Unrtad States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/2-80-095
Augubt 1980
Research and Development
A Field Test Using
Coal:dRDF Blends in
Spreader Stoker-Fired
Boilers
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine 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 (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental 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 through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-095
August 1980
A FIELD TEST USING COAL:dRDF BLENDS
IN SPREADER STOKER-FIRED BOILERS
by
Gerald H. Degler
H. Gregory Rigo
Boyd T. Riley, Jr. (Consultant)
Systems Technology Corporation
Xenia, Ohio 45385
Contract No. 68-03-2426
Project Officer
Carlton C. Wiles
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publi-
cation. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendations for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health and
welfare of the American people. Noxious air, foul water, and spoiled land are
tragic testimonies to the deterioration of our natural environment. The
complexity of that environment and the interplay of its components require a
concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution;
it involves defining the problem, measuring its Impact, and searching for solu-
tions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal and community
sources, to preserve and treat public drinking water supplies, and to minimize
the adverse economic, social, health, and aesthetic effects of pollution. This
publication is one of the products of that research and provides a most vital
communications link between the researcher and the user community.
In recognition of the fact that more than 50 percent of the roughly
42,000 industrial boilers in the U.S. are coal-fired boilers, the Environmental
Protection Agency undertook a project to investigate the technical and
environmental implications of using densified (pellet form) refuse derived
fuel as a substitute for stoker coal. This report presents the results from
co-firing 258.5 Mg (285 tons) of dRDF when firing at various volumetric blend
ratios of coalrdRDF, i.e., 1:1, 1:2, and 0:1.
The investigation specifically addresses the performance of the fuel
handling and feeding system, the boiler, and the resulting emissions from
each blend firing. Since a spreader stoker-fired 7.6 kg/sec (60,000 Ib/hr)
boiler was co-fired for 230 hours (132 hours continuously) without major
difficulty, the results are sufficiently encouraging to suggest a larger term
demonstration of co-firing coal and dRDF.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
This study program was initiated with the overall objective being to
characterize and demonstrate the technical, economical, and environmental
feasibility of combusting densified forms of refuse derived fuels (dRDF)
blended with coal in spreader stoker-fired boilers.
The testing was conducted at the Maryland Correctional Institute Power
House located in Hagerstown, Maryland. A total of 258.5 Mg (285 tons) of
pelletized 1/2-inch-diameter by 3/4-inch-long dRDF was co-fired with coal in
spreader stoker boilers rated at 7.6 and 9.9 kg/sec (60,000 and 75,000 Ib/hr)
of 1034 kPa (150 psig) saturated steam.
The field tests were designed to investigate (1) the material handling
characteristics of dRDF, i.e., storage in warehouses, drop boxes, and open
slabs; conveying; and feeding out of bunkers; (2) boiler performance, i.e.,
boiler efficiency, spreader limitations, grate speeds, underfire and overfire
air requirements, steam production, flame impingement, slagging, fouling,
clinkering, and combustion gas analysis; and (3) environmental performance,
i.e., size, mass, opacity, and resistivity of particulates; gaseous (SOX,
NO , Cl, F, He) emissions; and trace organic and inorganic emissions.
X
With the steam demand limiting the test boiler to a 30-55 percent load,
the 258.5 Mg (285 tons) of dRDF were satisfactorily co-fired with coal for
230 hours (132 hours continuously). The results indicate that coal:dRDF
blends up to 1:2 can be handled and burned in conventional spreader stoker-
fired boilers without major equipment modification. The fuel blends were
handled satisfactorily, although some pellet deterioration (due to excessive
handling and rain damage) caused much dusting and slightly impeded the pellet
flow. After adjustments of the air controls, the spreader-feeders, and the
grate pulse interval, the blends generally burned as well as coal alone.
Moreover, as more dRDF was substituted for coal, the flame volume increased,
the opacity decreased, the fly ash carbon burnout improved, and the turndown
ratio of boiler operation increased. Relative to the particulate emissions
from coal-only firing, the emissions from the blend firing decreased slightly
in mass flux, dropped significantly in particulate size and stack opacity,
and had resistivities within the range for satisfactory electrostatic precipi-
tator performance. Also as dRDF substitution increased, chlorine and trace
metals (specifically Pb, Sb, Br, and Mn) increased, and SOX decreased
correspondingly.
This report was submitted in fulfillment of Contract No. 68-03-2426 by
Systems Technology Corporation under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period June 29, 1976 to December 30,
1977.
IV
-------
CONTENTS
Foreword ill
Abstract iv
Figures viii
Tables xii
Acknowledgment xiii
1. Introduction 1
Feasibility of burning dRDF 1
Types of refuse fuels 1
Previous dRDF test programs 2
Current dRDF test programs 2
Site selection 2
Test program outline 4
2. Summary and Conclusions 5
Introduction 5
Test objective 5
Site selection 5
Test design 5
Test results 5
Material handling 5
Pellet storage 5
Pellet feeding 7
Pellet properties 7
Boiler performance 7
Spreader-feeder performance 7
Combustion of dRDF 8
Fouling 8
Clinkering 8
Corrosion 8
Boiler operation 9
Ash handling 9
Mass and energy balance 9
Environmental performance 10
Data normalization 10
Particulate emissions 10
Gaseous emissions .... 11
Trace organic and
inorganic emissions 11
Conclusions 12
Fuel handling system 12
Boiler performance 12
Environmental performance 12
Summary 12
-------
3. Material Handling 13
Fuel mechanical properties . 13
Coal 13
Pelletized dRDF 15
Storage of dRDF 15
Open containers 18
Warehouse 18
Remote slab 18
On-site slab 20
Fuel handling system 20
Description 20
Operation 22
Performance 22
Alternative fuel blending method 23
Summary 24
4. Boiler Performance 25
Boiler description 25
Boiler conditions 25
Fuel properties 27
Coal properties 27
dRDF properties 30
Blend properties 30
Conclusions 32
Fuel handling and response in
boiler systems 32
Fuel distribution 35
System description .... 35
Cold flow test 36
Hot flow test 40
Normal boiler operation 43
Ash handling 44
System description 44
Grate 45
Bottom ash 45
Ash silo 49
Reinjection and collector
fly-ash flows 49
Air and gas handling 53
System description 53
Underfire air setting 56
Overfire air setting 56
Induced draft fan 60
Furnace performance 61
Heat release rate 61
Flue gas temperature 61
Fouling, gas, and wastage 61
Firing phenomenon .... 72
Boiler controls 72
Mass and energy balance 77
Low-load performance 82
Summary 82
vi
-------
5. Environmental Performance 84
Introduction 84
Field sampling setup 84
Test chronology and procedures 84
Opacity 89
Particle mass flux 89
Size distribution 91
Fly ash resistivity 91
S02 94
Oxides of nitrogen 95
Halogens 95
Oxygen 95
Hydrocarbons 95
Trace organic emissions 95
Trace inorganic emissions 95
Data analysis and normalization 97
Data analysis 97
Data normalization 98
Particulate emission test results 101
Opacity 101
Particulate concentration 102
Size distribution 106
Fly ash resistivity 106
Overall ESP performance 106
Gaseous emissions test results 110
S02 110
Oxides of nitrogen .' . . . 113
Halogens 113
Hydrocarbons 113
Trace compound emissions test results 120
Trace organic emissions 120
Trace inorganic emissions 121
Summary 126
References 128
Appendices
A. Emissions, fuel, and ash data summaries 129
B. Summary sheets for ASME abbreviated efficiency
tests and Boiler No. 1 and 2 specifications 138
C. Procedure for estimating stack velocity 148
D. Cascade impactor data 151
E. Discussion of Monsanto1 s ESP Test Data 157
F. Heavy metals emissions data summaries 163
G. Physical and chemical characterization of
dRDF/coal fly ash 173
H. Preceding coal:dRDF studies 207
vii
-------
FIGURES
Number
1 Comparison of the size distributions for the
December and Hay coals 14
2 Characteristic length distributions for pellets
burned in March 16
3 Characteristic length distributions for pellets
burned in May 16
4 Comparison of arithmetic (interpolated)
and field-measured blend bulk densities 17
5 Comparison of pellet length distributions for
different types of storage 19
6 Effect of storage duration and method on pellet
moisture content 19
7 Comparison of a deteriorated pellet (left) and
a well-formed pellet (right) 20
8 Front- and side- view drawings of the
temporary fuel handling system 21
9 Cross section of Boiler No. 2 26
10 Coal and dRDF size distributions compared
with recommended size spectra 29
11 Cross section of Hoffman Combustion Engineering
spreader-feeder 35
12 Spreader-feeder injecting a dRDF:coal blend
into furnace 37
13 Uniform distribution of coal and dRDF pellets
near the furnace rear wall 38
14 Isolation of grate dRDF to determine its
spread density 39
viii
-------
FIGURES (continued)
Number
15 Ash reinjection and overfire air ports in
rear wall of Boiler No. 1 42
16 Effect of blend and load on grate pulse interval
or relative grate speed 46
17 Sieve analysis of bottom ash samples for coal,
blend, and dRDF firings 47
18 Drawing of a typical ash collection drain tube to
monitor relative ash flow in collector and reinjector . . 50
19 Ash flows in reinjector hopper drain tube for
coal, blend, and dRDF firings 51
20 Ash flows in collector hopper drain tube for
coal, blend, and dRDF firings 52
21 Carbon content of reinjector ash for coal,
blend, and dRDF firings 53
22 Reinjector ash size distributions for coal,
blend, and dRDF firings 54
23 Collector ash size distributions for coal,
blend, and dRDF firings 55
24 Carbon dioxide levels in furnace vs. time as
determined with a water-cooled probe 57
25 Relationship of furnace excess air level with
blend and load ; 58
26 Grate seal leakage forcing flames toward center
of furnace during a 1:1 blend firing in Boiler No. 2 ... 59
27 Heat release rates per unit grate area for
coal, blend, and dRDF firings 62
28 Heat release rates for coal, blend, and
dRDF firings 63
29 Effects of blend and load on flue gas temperature 64
30 Variations in ash fusion hemispheric temperatures
under reducing atmospheric conditions during
December runs 65
IX
-------
FIGURES (continued)
Number Page
31 Variations in ash fusion hemispheric temperatures
under reducing atmospheric conditions during March runs . . 66
32 Variations in ash fusion hemispheric temperatures
under reducing atmospheric conditions during May runs ... 67
33 Effects of blend and load on flue gas temperature
before and after a test 68
34 Effects of dRDF and load on flue gas temperature
before and after a test 69
35 Drawing of typical clamp-on corrosion test shield 70
36 Furnace flames viewed at 3.3 m (10 ft) above the
grate during blend and dRDF firings 73
37 View from top tube hatch in Boiler No. 2 to show
firing with a 1:1 blend 74
38 Effects of blend and load on flame temperature
measured with an optical pyrometer ........ 75
39 Pressure chart recordings for coal, blend, and
dRDF firings 76
40 Effects of blend and load on carbon content of
bottom ash 78
41 Effects of blend and load on carbon content of
collector ash 79
42 Effects of blend and load on carbon content of
fly ash 80
43 Effects of blend and load on input/output efficiency 82
44 View of MCI power plant showing stack sampling
shed and temporary fuel handling system at right 85
45 Layout of Boilers No. 1, 2, and 3 with sampling
locations indicated 87
46 Schematic of EPA Method 5 sampling train setup 90
47 Typical dust loading of MRI cascade impactor
stages during a 1:1 blend firing 92
-------
FIGURES (continued)
Number Page
48 Schematic of MRI cascade impactor sampling
train setup 93
49 Schematic of WAHLCO resistivity probe assembly 94
50 Schematic of Battelle Tenex sampling train setup 96
51 Graphic representation of probably similar (A & B) and
potentially dissimilar (A & C) regression lines
through data set 99
52 Effects of blend and load on stack opacity 103
53 Effects of blend and load on particulate mass
emission rate 104
54 Effect of blend on color of stock aerosol 105
55 Typical MRI cascade impactor results for blend firing .... 107
56 Average size distribution for coal, blend, and
dRDF firings during March tests 108
57 Average size distribution for coal, blend,
and dRDF firings during May tests 108
58 Effects of blend and load on aerosol resistivity 109
59 Cell configuration in the portable ESP Ill
60 Effects of blend and load on sulfur dioxide emissions .... 112
61 Effects of blend and load on nitrogen oxide emissions .... 114
62 Effects of blend and excess air on nitrogen oxide
emissions 116
63 Effects of blend and load on chlorine emissions 117
64 Effects of blend and load on fluorine emissions 118
65 Effects of blend and load on hydrocarbon emissions 119
XI
-------
TABLES
Number Page
1 Summary of Previous Co-Firing Tests 3
2 Percentage of Pellets in Hand-Sorted Samples 23
3 Average Properties of Coal on Both an As-Received and
a Moisture-Free and Ash-Free Basis 28
4 Average Properties of dRDF on Both an As-Received
and a Moisture-Free and Ash-Free Basis 31
5 As-Fired Properties for Blends in March Tests 33
6 As-Fired Properties for Blends in May Tests 34
7 Area Density of Pellets Removed from the Cold
Flow Test 40
8 Metal Wastage Rate Data for Eight Specimens 71
9 Heat Balance Summary Based on As-Received Fuel 77
10 Ash Mass Balance 81
11 Chronological Listing of Test Conditions 88
12 Fuel Elemental Composition Normalization Factors
for Adjusting Emissions to a Standard Fuel 101
13 Effect of Blend on Aerosol Resistivity 110
14 Relationship of NOX Concentration and Excess Air
Percentage for Coal, Blend, and dRDF Firings 115
15 POM Concentrations for Coal and Blend Firings 120
16 Trace Metal Concentrations Found in Coal and dRDF Fuel . . . 122
17 Average Heavy Metal Emissions in Ash from
Blend Firing Tests 124
18 Blend Heavy Metal to Coal-Only Heavy Metal Ratios
in Ash Samples 125
xii
-------
ACKNOWLEDGMENT
On behalf of Systems Technology Corporation, the authors gratefully
acknowledge the direction and cooperation of the EPA Project Officer,
Mr. Carlton C. Wiles of the Municipal Environmental Research Laboratory,
Cincinnati, and the timely and valuable support of Mr. Robert Olexsey of the
Industrial Environmental Research Laboratory, Cincinnati.
The authors are also grateful to Mr. James Farrell of the State of
Maryland Department of General Services for his assistance and comments
during the testing at the Maryland Correctional Institute in Hagerstown,
Maryland. In addition, special thanks are given to Mr. Lou Baltozer and his
powerhouse staff for their generous assistance that was essential to the
successful completion of the test program, and to Mr. Argo Kraus, retired
Detroit Stoker Field Engineer, for his resolution of testing difficulties
during the initial firing of the coalrdRDF blends.
Xlll
-------
SECTION 1
INTRODUCTION
FEASIBILITY OF BURNING DENSIFIED REFUSE DERIVED FUEL (dRDF)
During the first half of the 1970's, several short-duration, coal-dRDF
test burns had indicated that it may be feasible to use dRDF as a substitute
for stoker coal. However, while these tests have provoked the interest of
the resource recovery community, they have failed to answer, the questions
most critical to determining this feasibility:
1. Can dRDF be burned within existing environmental constraints?
2. Does dRDF burning have any detrimental effects on a boiler system
or its performance?
3. Is dRDF an economical substitute for coal?
The present study was designed to explore the answers to the.boiler perfor-
mance questions (Section 4) and the environmental questions (Section 5).
The economics of producing dRDF was addressed by the National Center for
Resource Recovery, Inc. (NCRR), under Grant Number 804150.
TYPES OF REFUSE FUELS
Beginning in the 1950's in Europe and in the late 1960's in the United
States, the technical community has had an increasing interest in the fuel
value of urban solid waste. As a result of this interest and the impetus
caused by the energy crisis of 1973, four basic types of solid waste fuels
have been developed: (1) unsorted urban refuse, (2) fluff refuse derived
fuel (fluff RDF), (3) powdered refuse derived fuel (powdered RDF), and
(4) densified refuse derived fuel. Unsorted urban refuse, the oldest type,
is thermally processed in mass-burning incinerators. This type of facility
usually excludes such bulky objects as applicances, rolled carpets, and
furniture. Fluff RDF is produced by shredding mixed urban refuse and passing
the milled material through a series of material separation steps to remove
many of the noncombustibles. Powdered RDF is usually produced by an acid-
embrittling and hot-milling drying operation. Densified RDF is produced from
either fluff or powdered RDF with equipment such as pelletizers, brickquetters
and cubetters. The dRDF is intended for plants which generally burn lump-
sized coal, such as industrial or institutional stoker-fired plants, rather
than plants which burn pulverized coal.
-------
PREVIOUS dRDF TEST PROGRAMS
Perhaps the first summary of the early programs to investigate the
feasibility of burning dRDF with coal was a report by R. T. Stirrup, Fellow
of the Institute of Public Cleansing and Director of Public Cleansing, City
of Southford, England. Published in 1965, this report covers applied research
in England and Europe during the 1956-1960 period. Specifically, it describes
programs which prepared briquettes out of mixed refuse and co-fired the
briquettes with coal to generate steam in short-term tests. One of these
programs generated 3 pounds of steam for each pound of briquettes burned.
Since the early 1970's, several similar short-duration programs were
conducted in the United States. Table 1 lists these programs, and Appendix H
details the results of each.
CURRENT dRDF TEST PROGRAM
Since most of the previous programs had test firings lasting less than
12 hours, the EPA contracted with Systems Technology Corporation (SYSTECH)
to conduct a comprehensive technical and environmental test program to
determine the feasibility of co-firing dRDF and coal in spreader stoker-fired
boilers.
SITE SELECTION
As an integral part of the Environmental Protection Agency (EPA) planning
for the test program, the EPA awarded a grant to NCRR in Washington, D.C., to
produce at least 907 Mg (1000 tons) of dRDF. Consequently, to keep costs
within budget limits, the principal requirement in selecting a boiler plant
for the test program was a site within a reasonable trucking distance of
Washington, D.C.
A second requirement was that the site have a spreader stoker boiler
which would be representative of many similar stoker-fired boilers. In
addition, it should have a. variable grate speed, an adequate fuel storage
capacity, a feeding system, and other facilities readily adaptable to the
testing requirements. A third requirement was a boiler plant with management
sufficiently interested and cooperative to ensure the successful performance
of the test program.
Accordingly, four boiler plants within a 241-km (150-mile) radius of
Washington, D.C., were established as candidate sites. After SYSTECH
engineers visited and evaluated each plant, the Maryland Correctional
Institute (MCI) Boiler House in Hagerstown, Maryland, was selected as the
testing site. This plant met the three requirements as follows:
1. The MCI plant had three Erie City Iron Works boilers rated at
3.1, 7.6, and 9.9 kg/sec (2.5,000, 60,000, and 75,000 Ib/hr) of
1034-kPa (150-psi) saturated steam. The steam generation capacity
was sufficient to ensure continued plant operation if a boiler
should go off-line because of malfunctions due to dRDF burning.
-------
TABLE 1. SUMMARY OF PREVIOUS CO-FIRING TESTS
1.
2.
3.
It.
5.
6.
7.
8.
9.
*10.
*11.
*12.
Location of test
Fort Wayne, Ind.
Municipal Power Plant
Sunbury Steam Electric
Station, Pennsylvania
Power & Light
Piqua, Ohio,
Municipal Power
Plant
Wright-Patterson
Air Force Base
Eugene Water &
Electric Board
University of
Wisconsin
Oshkosh, Wisconsin
Appleton Diversive
Menasha Paperboard
Mill
Chanute Air Force Base
Waupun, Wisconsin
Green Bay, Wisconsin
Stockertown,
Pennsylvania
Test sponsor
(Producer)
National Recycling
Center
(Elo & Rhodes)
Black-Clawson
Fibreclaim, Inc.
Air Force
Black-Clawson
Fibreclaim, Inc.
Sandwell
International, Inc.
(Vista)
Wisconsin Solid Waste
Recycling Authority
(Vista)
Wisconsin Solid Waste
Recycling Authority
(Grumman)
Wisconsin Solid Waste
Recycling Authority
(Grumman)
U.S. Army CERL
(Vista)
Wisconsin Solid Waste
Recycling Authority
Ft. Howard Paper
(Grumman)
Hercules Cement
(Vista)
Date
of test
1972
1975
1975
1975
1974
1976
1976
1976
1975
1976
1976
1975
Type of Vol blend
dRDF coal: dRDF
cubette 3:1
1 1/2" x 1 1/2" x 2"
5/8" pellets
3/8" pellets 1:1
3/8" pellets 1:1
1:2
3/8" pellets
1 1/8" pellets 1:1, 1:3,
and 0 : 1
3/4" pellets
3/4" pellets 3:2
1 1/8" pellets 1:1
0:1
3/4" pellets 20%, 30%, and
40% by heating
value
3/4" pellets 1:3
1:2
1 1/8" & 5/8"
pellets
Amt dRDF Test
fired duration
36 Mg
73 Mg 2 days
20 Mg 7 hr
36 Mg 34 hr
6 hr
19 Mg 1 1/2 hr
19 Mg
36 Mg 8 hr
19 Mg
136 Mg
19 Mg
36 Mg
182 Mg 7 days
* Appendix H does not include a discussion on these tests.
-------
2. The plant could accommodate all of the test equipment and proce-
dures. All flows in and out of the boiler were readily accessible.
The coal silo system could be easily bypassed to permit installing
a temporary coal-dRDF fuel handling system.
3. The plant management expressed sufficient interest and willingness
to cooperate in the test program.
TEST PROGRAM OUTLINE
The test program consisted of four separate field tests: (1) co-firing
coal and 20.9 Mg (23 tons) of dRDF in a series of short runs during
December 1976, (2) a coal base test in January 1977, (3) co-firing coal and
106.1 Mg (117 tons) of dRDF in a series of longer duration tests during
March 1977, and (4) co-firing coal and 127.9 Mg (141 tons) of dRDF also in a
series of longer duration tests coupled with electrostatic precipitator (ESP)
evaluations during May 1977.
Throughout each test, SYSTECH engineers monitored and evaluated the fuel
handling system, the boiler performance, and the stack emissions.
While Section 2 presents a summary and conclusions from the program,
Sections 3, 4, and 5 describe each phase and aspect of the test program and
evaluate the results for each.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
INTRODUCTION
Densified refuse derived fuel may be considered one of the more market-
able products recovered from municipal solid waste. When densified in the
form of pellets, cubettes, or briquettes, it may be handled, transported, and
fed separately or blended with coal and burned in existing-stoker-fired
boilers without major equipment modification.
Over the past few years, several limited tests have tentatively confirmed
that dRDF is a viable coal substitute. While these tests produced positive
results, boiler monitoring and emission tests were performed only in a few
instances.
As a. result of these encouraging tests, the Environmental Protection
Agency sponsored two parallel efforts: one to determine the economics of
preparing dRDF and the second to assess the technical and environmental
implications when the fuel is used as a coal substitute. This report presents
the technical and environmental evaluation of co-firing tests conducted at
the Maryland Correctional Institute power plant in Hagerstown, Maryland. The
dRDF used in these test were pellets prepared by the National Center for
Resource Recovery under a research grant.
TEST OBJECTIVE
The objective of the study was to determine, characterize, and demon-
strate the technical and environmental feasibility of combusting dRDF with
coal in spreader stoker-fired boilers. The tests were to be conducted in a
stoker-fired boiler which would have a rating between 3.1 and 25.1 kg/sec
(25,000 and 200,000 Ib/hr) of steam and would be within 241 km (150 miles) of
NCRR in Washington, D.C. The study was to specifically address fuel handling,
boiler performance, and environmental effects when dRDF pellets, cubettes,
and briquettes were fired with coal.
