1.58
0.198
13.3
Worst case-'
0.060
0.806
9.49
0.084
0.0558
1.33
0.0806
0.579
0.211
1.99
0.107
0.861
0.339
1.69
0.0949
50.0
3.48
3.15
677
1,310
8.33
6.77
0.513
-
Worst case
ground level
concentration
(mg/m3)
0.000060
0.00081
0.0095
0.000084
0.000056
0.0013
0.000081
0.00058
0.00021
0.0020
0.00011
0.00086
0.00034
0.0017
0.000095
0.050
0.0035
0.0032
0.68
1.31
0.0083
0.0068
0.00051
-
Threshold
limit value
(mg/m3)
0.5
0.5
0.5
0.002
0.1-0.2
0.5-1.0
0.1
1
0.15-0.2
5-6
0.05
1
0.2
0.1
2
10-15
0.1-0.5
5
-
-
3
2.5
_
-
a/ Worst case
Pollutant mass flow rate in coal (gm/min)
Average dust collector outlet flow rate (acfm)
x 35.3
-------
APPENDIX A
PROCEDURES FOR HANDLING. PREPARATION AND ANALYSIS OF SAMPLES
127
-------
This appendix describes the procedures used to treat samples
collected in the Widows Creek test program. The procedures fall
into four categories:
1. Removal of samples from the collection apparatus.
2. Storage of samples.
3. Preparation of samples for chemical analysis.
4. Chemical analysis of samples for pollutant content.
COAL AND ASH SAMPLES
Composition Procedures
Samples of coal, superheater ash, and dust collector ash were
composited for each run. Each composite sample (~ 20 g) was pre-
pared by weighing equal quantities of the individual samples of
coal or ash from a given run. Composite samples were mixed thoroughly
prior to chemical analysis. If additional composite samples were re-
quired, analytical comparisons were made to ensure equivalence.
Because bottom ash samples consisted of a slurry of highly
nonuniform aggregate in sluice water, composite samples could not
be prepared by the technique described above. After the wet-bottom
ash samples from each run were combined, the sluice water was de-
canted off and the separated ash was air dried. Finally, the total
dried bottom ash from a given run was ground in a ceramic-ball mill
to less than 120 mesh.
Analysis for Trace Metals
Sample Preparation - Two samples of coal and each type of ash for
each run were obtained from representative portions of the composite
sample. At least five portions were taken from different parts of
the composite sample until the approximate desired sample weight
(~ 100 mg) is obtained.
Two portions of each type of solid sample from each run were
analyzed. The first portion of each sample type from one run was
analyzed directly; the second portion of each type from that run
was fortified with 1 ml of nitric acid solution containing approxi-
mately the same mass of each of the 20 metals as expected in the
128
-------
sample so that quantitative recoveries could checked. The sets of
duplicate samples of each type from the remaining runs were di-
gested and analyzed identically to check the precision of the
analytical methods.
The procedure for digestion of samples was as follows. After
each sample is accurately weighed into a Teflon cup, it was placed
in a Parr digestion bomb with 3 ml of 48% HF added. The bomb was
sealed and digested at 130°C for approximately 12 hr. After tine
was allowed for cooling to room temperature, 1.5 g of boric acid
was added and the solution was diluted to 25 ml. If the coal sam-
ples had not undergone total dissolution, they were centrifuged
and the solution decanted off for analysis. The undissolved mate-
rial remaining was again digested with HF. Analyses of these re-
digested samples indicated very little metal remaining after
the first digestion.
Sample Analysis - The digested material was analyzed by atomic ab-
sorption spectrophotometry. Four different methods of atomic ab-
sorption were used—conventional flame methods, carbon rod atomiza-
tion (or micro flameless methods), metal hydride generation with
nitrogen entrained-air hydrogen flame, and cold vapor. Table 13
indicated the particular AAS technique used to quantitatively
determine the concentration of each trace metal present in the sam-
ples. In each case, the selection of the atomic absorption tech-
nique was based on the expected concentration of the metal and the
sensitivities of the various AAS techniques.
Mercury was determined by cold vapor AAS. An aliquot of the
digested solution was placed in 20 ml of water and prereduced with
hydroxylamina hydrochloride. Stannous chloride was then added to
reduce Hg+^ to HgO and the Hg° was swept into a cold vapor atomic
absorption cell with nitrogen.
Arsenic, antimony, and selenium were converted to their re-
spective volatile hydrides prior to AAS analysis. This was done
by placing an aliquot into a 25% HC1 solution, sealing the solu-
tion in a screw-cap jar and rapidly adding 5 ml sodium borohydride.
The respective volatile hydrides that were formed were swept out
of the cell with nitrogen into a hydrogen flame. The concentration
was then determined by atomic absorption at the appropriate wave-
length for each element.
129
-------
In order to determine if the conventional flame atomic ab-
sorption methods were sufficiently sensitive for the remaining 16
metals, coal samples were analyzed using air acetylene or nitrous
oxide-acetylene flames. Coal samples were selected for this test
because the various metal concentrations are lower in coal than in
any other type of sample. The concentrations were determined for
each metal using composite standards containing all 20 metals of
interest and the same concentrations of hydrofluoric acid and boric
acid as the sample.
The trace metals that were below the detection limit for flame
atomic absorption were determined by carbon rod atomization. The
hydrofluoric acid used to digest the solid samples and the boric
acid added after digestion was a previously untested matrix for
the carbon rod technique. MRI has done extensive development work
to assure that this matrix is compatible for analysis by this tech-
nique. We have found that it is a workable matrix if care is taken
in the experimental procedure.
Each element was optimized for maximum sensitivity and minimum
background interference before the sample was analyzed. Background
and standards were checked frequently to assure minimum interfer-
ence. Several metals that were analyzed by flame methods were checked
by the carbon rod technique to determine the agreement between the
two techniques.
Analysis for Anions
Sample Preparation for Chloride and Fluoride - Blended duplicate
samples (1 g quantities) of each type were subjected to high-
pressure digestion. After addition of mineral oil to accurately
weighed samples, the samples were sealed in bombs, pressurized
with oxygen, and combusted. The chloride and fluoride were trapped
in 1 N sodium hydroxide, neutralized, brought to volume, and split
for chloride and fluoride analysis.
Sample Analysis for Chloride and Fluoride - Chloride analyses were
done with a chloridometer which coulometrically generates Ag+ ions
at a constant rate. The end point was detected amperometrically.
Concentrations were found from calibration of chloride concentration
versus time of Ag+ generation.
130
-------
Fluoride concentrations were determined potentiometrically
with a fluoride selective electrode. The neutralized samples were
mixed Itl with total ionic strength buffer solution. A linear
calibration of the log of fluoride concentration versus millivolt
response was obtained for standards from 0.05 to 100 ppm. Samples
above 100 ppm were diluted to bring them into the linear range.
Sample Preparation for Sulfate and Nitrate - The solid samples
were accurately weighed to 5 g and Soxhlet extracted for 24 hr
with water. It should be noted that with solid samples only water
soluble sulfate and nitrate were determined.
Sample Analysis for Sulfate and Nitrate - Sulfate concentration
was determined by barium titration to the thorin end point; and
nitrate determinations were made spectrophotometrically using the
phenoldisulfonic acid method. These two procedures were the same
as specified in the Federal Register for SO2 and NOX determina-
tions. 28/
Analysis of Organics
Two classes of organic compounds were of interest in this
study, polyeyelie organic materials (POM) and polychlorinated
biphenyls (FOB). Eight POM compounds with the highest carcinogenic-
ity, were screened for identification and quantitative analysis.
Sample Preparation for PCS and POM - Duplicate samples (5 g each)
of coal and each type of ash were Soxhlet extracted for 24 hr with
benzene. One of the two samples from one run was spiked with Aroclor-
1260 and benzoQaJpyrene before extraction to check recovery of the
sample during the preparation and analysis procedure. All benzene
extracts were evaporated to dryness and redissolved in a known
volume of benzene.
Sample Analysis for PCS and POM - PCB samples were analyzed on a
gas chromatograph equipped with a tritium electron-capture detector.
A 6-ft x 1/4-in. glass column packed with 1.5% OV-17 and 1.95% QF-1
on 100/120 mesh Gas Chrom Q was used. Aroclor-1260, tetrachlorobiphenyl
and hexachlorobiphenyl were used as standards. Seven of the major
131
-------
peaks from Arochlor-1260 were used for retention time matching
with the samples. The concentration of identified FOB compounds
were calculated on the basis of the Aroclor standards. Additional
identifications were made on a second column using OV-1, and GO/MS
verification of the higher concentration samples was attempted.