SITE SELECTION
After all the candidate boiler plants within 241 km (150 miles) of
Washington, D.C., were surveyed, the MCI plant was selected because it met
the above described criteria most satisfactorily. This plant had three Erie
City Iron Works boilers, one each rated at 3.1, 7.6 and 9.9 kg/sec (25,000,
60,000, and 75,000 Ib/hr) of 1034 kPa (150 psi) saturated steam.
-------
TEST DESIGN
The test was designed to combust 258.5 Mg (285 tons) of dRDF during
236 hours of firing various blend ratios of coal:dRDF. These tests were
conducted in a series of burns with volumetric coal:dRDF ratios of 1:1, 1:2,
and 0:1 and with test durations ranging from 20 minutes to 132 hours. The
series of coal:dRDF tests were preceded and followed by a coal-only test with
duplicate conditions. Also, because of required plant steam demand, all of
the tests were conducted at only 30 to 55 percent of boiler design capacity.
The initial tests were designed to ensure that dRDF could be safely burned
without jeopardizing the boiler's capability of meeting the steam demand.
These tests included monitoring the performance of the spreaders while
introducing dRDF into an unfired boiler and a series of short-duration burns
to determine the combustion properties and the boiler performance. Subsequent
field tests involved a study of (1) the material handling characteristics of
dRDF, i.e., storing, conveying, feeding out of bunkers, etc.; (2) boiler
performance, i.e., grate speeds, underfire and overfire air requirements,
steam production, spreader limitations, boiler efficiency, flame impingement,
slagging, fouling, clinkering, combustion gas analysis, etc.; and (3) environ-
mental performance, i.e., particulates, gaseous emissions, and trace organic
and inorganic emissions. Since only pelletized dRDF was available, testing
with cubettes and briquettes was not conducted.
TEST RESULTS
Material Handling
Throughout the field testing, 259 Mg (285 tons) of dRDF were received,
stored, and conveyed to the boiler without major difficulty or malfunction of
the fuel handling system. Difficulties were limited to dusting and pellets
hanging up in feed hoppers.
Pellet Storage—
At successive periods, the pellets were stored in 20-cubic-yard open
containers, in a warehouse (uncovered), and on an outdoor concrete slab
(tarpaulin covered).
Twenty-Cubic-Yard Containers—Since the pellets received during the
winter tended to steam, they eventually froze into a solid mass. Minimal
rodding, however, broke up the frozen pellets, and subsequent handling further
restored the individual pellet integrity without significant degradation to
the pellet.
Warehouse—Approximately 125 Mg (140 tons) of pellets were stored in an
unheated warehouse for 2 months. With the exception of mild odors and some
fungus growth, this storage proved to be the most effective in maintaining
pellet integrity over extended periods of storage time. Since the depth of
the piles was limited to 1.8 m (6 ft), increases in temperature due to
composting effects were negligible, and the pile temperature stabilized at
60°C (140°F).
-------
Open Slab—The pellets stored in the warehouse were subsequently moved
to an outdoor storage area. The pellets were stored in 1.8-m (6-ft) piles
on an outdoor slab and covered with a tarpaulin. Moisture accumulation under
the tarpaulin caused pellets at the top of the piles to deteriorate and cake.
Also, some pellets sustained minor damage, i.e., swelling and roughened
edges, because of water infiltration onto the slab.
Pellet Feeding—
The pellets were conveyed to the boiler feed hopper by a temporary fuel
blending and handling system. The coal and dRDF were volumetrically blended
in the various ratios by separately feeding coal and pellets from two hoppers
to a common bucket elevator which subsequently conveyed both coal and pellets
to a weigh lorry. The fuels were volumetrically blended by filling the feed
conveyors to capacity (level with the conveyor flights) and operating the
coal and pellet feed conveyors at speeds commensurate with the desired blend
ratio, i.e., a 1:2 coal:dRDF blend would require the dRDF conveyor to run at
twice the speed of the coal conveyor. Although this feeding system generally
worked well, it had some difficulties with deteriorated pellets. As the
amount of fines increased (due to excessive handling), the pellets would not
flow from the feed hoppers without rodding. These fines also caused consid-
erable dusting throughout the plant. This dusting was subsequently controlled
by installing a steam jet at the conveyor transfer point.
Pellet Properties—
The 1/2- x 3/4-inch pellets had an average bulk density of 425 kg/m3
(26.5 lb/ft3) and ranged from 400 to 466 kg/m3 (25 to 29 Ib/ft3). The
material density for intact pellets ranged from 1.22 to 1.34.x 1Q3 kg/m3
(76 to 84 lb/ft3) while that for deteriorated pellets averaged
0.98 x 103 kg/m3 (61 lb/ft3). The as-received properties were 12.10 to
15.12 MJ/kg (5200 to 6500 Btu/lb), 20 to 29 percent ash, 9 to 10 percent
fixed carbon, 12 to 13 percent moisture, 50 to 57 percent volatiles, and
1142°C to 1152°C (2088° to 2105°F) hemispheric reducing fusion temperatures.
NCRR projected that further processing of the shredded refuse to remove glass
and other inerts could produce a pellet with a heat content of 19.1 MJ/kg
(8200 Btu/lb) and an ash of 10 to 12 percent.
Boiler Performance
Spreader-Feeder Performance—
In a cold flow run (furnace not fired) to test the fuel distribution of
the Hoffman Combustion Engineering spreader-feeders, two different sized
pellets were distributed onto the grate: 1/2 x 3/4 and 1x2 (diameter x
average length in inches) pellets. Because the Hoffman spreader throat has a
maximum size restriction of 1 1/2 inch, the 1-inch-diameter by 2-inch-long
pellets tended to hang up and slug-feed the furnace. However, the
1/2-inch-diameter by 3/4-inch-long pellets generally were handled and fed well
with the larger pellets traveling to the rear of the grate and the fines
falling close to the spreader. During the initial combustion tests with
100 percent pellets, the spreader had to be adjusted to decrease the pellet
trajectory in the furnace by approximately 0.3 m (12 in.). In addition, the
maximum steam load that the boiler could carry was 6.8 kg/sec (54,000 Ib/hr)
or 70 percent of design capacity. This derating is the direct result of
volumetric limitations of the spreader feeder.
-------
Combustion of dRDF—
While operating at these partial boiler capacities, the combustion of
the various blends of coal:dRDF was generally as good as the combustion of
coal only. However, when the dRDF substitution was increased, the flame
length, intensity, and volume of the fireball increased correspondingly. As
the intensity of the fireball grew, the flame temperature, measured about
1.5 meters above the center of the grate, also increased from 1200°C (21928F)
for 100 percent coal firing to 1240°C (2264°F) for 100 percent dRDF firing.
When test firing the 1:1 blend and 100 percent dRDF, the fireball was
kept well away from the rear wall of the furnace by adjusting the overfire
air. Once these jets were adjusted for minimum smoke and maximum efficiency
for coal-only burning, they continued to meet the mixing and wall protection
requirements when burning blends and 100 percent pellets. As viewed from the
side of the furnace when firing both pellets and blends, the bed was well
burned out by the time it approached the front ash pit. The flame pattern
above the grate indicated that the fuel bed was maintaining proper porosity
with minimum clinkering or agglomeration. This operation was achieved when
burning a double screened, high ash fusion temperature 1370°C (2498°F) coal.
With little attempt to optimize the system, a 10 to 12 percent carbon dioxide
content in the flue gas at the boiler outlet was readily obtained.
Fouling—
Inspection of the furnace interior after the tests revealed that a light
coating of ash had accumulated on the tubes. Also, an interim boiler inspec-
tion revealed that one-third of the rear wall of the boiler was covered with
slag. This slagging was subsequently eliminated when a spreader was adjusted
to prevent pellet impingement on the rear wall. Subsequent inspections of
the boiler after being on-line for 8 days revealed that the slag had sloughed
off.
Clinkering—
During the initial tests frequent clinkering occurred on the grate when
firing a 1:1 blend. This clinkering was subsequently attributed to a low
hemispheric fusion temperature, 1204°C (2200°F), of the coal. When the
coal was changed to one having a higher ash fusion temperature, 1373°C
(2500°F), the clinkering stopped. While coal with low fusion temperatures
clinkered, the 100 percent pellets, which had a low fusion temperature of
1151°C (2103°F), did not clinker. This observation is valid within the
constraints of the test conditions, i.e., a 4-hour test burn at a boiler
capacity of 30 percent of rated design capacity and 100 to 130 percent excess
air.
Corrosion—
Eight clamp-on corrosion test specimens were installed on the supply
tubes of the rear screen wall 1.5 m (5 ft) above the fuel bed. After
478 hours of exposure to various blend and coal-only firings, normal wastage
(less than 5 mils per year) was evident on all specimens except the
1018 specimen. This test specimen, which had extremely high metal wastage,
was mounted in the area where the heavy slagging occurred because of the
maladjusted spreader.
-------
Boiler Operation—
Air Flow Controllers—During periods of load shedding, the fuel bed was
more susceptible to clinkering when coal:dRDF blends were fired. The
clinkering was eliminated by biasing the underfire air control to supply
approximately 70 percent excess air to the fuel bed. On the basis of these
results, boilers which are tight (minimum air leaks) should be capable of
satisfactorily burning coal:dRDF blends with 50 percent excess air.
Oscillating Grate Dwell-Shake—Throughout the test, the duration and
amplitude of the grate shake pulse was adjusted to advance the fire line at
the rear of the boiler approximately 15.2 cm (6 in.) per excitation. In all
advances, the pulse frequency was the principal controlling variable. At
40 percent load, the frequency of the pulse decreased from 11 minutes for
100 percent coal to 3 minutes for 100 percent pellets. When firing a blend,
the pulse duration tended to increase because the bulk density of the blend
ash was less than that of the coal ash.
Ash Handling—
Bottom Ash—The sieve analysis of. bottom ash samples taken during blend
firings indicated that conventional pneumatic ash handling systems should be
able to handle the bottom ash from blend firings as well as they do the
bottom ash from firing coal-only. On a few occasions fire occurred in the
bottom ash hopper during blend firing. Rodding of the clinkers in the ash
hopper revealed that the ash had a taffy-like consistency. Under similar
conditions, when firing coal only, the bottom ash was much easier to break up
by rodding.
The bottom ash removal system malfunctioned only during 100 percent
pellet firing. The bottom ash was so fine that it would not de-entrain
properly in the cyclone. The particles, which had been wetted by the steam
in the vacuum ejector, passed through the cyclone and eventually plugged the
ejector.
Dust Collector Ash—As dRDF was substituted for coal, the fly ash
particles became finer. The size of the particles in the dust collector
ranged from 200 micrometers for 100 percent coal firing to 90 micrometers
(sizes at the 50th percentile) for 100 percent pellet firing. Also, the
carbon content of the fly ash decreased significantly with increasing dRDF
substitution. The primary factor contributing to this occurrence was the low
fixed carbon content of dRDF (12 to 18 percent) compared to coal which had
65 to 85 percent fixed carbon.
Mass and Energy Balance—
Mass Balance—The mass balance indicated that an unusually large amount
of the fuel ash had accumulated in the collectors. Subsequent analysis of
the collector fly ash revealed that the high collector ash weights were due
to the presence of 50 to 70 percent carbon in the collector ash. Also, since
90 percent of the particles exiting the boiler were greater than 50 micro-
meters in diameter, these large particles were removed by the cyclone. The
carbon content of the bottom ash varied from 2 to 10 percent, and the carbon
content of the stack fly ash (not captured by the cyclone) was 30 to
40 percent. The analysis of the stack fly ash as a function of blend revealed
that its carbon content decreased as the dRDF substitution increased.
-------
Efficiencies—During the testing the boiler efficiencies were extremely
low, namely 55 to 60 percent. When the boilers were installed in 1963, they
produced a boiler efficiency of 79 percent at an excess air of 34 percent and
rated design capacity. These low efficiencies were primarily due to the low
boiler loads (less than 30 percent of rating), high excess air (80 to
115 percent), and extremely high losses of combustibles in the refuse (up to
25 percent). The analysis of the results indicated that the coal-only and
blend firing efficiencies had no discernable differences. However, this
observation may be unique to the boiler installation at MCI since the large
amount of unburned combustibles removed by the collectors is certainly an
anomally to expected boiler performance.
Environmental Performance
Data Normalization—
Since the co-firing tests spanned a 6-month period, the properties of
the coal and dRDF burned in the successive tests varied widely. The boiler
excess air in the tests also varied considerably. To eliminate the effects
of these variables, all the emissions data were corrected to 50 percent
excess air and then normalized to a reference coal and dRDF composition. All
the co-firing emissions data were then statistically compared with a coal-
only baseline plot of emissions concentration versus boiler load. If the co-
firing emissions data fell outside the 90 percent confidence limits for the
coal-only emissions data, they were considered to be significantly different.
Particulate Emissions—
Mass Concentration—The particulate mass concentration (grams/standard
cubic meter, g/scm) in the 1:1 and 1:2 blend firings was slightly less than
in the coal-only firing. However, the reductions were not significant at the
90 percent confidence level. The mass flux at a 40 percent boiler load for
1:1 and 1:2 blend firings averaged 0.45 g/scm corrected to 12 percent C0a.
The coal fired during these tests was a nominal size of 1 1/4 * 1/4 inch with
a maximum of 30 percent passing through a nominal 1/4-inch screen.
Particulate Size—As more dRDF was substituted for coal, the particulate
diameter decreased. In the May tests, the diameters for the coal-only firings
were 3 micrometers, and those for the dRDF-only firings were 0.8 micrometer
(at the 50 percentile point).
Particulate Resistivity—Because of the unusually high carbon content in
the fly ash during the coal-only firing, the resistivity was generally less
than 106 ohm-cm. As dRDF was substituted for coal, the carbon burnout in the
fly ash improved, and the resistivity increased to 2 x 10*" ohm-cm for the
1:1 blend firing.
Electrostatic Precipitator Performance—A mobile 5-cell electrostatic
precipitator (owned by the EPA Industrial Environmental Research Laboratory
in Research Triangle Park, North Carolina) was evaluated to determine the
effect of dRDF on ESP performance. The evaluation, however, could not be
conclusive because of short circuiting within the ESP. This short circuiting
developed when the dielectric blocks from which the electrodes were suspended
became coated with high carbon content aerosol. Appendix E discusses the
test results.
10
-------
Opacity—As dRDF was substituted for coal, the overall opacity of the
plume reduced significantly. At 40 percent boiler load, the opacity for
coal-only firing was 16 percent (based on a 1.22-m (4-ft) diameter stack).
At the same boiler load and excess air, the opacity was only 10 percent for
dRDF-only firing.
Gaseous Emissions—
SO2—Since the dRDF had a sulfur content of 0.4 percent, the S02 emissions
reduced with increasing dRDF substitution. The decrease was particularly
significant for the 1:2 and 0:1 (100 percent dRDF) blend firings. At
40 percent boiler load and the same excess air levels, the S02 dropped from
1300 ppm for coal-only firing to 250 ppm for dRDF-only firing. This reduction
in S02 follows exactly the reduction in sulfur content of the fuel (see
Table 6 in Section 4).
NOX—There were no significant changes in NOX as dRDF was substituted
for coal. At 40 percent boiler load and the same excess air levels, the NOX
concentrations ranged from 200 to 350 ppm with either fuel.
Chlorine—As dRDF was substituted for coal, the chlorine in the emissions
increased from 60 ppm for coal-only firing to 650 ppm for dRDF-only firing.
There appeared to be no appreciable change in chlorine concentrations as the
load changed from 20 to 50 percent of design capacity.
Fluorine—Fluorine concentrations also increased with increasing dRDF
substitution. However, the concentrations were very low, e.g., 8 ppm for
coal-only firing and 12 ppm for dRDF-only firing at a 40 percent boiler load
and constant excess air conditions.
Hydrocarbons—There were no significant changes in hydrocarbon emissions
when substituting dRDF for coal. At a 40 percent boiler load, the total
hydrocarbons ranged from 10 to 25 ppm. As the boiler load increased, the
hydrocarbon concentrations decreased significantly. This reduction is
probably attributable to the improved conbustion conditions at higher
boiler loads.
Trace Organic and Inorganic Emissions—
Organic Emissions—The overall emissions of polycyclic compounds for
coal-only and blend firings were well below the threshold limits proposed by
the National Academy of Science. Typical measured values were: 543 ng/m3
for anthracene/phenanthrene, 100 ng/m3 for methyl anthracene, and 137 ng/m3
for fluoranthene (all at 1:1 blend firing).
Inorganic Emissions—The analysis of the fly ash for trace metals
revealed that relative to coal-only firing the blend firing enriched some
metals but reduced others. For example, when firing a blend of 1:2 coal:dRDF,
the amount of lead in the stack particulates was 8217 ug/m3. This compares to
a lead concentration of 230 yg/m^ for coal-only firing. While dRDF was the
main contributor of Br, Mn, Pb, and Sb, coal was the primary source of As,
Ni, and V.
11
-------
Several elements, particularly As, Ga, Na, and Sb, tended to concentrate
in small particles. In addition, as the dKDF substitution increased, both
the solubility of the fly ash and the amount of small-size particulates in
the respiratory range increased. Consequently, each of these effects pose
potential hazards from (1) respiration of heavy metals associated with aerosols
and (2) leaching of high levels of heavy metals in landfills.
CONCLUSIONS
Fuel Handling System
Pelletized refuse can be stored, handled, and blended with coal in
conventional material handling equipment provided the pellet structural
quality is maintained. However, if pellets deteriorate because of excessive
handling and exposure to rain, they will hang up in bunkers and will generate
considerable dust as they are conveyed throughout the plant.
Boiler Performance
Boiler performance was evaluated at reduced load conditions, i.e., 30 to
55 percent of design capacity. At these boiler loads, some minor operational
difficulties were encountered with slagging and clinkering throughout the
testing. These difficulties were controlled by making simple adjustments.
Typical adjustments included (1) biasing the air controls to higher excess
undergrate air levels to prevent clinkering in the fuel bed during load shed,
(2) properly adjusting the spreader feeders to prevent dRDF impingement on
the side and rear walls of the furnace, and (3) setting the grate dwell and
pulse intervals to compensate for the reduced ash bulk densities when the
blends were fired. The boiler operation was restricted only when the spreaders
and ash handling system became capacity limited during dRDF-only firing. The
increasing dRDF substitution resulted in (1) improved carbon burnout in the
fly ash, (2) decreased plume opacity, and (3) improved low-load performance
(more than a 4:1 turndown without excessive smoking).
Environmental Performance
Compared to the particulate emissions from coal-only firing, the emissions
from the blend firing decreased slightly in particulate concentration, dropped
significantly in particulate size and stack opacity, and had resistivities
within the range for satisfactory ESP performance. Of the gaseous emissions,
SOx decreased and chlorine increased, both significantly. Analysis of the
trace inorganic elements in the fly ash when dRDF was fired revealed that
concentrations of Pb, Cd, Mn, Zn, and Sb were significantly higher than other
elements. Since the solubility of the fly ash increased with increasing dRDF
substitution, landfilling the dRDF residue could result in hazardous levels
of heavy metals in the leachate.
Summary
While the test was limited to firing at reduced boiler loads, the
preliminary results from these field tests indicate that coal and dRDF can be
co-fired at volumetric coal:dRDF ratios up to 1:2 with only minor adjustments
to the boiler and fuel handling systems. Subsequent testing should address
the long-term effects of corrosion and erosion on boiler tubes.
12
-------
SECTION 3
MATERIAL HANDLING
The primary factors affecting the flow of a solid fuel through a material
handling system are the following fuel properties: (1) size distribution,
(2) moisture content (inherent and free), (3) bulk and particle density,
(4) bulk compressibility, and (5) configuration and roughness. Previous
tests at Chanute Air Force Base and the Pennsylvania Power & Light Company
revealed that excessively handled dRDF becomes fluffy with increased fines
which then promote high angles of repose and bridging in bunkers and hoppers.
Since the handling and storage of the dRDF pellets at MCI required five
separate fill/dump operations, similar dRDF deterioration and bridging were
anticipated.
Therefore, to ensure that the plant would continually meet its steam load
requirement and not have to be shut down because of blended fuel bridging in
the bunker, the dRDF and coal blending and handling system was installed
independently of the main coal feed system. Since the stoker spreader
feeders might disperse the dRDF and coal blend differently than coal alone,
the stoker spreader operation was monitored to characterize feeding behavior
and dispersion of fuel on the grate for coal, a blend of coal and pellets,
and pellets alone in a cold, idle furnace.
FUEL MECHANICAL PROPERTIES
Since the size and density of both the coal and the dRDF directly
influence the performance of the fuel handling system, tests were conducted
to quantify these properties. The following paragraphs discuss these pro-
perties.
Coal
Hoffman Combustion Engineering recommends that the coal fed to its
spreader stoker have a nominal size of 1 1/4 x 1/4 inch with a maximum of
40 percent passing through a nominal 1/4-inch screen. A sieve analysis of
the December coal showed that the distribution was 100 percent less than
3/4 inch with 70 percent passing through a 1/4-inch screen. Test firing of
this coal showed that the high fraction of fines impeded proper plant
operation, i.e., the fines caused furnace pressure pulsations as they entered
the furnace, plugged the grate and air ports (larger size fines), and over-
loaded the fly ash system (smaller size fines). In view of the poor boiler
system performance with the December coal, a double screened stoker coal had
to be acquired before meaningful co-firing tests could be started.
13
-------
Accordingly, arrangements were made with the Delta Coal Company (the
supplier for MCI) to provide specification coal for the March and May tests.
Although not standard fuel for the Hagerstown plant, the coal acquired was
Swickley seam coal, a commercially available spreader stoker coal. The coal
was obtained from the supply prepared for an industrial spreader stoker plant
about 64 km (40 miles) from MCI.
The coal sieve analysis in Figure 1 compares the coal for the December
and May field tests. The May size distribution is also representative of the
coal for the March field tests. The January coal had a size distribution of
100 percent less than 1/2 inch with 70 percent passing through a 1/4-inch
screen with an occasional 76-mm (3-in.) piece. The March and May coal was
double screened, 1 x 0 Stoker coal with 30 percent passing through a 1/4-inch
screen. The bulk density of the March and May coal averaged 777 kg/m3
(48.5 lb/ft3), and the material density ranged between 1.35 and
1.43 x 103 kg/m3.
l-ROSLIN—flAMMLER EXPONENT
a DECEMBER
A MAY(AVG.)
0» 05 075 •
PARTICLE SZE D. WOCS
Figure 1. Comparison of the size distributions for the
December and May coals.
14
-------
Pelletized dRDF
The determination of the optimum pellet size was based primarily on
obtaining a pellet that would fall in the middle of the recommended coal
particle size range. The rationale for such a size was (1) to prevent the
pellets from segregating from the coal during handling, (2) to ensure that
the pellets would flow through the unmodified spreader without jamming, and
(3) to spread the pellets onto the grate in a pattern similar to that of
coal. Cold flow tests, i.e., feeding blends of coalrdRDF into the furnace
without combustion or air flow, were conducted on nominal 1/2-inch-diameter
by 3/4-inch-length and 1-inch-diameter by 2-inch-length pellets.
The test results indicated that the 1/2-inch-diameter pellets flowed
satisfactorily through the material handling equipment and the spreader
feeders. The 1-inch-diameter pellets tended to hang up in the spreader
feeders and to slug feed the furnace. Since NCRR had a die available to
produce the 1/2-inch-diameter pellets and they performed satisfactorily
throughout the cold flow tests, only the 1/2- x 3/4-inch pellets were used in
the combustion tests. Throughout the combustion tests, the pellets were
sampled from the furnace hopper for length analysis. Figures 2 and 3
illustrate the length distribution data.