Analysis of POM compounds was made on a gas chromatograph
equipped with a flame ionization detector. A 6-ft x 1/4-in. glass
column packed with 3% Dexsil 300 on Supelcoport was used. Only
five of the eight POM compounds of interest could be obtained as
standards, at the time of this study, i.e., 7,12-dimethylbenzfalan-
thracene, benzo[V]pyrene, 3-methylchoLanthrene, dibenzQa,hJanthra-
cene and dibenzoQa,_i]pyrene. The possible presence in these sam-
ples of the other three POM of interest (benzo[c]phenanthrene,
dibenzoQa,ijpyrene, and dibenzo[C,g]carbazole) was checked by
relative retention time matching from other investigations.!^/
Proximate and Ultimate Analysis
Proximate and ultimate analyses were performed on the coal
samples and proximate analysis on the bottom ash, superheater
ash, and dust collector ash samples.
FLUE GAS SAMPLES
Hazardous Pollutants
A summary of the analyses that were performed on discrete
samples obtained from various components of the hazardous pol-
lutant sampling train was shown earlier (Figure 14). The overall
sample preparation and analysis system was diagrammed in Figure
A-l.
The following paragraphs discuss the sample handling and
preparation procedures that were applied to each component of
the sampling train. The analysis procedures are identical to
those described for coal and ash samples in the preceding sec-
tions.
Sample Handling
The procedures used for removing collected samples from the
sampling train were as follows:
132
-------
U)
OJ
Acid Permanganate
Back-up Filter
Tenax Plug
HF
Digestion
NaOH
Digestion
H20
Extraction
Benzene
Extraction
& Column
Chromatography
Reduce with
SnCI2
Reduce with
Elemental Analysis:
AAS
Hg Analysis:
Cold Vapor AAS
As, Sb, Se Analysis:
AAS of Hydrides
Cl~ Analysis:
Chloridometer
F~ Analysis:
Specific Ion Electrode
SO42" Analysis:
Barium -Thorin
NC>3" Analysis:
Phenoldisulfonic Acid
PCB Analysis:
Electron Capture GC
POM Analysis:
Electron Capture GC
Figure A-l. Sample preparation and analysis diagram.
-------
. The probe tip and probe were rinsed with minimum quantities
of acetone and chloroform and the rinses combined in a
screw-cap (Teflon) glass bottle.
. The heated quartz filter was removed from the holder and
placed in a wide-mouth screw-cap jar.
. The water impinger solutions were transferred to a screw-
cap bottle; the impingers were rinsed with acetone and the
rinses transferred to a separate screw-cap bottle.
. The backup filter and the Tenax-GC® were placed in
separate wide-mouth screw-cap jars.
. The sodium carbonate impinger solution was transferred to
a screw-cap (Teflon) glass bottle.
. The acid permanganate absorbing solutions were transferred
into separate bottles. Each impinger was rinsed with a
nitric acid-potassium dichromate solution and the rinses
combined with the respective absorbing solution.
. A minimum quantity of 37, hydrogen peroxide solution was
used to dissolve residues on the acid-permanganate impin-
gers. This rinse was transferred to a screw-cap glass
bottle.
. The final (dry) impinger was rinsed with 3% hydrogen per-
oxide solution to analyze for additional catch (if any).
This rinse was transferred to a separate screw-cap glass
bottle.
. The silica gel was transferred to a wide-mouth screw-cap
bottle.
All glassware used to store and/or transport components of the
sampling train was precleaned by washing with detergent, rinsing
with tap water, soaking in warm acid, rinsing with distilled water,
and rinsing with acetone.
134
-------
Sample Preparation
The procedures used to prepare for analysis the samples re-
moved from each component of the sampling train are described be
low.
Probe vj - The probe tip and probe rinse (acetone/chloroform) was
evaporated to dryness and the residue weighed for particulate load-
ing. The residue was split into five parts, which are treated as
follows: a 1-g portion was digested (as described in the coal and
ash section) for chloride and fluoride analysis; the second 1-g
portion was extracted with benzene for organic analysis; two por-
tions were hydrofluoric acid digested for metals analysis, and the
final portion was water extracted for sulfate and nitrate analysis.
First Filter \2s - The quartz filter and loose catch was desiccated
and weighed for particulate determination and then the loose catch
was divided into four portions that were treated as described im-
mediately above.
Water Impingers v3x and \A/ - The water impinger solutions were
extracted with benzene for organic anlaysis; the three portions of
benzene used in the extraction were combined, evaporated to dryness,
and the residue redissolved in a known volume of benzene. Each im-
pinger was treated as a separate sample to evaluate trapping ef-
ficiency. Aliquots were taken from the aqueous portion of the
first impinger solution for chloride, fluoride, sulfate and nitrate
analysis (after pH adjustment). The remaining aqueous solution was
then evaporated and the residue weight taken for the particulate
det e rminat ion .
Second Filter \Ji) - The second quartz filter was desiccated and
weighed for particulate loading determination. The filter was then
benzene extracted and analyzed for POM and PCB.
Tenax-GC® \&) - The Tenax-GC® was mixed with benzene, sonified,
and centrifuged. The benzene was decanted off and analyzed for POM
and FOB.
135
-------
N32C03 (j) and Acid Permanganate Impingers (o) and (9) - These
impinger solutions were digested with acid permanganate and each
brought to 100 ml volume. They were then analyzed for mercury,
antimony, selenium, and arsenic.
Particle Size Samples
Each portion of sample obtained in the Brink cascade impactor,
i.e., a cyclone catch, five impactor stages (each with a distinct
particle size cutoff diameter), and a tissue-quartz backup filter,
was analyzed for cationic elements. Due to insufficient sample
quantities anticipated, analyses for organics, sulfate, nitrate,
chlorides and fluorides were not made.
Sample Handling
After a sizing test, the aluminum foil liners were transferred
to snap-cap plastic vials.
Sample Preparation - The combined Brink samples were washed and
sonified with acetone into Teflon Parr bomb'digestion cups. The
Teflon cups were weighed before digestion and after the digested
sample has been removed* The backup filters were digested with the
filter catch to avoid removal of the catch from the quartz filter.
The digestion procedure were modified to account for the small ash
weights by using proportionally less hydrofluoric acid, boric acid,
and a smaller final sample volume.
Sample Analysis - Trace metals were analyzed by carbon rod atomiza-
tlon atomic absorption because of the small sample volumes (1 to
10 ml) and the desire to analyze for 21 metals (including thallium).
The reproducibility and background absorption was carefully checked
for each of the trace metals. The various carbon rod atomizer
operating parameters for each metal determined were optimized for
a maximum absorption signal and a minimum background interference
as checked by hydrogen lamp.
802» N0x» and Carrier Gases - EPA Methods 6, 7, and 3 were followed
in the analysis of collected flue gas samples for sulfur dioxide,
nitrogen oxides, and dry carrier gases, respectively. SO2 concen-
tration was determined by barium titration of the appropriate ab-
sorbing solution to the thorin end point; NOX concentration by
136
-------
spectrophotometric analysis of the appropriate absorbing solution
using the phenoldisulfonic acid method; and dry gas composition
by the Orsat technique.
Inlet Air
The Tissuquartz filters from the high-volume samplers were
conditioned and weighed to determine the particulate loading in
the sampled inlet air.
137
-------
APPENDIX B
REVIEW OF CHEMICAL ANALYSIS METHODOLOGY
138
-------
This appendix presents a review of chemical analysis methods which
are applicable to the determination of hazardous constituents in utility
boiler flue gases. Following a general review of candidate analysis
methods, details are presented on the methods of choice for each pollu-
tant category. This is followed by a brief description of alternate
methods of analysis.
CANDIDATE ANALYSIS METHODS
Generally, it is not feasible to select one analytical method that
is superior to all others for the analysis of a group of elements or com-
pounds. Each method has areas of strength and weakness when compared to
other methods. The selection of any method is based on many factors in-
cluding the following: accepted performance criteria, e.g., accuracy,
precision and detection limits; cost-time factors; and the degree of so-
phistication and reliability of the required instrumentation. In addi-
tion, the effect of the bulk matrix composition of "real world" samples
on each of the above criteria must be considered.
General Approach
The quantity of sample available at certain sampling stations
is limited. For example, the effluent gases sampled at the outlet
of electrostatic precipitators have relatively low concentrations
of particulates (~ 0.1 gr/scf). Increased quantities of particulates
can be collected by extending the sampling time, but sampling
times are also limited by practical considerations such as crew
exposure, cost and durability of equipment. Because of possible
sample size limitation, the strategy for selecting analytical
methods is based on choosing analysis procedures that are most com-
patible with a common sample preparation method which can be ap-
plied to a large group of pollutants.
The hazardous pollutants can be divided into two major groups:
organic compounds (polynuclear organic materials and polychlorinated
biphenyls) and elemental inorganic compounds. The latter group can be
subdivided into those elements which form anions (chlorides and fluo-
rides) and those which form cations upon sample dissolution. A general
procedure for dividing the sample into these three groups is depicted
in Figure B-l. The organic compounds are removed from the sample by
extraction with benzene. Compounds of special interest, e.g., the
highly carcinogenic benzo[a]pyrene and polychlorinated biphenyl are
highly soluble in benzene whereas elemental pollutants are insoluble
139
-------
Sample
Extract
with
Organic
Solvent
1
Cations
Inorganic
Materials
1
Organic
Materials
An ions
I
C
Analysis:
PCB, POM
(Analysis: ^ /Analysis: ,7\
As4** Cd++ Hg-"". etc^ ^Cl". F". NOg", SO4"2J
Figure B-l. General procedure for sample treatment and analysis.
140
-------
in benzene except when present as organometallic complexes, which have
limited solubility in benzene. The loss of inorganic pollutants, if
any, can be determined by elemental analysis of a number of benzene ex-
tracts; it is not expected to be significant. The benzene insoluble
material is divided and analyzed for cationic and anionic pollutants.
The analysis methods that can be applied to elemental and organic
pollutants are discussed in the following paragraphs.
Elemental Analysis
A number of methods have been applied to the analysis of cationic
elements3*7'13*18' 34/ including:
Atomic absorption (AA) and emission spectrometry;
Neutron activation analysis (NAA);
X-ray fluorescence spectrometry (XFS);
Spark source mass spectrometry (SSMS); and
Electrochemical methods (EC--voltametry and potentiometry).
The feasibility of analyzing each of the elemental pollutants of
interest by the methods mentioned in the preceding paragraph is indi-
cated in Table B-l. A plus indicates the analysis is feasible; a minus
indicates the analysis is not feasible; and a circled plus indicates
MRI has the capability of performing the analysis in-house. The cri-
teria used to determine the feasibility of each analysis were:
1. Detection limits; The detection limit of the method for coal
and fly ash matrices was considered versus the expected concentration
of the element. If the detection limit was above the expected concen-
tration, the method was indicated as not feasible.
2. Sample requirements; If a method required a prohibitive sample
size (> 1.0 g of particulate) or large dilution factors for multielement
analysis, it was indicated as not feasible.
3. Accuracy and precision; The accuracy and precision of the method
must be sufficient to obtain a meaningful material balance for the ele-
ment in question. A relative standard deviation of ± 20% was selected
arbitrarily. Each of the methods indicated as acceptable meet this re-
quirement in the most extreme situation (i.e., near a signal level of
141
-------
Table B-l. FEASIBILITY OF ANALYTICAL METHODS?/
Elemental
Pollutant
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Thallium
Tin
Titanium
Vanadium
Zinc
Chlorine
Fluorine
AA
9 si
©— '
©
©
-
©
©
©
©
©
©I/
©
©— '
©£/
©
-
©
-
©
-
-
AA (micro) NAA^/
e +
© +
+
©
© +
© +
© +
© +
©
© +
© +
© +
© +
© +
© +
© +
© +
© +
© +
+
+
XRF OES£/ EC*/
+ ©
+ ©
© + . -
- + -
©
- + ©
+ ©
- + ©
+ ©
© + ©
+
© + ©
_ _ _
+
+
- + -
© +
© +
- + ©
© - ©
©
a/ The analytical procedures are as follows: AA, atomic absorption spec-
trophotometry; AA (micro), AA with carbon rod or other flame less
atomization; NAA, neutron activation analysis; XRF, X-ray fluorescence
spectrometry; OES, optical emission spectrometry; EC, electrochemical
methods.
b/ NAA methods include chemical pretreatment.
£/ OES methods include chemical pretreatment, photometric detection, DC
arc (inert atmosphere) and argon plasma sources.
d/ EC methods are anodic stripping voltametry and potentiometry (specific
ion electrode for fluoride).
£/ Includes SbH2, AsH^, and SeH3 generation and ^-t^-air flame.
f_/ Cold vapor.
142
-------
twice the noise level). At normal working levels, the feasible methods
generally are within a relative standard deviation of 5%. Because of
these criteria, SSMS is not listed in Table B-l. The NAA and OES rat-
ings are for methods which include chemical pretreatment. The OES meth-
ods are DC Arc (controlled atmosphere) or argon induction coupled plasma
methods with photometric detection and not routine survey methods.
Cost considerations are not reflected in Table B-l. Because several
chemical treatment procedures are required for NAA and OES analysis of
20 elements of interest, these methods are more costly than AA analysis.
A realistic cost for AA analysis, including man-hours, chemicals and in-
strument operation and upkeep, is approximately $10/sample for sample
preparation plus $5 to $10/element quantified. To achieve comparable
sensitivity, accuracy and precision, the cost for sample preparation(s)
prior to NAA and OES analysis is on the order of $50/sample. The costs
for electrochemical analyses are similar to AA where satisfactory meth-
ods do exist.
Other anionic pollutants of interest, in addition to elemental
anionic pollutants, include sulfates and nitrates. A variety of methods
have been applied to the detection of both sulfates and nitrates* Spec-
trophotometric methods are most successful for the analysis of nitrate.
Particulate sulfate analysis methods are the subject of a recent review
by Forrest and Newman.^5.' Included in this review are turbid line trie,
photometric electrochemical, microtitration X-ray emission, and other
methods.
Analysis of Organic Materials
Gas chromatography, liquid chromatography, and UV-visible spec-
troscopy have been applied to the analysis of POM and PCS. ~^' Iden-
tification of both types of organic compounds should be verified by mass
spectrometry if sufficient concentrations are present. Verification is
accomplished most conveniently by a mass spectrometer interfaced directly
with a gas chromatograph.
ANALYSIS METHODS OF CHOICE
Table B-2 indicates the methods of chemical analysis that are rec-
ommended for each of the pollutants (for convenience Table 13 has been
reproduced as Table B-2).
In the following paragraphs, the details of the analytical methods
are presented separately for each class of pollutants, i.e., elemental,
organic (PCB, POM) and minor anionic (sulfate, nitrate).
143
-------
Table B-2. CHEMICAL ANALYSIS METHODS
Pollutant Methods of analysis1
Trace elements (cations)
Antimony
Arsenic
Barium 1
Beryllium 1
Cadmium 2
Chromium 1
Cobalt 1
Copper 1
Lead 2
Manganese 1
Mercury 2£/
Nickel 1
Selenium 1-/
Tellurium 1
Tin 2
Titanium 1
Vanadium 1
Zinc 1
Minor elements (cations)
Calcium 1
Iron 1
Anions
Chloride 3
Fluoride 4
Nitrate 7
Sulfate 8
Organics
POM 6
PCB 5
a/
a/ The methods of analysis are as follows:
(1) Atomic Absorption Spectrometry (AAS),
conventional flame methods;
(2) AAS, micro flameless methods;
(3) AgN03 titration, electrochemical (EC) detection;
(4) EC, fluorine selective electrode;
(5) Gas chromatography (GC), electron capture
detection;
(6) GC, flame ionization detection;
(7) Spectrophotometric, phenol disulfonic
acid complex; and
(8) Barium perchlorate titration.
b_/ AAS, hydride generation methods.
c/ AAS, cold vapor method.
144
-------
Elemental Pollutants (cations)
Atomic absorption spectrophotometry (AAS) is the method of choice
for the elements that commonly form cations. Some of the advantages of
AAS are: the low limits of detection for most of the trace elements of
interest; the large number of elements which can be analyzed; the low cost
of highly reliable instrumentation; and, the accuracy and precision which
can be obtained without highly trained technical personnel. Eleven of
the 19 elements can be determined at the 10-ppm level with less than
1 g of sample. Conventional cold vapor and arsine generation techniques
allow the determination of mercury and arsenic, respectively, at the 10-
ppm level with less than 1 g of sample.
The recent development of micro methods using the carbon rod and
tantalum strip atomizers enable analysis to be carried out on as little
as 5 ul of solution, hence reducing the total volume of solution required
to analyze 19 elements from £ 25 ml to 1 to 5 ml. In addition, the solu-
tion detection limits are in the nanograms per milliliter rather than
the micrograms per milliliter range of conventional flame methods. The
low detection limit and low dilution factor make this technique the method
of choice for the elemental pollutants that cannot be determined by con-
ventional flame AAS because of insufficient sample size or sensitivity.
The sensitivities of conventional and flameless methods are compared in
Table B-3.