The March test pellets were stored in a warehouse and had a lesser
deterioration and data spread than the May test pellets which were stored on
an open slab. These discrepancies could be attributed to differences in the
storage conditions and/or the original pellet characteristics. The pellet
bulk density averaged 425 kg/m3 (26.5 lb/ft3) and ranged from 400 to 466 kg/m3
(25 to 29 lb/ft3). Bulk density was determined by filling and weighing a
1-ft3 container and then subtracting the container tare weight. The material
density for intact pellets ranged from 1.22 to 1.34 x 103 kg/m3 while that
for deteriorated pellets averaged 0.98 x 103 kg/m3. Material density was
determined by weighing a pellet and then determining the amount of volume
displaced by the pellet when immersed in a liquid.
Figure 4 shows the bulk density of various coalrdRDF blends. The fact
that the measured bulk density of the blend was higher than the arithmetic
bulk density can be attributed to the dRDF particles filling the interstices
in the coal and vice versa.
STORAGE OF dRDF
Throughout the pellet storage SYSTECH engineers were able to avoid
spontaneous combustion by following the storage procedures for lignite.
These procedures required keeping the storage period to a minimum and limiting
the pellet piles to maximum depths of 1.8 to 2.4 m (6 to 8 ft).
Each of the field tests required 91 to 125 Mg (100 to 140 tons) of dRDF.
Since the supply from NCRR was generally 11.8 to 14.5 Mg/wk (13 to 16 tons/wk),
the deliveries had to be accumulated for 8 to 10 weeks before each test could
be conducted. The pellets were transported from the NCRR test facility in
Washington, D.C., to the Hagerstown, Maryland, plant in tarpaulin-covered,
20-yd3, open roll-on containers. With deliveries from December 1976 through
May 1977, the pellets acquired totaled 255 Mg (281 tons).
15
-------
PABTXXE SOE D. INCHES
Figure 2.
Characteristic length distributions for
pellets burned in March.
a MAY 10 1:2
• MAY 11 1:2
t MAY 13
o MAY 13 1:1
A MAY 12 1:1
• MAY 10 1:0
PWTKXE SZE 0, HOtS
Figure 3. Characteristic length distributions for
pellets burned in May.
16
-------
MEASURED DENSITY
\ \ /- ARITHMETIC DENSITY
V
\
% d-RDF ( by volume)
Figure 4. Comparison of arithmetic (interpolated) and
field-measured blend bulk densities.
Since the slab beside the fuel handling system at MCI was too small to
accommodate the pellet accumulations, the pellets were stored as follows:
During December 1976 and January 1977, the pellets were kept in the roll-on
containers until they were removed for burning. During February and March of
1977, the pellets were stored in an unheated warehouse within a residential
community. With the advent of warmer weather and the possibility of offensive
odors reaching the nearby homes, the pellets were subsequently moved to an
open slab about a half mile from the MCI power plant and stored under a
tarpaulin cover during April and May of 1977. The following sections discuss
each of the storage conditions and their effects on the pellets.
17
-------
Open Containers
When the pellets were stored in the roll-on containers during December
1976 and January 1977, they occasionally froze near the edges. Since the
initially received pellets tended to steam, evaporating moisture trapped
under the tarpaulin eventually froze the pellets into a solid mass. However,
minimal rodding broke the mass into blocks which could flow freely. Subse-
quent movement broke up the blocks into individual pellets. Problems were
also encountered with trash remaining in the containers from previous garbage
loads. The contaminating trash included such materials as cans, cardboard
boxes, and wood and metal pieces which had to be removed from the fuel to
prevent their jamming the conveyors and/or the fuel spreader. While this
problem was minimal (only 10 percent of the containers had such waste), it
does demonstrate that care must be exercised in this area.
Warehouse
When the pellets were stored in the warehouse during February and
March, they were dumped from the containers and pushed by a front-end loader
into 1.8-m (6-ft) deep piles in the warehouse. In addition to mild odors,
fungus growths appeared on the peaks of the piles or wherever there was a
moisture vent. There did not appear to be any rodent or insect damage to the
stored pellets. After the stored pellets were removed, the warehouse was
easily cleaned.
Figure 5 indicates the extent of the pellet length reduction due to the
warehouse storage. However, Figure 6 indicates that this reduction cannot be
attributed to a moisture loss. A reasonable cause for the reduction could
be the additional pellet handling and/or differences in the NCRR production
procedures.
Remote Slab
»
When the pellets were stored during April and May on the open concrete
slab at the MCI power plant, they were dumped from roll-on containers and
then pushed by a front-end loader into 1.8- to 2.1-m (6- to 7-ft) deep piles
on the slab. Although a plastic tarpaulin protected each pile from the
weather, moisture accumulation under the tarpaulin caused pellet deteriora-
tion and caking (or capping) on the tops of the piles. A similar capping
occurred on the piles stored in the warehouse. This caking consisted of a
5- to 10-cm (2- to 4-inch) thick layer in which the mechanical integrity of
the individual pellets was greatly reduced. In addition, because of poor
slab drainage, run-off water infiltrated some of the piles and deteriorated
pellets on the slab surface. Such pellets swelled, and their initially
smooth sides became rough. Figure 7 compares a deteriorated pellet with a
good one. However, the deteriorated pellets were relatively few, and they
were still usable although their rough sides impeded their flow out of
storage bins. Rodding was required to assist the flow of these pellets from
the storage bin.
18
-------
2 8
a
I
LU
JT8
i
£
OJ O_
3/23 — STORED UNDER TARP
^ STORED IN WAREHOUSE
3/29 V— & HANDLED 2 EXTRA
3/311 TIMES
^
1"
NOMINAL PELLET LENGTH
Figure 5. Comparison of pellet length distributions for
different types of storage.
INCREASING AGE
o_
K
o
UJ
CO
I °-
UJ
cr
i-
CO
O
liN.v/nc«oiwva «V3C
JANUARY PELLETS
X
r >
i
• m^*
APRIL PELLETS
jl
/ ^\
IT T1
>
T
. _
I
i
DIFFICULTY W/FLOW •
PELLETS GOT WET
FINES VERY HIGH
3/23 3/29 5/9 5/10 5/11 5/12 5/f3 5/14
DAY OF BURN
Figure 6. Effect of storage duration and method on pellet
moisture content.
19
-------
Figure 7. Comparison of a deteriorated pellet (left) and
a well-formed pellet (right).
On-Site Slab
Some of the 35 Mg (40 tons) of dRDF remaining from the January and March
tests and stored in the warehouse were transferred to the slab beside the
fuel handling system at MCI for subsequent use during the May tests. These
pellets were stored on the slab in a 2.4-m (8-ft) deep pile. The pile was
periodically monitored with a thermocouple imbedded 1.5 m (5 ft) down from
the top of the pile. After the pile temperature rose to 60°C (140°F), it
dropped as the pile dried out to ambient temperature by the time of the May
test.
FUEL HANDLING SYSTEM
Description
Coal is delivered to the Hagerstown plant in trucks which dump through a
grizzly grate onto a drag chain feeder to the bucket elevator inlet. The
bucket elevator delivers the coal to a square concrete silo. A "weigh lorry"
on rails over the firing aisle and located above the stoker feeder hoppers
transports the coal from the silo to the stoker feed hopper and then weighs
and dumps the coal.
As noted previously, since the blended dRDF and coal fuel could have
bridged the silo and thereby caused a plant shutdown, a temporary fuel
blending and handling system was installed to bypass the silo during the
field tests.
20
-------
As shown in Figure 8, the system included a canopy-covered slab with
storage space for about 27 Mg (30 tons) of both coal and pellets. The
conveying equipment consisted of two 8-yd3, pyramid-bottom bins that fed the
fuel into cleated, pin-pan, Z-belt conveyors which in turn emptied the fuel
into the feed hopper of a bucket elevator. Orbital vibrators were installed
on the coal and pellet bins to facilitate free fuel movement. The bucket
elevator had a straight back with the other three sides inclined at 45 degrees
to feed the fuel into a 30-degree, 0.2-m (8-in.) square chute.
The starters on the Z-belt conveyors and the bucket elevator were inter-
locked. The activation of a single start button began the blend feed. The
feed from the bins could be stopped from a station next to the weigh lorry or
from outside the building next to the fuel bins. Normally the fuel feed was
stopped and the elevator allowed to empty before a complete shutdown.
The coal and dRDF were blended to the various coalrdRDF ratios by
changing drive pulleys to vary the speed of one of the Z-belt conveyors. The
volume of fuel loaded per foot of conveyor was maintained by scraping the
fuel load level with the top of the conveyor flights.
SWIVEL CHUTE
Z-BELT CONVEYOR (TYP)
FRONT VIEW
SIDE VIEW
Figure 8. Front- and side-view drawings of the
temporary fuel handling system.
21
-------
The temporary blending and handling system could transfer 2450 kg
(5400 Ib) of a 1:1 coal and dRDF blend into the weigh lorry in approximately
30 minutes or at a rate of 1.39 kg/hr (3.1 Ib/sec). The limiting factor on
this rate was the speed of the Z-belt conveyors.
Operation
Two crews operated the fuel handling system. One crew, consisting of a
truck driver and a front-end loader operator, retrieved the dRDF from storage
and delivered it to the test site. The second crew consisted of two men at
the test site, one was a front-end loader operator and the other was a
helper. This crew filled the coal and dRDF bins after each loading to the
weigh lorry.
Performance
In general, the temporary fuel blending and handling system performed
well throughout the test program. The mixture of 1 x 0 coal with less than
30 percent fines and 1/2-inch-diameter x 3/4-inch-long pellets fed well and
required less hoeing in the stoker feed hopper than coal alone. During the
January and March field tests, the blend fed much easier than coal alone.
However, near the end of the May field tests, the pellets would "rat-hole"
(assume a high angle of repose) in the 8-yd3 pyramid feed bins unless they
were rodded periodically. The "rat-holing" was due primarily to pellets
whose sides had become fluted because of water damage. Consequently, such
pellets tended to interlock and bridge.
Lengths of pipe or chunks of blacktop (picked up by the front-end
loader) intermingled with the pellets and jammed the Z-belt conveyors once or
twice each 8-hr shift. While thorough cleaning of the transportation equip-
ment and careful operation of the front-end loaders would have prevented the
inclusion of these materials, a grizzly grate and a magnet, such as used in
a coal plant, would have been a more practical means of preventing the
introduction of such refuse into the fuel flow.
At the outset of the field tests, dust released during coal and dRDF
fuel transfer from the bucket elevator chute to the weigh lorry was excessive.
While a shroud over the chute initially contained the dust, the subsequent
motion of the bucket elevator blew out the loose dust. The dust had a lint-
like consistency and settled throughout the plant. Consequently, a hood was
installed over the weigh lorry and coupled to what proved to be an inadequate
exhaust fan. Although this installation provided some relief, the dust was
still excessive until a steam jet was installed under the hood. This steam
jet wetted the dust particles and adequately suppressed the dust from
spreading throughout the plant.
To quantify the blending system performance, five fuel samples were
taken from the feed trough (which filled the lorry) on each of two represen-
tative field test days. The samples, approximately 4.5 kg (10 Ib) each, were
hand-sorted into coal and pellets, and the two sorts for each sample were
weighed. The remaining dust (primarily dRDF) from the hand sorts was weighed
with the pellets. The consistency of the weight percentage of the samples
22
-------
evidenced the repeatability and homogeneity of the coal:dKDF blending. The
relative standard deviation of the pellet fraction was less than 5 percent in
these samples. Table 2 summarizes the sampling results.
TABLE 2. PERCENTAGE OF PELLETS IN HAND-SORTED SAMPLES
Blend
Time of Sample
% Pellets (by weight)
9 May 1:1
1:1
1:1
1:1
1:1
10 May 1:2
1:2
1:2
1:2
1:2
10:15
11:30
12:50
2:00
3:15
9:15
10:10
12:30
1:35
-
0.54
0.46
0.44
0.47
0.53
0.66
0.63
0.65
0.62
0.64
Alternative Fuel Blending Method
Although the fuel blending and handling system functioned as designed,
some preliminary tests had to be performed before the drives on the Z-belt
conveyors could be set at the proper speed ratios. For these initial tests,
the operator of the front-end loader alternately loaded coal and dRDF into
the feed hopper of the bucket elevator. When the hopper was full, the bucket
elevator was started, and the gate to the feed chute was opened. The layered
fuel flowed satisfactorily from the hopper. Moreover, samples of the blend
taken from the weigh lorry were consistently mixed.
Consequently, the alternate coal and dRDF layering may be considered as
effective in blending the two fuels as the proportioning conveyor mixing
system.
23
-------
Summary
Except for flow problems experienced with deteriorated pellets, the
field tests demonstrated that conventional equipment can adequately blend
coal and dRDF in various coalrdRDF ratios and can handle both the coal-only
and the blended fuel. Such equipment, however, may require provisions to
suppress dusting at conveyor transfer points.
24
-------
SECTION 4
BOILER PERFORMANCE
BOILER DESCRIPTION
The MCI boiler plant in Hagerstown consists of three 1034-kPa (150-psig)
Erie City Iron Works boilers. Their design steam ratings are 9.9, 7.6, and
3.2 kg/sec (78,500, 60,000, and 25,000 Ib/hr). Figure 9 shows a cross section
of a typical boiler. Each unit is equipped with a Hoffman Combustion
Engineering "Firerite" spreader-feeder with an appropriate number of spreader-
feeders to distribute the lump fuel in the furnace. The large coal pieces
that do not burn in suspension are consumed on the surface of the front ash
discharge vibrating grates.
The Erie City Iron Works boilers have tube-and-tile furnaces. The
waterwalls are composed of nominal 8.26-cm (3 1/4-in.) diameter tubes that
are partially covered by refractory to approximately 2.4 m (8 ft) above the
grate surface. The gases exit from the furnace passing through a two-drum,
vertically baffled boiler bank consisting of rows of in-line 5.7-cm
(2 1/4-in.) diameter tubes arranged in two gas passes. The boiler unit flue
gases pass through a decantation two-stage multiclone collector. The fly ash
captured in the first-stage collector is injected into the furnace to complete
combustion of the fly char, and the fly ash in the second-stage collector is
pneumatically transported to disposal. The cleaned gases are induced through
a centrifugal fan and exhausted to a breeching (common to all boilers) and
then to the stack.
BOILER CONDITIONS
Before testing the boilers, all the associated instruments, such as the
steam flow meters and the pressure and temperature gauges, were calibrated by
Johnson Controls. The boiler settings, grates, and grate seals were visually
inspected to determine their general condition and to seal obvious leaks.
Refractory cement was applied at various locations on the boiler setting to
seal leaks. In addition, the spreaders were adjusted for proper distribution
on MCI normal coal.
Most of the December through March tests were performed in Boiler No. 1,
the 9.9 kg/sec (78,500 Ib/hr) boiler. Boiler No. 2 was not used for the
initial testing because it was the only boiler equipped with both electric
and steam-driven facilities and had the capability of cold starting. Conse-
quently, most of the preparatory effort was devoted to putting Boiler No. 1
in good operating condition. However, the automatic control systems for
25
-------
Figure 9. Cross section of Boiler No. 2.
26
-------
Boiler No. 1 were in such poor condition that the boiler had to be operated
in the manual mode much of the time. The boiler tubes, grate, and refractory
were in generally good condition, but there was some grate seal leakage.
Since the restriction on Boiler No. 2 was eventually lifted and because
tests had to be scheduled according to the extremely limited availability of
an EPA mobile electrostatic precipitator which could be tied into only one
boiler, Boiler No. 2 was selected for the May tests.
During the start-up of the May test, air leakage around the grate seals
produced air levels higher than those normally existing in boilers of this
type. The leakage was detected when Orsat analyses were performed on combus-
tion gases sampled at various cross-sectional scans and heights above the
grate.
The following two sections on the boiler performance when substituting
dRDF for coal are as follows: the first examines the properties of the coal
and dRDF and their blends, and the second covers how the fuels responded in
the boiler subsystems.
FUEL PROPERTIES
The following discussion on the properties of coal, dRDF, and coal:dRDF
blends is preparatory to discussing their handling and burning in the field
test program.
Coal Properties
The coal used during the four field tests (December, January, March, and
May) came from different mines. This variation in the coal supply was the
result of the procurement procedures of the State of Maryland and the need to
use a specification coal during actual test runs.
During December the plant burned a coal which had a low ash fusion
temperature, high heating value, and 3 percent sulfur. In January, the coal
(supplied from the Pittsburg/Swickley seam) had a high ash fusion temperature.
However, the fines were excessive with 85 percent passing a 6.3-mm (1/4-in.)
sieve. In March and May, the coal met the ash fusion, heating value, and
size constraints of the stoker manufacturer. Laboratory analyses of the coal
samples taken throughout the test program are presented in Table 3. Exami-
nation of these coal properties reveals that the coal burned in March was
different from that fired in May.
During each daily test 2.2- to 4.5-kg (5- to 10-lb) coal samples were
taken periodically for subsequent analysis. At the end of the day the
samples were mixed, and the composite was divided by sectioning. The final
samples were placed in two separate containers for shipment to laboratories.
First, a 300- to 500-gram sample was placed in a 1/2-liter rigid polyethlyne
jar with a vapor proof lid and sealed. This sample was sent to the SYSTECH
laboratories for moisture determinations. Second, samples were placed in
4-mil polyethylene bags and taped shut. Subsamples from these macro samples
were sent to Broeman Laboratories and Commercial Testing and Engineering
27
-------
TABLE 3. AVERAGE PROPERTIES OF COAL ON BOTH AN AS-RECEIVED AND
A MOISTURE-FREE AND ASH-FREE BASIS
As Received
% Moisture
% Ash
% Volatile
7, Fixed Carbon
Btu/lb
MJ/kg
Dry Basis
% C
7, H
% N2
% Cl
% S
% Ash
% 02
Btu/lb
MJ/kg
Fusion
Initial
1st Softening
2nd Softening
Fluid
Mineral Analysis
Phos. Pent Ox.
Silica
Ferric Ox.
Alumina
Titania
Sodium Ox.
Potasium Ox.
Lime
Magnesia
Sulfur Tri Ox.
Undetermined
December
Average
2.12
10.78
29.42
57.68
13,471
31.33
77.15
4.77
1.26
.26
3.57
11.01
1.98
1128°C
1192-C
1249°C
1371°C
.48
35.43
34.94
22.39
.56
.25
.99
1.63
.28
1.23
1.82
January
Average
7.03
14.30
16.16
62.50
11,797
27.44
70.50
3.80
1.57
1.80
15.40
6.93
1332CC
1368°C
1379°C
1414°C
.92
43.50
21.00
20.70
1.42
2.70
2.37
.38
.58
.83
...
March
Average
3.78
10.23
22.43
63.55
12,959
30.14
74.15
4.38
1.59
.11
1.72
10.63
7.42
1274°C
1308° C
1335°C
1371°C
.41
52.02
12. 74
25.64
.70
.47
1.87
2.18
.36
1.66
1.95
May
Average
1.27
21.95
22.55
54.23
11,706
27.23
67.40
4.33
1.35
.05
1.22
22.23
3.42
1482+° C
1482+° C
1482+° C
1482+° C
.33
59.62
5.80
27.43
.90
.31
2.32
.54
1.15
.32
1.32
December January March May
Average Average Average Average
COAL MOISTURE AND ASH FREE
33.78 20.60 26.14 29
66.22 79.40 73.84 70
86.70 83.34 82.98 86
5.36 4.49 4.90 5
1.42 1.85 1.77 1
.29 .13
4.01 2.12 1.93 1
2.23 8.19 8.38 4
15,446 14,996 15,069 15,
35.93 34.88 35.05 35
NOTE: Hemispheric (Second Softening)
temperatures in excess of 1204
(2200'F) are preferred.
.13
.87
.73
.56
.72
.07
.43
.35
246
.46
°C
Corporation for fuel property analyses. These analyses included standard
ASTM ultimate and proximate determinations, ash chemistries, and reducing
atmosphere fusion temperatures. The analyses were ultimately expanded to
include a determination of chlorine. Table 3 lists the averages of the test
coal properties. The individual determinations from which the averages were
derived are contained in Appendix A.
Part of each macro sample was sized by using a Tyler portable sieve
shaker and a standard set of ASTM coal-sizing sieves. The sieves used were
nominal 1-, 3/4-, 1/2-, 3/8-, and 1/4-inch screens. Figure 10 shows the size
distributions of the coal samples overlayed on a probability display of the
recommended spreader stoker coal distribution.
The bulk density of the coal was determined by loosely filling a
0.03-m3 (1-ft3) container with coal, without its being agitated or tamped,
and by weighing the container on a 100-lb capacity platform scale. The bulk
28
-------
SIEVE ANALYSIS
l-ROSLIN—RAMMLER EXPONENT
o MARCH AVERAGE d-RDF
A MARCH AVERAGE COAL
• MAY AVERAGE d-RDF
A MAY AVERAGE COAL
PARTICLE SIZE D, INCHES
Figure 10. Coal and dRDF size distributions compared
with recommended size spectra.
density was determined by dividing the weight of the material by the volume
of the container. The density of the coal particles was determined by
weighing lumps of coal; measuring the amount of liquid displaced by each
piece when it was immersed in a water-filled, 500-mSL, graduated cylinder;
and dividing the piece weight by the total volume of coal.
Because the use of a fuel may be limited by its heat density (the amount
of energy per unit volume), the amount of air required to burn the fuel, or
the ability to introduce the fuel into the furnace, there is a particular
interest in the heat density (MJ/m3) of the fuel and the theoretical mass
(kg) of air required to burn a unit mass (kg) or unit energy value (MJ) of
fuel. The data obtained from the laboratory analyses of the fuel were
substituted in Equation 4.1 to determine the theoretical air requirements:
TA = 11.53C + 34.34(H2-02/8) + 4.29S kgair/kgfuel
(4.1)
where C, H, 0, and S are the respective weight fractions of the element in
the fuel ultimate analysis.
29
-------
The air requirement can also be expressed in terms of the mass per heat
content equivalent of fuel by using the available fuel properties information:
(kg, ,/kg . ) x 106
kg . /MJ • fue* air - (4.2)
.
The volumetric heat rate (megajoules per cubic meter) is the parameter
which determines the amount of energy that the volumetric feeders can intro-
duce into the furnace. This parameter is particularly important in a retrofit
application to determine the maximum amounts of dKDF which can be substituted
for coal and still have sufficient feeder capacity.
= J/kg x kg/m
106
dRDF Properties
Procedures similar to those for the coal sampling were used to collect
dRDF samples. Each time the dRDF was sampled, one sample was sealed to
permit moisture determination at the SYSTECH laboratories and another was
immediately processed for size and composition.
Table 4 lists the averages of the dRDF properties including ultimate and
proximate analyses, chlorine, reducing atmosphere ash fusion temperatures,
and ash chemistries. The table also includes the same information recalcu-
lated on a moisture-free and an ash-free basis. The individual determinations
from which the averages were derived are contained in Appendix A.
The theoretical air requirements (expressed as kilograms of air per
megajoule) and the volumetric heat rate (megajoules per cubic meter) were
computed by substituting the laboratory data in Equations 4.1 through 4.3.
Blend Properties
The theoretical combustion properties of coal:dRDF blends can be mathe-
matically computed once the heating value, the bulk density, the mass, and
the heat or volumetric blend ratio of the components are known.
Throughout the current study the mixture of coal and dRDF was given a
volumetric ratio designation. For example, a 1:1 blend would be an admixture
of 1 m3 of coal with 1 m3 of dRDF.
It is important, also, to know both the weight and the heat mixture of
the blended fuels in order to compare the test results with predictions made
by others as well as to size the fuel handling system, to estimate the ash
handling requirements, and to determine the best feed rate of blended fuel
entering the boiler.
30
-------
TABLE 4. AVERAGE PROPERTIES OF dRDF ON BOTH AN AS-RECEIVED AND
A MOISTURE-FREE AND ASH-FREE BASIS
As Received
% Moisture
% Ash
'/. Volatile
7. Fixed Carbon
Btu/lb
MJ/kg
Dry Basis
% C
% H
7. N2
% Cl
% S
Z Ash
X 02
Btu/lb
MJ/kg
Fusion
Initial
1st Softening
2nd Softening
Fluid
Mineral Analysis
Phos. Pent. Ox.
Silica
Ferric Ox.
Alumina
Titania
Sodium Ox.
Potasium Ox.
Lime
Magnesium
Sulfur Tri. Ox.
Undetermined
December
Average
13.40
19.97
56.54
10.10
6488
15.09
43.98
5.29
.35
.40
.40
23.19
30.80
March
Average
12.62
24.41
54.08
8.89
5534
12.87
39.17
4.47
.39
.45
.26
27.97
27.30
May
Average
12.22
28.75
49.27
9.76
5266
12.25
35.63
4.54
.85
.36
.28
33.02
25.33
December March May
Average Average Average
dRDF MOISTURE AND ASH FREE
85.04 85.80 83
14.97 14.21 16
54.29 54.42 53
6.53 6.20 6
.43 .54 1
.49 .62
.53 .36
38.02 ' 37.87 37
.38
.62
.36
.75
.25
.54
.42
.68
9785 8772 8956
1103CC
1142CC
1191°C
1246° C
.87
55.52
2,27
13.45
.66
6.82
1.30
10.75
1.14
6.03
1.19
1116°C
1151nC
1179°C
1213'C
.73
71.58
2.89
4.43
.99
5.66
.53
7.50
1.12
1.22
1.87
10963C
1152=C
1163°C
1218°C
.65
63.65
2.64
8.39
.69
7.53
.91
9.74
1.59
3.20
1.00
22.76 20.40 20
MOTE: Hemispheric (Second
Softening) temperatures
excess of 1204*0 (2200°
are preferred.
.83
in
F)
31
-------
The properties of the fuel for the various blends used in March and May
are listed in Tables 5 and 6.
The weight fraction of dRDF determined by physically separating the coal
from the dRDF and weighing each component had a relative standard deviation
of less than 5 percent. Hence, the blending system was producing a uniform
blend. The hand-sorted weight ratios are consistently higher than the
calculated weight ratios, probably because all fines not identified as coal
were considered dRDF. Since the coal contained between 15 and 30 percent
fines, the dRDF fraction was biased by considering all unidentified fines as
dRDF.
It would normally be expected that a linear interpolation should be
applicable for blends when the density of coal and the density of dRDF on a
volumetric basis are used to determine the relative density of the two
fuels. The fact that the fuel densities are uniformly higher than the
interpolated densities is understandable when both coal and dRDF are con-
sidered as ensembles of solid particles which have different void sizes.
When two solid fuels are admixed, the relative void between the two fuels
frequently becomes smaller because some of the coal particles will fill voids
in the dRDF and some of the dRDF particles will fill voids in the coal
structure. Consequently, the bulk density of an admixture would likely
be higher than the density determined by straight interpolation. This error,
however, is small and was determined to be less than 3 percent for all
blends.
Conclusions
In the current tests, the substitution of dRDF for coal increased the
ash and decreased the volumetric heat density of the blend. However, the
air:fuel ratio (kilograms air per megajoule equivalent) actually declined as
dRDF was substituted for coal. If the coal:dRDF blends can be fired at the
same excess air level as coal alone, the forced and induced draft fans
should be adequate, assuming there is little efficiency deterioration. The
decrease in heat content for a given fuel volume could cause the furnace fuel
feeders to become limiting. Increases in ash content could overload ash
handling equipment or air pollution control equipment.
FUEL HANDLING AND RESPONSE IN BOILER SYSTEMS
The three boilers at the Hagerstown plant are similar in configuration
although different in capacity. Each boiler can be divided into five sub-
systems: fuel feeding distribution, furnace, controls, air and gas handling,
and ash handling. The fuel distribution system moves the fuel from outside
the boiler into the furnace where it is combusted and converted to heat
energy and ash. The ash handling system removes the solid residue from the
furnace and air pollution control equipment. The air and gas handling system
moves air to the furnace for combustion, draws the combustion products through
the heat transfer sections, and exhausts the gases to the atmosphere. The
furnace and boiler system removes heat from the combustion products and
converts it into steam. The control system coordinates the activities of the
fuel and air handling systems in response to energy requirements of the
furnace system as coupled with plant demand.
32
-------
TABLE 5. AS-FIRED PROPERTIES FOR BLENDS IN MARCH TESTS*
AS FIRED
VOLUMETRIC BLEND
1:0
1:1
1:2
0:1
PARAMETER
Btu/#
MJ/kg
Moisture
Volatiles
Fixed Carbon
Ash
C
H
N
0 <
s
Cl
%dRDF
by Volume
by Weight
by Heat
12714
29.5
4.9
30.4
54.2
10.5
71.5
4.7
1.3
5.6
1.5
.06
0
0
0
9180
21.3
7.4
44.9
29.8
17.9
51.3
4.6
.8
16.8
.8
.24
50
35
20
8706
20.2
7.7
46.8
26.5
18.9
48.6
4.6
.8
18.3
.7
.3
67
52
36
6034
14.0
9,6
57.8
8.1
24.5
33.4
4.5
.4
26.8
.2
.4
100
100
100
*Unless noted, all values are weight percent on a wet basis.
33
-------
TABLE 6. AS-FIRED PROPERTIES FOR BLENDS IN MAY TESTS*
VOLUMETRIC BLEND
PARAMETER
Moisture
Volatiles
Fixed Carbon
Ash
Btu/lb
MJ/kg
Carbon
Hydrogen
Oxygen
Nitrogen
Sulfur
Chlorine
FUSION TEMP. -C
Init. Def.
1st Soft.
2nd Soft.
Fluid
THEORETICAL AIR
1:0
1274
1308
1335
1371
AS FIRED
1:1
1:3
22.6
54.2
22.0
11706
27.2
66.5
4.3
3.4
1.3
1.2
.05
6.6
31.7
38.4
23.2
8988
20.9
54.1
4.1
9.8
1.1
.86
.15
1:2
0:1
7.9
38.3
29.1
24.7
8382
19.5
47.3
4.1
14.2
.9
.66
.21
16.6
48.6
9.0
25.9
5130
11.9
30.9
3.8
21.8
.6
.23
.33
1116
1151
1179
1213
kg/kg fuel
kg/MJ in fuel
% Weight Rate dRFD
% Heat Rate dRFD
9.04
.351
0
0
7.26
.347
35
20
6.27
.332
52
37
3.93
.331
100
100
*Unless otherwise noted, all values are weight percent on a wet basis
34
-------
Fuel Distribution
System Description—
Each of the three boilers is equipped with Hoffman Combustion Engineering
Type A "Firerite" underthrow spreader-feeders. Figure 11 is a cross section
of a typical spreader-feeder mounted on the stoker front plate of the boiler.
The solid fuel is placed in the coal hopper and flows by gravity to a recipro-
cating feed plate which advances the lump fuel over the distributor blades.
As the reciprocating feed plate moves forward, the lump fuel falls into the
rotor and is carried between the rotor and the rotor casing through 220 degrees
of arc before being thrown into the furnace.
Adjustments on the spreader include varying the rotor speed to change
the length of the particle trajectory and varying the position of the rotor
casing (circular tray) to alter the arc of the particle trajectory. Slowing
the rotor shortens the particle throw. When the rotor casing is moved
forward (toward the furnace), the fuel particles are thrown into a higher
trajectory. Although the two adjustments jointly determine the longitudinal
fuel pattern, rotor speeds are adjusted grossly while the casings are adjusted
finely so that the fuel introduced by each spreader-feeder lands at approxi-
mately the same distance from the back furnace wall.
FEEDER BODY
DISTRIBUTOR BLADE
FACE PLATE
WATER GASKET-,
COAL HOPPER
FEED PLATE
ARCH
PATH OF COAL
END OF COAL SPREAD
TOP OF GRATE
Figure 11. Cross section of Hoffman Combustion Engineering
spreader-feeder.
35
-------
Cold Flow Test—
The cold flow test (feeding fuel to an idle boiler with no fire) was
designed to demonstrate that coal:dRDF blends could be successfully handled
and distributed into the furnace. The first goal of the cold flow test was
to operate and verify the performance of the new fuel conveying system that
had been installed to bypass the existing coal silo. This system conveyed
dKDF and coalrdRDF blends into the coal weigh lorry which in turn supplied
the stoker coal hoppers of the three boilers. Initially the operation of
the conveyor system was unacceptable because the coal and pellet conveyors
ran too slowly. Subsequently this problem was easily corrected.
To test the ability to feed pellets, two basic pellet sizes were fed
into the furnace: nominal 1-inch-diameter by 2-inch-length pellets were used
for the first experiment, and nominal 1/2-inch-diameter by 3/4-inch-length
pellets were used for the second. In each experiment the test sample was
44 kg (97 Ib). In the first experiment, with the reciprocating feed plate
adjusted for maximum stroke, 4 1/2 minutes were required to feed the nominal
1-inch-diameter by 2-inch-length pellets, an an average rate of about
9.7 kg/min (21.3 Ib/min). During this experiment unusual feeder noise
indicated some pellet breakage was occurring. The noise was attributed
primarily to the rotor casing (circular tray) clearance limitations. Although
there were no difficulties during this test, it was concluded that if the
1-inch-diameter by 2-inch-length pellets were simultaneously fed to all four
stoker spreader-feeders (which are driven by a single motor and drive shaft),
they could cause shear pin failure.
In the second experiment, with the feeder adjustment similar to that
used during normal coal firing, the 1/2-inch-diameter by 3/4-inch-length
pellets were fed into the furnace in about 3 minutes, an average rate of
about 11.3 kg/min (25 Ib/min). During this experiment no unusual noise or
other events were noted.
In both cold flow experiments the fuel distribution patterns (both
lateral and longitudinal) were excellent, and the pellets were generally fed
to the desired furnace locations. Figure 12 shows fuel being fed through the
spreader. Fine material accumulated at the bottom of the spreader opening
because the feed fan air jet was turned off during the test to minimize
dusting inside the boiler. Figure 13 shows coal and pellets spread on top of
each other with the fuels rather uniformly dispersed. The spreader performed
as intended in that the larger pellets with the greater mass traveled
to the rear of the furnace with the fines falling closer to the spreader.
The area densities of the 1/2-inch-diameter by 3/4-inch-length pellets
spread into the furnace were determined as follows: A rough square of the
fuel on the grate was isolated with a flat shovel; see Figure 14. The
isolated pile of fuel was then collected and weighed. The average density
across the grate section was determined by dividing the weight of the collected
pellets by the area covered by the pile. For each of four distances from the
back wall, Table 7 lists the spread density and condition of the pellets.
36
-------
Figure 12. Spreader-feeder injecting a dRDF:coal blend into furnace.
37
-------
00
Figure 13. Uniform distribution of coal and dRDF pellets near the furnace rear wall.
-------
Figure 14. Isolation of grate dRDF to determine its spread density,
-------
TABLE 7. AREA DENSITY OF PELLETS REMOVED FROM THE COLD FLOW TEST
Distance (m) Spread density
from back wall (kg/m2) Pellet Condition
1.
1.
1.
2.
07
52
83
90
(42
(60
(72
(114
inches)
inches)
inches)
inches)
29
12
12
21
.5
.9
.0
.3
(6.
(2.
(2.
(4.
06
65
46
35
Ib/ft2)
lb/ft2)
Ib/ft2)
lb/ft2)
whole
whole
whole
large number
fines
of
The cold flow test indicated that the 1/2-inch-diameter by 3/4-inch-
length pellets produced a spread density which was high near the front and
rear walls and low in the middle of the grate. Approximately twice as much
fuel was located at the rear wall than in the middle of the grate. Subsequent
testing during the hot flow test confirmed that the pellet throw was too long.
Hot Flow Test—
It was decided to fire 100 percent pellets for a short period of time to
determine their impact on furnace performance. A total of 945.7 kg (2085 Ib)
of 1/2-inch-diameter dRDF pellets were weighed in the lorry. When the stoker
coal hopper of the No. 2 boiler was empty, the pellets in the lorry were
discharged into the hopper. The steam pressure was 1055 kPa (153 psi),
and the steam flow was about 3.8 kg/sec (30,000 Ib/hr). The cam on each of
the Hoffman spreader-feeders was adjusted to the maximum feed stoke. The
spreader rotor rpm was 60, the same as that used for coal. After the pellets
were fed into the furnace, it was immediately obvious that the fuel trajectory
was too long, causing fuel and flame to impinge on the rear furnace wall.
The trajectory was reduced some 0.3 m (12 in.) by slightly retracting the
circular tray. Fifteen minutes after starting the pellet feed, the steam
pressure had decreased to 1027 kPa (149 psi). This pressure generally pre-
vailed for the remaining 5 to 7 minutes of the test. The pressure had dropped
because the spreader-feeders were incapable of supplying a sufficient quantity
of fuel to the furnace to maintain the steam pressure. The average firing
rate for the three feeders collectively was 48 kg/min (105 Ib/min) or about
15.9 kg/min (35 Ib/min) for each feeder. While the*steam pressure decreased
during the 20-minute firing with a 0:1 blend, the steam flow remained about
3.8 kg/sec (30,000 Ib/hr) which was 50 percent of boiler rating. This
dropping steam pressure indicated that the boiler system could not carry the
50 percent load when firing 100 percent pellets.
Following the successful 0:1 test, a 1:1 blend test was fired. During
this test there was frequent clinkering on the grate. One cause of this
clinkering is that when a fuel with too low art ash fusion temperature is
placed on the grate and agitated, hot burning particles with molten edges
roll against and under each other as the grate is vibrated. The burning
particles, which are now within the ash bed, heat the ash so that the softened
ash sticks together, restricting cooling air flow, and begins to form a skin
clinker layer across the bed surface. This skin clinker becomes progressively
40
-------
less permeable than the open portions of the ash and fuel bed; consequently,
less air is supplied to the fuel/ash bed prompting progressively deeper
clinkering. As new fuel is fed into the furnace and falls on top of the skin
clinker, the airflow is retarded still more, and the clinkering condition is
further aggravated, and smoking becomes severe.
The ash fusion temperature analysis of the coal revealed that the
hemispheric reducing ash fusion temperature was 1191°C (2176°F). Since
reducing hemispheric ash fusion temperatures in excess of 1204°C (2200°F) is
preferred for proper combustion conditions, it is not surprising that
clinkering was encountered. Since continuation of this clinkering during
blend firing would not permit evaluation of the effect of pellets, a different
coal had to be procured for subsequent tests.
Even though the pellets have an ash fusion temperature less than that of
coal, the furnace operated satisfactorily when the 100 percent pellets were
burned. Assuming that the ASME definition for lignitic type ash (defined as
ash in which the sum of CaO + MgO is greater than Fe203)i applies to dRDF ash
(which has a CaO + MgO to Fe203 ratio of 4), then dRDF burns similarly as a
lignite coal. Therefore the boiler design rules for lignite, rather than
those for bituminous coal, more nearly apply when burning dRDF.
During January the testing switched from Boiler No. 2, which is a
7.6 kg/sec (60,000-lb/hr) boiler with three spreader-feeders, to Boiler
No. 1, which is a 9.9 kg/sec (78,500-lb/hr) boiler equipped with four
spreader-feeders. In the January tests the boiler was fired with a 1:1 blend
for 4 hours and with 100 percent pellets for 2 hours. During the 1:1 firing,
the overfire air and fly ash reinjection air pressures were adjusted to
produce the desired flame mixing and low smoke opacity. Further discussion
on these adjustments are provided later in the text. The stack plume was
relatively clear during both the 1:1 and 0:1 blend firings. While firing
with the 1:1 blend, there was no reoccurrence of the clinkering which was
experienced during the December tests. This improved performance was due to
the coal having a hemispheric fusion temperature of about 1373°C (2500°F).
The December coal fusion temperature was 1204°C (2200°F).
When the blend ratio was changed to 100 percent dRDF, the furnace
remained clear of smoke streamers for the first 20 to 30 minutes of firing,
and there was no significant smoking. Subsequently two rows of ash were
observed on the fuel bed in line with the fly carbon reinjection nozzles.
Figure 15 shows the location of three of the fly carbon reinjection nozzles
at the back of the furnace. Each of the two rows was in line with the double
fly ash reinjection ports (one reinjector for the dust caught in the boiler
bank passes, the other for the coarse stage multiclone). This ash layer
phenomenon might be alleviated by spacing the fly ash reinjection nozzles
more uniformly across the rear wall of the furnace or increasing the number
of reinjection lines. The reinjection ash rows moved well through the furnace
Winegartner, E. C. Coal Fouling and Slagging Parameters. American
Society of Mechanical Engineers, 1974.
41
-------
Figure 15. Ash reinjection and overfire air ports in rear wall of Boiler No. 1.
-------
as the grate vibrated, and they remained sufficiently porous for airflow.
There was also a base of burned-out ash under the ash rows.
The furnace volume appeared more than adequate for the combustion
taking place, although a maximum of 6.8 kg/sec (54,000 Ib/hr) of steam could
be generated when firing 100 percent pellets. This unmodified vibrating-
grate, stoker-fired boiler supported 70 percent of the nominal rating for
2 hours when firing 100 percent pellets. The magnitude of this derating is
the amount predicted by the volumetric limitations of 1.47 m3/hr (52 ft^/hr)
for the feeder. The forced draft fan capacity was sufficient to meet the
underfire air requirements at this rating.
When a 1:1 blend was fired, there was no difficulty in maintaining the
1034 kPa (150 psig) boiler steam pressure. However, when 100 percent pellets
were introduced, the steam pressure fell to about 1000 kPa (145 psig) while
supplying 6.8 kg/sec (54,000 Ib/hr) steam flow. This pressure drop was due
to volumetric feeding limitations of the spreader-feeders. After the initial
loss in pressure, the boiler operated at this reduced pressure for the rest of
the test.
The principal change required during the 1:1 and 0:1 blend tests was to
reduce the spreader rotor speed so that the throw of the pellets would be
approximately 0.15 m (6 in.) less than that for 100 percent coal. This
adjustment was necessary to prevent furnace rear wall fuel impingement.
During the combustion tests with 1:1 blend and 100 percent dRDF, the
fireball was kept well away from the rear and front walls of the furnace by
adjusting the overfire air. Once these jets were adjusted for minimum smoke
and maximum turbulence efficiency for coal-only burning, they continued to
meet the mixing and wall protection requirements when burning blends and 100
percent pellets. As viewed from the side of the furnace when firing both
pellets and blends, the bed was well burned out by the time it approached the
front ash pit. The flame pattern above the grate indicated that the fuel bed
was maintaining proper porosity and that the combustion was good. With
little attempt to optimize the system, a 10 to 12 percent carbon dioxide
content in the flue gas at the boiler outlet was readily obtained.
Normal Boiler Operation—
The cold and hot flow tests showed that dRDF could be properly distributed
into the furnace. The blend also had a distribution pattern on the furnace
grate which was similar to that of coal. This finding is not particularly
surprising since the size distributions and material densities (not bulk
densities) of the coal and pellets were similar. Therefore, with the same
velocity and angle of injection into the furnace, pellets and lumps of coal
with equal weight would be expected to travel approximately the same distance.
There was a severe slag accumulation on the rear furnace wall in line
with Spreader-feeder No. 1 during the May testing. To stop the slagging, the
throw of all pellets was reduced approximately 15.24 cm (6 in.), and the
circular tray in Spreader No. 1 was retracted slightly to reduce the arc of
the pellet trajectories. After this adjustment the pellets still carried
43
-------
the same distance from the feeder and landed at the same place on the grate.
However, the 15.24-cm (6-in.) throw reduction eliminated the slag accumulation
on the furnace wall and reduced smoking. In December, January, and March,
the material impinging on the side walls was not excessive. In May, however,
Spreader No.l had to be adjusted because of its throwing too far and spraying
the left side wall of the furnace. Careful measurement of the spreader
showed that the circular tray had skewed left. Once the tray was properly
aligned, the fuel impingement on the left side wall was eliminated.
During January and March, Boiler No. 2 had a recurring problem of
clinkering on the left side when burning coal. Reports received after the
field testing stated that this clinkering had been eliminated by readjusting
the spreader circular tray.
Proper adjustment of the spreader-feeders is critical to the successful
combustion of coal:dRDF blends. During part of the March tests, clinkering
on the grate in front of Spreader No. 2 recurred, but the rest of the fuel
bed remained free burning. Clinkers formed on top of the piled burnt-out ash
at the front and moved out of the furnace with some difficulty. The rear wall
of the furnace remained clear of slag throughout these tests. During an
unexpected furnace outage caused by a control loop failure (the dRDF in the
furnace was not related to this failure), the furnace spreaders were inspected.
While Spreaders No. 1, 3, and 4 were clear of pellets, Spreader No. 2 had a
heavy accumulation of partially pyrolyzed pellets in the feed throat. Careful
measurement of the tray position from the inside of the boiler indicated that
the pellets were being thrown at too high an angle out of the spreader so
that they ricocheted off the refractory feed throat and accumulated at a
point in the furnace approximately two-thirds of the way back. This malad-
justment was solved by retracting the circular tray approximately 3.18 mm
(1/8 in.). The furnace was then brought back on line from a cold start with
a 1:2 blend and run continuously at loads of 3.2 kg/sec to 4.2 kg/sec
(25,000 to 33,000 Ib/hr) for 48 hours with no further clinkering.
Ash Handling
System Description—
Each of the three boilers at the MCI plant is equipped with a dry
pneumatic ash handling system. As the solid fuel burns on the vibrating
grate, the grate is periodically pulsed to advance the ash to the front of
the boiler where it falls into a refractory-lined ash pit. The bottom ash is
manually hoed from the ash pit into the pneumatic ash handling system. After
the ash is conveyed to a cyclone separator where it is de-entrained from the
carrier air, it falls into an ash storage silo. The vacuum source for the
pneumatic system is a steam ejector.
Ashes are also collected from under the boiler tube bank and from the
particulate collectors. The ash from the tube tank and primary cyclone is
reinjected into the furnace through nozzles in the rear of the furnace. The
secondary multiclone collector ash is pneumatically transported to the
de-entrainment cyclone and ash silo.
44
-------
To discuss the findings on relative boiler ash flows when substituting
dRDF for coal, the following aspects of the boiler ash handling system are
addressed: grate, bottom ash, ash silo, and reinjection and collector
fly-ash flows.
Grate —
As the fuel is spread into the furnace, portions of it burn in suspen-
sion, and the remainder falls to the grate. The primary combustion air is
introduced into the furnace through the grate supporting the burning fuel and
then into the furnace. The grate is maintained with a constant 5-cm (2-in.)
ash covering by periodically vibrating it to advance the accumulated ash. In
order to achieve a steady-state condition, as much ash must be removed as is
added to the furnace with the fuel. The two principal means of control of
the ash bed depth are the dwell period between grate pulses and the duration
of the grate pulse. In addition to frequency, the amplitude of the agitation
can also be adjusted to account for differences in ash bulk density.
Figure 16 shows the dwell between grate pulses for various blends as a
function of boiler load. When the load increased, the frequency of pulses
was increased to maintain the same depth of bed on the grate. Also, as the
dRDF substitution ratio increased, the frequency and/or duration of pulse was
increased to maintain the uniform bed depth. In all instances, pulse
frequency was the principal controlling variable. The duration of pulse was
adjusted to cause the fire line at the rear of the boiler to advance approxi-
mately 6 inches per excitation. The amplitude of the pulse was also adjusted
to help clear the rear fire line from the furnace wall. Minimal adjustment
to the amplitude was required. When firing a blend, the pulse duration
tended to increase because the bulk density of blend bottom ash was less than
that of the coal bottom ash. The outliers shown in Figure 16 were so indi-
cated because the ash bed depth and fire line were not representative of
normal boiler operation.
Bottom Ash—
As the ash was shaken from the grate into the ash pit, it was allowed to
accumulate for approximately 8 hours. After this period the ash doors were
opened, and the ash was hoed into the boiler's pneumatic ash handling system.