The minimum sample quantity required to accurately measure each ele-
mental pollutant (relative standard deviation of 5 to 107.) by flameless
AAS is given in Table B-4. in addition to required quantities of fly
ash (for elements in the particulate state), the table also lists sample
times for collection of sufficient amounts of vaporous elements.
The required quantities of fly ash were calculated with the follow-
ing equation:
Sample Required (mg) = Sensitivity (nfi/ml) x 50 x 5 ml
Expected concentration (ug/g)
where 50 = sensitivity multiplier which represents sufficient signal-
to-noise ratio for a relative standard deviation of 5 to
107o
5 ml = volume of solution into which the sample is digested prior
to analysis
145
-------
Table B-3. ANALYSIS SENSITIVITY OF ATOMIC ABSORPTION SPECTROMETRY
Sensitivity*/
Conventional flame Flame less methods
Trace element (cation)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Se lenium
Tellurium
Thallium
Tin
Titanium
Vanadium
Zinc
methods (ug/ml)
0.6
1.3
0.3
0.02
0.02
0.09
0.09
0.04
0.16
0.04
2.0
0.07
0.6
0.3
0.26
2.0
2.2
1.2
0.012
(ng/ml)
6.0
0.1
0.18
0.02
1.0
1.2
1.4
1.0
0.1
20
2.0
20
12
20
0.02
a/ The concentration of a solution which produces a 17. absorption.
146
-------
Table B-4. REQUIRED SAMPLE QUANTITIES FOR ANALYSIS OF ELEMENTAL POLLUTANTS BY FLAMELESS AAS
Trace element
(cation)
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Tellurium
Thallium
Tin
Titanium
Vanadium
Zinc
Analysis
sensitivity^
(ng/ml)
6.0
0.1
300
0.18
0.02
1.0
1.2
1.4
1.0
0.1
20
2.0
20
20
20
12
2,200
20
0.02
Fly ash
Elemental
concentration
(ug/g)
50
320
500
24
0.3
150
48
130
95
500
1.5
150
22
10
3
9.0
3,800
260
120
Vapor-/
Required
sample
quantity
(nig)
30
0.078
150
1.9
17
1.7
6.2
2.7
2.6
0.050
3,300
3.3
230£/
500
1,700
330£/
140
19
0.042
Concentration
in flue gases
(mg/m3)
0.33
2.1
-
-
-
-
-
-
-
-
0.01
0.15
-
0.02
-
-
-
Required sample
time (min)
8.66
0.022
-
-
-
-
-
-
-
-
940
63
-
470
-
-
-
a/ Sensitivity is defined as the concentration required to produce a 1% absorption.
W Data are given only for elements likely to be in vaporous form..!!'
£/ Approximately 10 mg of sample would allow analysis of these elements at a signal-to-noise level
slightly greater than 2.
-------
The concentration of an element in fly ash is calculated by mul-
tiplying the average concentration in coal (Table B-4) by 10.
Although the calculated minimum quantity of sample shown in Table
B-4 is sometimes 5 mg or less, at least 10 mg and preferably 50 mg of
each sample is desirable to reduce problems involved in sample handling
and contamination.
The required time for collection of sufficient amounts of vapor-
ous elements was calculated with the following equation:
_. „ . . , . N Sensitivity (ng/ml) x 50 x 200 ml x K
Time Required (min) = ——— * N ° . / —r-*-
0.75 cfm x Expected concentration (mg/m3)
where 50 = sensitivity multiplier which represents sufficient
signal-to-noise ratio for a relative standard devia-
tion of 5 to 107.
200 ml = volume of impinger liquid which collected the sample
0.75 cfm = flue gas sampling rate
10"6 mg/ng
K
0.0283 m3/ft3
The concentration of vapor in the flue gases, assuming all the ele-
ment is present in the vapor state, is calculated by dividing the
average concentration in coal (Table B-4) by 15.*
Conventional flame AAS is used when sufficient sample quanti-
ties and detection limits can be obtained. Micro methods are more
subject to contamination and may not include a representative portion
of large (£ 1 g) samples.
Interferences do occur in both conventional and micro AAS methods.
To control chemical interferences, one must optimize instrumental param-
eters including selection of the proper flame mixture, temperature, and
region. Chemical interferences due to matrix effects can enhance or
depress absorption. These interferences can often be eliminated by add-
ing releasing or chelating agents.
* This factor is derived from the approximation that 200 ft-* of flue
gases at 300°F are produced by the combustion of 1 Ib of coal.
148
-------
Physical interferences, e.g., light scattering by particles in the
atomization process, can be a source of error, especially in carbon rod
atomization. Also, molecular absorption and emission can result in er-
roneous results if not corrected. Background correction techniques us-
ing nonabsorbing lines or hydrogen continuum lamps is applied for
trace element analysis, especially when using flame less methods.
The sample preparation procedures for AAS depend upon the sample
matrices. For this study, the types of matrices anticipated are whole
coal, coal ash, and acid-permanganate solution. A suitable preparation
for these matrices is the acid-pressure decomposition technique developed
by Bernas.ftP.' This procedure has been applied to 12 of the 19 toxic
pollutants of interest including refractory forming elements, e.g., Ba,
Ti, and V. This procedure coupled with AAS has been applied to the
analysis of granite, coal, coal ash, glass, and fish tissue for 18 ele-
ments and resulted in a relative standard deviation of approximately
5% for trace elements. The procedure has the following advantages:
elimination of interelement and ionization interferences; elimination
of volatilization and retention losses; relatively low cost per analy-
sis because of the reduction in time and supervision required and the
elimination of expensive platinum ware.
In this method, the samples are decomposed in a Teflon cup encap-
suled in a stainless steel bomb with a decomposition medium of hydro-
fluoric and boric acid. The samples are digested for 0.5 to 3 hr at
110 to 170°C; the more rigorous conditions are applied to samples of
higher organic content. Coal samples will be dry ashed prior to acid
decomposition.
Two methods of sample preparation are used for the acid-
permanganate solutions used to collect elements that might be present
as vapors in the effluent gas stream. The first, is used for mercury
determination by the cold vapor technique, includes digestion with fresh
acid-permanganate followed by reduction with hydroxyl amine (for excess
permanganate) and stannous chloride. The second portion of the acid-
permanganate solutions is treated with sodium borohydride to form
the hydrides of arsenic, selenium, and antimony prior to their deter-
mination by AAS using an argon-hydrogen entrained air flame.
Elemental Pollutants (anions)
Fluorine and chlorine cannot be determined by AAS. Samples
are analyzed for fluorides following digestion under pressure in a sodium
hydroxide medium. After buffering, fluorides are determined with a
149
-------
fluoride selective electrode. The practical limit of detection is 10 u-g
of fluoride per gram of sample. Chlorine is determined by igniting the
samples in a bomb and titrating the aqueous washings with silver nitrate.
The practical limit of detection is 0.01 to 0.03%.
Organic Pollutants
The organic compounds in the particulate matter can be separated
from the total particulate by extraction with benzene. The separated
organic material, the organic material collected in the Na2(X)3 impin-
gers and the Tenax-GC"plug are analyzed for POM and PCB.
POM Analysis - The POM in the benzene soluble fraction has been sepa-
rated from PCB, aliphatic and heterocyclic compounds by column chro-
matography with activated silica gel as the adsorbent._Li/
Following the isolation of the total POM, gas chromatography with
electron capture detection is used to quantify individual POM. Solu-
tion detection limits are in the 0.2 to 0.5 p,g/ml range. Lao et al-^'
used a Dexsil-300 packed column to separate 70 POM. This packing has
been used in our laboratory for benzQa]pyrene analysis and can be used
to separate the 8 POM listed in Table 4. Quantification of 5 POM* will
be made by comparing sample peak areas with areas obtained from syn-
thetic mixtures. From Lao's work, benzo[c]phenanthrene and dibenzoQa,h]-
pyrene can be identified by relative retention times. However, there is
no information on GC analysis of dibenzo[a,j>]carbazole or Dexsil-300,
making identification impossible without a standard.
To verify the presence of POM, selected samples are analyzed by
gas chromatography-mass spectrometry (GC-MS). At least 10 |j,g of a com-
pound per milliliter of extract are required for identification of
individual POM with a high degree of confidence by GC/MS.
PCB Analysis - PCB is determined by gas chromatography with elec-
tron capture detection. Solution detection limits are in the sub parts
per million range. Other electron capturing materials can interfere
with PCB analysis. The chromatographic method described by Armour and
Burke!2/ was used to separate and quantify the PCB in the benzene ex-
tract. The interfering electron capturing materials are eliminated by
this method. Total PCB is determined by comparing total sample peak
areas with the area obtained from a representative PCB standard.