When the grate was free of clinkers, the ash which accumulated in the hopper
was free flowing and easily handled. Figure 17 shows the results of manual
sieving of various bottom ash samples and indicates that the coal and dRDF
ashes are equally well handled by a conventional pneumatic ash handling
system.
Because the spreader paddles were worn, a significant amount of fuel
"dribbled" onto the front of the grate and subsequently into the ash pit.
Since this fuel had not ignited on the grate, mixing the unignited coal with
the hot bottom ash often resulted in fires in the ash pit. When the ash pit
had fire, the blend ash tended to be more plastic than coal-only ash and had
a taffy-like consistency. The coal ash under similar conditions was much
easier to break up.
The ash flow from the various blends was monitored by first cleaning the
ash pit completely. Then, after the ash was allowed to accumulate in the pit
45
-------
cc
3
_J
OJ
Q
12 -
CO
LU
Z 8-
LU
*
<
I
CO
LU
4 -
AD
OUTLIERS
O.i
0.2 0.3 0.4 0.5
BLEND
I --0
I •• I
I -• 2
0 I
0.6 0.7
BOILER
I 2
o •
A A
O •
*
0.8
£ - FRACTION OF RATING
Figure 16. Effect of blend and load on grate pulse interval
or relative grate speed.
46
-------
BOTTOM ASH SIEVE ANALYSIS • MAncH 21
o MARCH 22
BLB4D 1:0 A MARCH 28
A MARCH 31
o APRIL 1
BOTTOM ASH SIEVE ANALYSIS
BLEND 1:1 o MARCH 23
n.(KMll«-«»««l.BE
-------
BOTTOM ASH SIEVE ANALYSIS
BLEND ! :0
*- HQst.m-tuMMt.efl E
a MAY 4
• MAY 3
o MAY 18
A MAY 17
x MAYS
BOTTOM ASH SIEVE ANALYSIS
BLEND 1:1
vROSLIN—RAMMLEM EXPONENT
-M-
• MAYS
a MAY 12
o MAY 13
* MAY 13P.M.
PARTKXE SZ£ 0.
BOTTOM ASH SIEVE ANALYSIS
BLEND 1 2
BOTTOM ASH SIEVE ANALYSIS
BLEND 01
'•RCSUN-RAMWLEfl EXPONENT
I
" i
1 £
- 0 *
i5 37* S 'S
P,V*TICL£ S2E 0- I
PAflTTOf SCE 0, INCHES
Figure 17 (concluded)
48
-------
for 1 hour, it was manually shoveled from the ash pit into tared 55-gallon
drums and weighed. As would be expected, the quantity of ash increased with
increases in dRDF.
Ash Silo—
The ash silo is used for dry ash storage. The bottom ash and the fly
ash are both transferred to the silo by the pneumatic ash handling system.
The ash is withdrawn from the bottom of the silo by a rotary drum mixer where
the ash is blended with water for dust control before discharge into a truck.
The only problem with this system occurred during 100 percent pellet
firing when the bottom ash was so fine that it would not de-entrain properly
in the cyclone. Rather, the paper ash particles, which were wetted by the
steam in the ejector, carried through the cyclone and eventually plugged the
ejector. Although the ejector was easily cleaned, continued maintenance of
the ejector will likely require developing some other means for keeping it
clear.
Reinjection and Collector Fly-Ash Flows—
To measure the change in rate of fly-ash reinjection into the furnace
when changing from coal to dRDF, a secondary measurement technique was used.
Because the reinjection and collector dust streams were piped directly into
the plant's pneumatic ash handling system, it was impractical to isolate the
flow from one collector and weigh it as had been done for the bottom ash.
Consequently, the flow was measured by installing a drain tube near the
bottom of the collector and reinjector hoppers. Figure 18 illustrates this
sampling technique. As fly ash accumulates in the hoppers, it flows down the
sides. A portion of that flow will fall into the drain tube. The amount of
ash collected in the drain tube over a fixed period of time indicated the
relative fly-ash flow in the hopper.
The fraction of ash falling to the ash pit was relatively independent of
the type of fuel being used (e.g., it did not matter if the fuel was coal or
dRDF), but it was strongly dependent on the ash content of the blend.
However, the weight of the carbon fly ash accumulating in the reinjector and
collector drain tubes (see Figures 19 and 20) did not appear to have any
correlation to the ash content of the fuel. Since the dRDF had a higher ash
content than the coal, it was anticipated that the ash flow into the drain
tube would increase with increasing dRDF substitution. This occurrence
however, was not indicated by the data.
The lack of variation in the ash accumulation in the drain tube as a
function of fuel ash content was perhaps due to incomplete burning of carbon
or the difference between coal and dRDF carbon fly ash particle gas
de-entrainment characteristics. The data shows, however, that combustion
improved as dRDF was substituted for coal. Figure 21 shows the carbon content
of the reinjector ash as a function of the boiler load and coal:dRDF ratio.
While the data has much scatter, a general trend toward improved carbon
burnout with dRDF substitution is evident. Consequently, since the decreasing
amount of carbon with increasing dRDF substitution tended to offset the
increasing ash content of the fuel, the drain tube ash flow measurements for
various coal:dRDF blends could be deceptive.
49
-------
. 2" SCH 40 PIPE - .31 m LONG
PIPE CAP
Figure 18. Drawing of a typical ash collection drain tube to
monitor relative ash flow in collector and reinjector.
50
-------
1000
800
600
m
m
400
cc
a
cc
O
5 200
LJJ
CC
0
0.00
• . 2
* • • •
*o V * '
• ° 8 a° 8
0.10 0.20 0.30 0.40 0.50
FRACTION OF RATING
BLEND
1:0 —
1:1 —
1:2
0:1 —
BOILER
I 2
0.60 Q70
0.80
Figure 19. Ash flows into reinjector hopper drain tube for
coal, blend, and dRDF firings.
51
-------
z
5
O
1000
800
BLEND
1:0
1:1
1:2
0:1
BOILER
I 2
<
01
CD
(-
Z
<
600
400
o
O 200
O
* »
0
QOO
0.10 020 0.30 0.40 050
FRACTION OF RATING
0.60
0.70
080
Figure 20. Ash flows into collector hopper drain tube for
coal, blend, and dRDF firings.
52
-------
REINJECTOR ASH
100
80
%C
60
40
20
BLEND
I = 0
I = I
I = 2
. OH
BOILER
I 2
o •
A A
a •
A •
•A
o
•
a
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
~ FRACTION OF RATING
Figure 21. Carbon content of reinjector ash for coal,
blend, and dRDF firings.
As dRDF is substituted for coal, the ash distribution becomes finer.
Figures 22 and 23 show that the 1:1 and 1:2 size distributions are similar
to, but finer than, the 1:0 size distributions. The ash from firing a 0:1
blend produced the highest amount of fines. A possible explanation of this
phenomenon is that as more paper is burned, there is an increased amount of
fine ash which is easily lofted by the underfire air.
Air and Gas Handling
System Description—
Each boiler has three separate air systems: underfire air, overfire
air, and an induced draft fan. The underfire airflow was controlled from the
fuel flow to the furnace. The overfire air was adjusted manually to achieve
the most smoke-free operations. The induced draft fan was controlled by a
draft sensor in the upper part of the furnace to maintain a specified
negative pressure in the firebox.
The underfire air and induced draft fans are equipped with both electric
motors and steam turbine drives. The overfire air fans are driven by an
electric motor.
53
-------
REINJECTOR
MCI FLY ASH
COAL (1:0) MAY. 1977
n-ROSLIM-MAMMLER E1POMENT
• 5/3
A 5/4
c 5/16
x 5/17
+ 5/17
:L£ SIZE O MICRONS
REINJECTOR
BLEND 1 2
PARTICLE SIZE D MICPOMS
REINJECTOR
BLEND 1:1
MCI
MAY 1977
S
« 5 12
x 5'12
REINJECTOR
d-ROF |0 1)
5 14 77
MCI
MAY 1977
• 5/10
x 5/10
Figure 22. Reinjector ash size distributions for coal,
blend, and dRDF firings.
-------
COLLECTOR
COAL (1.0)
MCI FLY ASH
COLLECTOR
MCI
MAY. 1977
COLLECTOR
MCI
MAY, 1977
n-ROSLIN—HAMMLER EXPOMCNT
PADTICLE SIZE D. HICHONS
5/12
5/13
COLLECTOR
MCI FLY ASH
d-ROF(Ol) MAY. 1977
n ROSLIN-RAMMLEP. EXPONENT
^u
•t r
K *
PARTICLE SIZE D. MICRONS
Figure 23. Collector ash size distributions for coal,
blend, and dRDF firings.
55
-------
Underfire Air Setting—
When the underfire air fans were set for manual control to follow the
steam flow, a clean ash bed was maintained; but when they were placed on
automatic control, clinkering occurred. The controls were designed so that
the amount of fuel fed to the furnace is varied to maintain boiler pressure.
In addition, the airflow is controlled proportionate to the fuel flow.
Consequently, during a load shed, fuel on the grate could deliver an extra
0.6 to 1.3 kg/sec (5,000 to 10,000 Ib/hr) of steam due to the lag time
associated with this control mode. As a result, the ash bed could change
from oxidizing to reducing conditions which promote clinkering. This
condition occurred when the airflow followed the fuel feed rate rather than
the amount of fuel in the furnace. To prevent this problem during the May
testing, the underfire was biased upward as though the boiler were carrying
1.3 kg/sec (10,000 Ib/hr) more steam than it actually was. When the controls
were set for 50 percent excess air, rapid load sheds induced clinkering.
After the air was biased upward to provide 20 percent more excess air, the
load sheds caused no problems.
The need for this biasing was attributed to the appreciable amount of
underfire air leakage around the grate. When the boiler was inspected
before testing, the seals between the grate and the windbox appeared to be
tight. However, when a water-cooled lance was used to probe the furnace
during operation, the excess air levels at 1 m (3 ft) and 3 m (10 ft) above
the center of the grate were 30 and 50 percent, respectively. In the center
of the duct leaving the boiler, the excess air levels were about 120 percent.
Figure 24 shows the carbon dioxide level inside the furnace compared with
that at the furnace outlet. Figure 25 shows the excess air in the breeching.
Figure 26 shows the leaning of the flame toward the center of the furnace
which confirms the hypothesis of excess underfire air leakage.
In view of the above, it appears that a coal:dKDF blend can be fired at
the same settings as coal only if adequate air is available to prevent
clinkering during load sheds. In addition, excess air levels of 50 percent
should be attainable in a tight boiler since the excess air at the grate in
the test boiler with significant air leakage around the grate was 30 percent.
When firing 100 percent dRDF in March, no problems were encountered when
the air control was set on manual at an air flow typical of that required
when burning a specification stoker coal and supplying a desired steam flow.
In May, however, when pellets were first introduced, ignition problems
developed when the underfire air was automatically controlled. A malfunction
of the air control system resulted in a 600 percent excess air condition on
the fuel bed which nearly extinguished the flame. Once the ignition problem
was diagnosed and the air controller was put on manual (it could not be
biased enough to stay on automatic), the fire returned to normal, and the
feeders operated at a low fraction of capacity with steam pressure returning
to normal.
Overfire Air Setting—
Combustion air is normally provided through the grate as underfire air
so that the fuel is supplied with sufficient oxygen to ensure its complete
combustion. However, because of the different combustion rates of various
56
-------
10
PROBE 6 FT
ABOVE GRATE
(CENTERED IN FURNACE)
%CO2 8
PROBE 10 FT
ABOVE GRATE
(1/s INTO FURNACE
FROM WALL)
BOILER LOAD - 30% OF DESIGN RATING
BREACHING
o o
o o
11:00
12:00
13:00
TIME
Figure 24. Carbon dioxide levels in furnace vs. time as
determined with a water-cooled probe.
57
-------
140
40
'• • A 1 A« AO A 00
O.I
0.4 0.5
- FRACTION OF RATING
BLEND
I =0
I •• I
I = 2
0 = 1
BOILER
I 2
o •
A A
a •
*
DESIGN EXCESS AIR
LEVELS
0.6
0.7
Figure 25. Relationship of furnace excess air level with
blend and load.
58
-------
10
Figure 26. Grate seal leakage forcing flames toward center of furnace
during a 1:1 blend firing in Boiler No. 2.
-------
lumps of coal and the nommiformity of coal distribution onto the grate, some
of the air coming up through the grate into the combustion zone is channeled,
and it bypasses the areas where it is needed. With such channeling, pockets
of pyrolysis gas form above the grate. If these pockets of hydrocarbons pass
through the flame zone without being exposed to adequate oxygen to complete
combustion, smoking will occur. To prevent smoking, the gases must be mixed
so that they do not drift uniformly upward and out of the furnace without
combustion. These pyrolysis gases are mixed with overfire air jets. The
proper use of overfire air jets is particularly critical when a boiler is
being operated at a low fraction of load.
In order to find the proper setting for coal:dRDF operation, the overfire
jet pressure was increased in 12.7 mm (0.5 in.) of water increments starting
at 12.7 mm (0.5 in.) of water until the opacity meter showed no reduction in
smoke with increasing jet pressure. Because of time constraints and the
amount of fuel available, this procedure was iterated three times for the
front and rear rows of overfire air jets for each blend during the first day
of each test. After the minimum smoking settings were determined, they were
used in the subsequent tests.
The differences between the settings for coal-only and blend firing were
as follows: When firing blend, the air pressure applied to the front
5.08 x 1.9 cm (2 x 3/4 in.) rectangular overfire air jets was reduced from
178 to 114 mm (7 to 4.5 in.) of water while the air pressure applied to the
rear 5.08 x 1.9 cm rectangular overfire air jets was increased from 38.1 mm
(1 1/2 in.) to approximately 76.2 mm (3 in.) of water. The pressure for the
rear jets was increased because the high volatile content of the dRDF produced
a large fireball which enveloped more of the furnace volume than the fireball
from coal-only firing. The pressure for the front jets was decreased because
the fuel was burning more in suspension and further to the back of the furnace.
Consequently, there was only burnt-out fuel in the forward portions of the
grate which required less overfire air than normally needed for coal.
Regardless of the blend employed, including 100 percent pellets, the
amount of overfire air required was the same.
Induced Draft Fan—
The induced draft fans were adequate for all the tests performed.
However, the boilers were not fired above 70 percent of rating on blends
because of limitations in steam demand and make-up water equipment. Attempts
to operate at 100 percent of boiler design capacity by blowing steam were
aborted when it was determined that the make-up water equipment had insuffi-
cient capacity to handle the make-up water flow. Consequently, it could not
be determined whether the existing induced draft fans have sufficient capacity
for firing various blends up to 100 percent of rating.
A study of the fuel properties listed in Table 6 reveals that the
air:fuel ratio for coal-only and coal:dRDF blends is approximately the same
with blends requiring slightly less air per megajoule heat release than coal.
This is particularly significant because if coal and blends can be fired at
the same excess air levels, the amount of gas passing through the boiler to
the collectors and fans will be the same when the boiler efficiencies are
60
-------
equivalent. However, as discussed subsequently, the efficiency of a boiler
firing dRDF is expected to be slightly lower than the efficiency of a boiler
firing coal because of the higher hydrogen content and bound moisture of the
blend.
If the feeders do not limit the substitution ratio which can be used
without derating the boiler, then the fan capacity could become the limiting
factor. The induced draft fan, rather than either the overfire air blower or
the underfire air fan, would most likely have an insufficient capacity.
Furnace Performance
After the fuel has been mixed with air and ignited, it burns to release
the chemical energy in the fuel. The heat is recovered in two different
sections of the boiler to produce steam. First, some heat is transferred by
radiation to boiler tubes in the furnace walls. Second, additional heat is
withdrawn from the gases as they pass through the convection section of the
boiler. The overall performance of a furnace-boiler combination depends to a
large extent on the radiant heat transferred in the furnace and the removal
of sufficient heat from the combustion products so that the fly ash is solid
and not molten in the convection section. The heat transfer characteristics
of the rest of the boiler are governed by the gas properties and the mass
flow rate.
Heat Release Rate—
The design heat release rates for Boilers No. 1 and No. 2 were:
961 MJ/m3/hr (25,800 Btu/ft3/hr) and 616,000 MJ/m2/hr (543,000 Btu/ft2/hr)
for Boiler No. 1 and 1002 MJ/m2/hr (26,900 Btu/ft3/hr) and 662,000 MJ/m2/hr
(583,000 Btu/ft2/hr) for Boiler No. 2. Figures 27 and 28 show the heat
release rates attained during the tests. The maximum heat release rates were
low because of low steam demand requirements. Even in the coldest time of
the winter, only 7.1 kg/sec (56,000 Ib/hr) of steam were required to meet the
heating needs of the Maryland Correctional Institute. Consequently, the
tests do not show the effect of coal:dRDF blends on boiler performance at or
near design heat release rates.
Flue Gas Temperature—
Of particular concern to a boiler operator is the temperature of the
gas leaving the boiler. This temperature indicates the amount of potentially
available energy that is lost to the environment. Figure 29 is a graph of
the flue gas temperatures in the furnace when firing coal-only and coal:dRDF
blends. The exhaust gas temperature characteristics of Boilers No. 1 and
No. 2 differ. The exhaust temperatures also differ when firing coal-only or
when firing coal:dRDF blends for the same boiler load.
Fouling, Slagging, and Wastage—
Coal:dRDF blend firing resulted in occasional slagging, slight fouling,
and perhaps slightly higher than normal wastage. Since these are areas of
major concern to potential fuel users, a detailed discussion is presented
although the total evaluation effort is not complete or conclusive because of
the relatively short duration of testing and the low boiler loads maintained
throughout the tests.
61
-------
f-
m
500,000
400,000
300,000
200,000
! 00,000
BLEND
BOILER
I 2
o •
A A
n •
*
000
0.10
0.20
0.30
0.40
0.50
0.60
- FRACTION OF RATING
Figure 27. Heat release rates per unit grate area for
coal, blend, and dRDF firings.
62
-------
BOILER
25000
20000
15000
s—
\
13
H- 10000
m
5000
n
BLEND 1 2
| : 0 0 •
| : | A A
1 : 2 n •
• 0 : 1 *
o
A
* • ° A A
a
a
i i i i j i
0.00 0.10 0.20 0.30 0.40 0.50 0.60
jL- FRACTION OF RATING
Figure 28. Heat release rates for coal, blend, and dRDF firings.
63
-------
"t O
co in
CM
O O
CD O
CM ID
LU
ID O> O
pi CM in
< cj ri-
ce
LU
Q.
LU
CO •* O
< o o-
CNJ rj-
o o
N. in
i- CO
1 - MFG PROJECTION
0.00 0.10 0.20 0.30 0.40 0.50
•£ ~ FRACTION OF RATING
0.60
0.70
0.80
Figure 29. Effects of blend and load on flue gas temperature.
Since the boilers were operated at low loads throughout the tests,
furnace temperatures were not high. Even under these conditions, the boiler
deficiencies were corrected with simple adjustments. The findings imply that
a boiler owner should not anticipate immediate failures. The long-term
effects, however, are as yet unknown.
When pellets impinge on a wall, they tend to stick and burn because of
the low fusion characteristics of the dRDF. For fuels with a higher ash
fusion temperature, the material rebounds and falls to the grate where it is
burned. Severe slagging occurred during a portion of the May tests when
one-third of the rear wall of the furnace opposite Spreader No. 1, which was
maladjusted, was covered with slag. The remaining two-thirds of this wall
remained clear. This slag was generally loose and could be easily removed.
Fouling also accumulated rapidly on the leading rows of the convection section
of the boiler in line with the maladjusted spreader. The remaining two-
thirds of the convection section remained clean. The fouling was very loose,
was easily removed by rapping, and had a porous structure. At the end of the
blend firing, the/leading tube elememts had a velvet-like ash accumulation.
The same type of coating was found at the conclusion of the low-grade coal
burning in December. The potential hazard of this coating cannot be assessed.
64
-------
After the spreader throw on Spreader No. 1 was adjusted, the accumulation
ceased. By the time the boiler was brought off line 8 days later for inspec-
tion, the slag had sloughed off. Inspection of the furnace interior revealed
that a glassy slag layer had accumulated on the lower portion of the side
wall where a grate clinker had contacted the wall. This material was easily
removed.
The reducing atmosphere hemispheric fusion temperatures for the ash
processed through the system for the December, March, and May tests are
graphed in Figures 30 through 32. The low December ash fusion temperatures
for coal explain the clinkering observed. The March and May results show
that blending coal and dRDF depresses the fusion temperature of the coal.
Interestingly, the dRDF-only bottom ash shows a higher fusion temperature
than coal. If the glass fragments blow out of the furnace while the fuel is
in suspension and leave only paper ash, a higher fusion temperature is
possible.
O u_
o o
? O
CO «
t- CM
Q.
LU
™
8
0) O
Tf O.
CO Q
O> O
O O'
»- eg
71
EOT TOM ASH
COAL
1:0
1:1
BLEND
Figure 30. Variations in ash fusion hemispheric temperatures under
reducing atmospheric conditions during December runs.
65
-------
0.
01
o u.
o o
(Q O
" CM
o
CD
CM
3 8
CM PJ
»- CM
§ 8_
go o
o'
CM
COAL
BOTTOM ASH
RE1NJECTOR ASH
DRDF
1:0
1:1
BLEND
1:2
Figure 31. Variations in ash fusion hemispheric temperatures under
reducing atmospheric conditions during March runs.
66
-------
O
CM O
00 O
^ O
l~- O
ro m'
•*- c\
H .DR«,E
|g
-------
In Figures 33 and 34 the flue gas temperature plots before and after the
boiler was fired with blends show that there was some change in the character-
istics of the heat transfer surface. Because the "after" data lie outside
the confidence band for the "before" data, the heat transfer surface had
deteriorated somewhat. The blend ash was more insulating and/or more fouling
than the coal-only ash. The increase in exhaust gas temperature was greater
after 48 hours of blend firing in March than after 136 hours of blend firing
in May. It should be noted however that because different coals were used
for the two tests, the fouling may be related to the mixture properties and
not just to the presence of dRDF.
03
LU
O
LU
CC
H-
cc
LU
CL
LU
LO O
T- O
m CD
o o
CD O
"* 0
O O
OJ -t
o> o
rf O
T- CO
8-L
• o MARCH 31 & APR. 1
0.00 0.10 0.20 0.30 0.40
^ - FRACTION OF RATING
0.50
0.60
Figure 33. Effects of blend and load on flue gas temperature
before and after a test.
68
-------
s> s-
I? 8
co co
§ 8
CM m
C3 -to
MJ O O
g OJ -t
LL
O
UJ
tr
I
£
0.
UJ
o>
2 8
CNJ
00 §
co O
CO
00
• MAY 4 & 5
o MAY 16 & 17
0.0 0.10 0.20 0.30 0.40
£ —FRACTION OF RATING
0.50
0.60
Figure 34. Effects of dRDF and load on flue gas temperature
before and after a test.
During the May tests, eight corrosion test specimens were installed on
the downcomers in the rear screen wall 1.52 m (5 ft) above the fuel bed.
These specimens were clamp-on erosion shields similar to the one shown in
Figure 35. Since the shields were bolted in place before bringing the
boiler on line, they were exposed first to coal-only firing (1:0); then to
1:1, 1:2, and 1:1 blend firings; next to 0:1 (100 percent pellets) firing;
and finally to coal-only firing again. At the end of 478 hours of exposure,
the specimens were removed from the furnace and cleaned by the procedures
described in Corrosion Engineering.2 The weight loss of the specimens was
converted to a wastage rate in mils per year by Equation 4.4.
2Fontana, M. G., and M. D. Green.
N.Y.C., 1967.
Corrosion Engineering. McGraw-Hill,
69
-------
25.4 cm (10 in)
SHIELD FABRICATED FROM
111,304,309,316 S.S.
Figure 35. Drawing of typical clamp-on corrosion test shield.
MPY =
534 W
DAT
(4.4)
where W = weight loss, mg
D = density of specimen, g/cm
ip
A = area of specimen, in.
T = exposure time, hr
The wastage rate was less than 127 micrometers (5 mils) per year for
carbon steel while the wastage (weight loss) of some stainless steel specimens
(309 and 310) was not even detectable.
One specimen of 1018 cold rolled steel was in line with the maladjusted
spreader. It had a relatively high wastage. Since the surface of this
material was exposed to burning material, it was in a strong'reducing atmos-
phere. A twin test specimen not subjected to fuel impingement had a wastage
rate of only 76 micrometers (3 mils) per year. Table 8 lists the data used
to calculate the wastage rates. While the data provided by this corrosion
test provides some guidance, further corrosion testing needs to be carried
70
-------
TABLE 8. METAL WASTAGE RATE DATA FOR EIGHT SPECIMENS
Specimen
No.
1
2
3
4
5
6
7
8
Material
1018
304
1018
304
309
310
309
310
Location
1
2
3
4
5
6
7
8
W
mg
19310
3620
2007
4111
n/d
50
240
n/d
ID
g/cm3
7.86
8.02
7.86
8.02
8.02
8.02
8.02
8.02
(&)
100.00
98.75
98.75
97.50
99.38
100.00
98.75
98.75
T
hr
478
478
478
478
478
478
478
478
Mils/y
2700*
5.00
3.00
6.00
n/d
0.07
0.34
n/d
mm/y
.69
.13
.08
.15
—
.002
.009
—
On.Line
Off Line
1400
1200
April 27, 1977
May 17, 1977
Elapsed Time
478 Hours
*Specimen in line with spreader spraying fuel directly on the back
furnace wall.
o
8
12 16 20
NO. OF TUBES -
1.52 m
(5 ft.)
24 28 30
71
-------
out over longer periods of time (6 months to 1 year) with the boiler operating
at or near rated capacity.
Firing Phenomenon—
During the combustion of a series of different mixtures, the coal:dRDF
blends seemed to perform as well in the furnace as coal. The principal
differences at equivalent loads were that as the dRDF substitution increased,
the height, intensity, volume, and violence of the fireball increased
correspondingly. Moreover, because of the intermingling of paper platelets
with the fuel, more sparklers or live pieces of glowing char were carried by
the combustion products toward the heat transfer elements. These sparklers
increased as additional dRDF was introduced into the furnace.
Figure 36, a set of photographs taken through a port in the side wall of
the furnace 3.3 m (10 ft) above the grate, shows the extent to which the
flame filled the furnace volume. As seen in the progressive photographs, as
the dRDF subsitution ratio increased, the number of sparklers and the height,
intensity, violence, and volume of the fireball increased.
Figure 37, a photograph taken through the tube removal hatch in the top
of the furnace, shows the fire distribution for a 1:1 blend. As seen in this
figure, the furnace front has a relatively good burnout, and the furnace rear
has a square fire line. The two zones of high-intensity flame in the fire
bed indicate that the fuel was spread in two distinct waves: one at the rear
of the furnace and the other in the middle of the furnace. This nonuniform
spreading was not detected during the cold flow test. However, it did not
seem to have any impact on the degree of burnout. The lighter fire to the
front left was the result of starving Spreader No. 2 to rectify a worn r6tor
deficiency which caused an excessive left throw of the fuel.
The radiant heat transfer characteristics of the fireball probably
improved as more dRDF was substituted for coal. Figure 38 shows flame
temperature versus load as a function of blend. The flame temperature was
measured by a Leeds & Northrup optical pyrometer focused on the middle of the
fireball at the center of the furnace. Since an optical pyrometer measures
the product of emissivity and temperature, an increased reading would indicate
that the furnace radiant heat transfer characteristics have improved (assuming
excess air remains constant). Consequently, the data indicates that substi-
tuting dRDF for coal would likely have a slightly beneficial effect on the
heat absorbing capacity of the furnace.
Boiler Controls
The boiler controls in the Hagerstown plant are typical of those installed
in heating plants in the mid 1960's. They consist of a master controller
which modulates the fuel supply in response to changes in boiler pressure.
As the steam demand on the facility increases, the pressure in the steam
distribution system decreases, and the amount of fuel introduced into the
furnace is increased. The underfire air is modulated in response to fuel
flow. In response to load changes, the cam connecting the underfire air to
the fuel flow causes the air to modulate to "optimum" firing conditions.
Consequently, when there is a load shed, the airflow drops while there is
72
-------
1:1 Blend Firing
1:2 Blend Firing
0:1 dRDF Firing
Figure 36. Furnace flames viewed at 3.3 m (10 ft) above the
grate during blend and dRDF firings.
73
-------
BOILER REAR WALl
COMBUSTION
GAS
SAMPLING
PROBE
Figure 37. View from top tube hatch in Boiler No. 2
show firing with a 1:1 blend.
to
74
-------
co
LU
I
LL
U-
o
UJ
cc
I
CM
CD O
*- O
CD •*
*- CM
O
•* O
co o
o> o
o o
BLEND
1:0 -
1:1 -
1:2
0:1 -
BOILER
1 2
o •
& A
O •
*
0.00 0.10 0.20 0.30 0.40
•£ — FRACTION OF RATING
0.50
0.60
Figure 38. Effects of blend and load on flame temperature
measured with an optical pyrometer.
still fuel in the furnace for the higher load level. During such a load
change the fuel bed has reducing conditions. Conversely, during a load gain
the air input leads the fuel increase, and the fuel bed has oxidizing con-
ditions. This type of control network can cause problems when the fuel on
the grate is a low fusion coal or a coal:dRDF blend. During the co-firing
tests, some clinkering occurred during load sheds. This clinkering was
overcome in the tests by biasing the underfire air control upward. With a
combustion control system designed to maintain constant oxygen levels in the
flue gas, this clinkering should not occur.
The control system and feeders allowed the boiler to follow the load
without any discernable difference in pressure fluctuations in the header
when firing coal only and a 1:1 blend. Circular charts for the steam pressure
are shown in Figure 39. The modulations in steam pressure were minor for
coal only and a 1:1 blend; however, when a 1:2 blend was fired, the feeders
were volume-limited and, as a result, had a lag such that the pressure
modulated 7 to 14 kPa (1 to 2 psi) in a sawtoothed pattern.
75
-------
Figure 39. Pressure chart recordings for coal, blend, and dRDF firings.
76
-------
When 100 percent pellets were fired, the feeders were the rate-limiting
component. The 9.9-kg/sec (78,500-lb/hr) boiler could maintain 7.1 kg/sec
(56,000 Ib/hr) of steam per hour with 100 percent pellets and manual airflow
control. When the 7.6 kg/sec (60,000-lb/hr) boiler burned 100 percent pellets
while using the automatic air/fuel ratio controller, unburned fuel accumulated
at the base of the rear wall, the entire fuel bed had a very sparse fire, and
the steam pressure dropped 413 kPA (30 psi). This condition was due princi-
pally to high excess air levels. Because of the control system limitations,
the air control could not be adequately biased downward; consequently, there
was a 600 percent excess air level in the furnace. The difficulty with the
boiler control was easily resolved by taking the underfire air modulation
control off automatic to allow an operator to manually control the airflow
and to track the steam load rather than the fuel feed rate. With the air
control in the manual mode, steam pressure variations were about 14 kPa
(2 psi) with the air supply fixed for 12 percent C02 at the peak of a load
swing.
Mass and Energy Balance
Table 9 summarizes the boiler efficiency data which were calculated by
the AMSE Short Form loss method. The complete forms are included in
Appendix B.
TABLE 9. HEAT BALANCE SUMMARY BASED ON AS-RECEIVED FUEL
BLEND 1:0 1:1 1:2 0:1
PARAMETER
Fraction of Rating .17 .33 .30 .19
Excess Air (%) 104 82 99 113
LOSSES
Dry Gas 17.9 13.7 17.8 19.4
Fuel Moisture .1 .9 1.2 4.0
H20 for H2 Combustion 4.0 5.1 5.4 8.1
Combustibles in Refuse 18.3 25.3 16.6 3.0
Radiation 3.7 1.8 1.8 3.7
Unmeasured 1.5 1.5 1.5 1.5
TOTAL 45.5 48.3 44.1 39.7
EFFICIENCY 54.5 51.7 55.9 60.3
77
-------
The high carbon losses in the refuse are most unusual and account for
the extremely low efficiencies (normal spreader-feeder efficiencies are
between 74 and 80 percent). The carbon content of the refuse (bottom ash,
collector ash, and fly ash) for the various coal:dRDF blends is shown in
Figures 40, 41, and 42. The carbon content of the three ash streams varied
as follows: (1) bottom ash, 2 to 10 percent; (2) collector ash, 50 to
90 percent; and (3) fly ash, 30 to 40 percent.
BOTTOM ASH
%c
3LENC
I •• 0
BOILER
i 2
NOTE SiVPLES COLLECTED
FROM THE FRONT
OP THE GRATE
o
•
A A
O
0.20
0.30
0.40
0.50
0.60
0.70
0.80
- FRACTION OF RATING
Figure 40. Effects of blend and load on carbon content of bottom ash.
78
-------
COLLECTOR ASH
%C
100
80
60
40
20
000 0.10
A
A
BLEND
i : 0
BOILER
I 2
o •
A A
a •
SAMPLES COLLECTED
^ROV THE DRAI\ TUBE
0-20 0.30 0.40 0.50
«£ - FRACTION OF RATING
0.60 0.70
0.80
Fieure 41. Effects of blend and load on carbon content of collector ash.
79
-------
FLY ASH
%C
100
80
60
40
20
0
0.00 0.10
BLEND
I --0
I •• \
I : 2
0 = I
BOILER
I 2
o •
A A
D •
NOTE'- SAMPLES COLLECTED
AT THE STACK
I
020 0.30 0.40 0.50
£- FRACTION OF RATING
0.60
0.70 0.80
Figure 42. Effects of blend and load on carbon content of fly ash.
80
-------
The mass balance for the various blends is presented in Table 10. This
balance indicates that an abnormally high amount of ash was removed by the
collector. Analysis of fly ash samples taken from the collector revealed
that 85 to 90 percent of the particles were larger than 50 micrometers.
Since collectors have much higher efficiencies for particulates in this size
range, the higher collector weights may be justified. However, such an ash
weight distribution is not typical of expected boiler performance.
TABLE 10. ASH MASS BALANCE
Fuel
Blend Flow
Test Coal:dRDF kg/hr
=/
Ash
in
Fuel
Ash in
Fuel
kg/hr
Bottom Ash
kg/hr
Carbon With
Free Carbon
Fly Ash
kg/hr
Carbon With
Free Carbon
Collector Ash*
kg/hr
Carbon With
Free Carbon
May 4, 1977 1:0 872 21.9 191 82 89
May 13, 1977 1:1 1489 23.3 347 232 238
May 11, 1977 1:2 2035 23.4 476 324 341
7.7
6.8
10.2
104
110
145
219
369
300
* The collector weight was determined by difference.
In addition, the carbon content of the bottom ash in the various blends
varied little. When the dRDF substitution was increased, the fly ash burnout
was improved.
Figure 41 indicates that the carbon content of the collector ash dropped
significantly when the amount of dRDF was increased in the 1:1 to the
1:2 blend firings. Since this drop effectively offsets the wet flue gases
losses, the boiler efficiency did not change appreciably as more dRDF was
substituted for coal. As a result, the boiler efficiency had only minor
differences over the various coal:dRDF blend ratios tested. This fact is
confirmed when the input/ouput efficiency data are plotted as in Figure 43.
To determine the input/output efficiency, the fuel in the feed trough was run
out and the steam integrator read. After the feed trough was refilled with
weighed fuel, it was again emptied and the final integrator reading determined,
Dividing the fuel heat content into the heat content of the steam yields the
efficiency. Even though the steam meter was calibrated, the differences in
efficiency when comparing those calculated by the loss method with those
computed by the input/output method were significant enough to suggest a
constant multiplier error in the steam meter. Hence, efficiency values are
not provided in Figure 43. The input/output efficiency data shown in
Figure 43 confirm the heat loss calculations in that there is no distinguish-
able difference in the efficiencies when firing blends or coal. This
conclusion is unique to the boiler installation at MCI since the large amount
of ash and unburned carbon losses attributed to the collector is certainly an
anomally to expected boiler performance.
81
-------
INPUT/OUTPUT
EFF.-%
1 DIV= 10%
BOILER No. 2
BLEND
1:0 • MAY
1:0 o MARCH
1:0 a JANUARY
1:1 A MAY
1:2 a MAY
0:1 * MAY
O Q Q
Q
L
00 0.10 0.20 0.30 0.40 0.50
•^ - FRACTION OF RATING
0.60
Figure 43. Effects of blend and load on input/output efficiency.
Low Load Performance
A positive result of the tests was the substantially improved low-load
performance and the decreased plume opacity (indicative of better burnout)
when dRDF was substituted for coal.
Normally a 3:1 or 4:1 turndown ratio on spreader-feeders is considered
the practical limit to avoid severe smoking. With the use of dRDF, this
turndown ratio was increased. Although of lesser importance for base-loaded
industrial plants, the demonstration of the boiler's ability to operate at
extremely low loads is particularly advantageous for heating plants and
institutional facilities which must support a very small summer base load yet
have sufficient capacity to meet severe winter heating requirements.
SUMMARY
The Hagerstown experience has increased the knowledge of blend behavior
in a spreader-feeder. The fuel entered the furnace satisfactorily, burned
82
-------
well, and met plant energy requirements. The operative and control dis-
crepancies were all corrected by simple adjustments. Some biasing of the air
controls was required to prevent slagging on the fuel bed during load sheds.
The only other limitations on the boiler operation occurred when the boiler
was operated on 100 percent dKDF. During this test both the spreader and the
ash handling systems became capacity-limited.
Proper adjustment of the spreader-feeders is critical to prevent slagging
and fouling. Some slagging and fouling occurred (in excess of what would occur
when firing with coal only) on the walls slightly above the grates but was
readily removed. The corrosion experiment resulted in wastage comparable to
what might be expected for coal-only firing. This test was too short in
duration, however, to permit any definite conclusions on material wastage.
While the boiler performance when firing coal:dRDF blends up to 1:2 (by
volume) was generally similar to that when firing coal only, final conclusions
must await further long-term demonstration testing in which boiler loads can
be established at rated capacity and satisfactory boiler operating character-
istics can be maintained.
A follow-on demonstration test is currently scheduled to be carried out
in a spreader-feeder fired boiler with a rated capacity of 18.9 kg/sec
(150,000 Ib/hr) and superheat capability.
83
-------
SECTION 5
ENVIRONMENTAL PERFORMANCE
INTRODUCTION
If a boiler plant can comply with emission regulations when firing coal,
the question arises whether it can still do so when substituting dRDF for
some of the coal. The low sulfur content of the dRDF could degrade pre-
cipitator performance which, when coupled with the high ash content of dRDF,
could result in the emissions exceeding the regulation limits. Also, the
high chlorine content of the dRDF raises concern about the long-term corrosion
effects on the boiler system.
To answer this question, the current study assessed changes in emissions
as dRDF was substituted for coal. Although the quantitative results for the
coal:dRDF emissions are significant, the principal conclusions are drawn by
comparing the coal:dRDF blend emissions with coal-only emissions. Therefore,
base lines were established before and after each blend run by duplicating
all test conditions for coal-only firing.
This section discusses the method of sampling and data analysis. It
also covers the impact of substituting dRDF for coal on particulate emissions,
gaseous emissions, and trace compound emissions in the order given.
FIELD SAMPLING SETUP
Since at least four 2-week tests were to be conducted at the Hagerstown
plant, a weatherproofed test shed enclosed in sheet metal and readily
accessible to the stack on two sides (see Figure 44) was constructed for the
environmental testing. Placed 4.9 m (16 ft) above the roof line, the shed
encompassed one quarter of the single stack that served all three boilers.
Four 0.1-m (4-in.) half couplings were installed in the stack to serve as
sampling ports, two on the north side and the other two on the west side,
with each pair arranged vertically with a 0.61-m (2-ft) separation. A stair-
well from the plant catwalk system provided access to the shed. A trans-
missometer was installed in another pair of ports spanning the stack on the
north-south axis at about 0.9 m (3 ft) above the roof line and 4.6 m (15 ft)
below the other sampling ports.
One port in the stack was used to insert a sintered steel filter through
which stack gas samples were extracted, cooled in a condenser/knockout box,
and piped through heated umbilicals to SYSTECH's sample analysis trailer.
This trailer housed facilities for wet chemistry analyses. In the trailer,
the samples were distributed through a manifold to provide flue gas to the
84
-------
Ill ill
^PHK.^-^
Figure 44. View of MCI power plant showing stack sampling shed
and temporary fuel handling system at right.
85
-------
gas analyzers. A Theta Sensor, Inc., trigas analyzer was used for continuous
oxygen, nitrogen dioxide, and sulfur dioxide measurements. A slip steam was
also distributed to an AID gas chromatograph to determine both total hydro-
carbons and the composition of hydrocarbons lighter than C6. The manifold
was also used to distribute calibration gases.
Another sampling port was used to insert a 1/4-inch-diameter steel pipe
into the center line of the boiler outlet. The pipe was connected with
copper tubing to an Orsat analyzer located on the boiler house floor. This
sample system was used to determine the characteristics of the flue gas
leaving the boiler. Figure 45 illustrates the overall arrangement of the
boilers, breeching, and locations where the various ash, fuel, and flue gas
samples were taken.
TEST CHRONOLOGY AND PROCEDURES
Table 11 illustrates the test chronology for the entire program. The
consumption of dRDF throughout the program was 20.9 Mg (23 tons) in December,
106.1 Mg (117 tons) in March, and 127.9 Mg (141 tons) in May for a grand
total of 254.9 Mg (281 tons). During the May test, Boiler No. 2 was con-
tinuously fired with coal:dRDF blends for 132 hours. The test time breakdown
for the blend firings was 58 hours for the 1:1 blend, 53 hours for the
1:2 blend, and 29 hours for the 0:1 blend (100 percent dRDF). The table
lists the fraction of the boiler load rating, the coal:dRDF ratio, and the
number of emission data measurements for each test blend. The test program
was designed so that three sets of emissions data would be acquired for each
test blend. While good boiler testing practices would dictate that the
boiler be stabilized for 24 hours on each blend before collecting emissions
data, the limited supply of dRDF necessitated that the stabilization period
be limited to overnight (approximately 12 to 15 hours). Each test blend was
subjected to the following battery of emission tests:
3 each - EPA Method 5 - Particulate mass flux, Cl,
F, S02, S03, and trace organic
and inorganic compounds
3 each - Cascade Impactors - Particulate size distribution
2 each - EPA Method 7 - Nitrous oxides
6 each - Orsat - C02, 02, and CO
8 each - Orsat - CO
2
4 each - AID Gas - Total hydrocarbons
Chromatograph
4 each - Wahlco Probe - Resistivity
1 each - Tedlar Bag - Record sample
86
-------
00
NO it SAMPLING PORTS FOR PARTICIPATE (FLUX
AND SIZ.E)SOX,NO,,02,CO CON-LINE AND
ORSATj HYDROCARBONS, CL,F, METALS AMD
RESISTIVITY ARE LOCATED IN THE STACK
26 FT. ABOVE "PHE BREECHING CENTER-
LINE.
A3H SAMI'Lf 5<
' JJ ID FANS
Figure 45. Layout of Boilers No. 1, 2, and 3 with sampling
locations indicated.
-------
TABLE 11. CHRONOLOGICAL LISTING OF TEST CONDITIONS
Date
12/8/76
12/10
12/13
12/14
1/20/77
1/21
1/24
1/25
3/19
3/21
3/22
3/23
3/24
3/28
3/29
3/30
3/31
4/1
5/3
5/4
5/5
.w 5/10
''4 5/11
5/12
5/13
5/14
5/16
5/17
Boiler Load
Boiler No. Fraction of Rating Blend Data Replication
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
.45
.45
.43
.53
.47
.53
.50
.51
.40
.40
.43
.44
.38
.36
.26
.27
.35
.25
.20/.22
.31/.51
.36/.30
.36/.34
.347.28
.397.30
.277.26
.217.36
.17
1:1
1:0
1:1
1:1
1:0
1:0
1:0
1:0
1:0
1:0
1:0
1:1
1:1
1:0
1:2
1:2
1:0
1:0
1:0
1:0
1:0
1:2
1:2
1:1
1:1
0:1
1:0
1:0
1
2
2
2
2
1
1
1
2
2
1
2
1
1
2
1
2
1
1
2
2
2
2
2
2
2
2
2
Note: Boiler No. 1 is rated at 9.9 kg/sec (78,500 Ib/hr) and
Boiler No. 2 is rated at 7.6 kg/sec (60,000 lb.hr).
88
-------
Continuous - Theta Sensors, Inc. - C02, S02> and N02
Trigas Meter
Continuous - Leads & Northrup - Opacity
T r ansmi s some t e r
As Req'd - Draeger Tubes - CO, C02, and SOX
The Battelle Tenex traps were only run three times throughout the test,
i.e., at 1:0, 1:1, and 1:2 test blends. When the measured emissions produced
unusual results, the Tedlar bag sample was used to further clarify the data
through gas-chromatograph analysis.
The Draeger tubes were used periodically throughout the program whenever
a quick, approximate concentration of a particular pollutant gas was desired.
The following paragraphs detail the test procedures for collecting,
monitoring, and analyzing each emission.
Opacity
Opacity was measured by a Leads & Northrup single-pass transmissometer
spanning the stack and calibrated with neutral density filters. The trans-
missometer was calibrated with the neutral density filters by installing the
unit in a pipe section whose length was equivalent to the stack diameter.
Particle Mass Flux
The particle mass flux was measured by an EPA Method 5 train, which is
schematically illustrated in Figure 46. Because of the low gas flow rates,
nominal 12.7-mm (1/2-in.) nozzles were installed to produce nominal
0.0047-m3/sec (1-cfm) flow rates through Greenburg-Smith impingers. The
3-m (10-ft) diameter stack was traversed from two sides by a single 3.7-m
(12-ft) stainless steel probe.
The stack was sized for acceptable flow rates with all three boilers
operating simultaneously. During testing, however, only a single boiler was
on line at part load. Consequently, the velocity of the stack gases was less
than 2.1 m/sec (7 ft/sec) which is below the detection limit of an S-type
pitot tube. Although sophisticated velocity monitoring equipment was con-
sidered, the large quantities of dust made its application inappropriate. As
an alternative method for measuring the flow rates, the flue gas composition
at the stack, the boiler load and efficiency, and the ultimate analysis of
the fuel were used to calculate the velocity. The mathematics relating these
parameters and the apparent pitot reading are presented in Appendix C.
Because of the velocity conditions at the stack, the extent of aniso-
kineticism could not be determined precisely. In any event, the error due to
anisokinetic sampling in these experiments is likely negligible because of
89
-------
ffn
'/,
PROBE
HEATED FILTER HOLDER
AREA
THERMOMETER
CHECK
VALVE
i__ :_i_zi
REVERSE
TYPE
PITOT
TUBE
1
PI TOT
MANOMETER
IMPINGERS
THERMOMETERS
ICE BATH
ORIFICE
BY-PASS
VALVE
VACUUM
GAUGE
T
MAIN
VALVE
VACUUM LINE
MANOMETER
DRY TEST METER AIR TIGHT PUMP
Figure 46. Schematic of EPA Method 5 sampling train setup.
-------
the fine particle sizes and low gas velocities.3 As a general rule, when
an aerosol is less than 5 micrometers in diameter, there is no need for
isokinetic sampling. At Hagerstown, approximately 65 percent of the par-
ticulates were less than 5 micrometers. In view of the large amount of less
than 5-micrometer-diameter particulates present in the flue gas and the
utilization of a calculated stack velocity, the particulate emissions should
be representative of actual plant operation.