Standards cannot be obtained for benzoQcJphenanthrene, dibenzoQa,h]-
pyrene, and dibenzo[_c,£]carbazole.
150
-------
Sulfates and Nitrates
Water soluble sulfates and nitrates can be extracted with hot
water and the extract divided into two portions. Sulfates are deter-
mined by microtitration with barium perchlorate using a Thorin indi-
cator. Nitrates are determined spectrophotometrically after complex-
ing with phenol disulfonic acid.
ALTERNATE ANALYSIS METHODS
This section contains a brief description of methods of analysis
that meet the accuracy, precision, detection limit, and sample size re-
quirements for the analysis of a number of pollutants pertinent to this
study. In some cases, the methods listed below are equal to the method
of choice listed in previous paragraphs, but have a higher cost per analy-
sis or require more expensive instrumentation.
Neutron Activation Analysis (NAA)
The application of NAA to the elemental pollutants of interest has
been shown to be feasible in coal and coal ash matrices. This technique
has the advantage of being a sensitive multielement technique. Generally,
the detection limits and accuracies obtained for these elements by NAA
are comparable to those obtained by AAS. However, to achieve optimum
sensitivity and to remove interfering radioactivity from other elements,
chemical separations are usually required. The sample preparation tech-
niques are usually specific for one or a few elements so that several
methods would be required to improve the sensitivity and specificity for
the 18 elements of interest.
The instrumentation required for NAA is sophisticated and expen-
sive. Included in the basic instrumentation are an irradiation source,
i.e., a nuclear reactor, a pulse height analyzer and a high resolution
lithium-drifted germanium detector system.
Current work at Oak Ridge National Laboratory is directed toward
the detection of elemental pollutants by nondestructive methods, i.e.,
without chemical treatment. Computer programs are being developed to
identify X-ray photopeaks, assign and catalog energies, identify nu-
clides, etc. The goals are to eliminate separations, minimize inter-
ferences from high concentration components and reduce post-irradiation
counting time. However, the state of the art for NAA analysis of the
18 elements of interest requires the most costly chemical separation
techniques at the micrograms per gram.
151
-------
X-Ray Fluorescence (XRF)
X-ray fluorescence is a multielement technique that can be applied
to most of the elements of interest; the analysis of beryllium and boron
by XRF is not feasible. The detection limits (50 to 1,000 ppm) without
extensive preconcentration steps are generally 1 to 2 orders of magni-
tude higher than AAS and NAA. Conventional XRF techniques require sample
sizes of 1 to 10 g.
X-ray fluorescence equipment is available for simultaneous deter-
mination of as many as 10 elements. The technique can be totally auto-
mated with a resulting low cost for routine analysis if the required
sample size and element concentrations are available.
The combination of XRF with electron microprobe analysis does offer
the unique capability of determining the size and elemental composition
of particles. In this technique, a focused beam of electrons is moved
across the surface of the sample to be examined. The electrons cause
the emission of low energy secondary electrons. The images of the par-
ticles in the sample are depicted by these electrons. Resolutions of
200 A (0.02 u) can be obtained. The focused electron beam generates
X-rays in addition to low energy electrons. Measurement of the X-rays
results in the determination of the elements present. Particles as
small as 0.5 u can be analyzed at concentrations as low as one part per
thousand.
Spark Source Mass Spectrometry (SSMS)
SSMS is best suited to survey an unknown sample for all possible
elements including trace and major constituents. The detection limits
are in the parts per billion range for all elements. The required sample
size is from 10 to 100 mg. Detection of positively charged ions can be
made photographically or electronically. The accuracy of SSMS when
operated as a survey technique, i.e., when no standard is available, is
within a factor of 10 of the correct answer. The accuracy can be im-
proved to ± 30% if standard reference materials are available or by
using synthetic standards if matrices can be adequately matched.
Isotope dilution is the most accurate SSMS technique. This tech-
nique can be applied to elements with two or more stable isotopes. The
sample must be spiked with an isotope in the same chemical oxidation
state. NBS results indicate that the isotope dilution technique can be
applied to 10 of the 18 elemental pollutants at the 0.1 ppm level for
1-g samples. The dynamic range for a single isotope spike is limited
to 30. This range limitation makes isotope dilution SSMS best suited
for testing samples for compliance to preset standards.
152
-------
Electrochemical Methods
Of the many electroanalytical methods available, anodic stripping
voltametry appears to be the best suited for trace pollutant analysis.
Electrochemical methods have the advantages of inexpensive instrumenta-
tion and minimum sample manipulation. However, they have narrower ap-
plicability than AAS and NAA. Of the elements listed in Table B-2, the
stripping technique is especially suited for the analysis of Cd, Cu, Pb,
and Zn; detection limits for these elements are in the sub parts per
billion.
153
-------
APPENDIX C
CALCULATION OF BOILER STEAM EFFICIENCY
154
-------
This appendix outlines the procedures for calculating boiler steam
efficiency using the abbreviated heat loss method specified in ASME
Performance Test Code 4.1 (1964).
The efficiency of steam generating equipment determined within the
scope of the ASME Code is the gross efficiency and is defined as the
ratio of the heat absorbed by the working fluid to the heat input. This
definition disregards the equivalent heat in the power required by the
auxiliary apparatus external to the envelope (Figure C-l).
For conducting an abbreviated efficiency test that considers only
the major losses and only the chemical heat in the fuel as input, the
ASME data summary form and calculation form (Tables C-l and C-2, respec-
tively) are used. The mathematical scheme used to calculate boiler steam
efficiency by the abbreviated heat loss method consists of the following
10 equations:
C Wr x Hr
Cb ~ 100 " 14,500
where Co = pounds of carbon burned per pound of "as fired" fuel
C = percent carbon in "as fired" fuel (ultimate analysis)
Wr = pounds of dry refuse per pound of "as fired" fuel
Hr = heating value of total dry refuse (Btu/lb)
11 C02 + 8 02 + 7(N2 + CO)
W8 3 (C02 + CO) X
where W. = pounds of dry gas per pound of "as fired" fuel
C02» 0£» N£, CO = composition of dry flue gas (% by volume)
S = pounds sulfur per pound "as fired" fuel (ultimate
analysis)
CD = pounds of carbon burned per pound of "as fired" fuel
,02 - 0.5 CO
N2 -02
O "1
+ 0.5 CoJ
where EA = excess air (%)
O-i CO, N2 = composition of dry flue gas (% by volume)
155
-------
f D FAN
DliCHARGE
ITE*« COS
COOLING «T» LCAf
• ATE* MJICTIOM OFF
UK PIT WATid
Figure C-l. Steam generating unit diagram.
156
-------
SUMMARY SHEET
ASME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
PTC 4 1-0(1964)
TEST NO BOILER NO
DATE
0»NCR Or PLANT LOCATION
If <1 CONDUCTED BY OBJECTIVE OF TEST
DURATION
BOILER MAKF & TYPF RATED CAPACITY
STOKER TYPE & SIZE
PULVFRI7TR TYPF & SIZE BURNER. TYPE
ruFL uscn MINE COUNTY STATE
& SIZE
SI7E AS FIRED
PRESSURES & TEMPERATURES FUEL DATA
1
2
}
__4
5_
6
;
8
1
10
II
12
13
14
STEAM PRESSURE IN BOILER DRUM
STEAM PRESSURE AT S H OUTLET
STEAM PRESSURE AT ft H INLET
STEAM PRESSURE AT R H OUTLET
STEAM TEMPERATURE AT S H OUTLET
STEAM TEMPERATURE AT R H INLET
STEAM TEMPERATURE AT R H OUTLET
WATER TEMP ENTERING (ECON HBOILER)
STEAM QUALITY '.MOISTURE OR P P M
AIR TEMP AROUND BOILER (AMBIENT)
TEMP AIR FOR COMBUSTION
TEMPERATURE OF FUEL
CAS TEMP LEAVING (Boilor) (Econ ) (An Hli )
CAS TEMP ENTERING AH (II cc.ndil.om to bo
corrected to ouorontao)
PIIO
pno
PIIO
ptia
F
F
F
F
F
F
F
F
F
UNIT QUANTITIES
IS
16
17
18
19
20
21
22
23
It
25
26
27
28
29
30
31
ENTHALPY OF SAT LIQUID (TOTAL HEAT)
ENTHALPY OF (SATURATED) (SUPERHEATED)
STM
ENTHALPY OF SAT FEED TO (BOILER)
(ECON.)
ENTHALPY OF REHEATED STEAM R H. INLET
ENTHALPY OF REHEATED STEAM R H
OUTLET
HEAT ABS'Lfl OF STEAM (ITEM 16-ITEM 17)
HEAT ABS LB R.H STEAMIITEM 19-ITEM 18)
DRY REFUSE (ASH PIT » FLY ASH) PER LB
AS FIRED FUEL
Biu PER LB IN REFUSE (WEIGHTED AVERAGE)
CARBON BURNED PER LB AS FIRED FUEL
DRY CAS PER LB AS FIRED FUEL BURNED
HOURLY QUANTITIES
ACTUAL WATER EVAPORATED
REHEAT STEAM FLOW
RATE OF FUEL FIRING (AS FIRED .1)
TOTfH MEAT INPUT I1"1" M * lum 41)
HEAT OUTPUT IN BLOW DOWN WATER
HEATL """" M-lw MMIt.m27.lion 2D» I"- 30
OUTPUT 1000
Bht/lb
Biu/lh
Btu/lb
Biu/lb
Biu/lb
Btu'lb
Blu'lb
Ib/lb
Slu/lb
Ib/lb
Ib/lb
Ib/ht
Ib/hr
lb>hr
kBAr
kB/hr
IB/n,
FLUE CAI ANAL (BOILERHECQN) (AIR HTR) OUTLET
32
33
34
35
36
CO,
0,
CO
N 'BY DIFFERENCE)
FXCFSS AIR
•• VOL
•i VOL
'.VOL
% VOL
*•
COAL AS FIREO
PR OX. ANALYSIS
37
38
39
40
MOISTURE
VOL MATTER
FIXED CARBON
ASH
TOTAL
41
42
Btupor Ib AS FIRED
ASH SOFT TEMP •
ASTM METHOD
is-i
COAL OR OIL AS FIRED
ULTIMATE ANALYSIS
43
44
45
46
4,7
40
37
CARBON
HYDROGEN
OXYGEN
NITROGEN
SULPHUR
ASH
MOISTURE
TOTAL
COAL PULVERIZATION
48
49
SO
64
GRINDABILITY
INDEX*
FINENESS 7.THRU
50 M'
FINENESS fi THRU
200 M-
SI
J2
S3
44
41
OIL
FLASH
Sp Gro.
POINT F-
iiy Dog API1
VISCOSITY AT SSU-
BURNER SSF
TOTAL HYDROGEN
% "1
Biu P.I
Ib
CAS
54
SS
96
97
58
59
60
61
CO
CH. METHANE
C,H, ACETYLENE
C,H. ETHYLENE
CiH. ETHANE
H,S
CO,
HI
HYDROGEN
TOTAL
62
63
41
TOTAL HYDROGEN
S .1
SVOL
DENSITY 68 F
ATM PRESS
Bin PER CU FT
Biu PER LB
INPUT.OUTPUT ITEM 31 • 100
EFFICIENCY OF UNIT *i ITEM 29
HEAT LOSS EFFICIENCY
65
66
67
68
69
70
71
72
HEAT LOSS DUE TO DRY CAS
HEAT LOSS DUE TO MOISTURE IN FUEL
HEAT LOSS DUE TO H,O FROM COMB OFH,
HEAT LOSS DUE TO COMBUST IN REFUSE
HEAT LOSS DUE TO RADIATION
UNMEASURED LOSSES
Bm/lb
A.F. FUEL
TOTAL
EFFICIENCY = (100 -It.™ 71)
' Not Roqumo' (OP Efficiency Toiling
ft ol A. F
FUEL
1 For Point el Mooiuromonl Soo Pat. 7 2 8.1-PTC 4 1. 1964
Table C-l
157
-------
CALCULATION SHEET
PTC4.1-b (1964)
ASME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST Revised September. 1965
30
24
25
36
65
66
67
6B
69
70
71
72
OBNFR OF PLANT TEST NO BOILER NO DATE
ITEM IS ITEM 17
HFAT OUTPUT IN BOILER BLOW.OQWM WlTFP =1 R OF WiTFR pi OW.DOWM PljP HP > .. . . _-
1000
If impractical to weigh refuse, this
item can be estimated as follows
r,Dv Dpru5E PER LB OF A5 FIRED FUEL . '• *SH IN »' FIRED COAL HOT[. |r r
100-'. COMB IN REFUSE SAMPLE PIT REFUSE
fcB/hr
e * •
LUE DUST & ASH
DIFFER MATERIALLY
r- -| "" IN COMBUSTIBLE CONTENT. THEY
ITEM 43 HTEM22 ITEM 23 1 SHOULD BE ESTIMATED
CARBON BURNED SEPARATELY SEE SECTION 7.
FUEL 1BO |_ l000 J UtaHimli
DRY CAS PER LB n co, • ao, • 7 IN, • coi
BURNED JIC0' ' C01 / V
ITEM 32 ITEM 33 ( ITEM 35 ITEM 34 ] ITEM 24
."" '8X ' 'I . ' .... / «
ONS
••4-»
ITEM 47 |
/ITEM 32 ITEM 34 \ [_ 267 J
,*(.. . ' ...)
EXCESS °i - — ITEM 33 - ITEM 34
2682N, - (o, - CO. ,
2 .2682 (ITEM 35) -(ITEM 33 - lirlz: )
HEAT LOSS EFFICIENCY
HEAT LOSS DUE LB DRY CAS ITEM 25 (IT EM 111 MTFMIll
TODRVGAS = PERLBAS «C x (M., _ i.,,) » * ™»0 14 ' J> ' B '".
FIRED FUEL ' Un,t
-(ENTHALPY OF LIQUID AT T AIR)) = HE^LZL x [(ENTHALPY OF VAPOR
100
AT 1 PSIA A T ITEM 13) -(ENTHALPY OF LIQUID AT T ITEM 11)1 • •
HEAT LOSS DUE TO H,0 FROM COMB OF H, -. 9H, X {(ENTHALPY OF VAPOR AT 1 PSIA « T CAS
LVC) - (ENTHALPY OF LIQUID AT T AIR)]
- 0 « ITEM f*. « [(ENTHALPY nt U1PQD AT 1 P$IA * T ITfU 1 J) _ (FMTHAI PY OF MQU"> AT
100 T ITEM 11)] =
HEAT LOSS DUE TO ITEM 22 ITEM 23
COMBUSTIBLE IN REFUSE = x
HEAT LOSS DUE TO TOTAL BTU RADIATION LOSS PER HR
RADIATION- LB AS FIRED FUEL - ITEM 21
UNMEASURED LOSSES *•
TOTAL
EFFICIENCY = (100 - ITEM 71)
Blu/lb
AS FIRED
FUEL
•
. ..
LOSS x
HHV
100 *
" X 100:
41
_ X 100 a
41
41
5-«.
" x 100-
41
^i x 100 ,
41
LOSS
n
••
•
f FBI ngoroui dtierminotton of ••catt mr tea AppandiR 97- PTC 4 1-1964
• If lottn or* not mvosurad, via ABM A Standard Radiahei Lo»t Chan. Fig 8, PTC 4 1-1964
•• Unmrotufcd louai lislad m PTC 4 1 but not tabulated above may by provided for by atligning a mutually
agreed upon value lor Item 70 •>
Table C-2
158
-------
= Wg x cp x (tl - ta ) (4)
where Lg = heat loss due to heat in dry flue gas (Btu/lb of "as
fired" fuel)
W = pounds of dry gas per pound of "as fired" fuel
Cp = mean specific heat of dry flue gas = 0.24 Btu/lb °F
ti = temperature of gas leaving boiler (°F)
t& = reference temperature of air for combustion (°F)
= m x (ht - ht) (5)
where L_ = heat loss due to moisture in "as fired" fuel
mf = pounds of water per pound of "as fired" fuel
ht. = enthalpy of vapor at 1 psia and t^
ht = enthalpy of vapor at 1 psia and ta
•a
= 9 H2 x (h|.i -
(6)
where L,. = heat loss due to moisture from combustion of hydrogen
(Btu/lb of "as fired" fuel)
H2 = pounds of hydrogen per pound of "as fired" fuel (ulti-
mate analysis)
ht and ht = as defined in Eq. (5)
Ju a
L = W x H ...