Size Distribution
The size distribution of the aerosol emitted from the Hagerstown plant
was monitored in the stack downstream of the multiclone collectors. All
measurements were made with an MRI cascade impactor which was operated
according to the manufacturer's recommendations. The method of forming a
nonrebounding substrate for the impactor plates was developed at SYSTECH
according to other researchers' experience with the MRI cascade impactor.
After the impactor plates were dipped in benzene in which Apiezon-H grease
was dispersed, they were baked overnight at 232°C (450°F). The resulting
coating was extremely uniform. "Blank" test runs were performed in which the
MRI cascade impactor was inserted in the stack with a filter installed before
the impactor. These tests confirmed that the coating on the impaction disks
did not come off on the 0-rings or during handling. The substrate forming
method, therefore, proved to be an acceptable procedure. Consistent particle
size distributions, as shown on the stages in Figure 47, further substantiate
the validity of this method.
The MRI cascade impactor was always inserted 1.5 m (5 ft) in from the
west side wall. After the impactor was heated, it was connected to the train
shown in Figure 48, checked, and inserted into the stack. When the sampling
was completed, the impactor was removed from the stack and disassembled. The
impactor plates were placed in tared petri dishes. After the plates were
returned to the laboratory for final weighing in a clean environment, the net
weight gain per stage was used to determine the cumulative mass distribution.
The characteristic aerosol diameter of each stage was obtained from the
factory-supplied calibration curves for the test conditions and unit density
aerosols.
Fly Ash Resistivity
The particle resistivities were measured on site with a WAHLCO probe. A
sketch of this probe is shown in Figure 49. Dust samples were cyclonically
captured from the stack gases in a collector cup. The captured particles
collectively became a resistor between electrodes A and B. After a constant
voltage was applied to the electrodes and the current flow was measured, the
resistivity was computed by substituting the voltage and current values in
3Watson, H. H. American Hygiene Association Quarterly, Volume 15
1954. p. 21.
91
-------
NJ
Figure 47. Typical dust loading of MRI cascade impactor stages during 1:1 blend firing.
-------
CHECK
VALVE
CASCADE
IMPACTOR
HEAD
GAS
FLOW
/s
VACUUM LINE
MANOMETER
DRY TEST METER AIR TIGHT PUMP
Figure 48. Schematic of MRI cascade impactor sampling train setup.
-------
CYCLONE BODY
COLLECTING
CUP AND
ELECTRODE
TO ASPIRATION
ASSEMBLY
CAP WITH STATIC
DISCHARGE SHIELD
PIN AND PURGE
AIR FITTING
@ - ELECTRODE PIN
(S>- DISCHARGE PIN
Figure 49. Schematic of WAHLCO resistivity probe assembly.
Equation 5.1. Particle resistivities were continuously checked through the
blend and coal-only tests.
Resistivity = (voltage/current) L
(5.1)
where L = constant
S02
Sulfur dioxide levels in the flue gas were determined by two different
techniques. With one technique, S02 levels were continuously monitored by a
trigas meter manufactured by Theta Sensor, Inc. This electro-chemical sensor
was calibrated with standard S02 gases. With the second technique, wet
chemistry determinations were made by analyzing the sulfur level in the first
impinger of the EPA Method 5 train. A 10 percent hydrogen peroxide/water
solution in the impinger was titrated with barium perchlorate to yield a
measure of the SOX in the flue gas. Both samples were collected at the same
location; i.e., the stack. When the results from the two techniques were
cross correlated, the corresponding values proved to be similar. Since the
SOa concentration is an order of magnitude less than the S02 concentration in
a flue gas stream, this cross correlation was considered to be valid.
94
-------
Oxides of Nitrogen
The emissions of oxides of nitrogen were determined by two different
techniques. In one technique, the Theta Sensor trigas monitor was employed
and in the second, an EPA Method 7, a phenol disulphonic test method, was
used.
Halogens
During each particulate mass flux evaluation, the halogen emissions were
simultaneously determined by replacing the water in the first impinger in the
EPA Method 5 train with a 10 percent hydrogen peroxide solution. After the
Method 5 testing was completed, the first impinger was then analyzed for
chlorine and fluorine by using specific ion electrodes.
Oxygen
In addition to monitoring SO and NO , the Theta Sensor trigas monitor
had the capability of continously monitoring oxygen. All three of these
parameters were continuously recorded on a strip chart. Standard calibration
gases were used to calibrate the instrument at the beginning and end of each
test day. In addition, Orsat readings were taken at the stack, and the 02
readings were cross correlated with the 02 readings on the Theta Sensor trigas
monitor. Identical readings verified the integrity of the sampling line.
Hydrocarbons
To determine the emissions of hydrocarbons from the Hagerstown plant,
continuous gas samples extracted from the stack were passed through a sintered
steel filter, a condenser knockout box, and a heated umbilical to the SYSTECH
trailer where they were analyzed in the AID flame ionization detector-equipped,
field-portable gas chromatograph. This instrument was operated in the total-
izing mode and calibrated with a methane gas. Consequently, the total hydro-
carbons are expressed in terms of a methane equivalent. The gas samples were
fractionalized with a molecular sieve column capable of distinguishing between
various hydrocarbons lighter than C6.
Trace Organic Emissions
The trace organic emissions were collected by a Battelle Tenex sampler.
Figure 50 illustrates how the sampler was connected after the filter in the
EPA Method 5 train. The sampler was maintained at 50°C (122°F) by a recir-
culating water bath. After the samples were taken, the probe washes, filters,
and Tenex traps were all preserved and sent to the Battelle Columbus Labora-
tories for analysis.
Trace Inorganic Emissions
The inorganic compounds emitted from a boiler can be in either the
aerosol or the gaseous phase. To quantify these emissions, SYSTECH modified
the standard EPA Method 5 train so that while the train used the normal
hardware, the first impinger was loaded with a 10 percent hydrogen peroxide
95
-------
CHECK
VALVE
BATTELLE TENEX TRAP
REVERSE
I TYPE
1 PITOT
VACUUM LINE
MANOMETER DRY TEST METER AIR TIGHT PUMP
Figure 50. Schematic of Battelle Tenex sampling train setup.
-------
solution, and the next two impingers were charged with a catalyzed ammonia
persulfate reagent. These impinger solutions were used to ensure the capture
of the gas phase metals.
The probe washings, filters, and impinger solutions were all analyzed in
SYSTECH's laboratories by using atomic absorption spectrometry with appro-
priate detectors and furnaces. Each filter was cut in half, dryed, and
desiccated. The half to be digested was then weighed for total particulate.
The remaining half was retained for voucher or repeat analysis. One to two ml
of concentrated HN03 were added to each filter and dried slowly on a heated
steam table. The filters were not dried completely to avoid losing volatile
metals such as Pb. Additional HN03 was then added to cover the sample. The
sample was refluxed until the digestion was complete. The reflux was removed
from heat, and after the sample had cooled, concentrated HC1 was added until
the sample was in solution. After the sample was filtered, the filtrate was
brought to a known volume for atomic absorption analysis. Filter blanks were
analyzed along with the samples to provide supporting data.
Fly ash samples were dried, desiccated, and then halved' and quartered to
obtain a representative 2-g sample. The bottom ash samples were ground up to
a minimum sieve size in a Wiley mill; then the ground samples were halved and
quartered, and a 2-g aliquot was taken for analysis. These samples were
digested by the same procedure as previously described for the filters. With
the high silica content, it was difficult to dissolve the entire sample.
Consequently, the digestion was considered complete when the sample had a
straw-like color after a minimum of 2 to 3 hours of refluxing.
For the Hg analysis, a separate digestion was necessary. Each sample
was weighed and put into a BOD bottle. The procedure for analyzing Hg
required a persulfate digestion as described in EPA Manual of Methods for
Chemical Analysis of Water and Wastes. All samples were analyzed in duplicate.
DATA ANALYSIS AND NORMALIZATION
Data Analysis
The data management and interpretation was complicated by the test
program and load limitations; for example, only one boiler would be on line
at a time. Since the single on-line boiler had to follow a modulating steam
load, the test matrix had to include the steam load as a variable. Conse-
quently, data were taken at various load points as the boiler met the varying
steam needs.
Since soot formation, flame temperature, boiler efficiency, hydrocarbon
emissions, etc., are all functions of the boiler firing rate, all data were
analyzed as a function of the boiler rating to minimize the effect of testing
in two different size boilers. Because of the limited number of data points
and the similarity of the data taken on Boilers No. 1 and 2, all data were
combined.
A method of data interpretation different from usual techniques had to
be employed to identify the effect of substituting dRDF for different test
97
-------
coals with varying coal chemistries. The selected method was based on the
principle of establishing the range of likely values for the reference
parameter and then applying that range to the dependent parameters to ensure
meaningful results.4 Accordingly, with this method a regression analysis was
conducted with 90 percent confidence intervals for the best fit lines based
on the coal-only (1:0) data set. Figure 51 illustrates the principle. After
the data for two different blends were independently regressed and the con-
fidence interval of the regression line for the first data set (A) was
established, the two data sets could not be called different if the regression
line for the second data set (B) was within this interval. However, if the
regression line for the second data set (C) was where it would be outside
this interval, the two data sets could be called different.
Because of the marked disparity in the number of data points for the 1:0
and 0:1 blend tests (25 versus 2 data points), the slope of the 1:0 blend
data as a function of load was used as the slope for all other emissions data
graphically presented in this report. A t-test procedure described by
Natrella5 was used to confirm the reasonableness of this approximation.
Results of this analysis indicated that at the 90 percent confidence level,
the slopes of the 1:0 and 1:1 data sets would not be considered different.
In fact, even though sloped lines are shown for each blend (based on the 1:0
data), a horizontal line (slope = 0) might also fit the data. The 1:1, 1:2,
and 0:1 data (designated as m:n) were fit with the 1:0 slope by realizing
that for the least squared error under the common slope constraint the
following is true:
and
Y = A + B1 _X
m:n m:n 1:0
A = Y - B_ nX
m:n m:n 1:0
(5.2)
v '
where Y and X are the average values of the emission and load measurements,
respectively.
Data Normalization
Since the excess air levels varied throughout the testing, all results
were adjusted to a common reference to remove the dilution effect. Accord-
ingly, Equation 5.3 was used to adjust all emissions to the 50 percent excess
air level.
CO,
4Murphy, T. D., Jr. Design and Analysis of Industrial Experiments.
Chemical Engineering, June 1977. pp. 169-182.
5Natrella, M. G. Experimental Statistics. National Bureau of Standard
Handbook 91. U.S. Government Printing Office, Washington, D.C., 1963.
98
-------
90% CONFIDENCE
INTERVAL FOR PLACEMENT
OF A REGRESSION LINE
Figure 51. Graphic representation of probably similar (A & B) and
potentially dissimilar (A & C) regression lines through
data set.
99
-------
where X2 = emission level corrected to 50 percent excess air
Xx = measured emission level at test conditions
C02 = carbon dioxide level at 50 percent excess air as
calculated from fuel properties and stoichiometry
C02 = measured carbon dioxide levels at test conditions
(1 + EA) = stoichmetric air plus excess air
Since carbon dioxide levels are direct measurements that correlate
fairly well with excess air levels, the carbon dioxide levels were used for
all adjustments. The expected carbon dioxide level at 50 percent excess air
can be determined from stoichiometry by using the fuel property data. The
reference carbon dioxide levels for all coal:dRDF blend and dKDF-only tests
in March were 12.2, 12.6, and 12.3 percent. These percentages correlate with
the 1:1, 1:2, and 0:1 blends, respectively.
The experiment was further complicated by varying fuel properties when
different coals and dRDF supplies were used. The ultimate and proximate
analysis for coal and dRDF shown in Tables 3 and 4 reflects the varying fuel
properties during the test period.
To facilitate detection of the emission changes when the coal:dRDF ratio
was varied, the fuel properties of the different coal and dRDF supplies were
normalized by using the properties for the coal delivered for the March tests
and the properties for the average composition of dRDF delivered for all
tests.
Emissions were normalized by dividing the fraction of each element in
the fuel per joule equivalent by the amount of that element in the reference
fuel. This correction assumes that a constant fraction of sulfur, for
example, is emitted as SOX regardless of the actual percentage of sulfur in
the fuel. While this correction ignores secondary effects such as the sulfur
chemically bound with the ash, it compensates for primary effects. The
correction factors for particulates, SOX, NOX,* and Cl are tabulated in
Table 12. The primary reduced data in Tables A-l through A-4 of Appendix A
were multiplied by the appropriate factors in Table A-5 before the statistical
analysis or plotting. The fluorine, opacity, and hydrocarbon emissions were
not corrected to the normalized fuel properties because either the requisite
data was not available (fluorine is not measured in an ultimate analysis) or
the controlling parameter was uncertain.
*The validity of correcting NOX organically bound nitrogen was question-
able since the NO formation was governed by combustion and flame cooling
rates, excess air levels, point of air addition, and recirculation effects as
well as fuel nitrogen. Because of the low volumetric heat release rates
encountered and high excess air levels, the NO formation for this test should
have been governed by the fuel nitrogen only.
100
-------
TABLE 12. FUEL ELEMENTAL COMPOSITION NORMALIZATION FACTORS
FOR ADJUSTING EMISSIONS TO A STANDARD FUEL
Element
Blend S Cl Ash N2
December /January
1:0
1:1
March
1:0
1:1
1:2
May
1:0
1:1
1:2
0:1
.73
.51
1.00
1.00
1.01
1.23
1.22
1.20
1.00
.40
.72
1.00
.90
.93
2.00
1.20
1.09
1.00
.72
1.14
1.00
.99
.99
.42
.56
.62
.77
1.03
1.35
1.00
1.03
1.06
1.04
.93
.87
.56
PARTICIPATE EMISSIONS TEST RESULTS
In the framework of compliance with emission regulations, the blend
effects on opacity, particulate concentration, size distribution, fly ash
resistivity, and overall ESP performance were evaluated. The following
sections discuss these effects in the order given.
Opacity
Figure 52 shows the opacity readings averaged over 8-hour test intervals
as a function of boiler load and coal:dRDF blend. The comparison of the
confidence interval about the coal-only regression line with the best fit
curves for the 1:1, 1:2, and 0:1 firings indicates that the overall opacity
was reduced as the dRDF substitution was increased.
Because of the large diameter of the stack relative to the amount of gas
discharged, the opacity appeared lower to ground-level observers than the
meter reading indicated. As the plume left the stack, it was lazy, and it
immediately fanned. While the opacity meter spanned a 3-m (10-ft) path, a
ground-level observer could see only about a 0.3-m (1-ft) path. To adjust
the measured data to indicate the opacity which would be seen in a more
closely sized stack (4 feet versus the actual 10 feet), a second scale was
101
-------
included on Figure 52. Beer's Law was used to apply the path length correction
to the measured light attenuation. Equation 5.4 is the form of Beer's Law
which governs the transmittance of light across an attenuated gap.
I = I0EXP (-KL) (5.4)
and
where I = perceived source intensity
Io = source intensity
K = extinction coefficient
L = path length
0 = opacity
Assuming that shortening the path length affects only the extinction
coefficient, Equation 5.5 is the manipulation of Equation 5.4 to show the
impact of an altered path length on the measured opacity.
02 = 1 - EXP
(5.5)
where 02 = opacity at L2
Q! = opactiy at Li
Since the Hagerstown boiler plant has a significant amount of carbon
carry over when coal is fired, most of the reduction in the plume opacity can
be attributed to improved combustion conditions within the boiler when the
blends were fired.
Particlate Concentration
Figure 53 shows the particulate emission rate as a function of load and
blend. Except for the 0:1 (100 percent dRDF) firing, the particulate mass
flux in the flue gas was reduced with increasing dRDF substitutions, i.e.,
from the 1:1 to the 1:2 blend firing. However, the reductions are not sig-
nificant at the 90 percent confidence level. While the 0:1 firing produced
results that differed from the data presented within the 90 percent confidence
level, the limitation of only two data points for the 0:1 firing precludes
definitive conclusions.
The data show that when 100 percent dRDF was fired, the increase in the
fuel ash more than offset the reduction in the fly ash carbon. This improved
burnout is confirmed by the filters shown in Figure 54. These filters, which
were removed from the EPA Method 5 train, show a color shift from the 1:0 to
the 0:1 blend firing. The analysis of carbon in the fly ash, as shown in
Figure 42, also indicates a reduction in carbon content with increased boiler
load and dRDF substitution for coal.
102
-------
100 r
80
% 60
CL
O
Q
UJ
DC
40
20
- 38 *
-31
24 V
cc
O
LL
18 £
O
"13 I
- 9
- 4
BLEND
1:0
1:1
1:2
0:1
BOILER
1 2
O •
& A
a •
*
0.10 0.20 0.30 0.40 0.50
£ — FRACTION OF RATING
0.60
Figure 52. Effects of blend and load on stack opacity.
103
-------
O> OC
a
o
CM IO
CD
05
CO
CM
CM i-
BLEND
1:0
1:1
1:2
0:1
BOILER
1 2
O •
A A
O •
0.00
0.10
0.20 0.30 0.40
«£ — FRACTION OF RATING
0.50
0.60
Figure 53. Effects of blend and load on particulate mass
emission rate. ,
104
-------
EPA Method 5
Filters
Coal/d'RDF
1:0
0:1
Figure 54. Effect of blend on color of stack aerosol.
105
-------
Size Distribution
Typical of all data results, Figure 55, a size distribution plotted on
Rosin-Rammler paper, shows that the cumulative mass vs size distribution
plots as a straight line. The two probability plots of the MRI cascade
impactor data in Figures 56 and 57 are the averages of all related values for
the March and May tests, respectively. Appendix D lists the data for each of
the MRI cascade impactor runs.
These plots indicate that as dRDF is substituted for coal, the aerosol
size distribution shifts to the fines. This was expected for two reasons:
First, improved burnout of the aerosol produces smaller particulates for the
same amount of ash. Second, the large number of fine paper platelets in the
dRDF causes the number of particles formed from burning a unit of fuel to
increase.
Fly Ash Resistivity
The resistivity of the fly ash leaving a source is of interest because
the performance of a precipitator is governed by the total particulate concen
tration, the aerodynamic characteristics of the particles, and the resistivit
During design and operation the resistivity determines the power which can be
applied to collect the aerosol as well as the extent to which the aerosol can
be re-entrained into the flue gas stream after "collection."
The coal-only firing produced a resistivity generally less than 106 ohm-
This unusually low resisitivity is a direct result of the high carbon content
in the fly ash. At this level of resistivity, particles collect at the wall
of the precipitator, rapidly lose their charge, and re-enter the gas stream.
Consequently, an ESP does not perform well when the particle cloud has a low
resistivity. Figure 58 presents the resistivity results for coal-only and
blend firing. The 1:1 blend fly ash had resistivities about 1010 ohm-cm.
This resistivity is within the range of 108 to 1010 ohm-cm required for
efficient precipitator performance. The fly ash from the 1:2 and 0:1 blends
may have resistivities which are too high (1012 ohm-cm) for good ESP
collection efficiencies. Table 13 illustrates the resistivities of the
various blends for the March and May tests.
The resistivity was not plotted against carbon content or gas temperatui
since the data were too limited to make the results meaningful.
Overall ESP Performance
To test the aerosol control capability at the Hagerstown boiler plant
while firing coal and coal:dRDF blends, a field-portable, 5-cell ESP was
installed and tested between April 28 and May 17. This was the only period
when the precipitator could be used at the Hagerstown plant because of other
test demands. The precipitator is owned by the EPA Industrial Environmental
Research Laboratory in Research Triangle Park, North Carolina, and operated
by Monsanto Research Corporation. The precipitator system was housed in two
trailers, one for the five precipitator cells with independent power supplies
and the other for the precipitator electrical monitors and an aerosol
106
-------
:m
s »
§:
vROSLIN-HAMMLER EXPONENT
PARTtCLE SIZE 0, MICflOMS
Figure 55. Typical MRI cascade impactor results for blend firing.
107
-------
BLEND
MARCH
VROSLIN-RAMMLER EXPONENT
1:0 -
1:1
1:2 -HUH-
0:1 ---
SIZE D MtCJIOWS
Figure 56. Average size distribution for coal, blend, and dRDF firings
during March tests.
MAY
M[- t'Ttiu^-jjffrrt-'iM!
of::::::ti±i^iI^J~;^ ""'Hi:
*i-tiggx» r\ , ^mr
aiWjr: f - ,!
--r:::-;—H
,:n
*
fl» ii-
Figure 57. Average size distribution for coal, blend, and dRDF firings
during May tests.
108
-------
c;
i
io12-
10"-
10
I0 -
c/2 10 -
CO
bJ
cc
io8H
1ST TEST
-0*
BLEND
I '0
I : I
I = 2
0: I
BOILER
I 2
- o •
- A A
am
— *
O.I 0.2 0.3 0.4 0.5 0.6
jf, - FRACTION OF RATING
Figure 58. Effects of blend and load on aerosol resistivity.
109
-------
TABLE 13. EFFECT OF BLEND ON AEROSOL RESISTIVITY
Resistivity (
Blend March Tests May Tests
1:0
1:1
1:2
0:1
<107
8 x io9
—
—
6 x IO7
2 x io10
1 x IO12
1 x IO12
laboratory. As shown in Figure 59, the precipitator cells are of the plate
and frame construction with the charging electrodes installed in a pipe frame
support centered between parallel, smooth-surface collecting electrodes.
Since the charging electrodes are suspended from dielectric blocks
resting across the tops of the grounded electrodes, an aerosol accumulating
on the dielectric block tends to short circuit the cells. Because of such
short circuits during the testing, the number of operating cells progressively
decreased. Consequently, the results from this test cannot be considered
representative of the data which might be collected from a commercially
available ESP. Additional ESP testing is planned for the demonstration test.
Appendix E includes the data acquired by Monsanto during the May testing and
a discussion of the results.
GASEOUS EMISSIONS TEST RESULTS
The following sections detail the test results for the following gaseous
emissions: S02, oxides of nitrogen, halogens, and hydrocarbons.
S02
The sulfur oxide level in the flue gas is of particular interest to
precipitator designers since it is related to aerosol resistivity. Figure 60
shows the blend effect on the overall sulfur dioxide emission rate as a func-
tion of boiler load. The reduction in overall S02 emissions with the replace-
ment of higher sulfur coal by 0.6 percent sulfur dRDF is significant for the
1:2 and 0:1 blend firings. While the 1:1 blend firings showed a reduction in
S02 when compared to coal, the difference did not exceed the 90 percent con-
fidence limits. Very good agreement was obtained between the sulfur oxide
emissions as determined by the continuous monitoring electro chemical sensor
and the wet chemistry determinations. Reference is made to Tables A-2 and A-3
in Appendix A. The SO for test days May 10 through May 13 were: blend 1:1
110
-------
CHARGING
ELECTRODE
FRAME
INSULATING
FRAME
SUPPORT
SHORT CIRCUITS DEVELOPED
AT THESE LOCATIONS
GROUNDED
ELECTRODE
Figure 59. Cell configuration in the portable ESP.
Ill
-------
4500 r
2
0.
0.
£
CM
O
CO
4000
3500
3000
2500
2000
1500
1000
500
0
0
-
-
-
rrr^A*
\ • \ \ I • \ \~ \ * \ A°l "
\ \ \ tit* * ** t
— i t i t ^t i i > > * *
&
a
IF*
i i i I
00 0.10 0.20 0.30 0.40
BOILER
BLEND 1 2
1:0 o •
1:1 A A
1:2 fill 0 •
0:1 x
O
^_^o— — to- r
^UJ
0.50 0.60
— FRACTION OF RATING
* S02 WC is S02 as determined by wet chemistry.