UC r r (7)
where L = heat loss due to unburned carbon (Btu/lb of "as fired"
UC
fuel)
Wr and Hr = as defined in Eq, (1)
where LR = heat loss due to radiation (Btu/lb)
R_ = radiation loss as percent of gross heat input
HHV = heat of combustion (Btu/lb from proximate analysis)
159
-------
L = L + L + L + L „ + L0 (9)
g mj H lIC R
where L = total heat loss (Btu/lb)
Lg = heat loss due to heat in dry gas (Eq. (4))
L = heat loss due to moisture in fuel (Eq. (5))
LH = heat loss due to 1^0 from combustion of H2 (Eq. (6))
Lye = heat loss due to unburned carbon (Eq. (7))
LR = heat loss due to radiation (Eq. (8))
All above in units of Btu/lb
Tig = 100 - x 100 (10)
where T]g = gross efficiency (%)
L = total heat loss (Eq. (9))
HHV = as defined in Eq. (8)
Application of the heat loss method to the data from Widows Creek
Run 4 yields the following:
\ -
(2)
= 12.6 Ib dry gas/lb fuel
EA _ 100 _ f 5.8 - 0.5 (0.3) 1 _
^ ~ 10° X L0.2682 (81.2) - 5.8 + 0.5 (0.3)J ~ 35<01 (3)
L0 = 12.6 x 0.24 x (360 - 100) = 786 Btu/lb (4)
= 0.0138 x (1,222.7 - 67.9) = 15.9 Btu/lb (5)
= 9 (0.0447) x (1,222.7 - 67.9) = 465 Btu/lb (6)
= 0.197 x 300 = 59.1 Btu/lb (7)
160
-------
LR = Jr1 x 11,541 = 50.8 Btu/lb (8)
L = 786 + 15.9 + 465 + 59.1 + 50.8 = 1,377 Btu/lb (9)
Note that in estimating the value of 0.44 for the radiation loss
as percent of gross heat input (Eq. (8)), the following assumptions were
made:
1. One megawatt of generating capacity is equivalent to 10^ Btu/hr
of heat output.
2. The temperature differential between the outer furnace wall and
the surrounding air was 100°F.
The calculated value of 88.1% for the boiler steam efficiency is
very close to the design value of 88.457..
161
-------
APPENDIX D
CALCULATION PROCEDURE FOR MASS BALANCE PRECISION
162
-------
1. Introduction
Table D-l defines the variances in the mass balances calculated in
Section VIII of this report. The following subsections of this appendix
document (a) the general statistical equations used in the calculations
and (b) the methodology for calculation of V(M)T, V(M)A, V(M)p> and
V(M)s, respectively. (V(M)A, V(M)p, and V(M)g are component parts which
totalled together equal V(M)T.)
Table D-l. DEFINITIONS OF VARIANCES
V(M)T The total variance in the actual mass imbalance which includes:
(1) analytical variability, V(M)A,
(2) process variability, V(M)p, and
(3) variance due to nonrepresentative sampling, V(M)g.
V(M)A The theoretical variability in the mass imbalance if errors
arose only from imprecision of the laboratory analysis results.
V(M)p The theoretical variability in the mass imbalance if errors
arose only from run to run changes in pollutant concentration.
V(M)« The difference between the total variability in the mass im-
balance, V(M)T, and the sum of V(M)A and V(M)p. Ideally, this
"sampling" error represents the "errors of the sample," i.e.,
the inaccuracy of the result arising from the fact that only a
finite number of samples were examined (as opposed to the whole
population). We have no direct way of estimating this, however,
since we took only a sample size of 1 under each condition.
2. General Equations
Consider the basic equation for calculation of fractional mass im-
balance:
_ Output - Input
Input
163
-------
Let c^ be the concentration (ppm) in each stream i , where:
Stream
i Stream abbreviation
1 Coal C
2 Bottom ash BA
3 Superheater ash SA
4 Dust collector ash DCA
5 Outlet fly ash OFA
Let w^ be the flow rate through each corresponding stream i
The fractional mass imbalance can then be considered as:
M
wlcl
Let y, - — and x. = - (4 and 5)
K W]_ K Cj^
Thus, M = yjXj + y2x2 + y3x3 + y^ - 1 (6)
Then, letting zk = y^x^t
M = t]_ + z2 + z3 + z^ - 1 (7)
The following general equations were used in the calculation of
variances V(M)p and V(M)A for sums, products, and quotients, respectively:
V(M) = V(Zl) + V(z2) + V(z3) + V(z4) (8)
164
-------
with
V " 2xi cov
where the covariance, cov, is defined as:
x °Swiwi+i - £wi>
covCw^. ..) = - (12)
n (n - 1)
and
"Iclci+l - j
cov(Clci+1) • (13)
n (n - 1)
3. Calculation Methodology
a. V(M)T
The total variance in the actual mass imbalance, V(M)T, can
be calculated as follows:
3 (m22 + m32 + m42) - (m2 + m3
where m« » m3 > an^ m4 are tne fractional mass imbalances as deter-
mined for each run (Runs Nos. 2, 3, and 4, respectively).
165
-------
b. V(M)A
V(M)A represents the theoretical variability in the mass im-
balance if errors arose only from imprecision of the laboratory results.
Duplicate samples for Runs Nos. 3 and 4 were analyzed by the
same procedures. Since "within run" variation is considered here, vari-
ances between the duplicate samples were determined for each stream of
Run No. 3 and each stream of Run No. 4, by the formula:
.
ij n (n - 1)
where n = 2 (duplicate samples) and
i = stream number, 1 through 5
j = run number, 3 through 4
the average variance for each stream is:
V(c i ) - (16)
By substitution of appropriate values into Eqs. (8) through (13)
(in reverse order), Eq. (8) yields the result of V(M)A for each pollutant.
c. V(M)P
The calculation procedure for V(M)p parallels the procedure
for calculation of V(M)A • However, V(M)p is defined as the theoretical
variability in the mass imbalance if errors arose only from run to run
changes in pollutant concentration.
Instead of determining variances and covariances for each run
and stream, and then averaging for each stream, as we did for calcula-
tion of V(M)A , we considered the average values determined for Run
No. 3 for each stream and the average values determined for Run No. 4
for each stream. The variance for each stream, between runs, can be cal-
culated by:
166
-------
V(c.
n (n - 1)
where n - 2 and the c^'s are summed for Runs Nos. 3 and 4.
Again, by substitution of the appropriate values into Eqs. (8)
through (13) (in reverse order), Eq. (8) yields the result of V1(M)p
for each pollutant. To obtain a measure of run to run changes in pollu-
tant concentration, we subtracted out the analytical contribution. Thus,
V(M)p = V(M)P - V(M)A
d. V(M)S
V(M)§ represents the "sampling" error or errors due to the
fact that only a finite number of samples were taken (as opposed to the
whole population).*
The formula used is as follows:
V(M)S = V(M)T - V(M)A - V(M)P
In Section IX, we defined nonrepresentative sampling as V(M)g +
V(M)p .
167
-------
APPENDIX E
FACTORS FOR CONVERSION TO METRIC UNITS
168
-------
English unit
Metric equivalent
Conversion
factor (or equation)
barrels (oil)
Btu
Btu/hr
Btu/kw-hr
Btu/lb
cfm
cu ft
dscfm
dscfm/mw
°F
ft
8r
gr/106 Btu
gr/acf
gr/scf
horsepower
in.
in. Hg
Ib
Ib/hr
miles
ppm by volume
ppm by weight
tons (short)
tons/hr
m3
joule
watt
watt/kw
joule/kg
m3/min
m3
nm^/min
nm3/mw
°C
m
g
g/106 joule
mg/m3
mg/nm3
watt
cm
mm. Hg
gm
kg/min
km
V.IISL
M-g/gm
kg
kg/hr
0.1589873
1055.056
0.2930711
0.2930711
2326.012
0.028317
0.028317
0.028317
0.028317
5/9 (°F - 32)
0.3048
0.064799
0.0009478
2288.34
2288.34
745.6999
2.54
25.4
453.59
0.007558
1.609344
1
1
907.1847
907.1847
169
-------
REFERENCES
1. Goldberg, A. J., "A Survey of Emissions and Controls for 'Hazardous*
and Other Pollutants," U.S. Environmental Protection Agency, Of-
fice of Research and Monitoring, Air Pollution Technology Branch,
November 1972.
2. Lee, R. E., Jr., and D. J. von Lehmden, "Trace Metal Pollution in
the Environment," J. Air Pollut. Cntrl. Assn.. 23jC10):853-857,
October 1973.
3. Bo Iten, N. E. et al., "Trace Element Measurements at the Coal-Fired
Allen Steam Plant," Progress Report, June 1971 - January 1973, for
NSF by Oak Ridge National Laboratory, Contract No. W-7405-eng-26,
Report No. ORNL-NSF-EP-43, March 1973.
4. Joensuu, 0. I., "Fossil Fuels as a Source of Mercury Pollution,"
Sci., 172^:1027-1028, 4 June 1971.
5. Bertine, K. K., and E. D. Goldberg, Sci., .17J3:233 (1971).
6. "Particulate Polyeyelie Organic Matter," National Academy of Sciences,
Washington, D.C. (1972).
7. Bolten, N. E. et al., draft copy of Section V, "Trace Element Mea-
surements at the Coal-Fired Allen Steam Plant," for NSF by Oak
Ridge National Laboratory (1973).
8. Kaakinen, J. W., R. M. Jorden, and R. E. West, "Trace Element Study
in a Pulverized-Coal-Fired Power Plant," presented at the 67th
Annual Meeting of the Air Pollution Control Association, Denver,
Colorado, 9-13 June 1974.
9. Hillenbrand, L. J., R. B. Engdahl, and R. E. Barrett, Final Report
on "Chemical Composition of Particulate Air Pollutants from Fossil-
Fuel Combustion Sources," Contract No. EHSD 71-29, 1 March 1973.
170
-------
10. Burton, J. S., G. Erskine, E. Jamgochian, J. Morris, R. Reale, and
W. L. Wheaton, "Baseline Measurement Test Results for the Cat-Ox
Demonstration Program," Contract No. F19268-71-C-0002, U.S. En-
vironmental Protection Agency, June 1972.
11. Minerals Yearbook; 1970. Vol. 1, "Metals, Minerals and Fuels," U.S.
Bureau of Mines, pp. 423-455 (1972).
12. "Petroleum Facts and Figures," American Petroleum Institute, p. 67,
Washington, D.C., (1971).
12a. Minerals Yearbook; 1972. Vol. 1, "Metals, Minerals and Fuels," U.S.
Bureau of Mines (1974).
13. Magee, E. M., H. J. Hall, and G. M. Varga, Jr., "Potential Pollu-
tants in Fossil Fuels," EPA Report No. EPA-R2-73-249, Contract No.
68-02-0629, June 1973.
14. "Steam-Electric Plant Factors, 1972," National Coal Association,
Washington, D.C., December 1972.
15. Smith, W. S., and C. W. Gruber, "Atmospheric Emissions from Coal
Combustion—An Inventory Guide," Environmental Health Series, U.S.
Department of Health, Education and Welfare, Public Health Service
Publication No. 999-AP-24, April 1966.
16. Cuffe, S. T., and R. W. Gerstle, "Emissions from Coal-Fired Power
Plants: A Comprehensive Summary," Environmental Health Series,
U.S. Department of Health, Education, and Welfare, Public Health
Service Publication No. 999-AP-35 (1967).
17. Schroeder, H. A., "A Sensible Look At Air Pollution by Metals," Arch.
Environ. Health. 21:798-806, December 1970.
18. Ruch, R. R., H. J. Gluskoter, and N. F. Shimp, "Occurence and Dis-
tribution of Potentially Volatile Trace Elements in Coal," Illinois
State Geological Survey, Environmental Geology Notes, No. 61, April
1973.
19. Vandegrift, A. E., and L. J. Shannon, "Particulate Pollutant System
Study, Volume II—Mass Emissions," prepared for EPA by Midwest
Research Institute, May 1971.
20. National Emissions Data System (NEDS) computer file listing, dated
9 January 1974.
171
-------
21. Perry, John H. et al., Chemical Engineers Handbook, 4th Edition,
McGraw-Hill, New York (1963).
22. Personal communication from Mr. Robert E. Hall, Combustion Research
Section, U.S. Environmental Protection Agency, 3 May 1973.
23. Personal communication from Mr. William Axtman, American Boiler Man-
ufacturers Association, 25 February 1974.
24. The 1970 National Power Survey, Federal Power Commission, available
from U.S. Government Printing Office, Washington, D.C., pp. 1-5-2
(1971).
25. National Emissions Data System (NEDS) computer file listing, dated
28 August 1973.
26. Brenchley, D. L., C. D. Turley, and R. F. Yarmec, "Industrial Source
Sampling," Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan
(1973).
27. Cooper, H. B. H., Jr., and A. T. Rossano, Jr., "Source Testing for
Air Pollution Control," Environmental Science Services, Environ-
mental Research and Applications, Inc., Wilton, Connecticut (1971).
28. Federal Register. 36(247), 23 December 1971.
29. "Reference Method for the Determination of Suspended Particulates
in the Atmosphere (High Volume Method)," Federal Register. 36:28,
Appendix B, 22388-22390, 25 November 1971.
29a. Lao, R. C. et al., Anal. Chem.. 45j908 (1973).
30. Klein, D. H. et al., "Mass Balance for Thirty-Three Elements at a
Coal-Fired Power Plant," Oak Ridge National Laboratory (1974).
31. Marcus, M., "Evaluation of Sampling Trains," EPA Project No. 68-02-
1399, 20 February 1975.
32. "Compilation of Air Pollutant Emission Factors," U.S. Environmental
Protection Agency, Report No. AP-42, January 1975.
33. Gorman, P. G., J. Nebgen, and I. Smith, "Assessment and Development
of Control Technology Applicable to Removal of Mercury and Other
Potentially Hazardous Pollutant Vapors from S02-Bearing Waste
Gases (Power Plants)," Interim Report No. 2 (Draft), U.S. Environ-
mental Protection Agency, Contract No. 68-02-1097, 12 July 1974.
172
-------
34. Griefer, B., and J. K. Taylor, "Survey of Various Approaches Co the
Chemical Analysis of Environmentally Important Materials," Na-
tional Bureau of Standards, Department of Commerce, Report No.
NBSIR 73-209, Washington, D.C., June 1973.
35. Forrest, J., and L. Newman, J. of the APCA. 2_3:761 (1973).
36. Cantuti, V. et al., J. Chromatog.. 17,:60 (1965).
37. Charkraborty, B. B., and R. Long, Environ. Sci, Techno 1,, 1:829
(1967).
38. Bhatia, K., Anal. Chem.. 43j609 (1971).
39. Armour and Burke, JAOAC. 53^761-768 (1970).
40. Bernas, B., Anal. Chem.. 40jl682 (1968).
41. Moore, G. E. et al., J. Chromatog.« 2£:456 (1967).
173
-------
TECHNICAL REPORT DATA
rrad /aUfiiflioin on lite rci»ne before i uinplciing)
1 REPORT NO |2
EPA-650/2-75-066 [__
4 TITLE ANDSUBTITLE
Hazardous Emission Characterization of Utility
Boilers
3 RECIPIENT'S ACCESSION-NO
5 REPORT DATE
July 1975
6 PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Chatten Cowherd Jr. , Mark Marcus,
Christine M. Guenther, and James L. Spigarelli
8 PERFORMING ORGANIZATION REPORT NO
9 PERFORMING OR8ANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
10 PROGRAM ELEMENT NO.
1AB015; ROAP 21AUZ-002
11. CONTRACT/GRANT NO.
68-02-1324, Task 27
12 SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PE
Final; 11/73 - 7/75
PERIOD COVERED
14 SPONSORING AGENCY CODE
15 SUPPLEMENTARY NOTES
16. ABSTRACT
repOrt gjves results of a field sampling program aimed at quantifying
potentially hazardous pollutants in the waste streams of a representative coal-fired
utility boiler: a 125-MW boiler (fired with pulverized coal and equipped with a mech-
anical fly ash collector) at TVA's Widows Creek steam electric generating station.
The combustion products identified as potentially hazardous air pollutants included
22 trace elements , nitrates , sulfates , polycyclic organic compounds , and polychlor-
inated biphenyls. The waste streams sampled included pulverized coal, furnace
bottom ash, superheater ash, collection ash, and flue gases at the fly ash collector
inlet and outlet. Acceptable mass balance was achieved for about half of the elemen-
tal pollutants. Trace metal enrichment was measured. Study results include
recommended modifications of sample collection and preparation methods: larger
and more frequent samples of coal and bulk ash streams are expected to improve
sample representativeness; development of methodologies for estimating bulk ash
flows will permit internal checks on mass balances: and routine chemical analysis
of NBS standard coal and fly ash will improve quality assurance of the analytical
methods .
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution Field Tests
Utilities Quantitative Analysis
Boilers Trace Elements
Hazardous Materials Sulfates
Coal Polycyclic Compounds-Polychlorinated
Combustion Fly Ash
b IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Hazardous Emissions
Nitrates
Biphenyls
c COSATI I idd/Group
13B
ISA
11G
21D
21B
14B
07D
06A, 06F
07B
07C
2 DISTRIBUTION STATEMENT
Unlimited
19 SECURITY CLASS (Tins Rtport)
Unclassified
20 SECURITY CLASS (Tins page)
Unclassified
21 NO OF PAGES
185
22 PRICE
EPA Form 2220-1 (9-73)
175
------- |