Figure 60. Effects of blend and load on sulfur dioxide emissions.
112
-------
1334 ppm (meter), 1309 ppm (wet chemistry), and blend 1:2 825 ppm (meter),
825 ppm (wet chemistry). As would be expected, the reduction in sulfur
emission was proportional to the reduction of the sulfur in the fuel combusted.
Reference is made to Table 5 and Tables A-l through A-4 in Appendix A for
verification of the reduction in sulfur emissions as a function of the sulfur
in the fuel combusted.
Oxides of Nitrogen
The results from the on-line NOX analyzer determinations are shown in
Figure 61 and Table 14. The plot of NOX versus load as a function of blend
has considerable scatter. The random pattern of the line placement for blend
firing suggests that there is no apparent relationship between the blend
firings and the NOX emissions.
Figure 62 shows the NOX emissions replotted as a function of excess air.
All the results are similar in that they fall within the 90 percent confidence
level. Table 14 summarizes all the NOX concentrations as determined by both
wet chemistry and the Theta Sensor trigas analyzer.
Halogens
Of concern to many people is the emission of chlorine from the combustion
of dRDF. Although chlorine levels (weight basis) in dRDF are about the same
as those found in some coals burned throughout the United States and Europe,
the coal:dRDF blends emit more chlorine per megajoule than most coals because
of their lower heat content. The chlorine level per magajoule equivalent for
the coal:dRDF firing (1:2 blend) was seven times greater than the level for
coal-only firing.
Figures 63 and 64 show the emission rates for chlorine and fluorine,
respectively, as a function of coal:dRDF blend and boiler loads. Since the
chlorine emission regression lines for blend firing are well outside those
for coal-only firing, the blend firing apparently had substantially greater
chlorine emission rates. Since fluorine concentrations were omitted in the
laboratory analysis of the base fuel, the significance of the fluorine
emissions for blend versus coal-only firing cannot be established.
Hydrocarbons
Hydrocarbon emissions are particularly important because of their smog
forming potential. Figure 65 shows the hydrocarbon emissions as a function
of boiler load and coal:dRDF blend. The hydrocarbon emissions from the blend
firing was not significantly different from the coal firing. The single data
point at 100 percent dRDF firing (55 ppm) suggests that further testing may
be required at higher dRDF substitution ratios to verify this increase.
113
-------
400
300
a.
Q.
O
200
100
0.00
0.10
0.20 0.30 0.40
o£ — FRACTION OF RATING
0.50
0.60
Figure 61. Effects of blend and load on nitrogen oxide emissions.
114
-------
TABLE 14. RELATIONSHIP OF NOX CONCENTRATION AND EXCESS AIR
PERCENTAGE FOR COAL, BLEND, AND dRDF FIRINGS
Date
1:0 Blend (coal only)
3/21
3/22
3/28
3/31
4/1
5/3
5/4
5/5
5/16
1/20
1/20
1/21
1/24
1/25
NOX M(ppm)
362
257
197
277
300
230
141
105
424
358
358
228
NOX WC(ppm)
318
238
449
424
594
Percent
Excess Air
104
72
92
71
82
72
116
84
147
108
88
36
87
110
1:1 Blend Firing
3/24 255 102
5/12 326 360 109
5/13 405 386 101
1:2 Blend
3/29 232 94
5/10 223 192 138
5/11 305 273 132
0:1 Blend
5/14 272 274 133
5/14 255 247 106
Notes: 1. NOXM is - NOX as determined by the Theta Sensors, Inc. Tri-gas
Meter.
2. NOXWC is - NOX as determined by the EPA Method 7 Wet Chemistry.
115
-------
600
500
400
CL
Q.
O* 300
z
200
100
BLEND
1:0 —
1:1 --
BOILER
1 2
1:2
0:1 —
- o •
- A A
O •
0 20 40 60 80 100 120 140 160
EXCESS AIR — %
Figure 62. Effects of blend and excess air on nitrogen oxide emissions.
116
-------
0.
a
o
BOILER
BLEND 1 2
700
600
500
400
300
200
100
0
(
v.o ° •
* V.I A A
y. v.2 tttt a •
o-.i *
a
Q A
A * n"
A
A
T__1 • 8
\ !• \ V \« \ \ \ \° v A v \\b\\
\ ^ \ \ ^P ^ -^ i 1 1 O- ^ i ^ A ^ \ "*\ k \
300 0.10 0.20 0.30 0.40 0.50 0.60
— FRACTION OF RATING
Figure 63. Effects of blend and load on chlorine emissions.
117
-------
20 r
o.
Q.
10
BLEND
1:0 —
1:1 —
1:2
0:1 —
BOILER
1 2
— O •
- A A
0 •
NOTE: DATA FROM
MARCH & MAY
ONLY
_l_
I
.00
.10
.20 .30 .40 .50
it — FRACTION OF RATING
.60
Figure 64. Effects of blend and load on fluorine emissions.
118
-------
Q.
O.
6
60
50
40
30
20
10
0.00
0.10
BLEND
1.0
1:1
1:2 -t-+
0:1
BOILER
1 2
— O •
— A A
•
-------
TRACE COMPOUND EMISSIONS TEST RESULTS
While not yet regulated, the emissions of potentially carcinogenic
polycyclic hydrocarbons and hazardous heavy metals from stationary combustion
sources are coming under intense scrutiny. Therefore, the flue gases were
evaluated for polycyclic organic compounds and heavy metal emissions.
Trace Organic Emissions
Table 15 lists the results of the GC-Mass Spectrometer Analysis of the
Battelle Tenex samples. The overall emissions of polycyclic organic materials
(POM's) for the coal-only and the blend firings were very low. Morever, all
the monitored emissions were below the threshold limits proposed by the
National Academy of Science. Since the increase in emissions for the 1:1
blend firing is based on the data of a sample taken during boiler switchover,
this increase is questionable.
TABLE 15. POM CONCENTRATIONS FOR COAL AND BLEND FIRINGS
COMPONENT
Anthracene/Phenanthrene
Methyl anthracenes
Fluoranthene
Pyrene
Methyl Pyrene/
Fluoranthene
Benzo (c)phenanthrene
Chrysene/Benz (a)
anthracene
Methyl chrysenes
7 , 12-Dimethylbenz (a)
anthracene
Benzo fluoranthenes
Benz (a) pyrene
Benz (e) pyrene
Perylene
Methyl Benzopyrene
3-Methyl Chloranthrene
Indeno(l,2,3,-cd)
pyrene
Benzo (ghi)perylene
Dibenzo(a,h) anthracene
Dibenzo (c, x) carbazole
Dibenz (ai and ah)
pyrenes
Coronene
1:1
ppb
0.0736
0.0126
0.0164
0.0039
0.00043
<0.0002
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
1:2
ppb
0.0516
0.0052
0.0064
0.0027
<0.0002
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
COAL
1:0
ppb
0.00086
0.00032
0.00030
0.0018
<0.0002
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
Ng/m
543
100
137
33
4
<1.0
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
1:2
Ng/m3
380
42
54
23
<1.0
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
COAL
1:0
Ng/m3
6
3
3
15
<1.0
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
n/d
120
-------
Trace Inorganic Emissions
Table 16 lists the concentrations of the fuel ash trace metals found in
two coal and two dRDF samples. The concentrations were determined by first
preparing the specimens by oxygen plasma ashing and then analyzing them with
spark source mass spectrometry. The enrichment ratios indicate the greater
amount of metals in dRDF than in coal. The range of reported fuel metal
values clearly indicates the heterogeneous content of the dRDF and the
variability of the metalic material in coal.
The detailed heavy metal ash analyses are presented in Appendix F and
summarized in Table 17. This table lists the average emission rates for each
test battery. The March and May data were separated to eliminate the effects
of normalizing the data to the reference coal. Table 18 is a manipulation of
the data in Table 17 to present the emission rate data in terms of enrichment
functions. For example, Table 18 shows that 43.3 times more lead was emitted
in the total particulates when firing dRDF only than when firing coal only.
In addition, while some metals were enriched, others had reduced emission
rates.
Tables 17 and 18 also show the enrichment of certain metals in the
bottom ash, the reinjected fly ash, and the collected fly ash. The presence
of these metals in the ash implies an increased possibility of heavy metal
leaching when boiler ash is landfilled or used for various applications. The
significance of this leaching is unknown.
The amounts of the various metals found for all blends were generally
normal except for the amount of arsenic which was fairly high. Most notable
of the trends was the variation of several metals with the coal:dRDF ratios:
whereas the concentrations of Br, Mn, Pb, and Sb generally increased with
increasing dRDF substitution in the coal:dRDF blends, the concentrations of
As, Ni, and Vn decreased. Although these trends are not definitive, they are
probably true since they were also observed in the data for the fly ash
leachates.
While variations of the heavy metal concentrations with particle size
were poorly defined, the concentrations of As, Ga, Na, and probably Sb
generally increased with decreasing particle size. In contrast, the concen-
trations of Br and Mn markedly increased with increasing particle size.
Appendix F provides a detailed summary of the heavy metal data.
On the basis of previous work with coal aerosols and incinerator fly
ash,6 it would be expected that such metals as Br, Mn, Pb, and Sb would have
higher concentrations in RDF than in coal and that As, Ni, and V would have a
greater affinity to coal than to RDF. Similarly, the affinity of As, Ga, Na,
and Sb to small particles would be expected since these metals can be vola-
tilized during combustion and then adsorbed onto the more developed surface
6Kaakinen, J. W., et al. Trace Element Behavior in Coal Fired Power
Plants. Environmental Science and Technology, Volume 9. pp. 862-869.
121
-------
TABLE 16. TRACE METAL CONCENTRATIONS FOUND IN COAL AND dRDF FUEL
Element
Coal Sample
#1 #2
d-RDF Sample
#1 #2
ppm
ppm
ppm
ppm
dRDF:COAL
Enrichment Ratio
Per kg Ash Per MJ
Li
Be
B
F
Na
Mg
Al
Si
P
S
Cl
K
Ca
Sc
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Ga
Ge
As
Se
Br
Rb
Sr
Y
Zr
Nb
Mo
Ru
Rh
Pd
Ag
50
1
30
<5
= 1000
1000
High
High
300
= 1%
100
High
High
10
2000
40
100
100
High
20
100
10
5
20
<1
10
<1
3
100
300
10
100
5
10
<1
<0.3
<1
<0.6
54
4
23
=150
MC
MC
MC
MC
MC
MC
MC
MC
MC
14
MC
97
88
22
MC
7
19
13
16
32
<5
13
12
19
33
690
70
280
11
11
0
0
0
1
10
0.05
20
10
High
High
High
High
High
2000
2000
High
High
<1
2000
10
40
500
High
1
20
30
500
2
<0.3
4
<0.5
10
20
150
2
50
1
10
<0.2
<0.1
<0.2
1
<0.1
<0.1
<0.1
=36
MC
MC
MC
MC
MC
MC
MC
MC
MC
0.7
MC
8
260
>470
MC
2
15
41
300
4
0.1
4
0.4
4
3
74
1
13
2
4
0
0
0
0.4
0.097
0.03
0.38
0.30
0.12
0.13
1.60
7.95
0.11
0.29
3.09
38.10
0.12
0.07
0.35
0.07
0.64
0.17
0.23
0.04
0.17
0.19
0.67
0.20
3.00
0.20
0.88
.3
.1
1.1
.9
.4
.4
4.6
22.9
.3
.8
8.9
109.9
.3
.2
1
.2
1.8
.5
.7
.1
.5
.5
1.9
.6
8.7
.6
2.5
( continued )
122
-------
TABLE 16. (continued)
Element
Cd
In
Sn
Sb
Te
I
Cs
Ba
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Hf
Ta
W
Re
Os
Ir
Ft
Au
Hg
Tl
Pb
Bi
Th
U
Coal Sample
#1 #2
ppm ppm
<1
<1
<1
<2
<1
1
<3
100
10
20
1
4
<1
<0.6
<1
<0.4
<1
<0.4
<1
<4
<1
<0.4
<2
<2
<2
<1
<1
<1
<2
<2
<5
<1
<2
<1
<1
<1
2
STD
2
<0.8
<0.8
5
3
410
44
110
5
11
7
1
2
0.9
0
0
0
0
0
0
9
0
5
0
0
0
0
0
<0.8
0
8
0
17
8
d-RFD Sample
#1 #2
ppm ppm
<0.6
<0.06
20
1
<0.2
0.6
0.2
200
2
10
0.5
15
<0.3
<0.2
<0.3
<0.1
<0.3
<0.1
<0.3
<0.1
<0.3
<0.1
<0.3
<0.1
<0.3
<0.3
<0.3
<0.3
<0.3
<0.3
<0.3
<0.1
500
0.2
<0.1
<0.1
0.3
STD
8
9
0
0.1
0.2
330
3
3
0.5
0.5
0.4
0.1
0.1
0.1
0.2
0.1
0.1
0
0
0
0
0
0
0
0
0
0
0
<0.1
0
170
0.1
0.7
0.5
dRDF : COAL
Enrichment Ratio
Per kg Ash Per MJ
0.3
9.33
3.57
0.11
0.12
0.07
1.04
0.09
0.1
0.17
1.03
0.09
0.19
0.13
0.15
0.5
0.5
0.4
0.03
0.3
0.25
0.03
0.05
0.04
0.3
0.3
0.3
0.15
0.15
0.07
0.1
67
0.3
0.04
0.07
.9
26.9
10.1
.3
.3
.2
3
.3
.3
.5
3
.3
.5
.4
.4
1.4
1.4
1.2
.07
.9
.7
.08
.1
.1
.9
.9
.9
.4
.4
.2
.3
193
.9
.1
.2
123
-------
TABLE 17. AVERAGE HEAVY METAL EMISSIONS IN ASH FROM BLEND FIRING TESTS
N3
Threshold Limit Level
ug/m1
Total 1'articulate
stark - lig/fli
Bottom Ash - ti£/k£
Multiclone Reinjectlon
Fly Ash - ug/kg
Thresliold Limit Level
uR/m1
Total Partlculate
stark - ug/nr
Bottom Ash - ug/kg
Multiclone Relnjectlon
Fly Ash - us/kg
Collector Ash
ug/kg
Hl.KNI)
„_
1:0
1 :1
1:2
1:0
1:1
1:2
1:0
1:1
1:2
BLEND
1:0
1:1
1:2
0:1
1:0
1:1
1:2
0:1
1:0
1:1
1:2
0:1
1:0
1:1
1:2
0:1
mi . it i
Samples
Analyzed
3
1
1
1
2
1
no. of
Samples
Analyzed
7
3
3
2
2
2
2
1
1
1
2
2
3
3
2
I'h
200
228
3975
7660
<12.5
26.3
128
16.3
97.5
109
Pb
200
230
4237
8217
9953
<12.5
46.3
65.0
169
20.0
92.5
165
363
20.8
217
274
1012
Cd
20
< 4 . 4 3
79.4
233
• .75
< . 8
0.75
'.75
1.12
1.5
Cd
20
4.33
72.4
220
267
<.75
£-75
.75
2.13
1.00
1.9
7.88
<0.75
3.58
6.00
25.4
As
500
173
45.9
44.9
11.0
39.3
36.7
34.4
39.3
66
As
500
184
153
126
49.4
16.1
35.0
28.8
56.3
48.1
49.5
34.9
46.3
76.3
80.4
43.5
103.5
HH
100
17 . 85
19.6
12.3
<0.4
<0.4
<0.4
<0.4
<.68
<0.4
Hg
100
<5.57
15.7
11.4
94.7
<0.4
'0.4
<0.4
<0.4
<0.4
0,58
S.415
i.440
<0.4
S.545
SJ..O
<.0.65
MARCH
Or
100
35.1
33.5
47.6
22.5
25.6
42.5
8.75
12.6
15.0
MAY
Cr
100
50.7
35.4
55.4
79.7
8.75
23.7
20.7
50.0
21.3
21.3
23.6
69.4
17.7
19.7
34.9
185
Nl
1000
32.6
32.1
41.0
20.0
27.4
139
15.0
16.9
18.7
Nl
1000
49.5
35.9
50.9
29.4
20.0
33.2
22.1
35.0
26.2
23.7
23.2
40.9
87.5
25.2
31.9
75.4
Mn
47. ;
64.6
101
51.0
138
250
105
193
300
Mn
30.4
62.6
115
275
40.5
170
135
43.5
45.0
120
132
81.3
49.7
145
328
123
Zn
5000
592
6012
8569
31.2
73.3
188
31.2
134
194
Zn
5QOO
596
5Cu3
8317
8033
68.8
112.5
73.4
539
50.0
356
292
1118
60.4
343
608
Cu
S.51.7
96.1
82. 5
15.0
152
200
<12.5
15
17.5
Cu
50.1
82.4
134
203
<18.8
221
>136
~205
15.0
22.5
42.2
88.4
15.0
27.8
39.9
149
Sn
•-1.46
3.36
4.99
0.50
3.58
5.00
1.88
2.50
2.25
Sn
51,45
2,70
3.47
6,07
0,50
3.31
3.62
3.6
1.0
3.0
3.0
4.84
2.09
3.58
1.63
8.08
Sb
500
<87.2
<52.2
<87.3
<25
<25
'25
<25
<25
'25
Sb
500
<65.6
<48.5
59.1
<107
<25
<25
<25
<25
<25
<25
<26
<25
<25
<25
<25
<26.3
Ag
10
<8.72
12.0
17.1
'2.5
<2.5
'2.5
<2.5
'-2.5
<2.5
Ag
10
5.75+120.
'6.56
<6.51
19.4
29.7
62.9
<2.5
<2.5
88.8
<2.5
<2.5
6.0
<2.5
<2.5
<2.5
13.7
Vn
<:87.2
<52.2
59.6
<25
<25
'25
'25
i.25
25
Vn
'65.6
<48.5
'59.1
-------
TABLE 18. BLEND HEAVY METAL TO COAL-ONLY HEAVY METAL RATIOS IN ASH SAMPLES
to
U1
MARCH
Total Particulates
stack - ug/m3
Bottom Ash - pg/kg
Multiclone
Reinjection Fly Ash
pg/kg
Total Particulate
stack - pg/m3
Bottom Ash - pg/kg
Collector Ash - pg/kg
Multiclone Reinjection
Fly ash - pg/kg
BLEND
1:0
1:1
1:2
1:0
1:1
1:2
1:0
1:1
1:2
BLEND
1:0
1:1
1:2
0:1
1:0
1:1
1:2
0:1
1:0
1:1
1:2
0:1
1:0
1:1
1:2
0:1
Pb
1.0
17.4
33.6
1.0*
2.10
10.2
1.0
5.98
6.69
Pb
1.0
18.4
35.7
43.3
1.0*
3.70
5.20
13.5
1.0
10.4
13.2
48.6
1.0
4.63
8.25
18.2
Cd
1.0*
17.9
52.6
<
1.0*
1.49
2.00
Cd
1.0
16.7
50.8
61.7
1.0*
1.0*
1.0
2.84
1.0*
4.77
8.0
33.9
1.0*
1.33
2.53
10.5
As Hg
1.0 1.0*
.265 2.50
.259 1.57
1.0 <
3.57
3.34
1.0 <
1.14
1.92
As Hg
1.0 1.0*
.832 2.82
.685 2.05
.269 17.0
1.0 <
2.17
1.79
3.50
1.0 <
1.05*
.570
1.36
1.0 <
1.03
.726
.963
Cr
1.0
4.27
6.06
1.0
1.14
1.89
1.0
1.44
1.71
MAY
Cr
1.0
.698
1.09
1.57
1.0
2.71
2.37
5.71
1.0
1.11
1.97
10.5
1.0
1.0
1.11
3.26
Ni
1.0
.985
1.26
1.0
1.37
6.95
1.0
1.13
1.25
Ni
1.0
.725
1.03
.594
1.0
1.66
1.11
1.75
1.0
.288
.365
.862
1.0
.905
.885
1.56
Mn
1.0
1.35
2.12
1.0
2.71
4.90
1.0
1.84
2.86
Mn
1.0
2.06
3.78
9.05
1.0
4.20
3.33
1.07
1.0
2.92
6.60
2.47
1.0
2.67
2.93
1.31
Zn
1.0
10.2
14.5
1.0
2.35
6.03
1.0
4.29
6.22
Zn
1.0
9.51
14.0
13.5
1.0
1.64
1.07
7.83
1.0
5.68
10.1
1.0
3.12
5.84
22.4
Cu
1.0*
1.86
1.60
1.0
10.1
13.3
1.0*
1.2
1.4
Cu
1.0
1.64
2.67
4.05
1.0*
11.8
9.89*
10.9
1.0
1.85
2.66
9.93
1.0
1.5
2.81
5.89
Sn
1.0*
2.30
3.42
1.0
7.16
10.0
1.0
1.33
1.20
Sn
1.0*
1.86
2.39
4.19
1.0
6.62
7.24
7.20
1.0
1.71
0.780
3.87
1.0
3.0
3.0
4.84
Sb Ag
< 1.0*
1.38
1.96
< <
< <
Sb Ag
< 1.0*
1.0*
2.96
4.53
< 1.0**
<.435*
<.435*
15.4
< 1.0*
1.0*
1.0*
5.48
< 1.0*
1.0*
1.0*
2.4
Vn
<
<
1.0*
1.0*
1.0
Vn
<
<
1.0
1.0*
1.09
1.66
1.0
1.0
<1.0*
<1.0*
< Below the detection limit
* Extreme value deleted
-------
area of the small particles. The increasing Br and Mn concentrations with
increasing particle size cannot be explained.
The MRI collector stages for coal-only, blend, and dRDF-only firing in
May were accumulated and sent to Colorado State University for analysis.
Contained in Appendix G, the University's complete report presents considerable
information about both the chemical characteristics and the potential environ-
mental impact of dRDF-coal fly ash. The following trends are based on the
data of Table G-3 in Appendix G.
First, the amounts of most metals that are soluble increase with
increasing dRDF fraction of the original fuel. While this trend is apparent
for Ca, Cu, K^ Mg, Mo, Na, Si, Cl~, N0~3, and SO,,2", it may also exist for B,
Ba, Cd, and F . Ni and P, and possibly Cr and Sr, have a reverse trend.
Since many of the major matrix metals have increasing solubility with
increasing dRDF percentage, the addition of dRDF to coal might result in
greater bulk solubility (as well as greater trace metal mobilization) than
that evidenced in the fly ash of pure coal. This increased solubility may
require special procedures for landfill disposal.
Second, the metal mobilization increased with decreasing particle size.
While this trend is apparent for Cd, Cr, Cu, K, Mn, Mo, Na, Ni, Pb, and Cl ,
it may also exist for Ba, Be, P, and F . This trend may be due to the con-
densation of these metals from vapor onto the particulate surfaces or to the
more efficient formation of soluble oxides (i.e., calcining) in small particles.
With the first supposition, similar size dependencies would be expected for
both the bulk and the soluble metal concentrations. However, such dependencies
would not be expected if solubility is the direct result of chemical reaction
at a particle surface. In any event, the available data are not sufficient
to rule out either supposition.
In the analysis of the fractional solubility of coal-dRDF fly ash, Al,
Ba, Mg, P, Si, and Sr have a very low solubility (<10 percent); Be, Cd, K,
Mn, and Na have a moderate solubility (=20-80 percent); and Ca has a very
high solubility.
Several metals, particularly Mn, have an increasing fractional solubility
with decreasing particle size. Since Mn exhibits no dependence of concen-
tration on particle size, its solubility increase with decreasing particle
size is due to its more efficient calcining.
Summary
The following summarizes the major findings:
1. The specific concentrations of trace metals in dRDF-coal fly ash
are similar to those found in pure coal fly ash. The dRDF is the
primary source of Br, Mn, Pb, and Sb while the coal is the primary
source of As, Ni, and V.
2. Several metals, particulary As, Ga, Na, and Sb, tend to concentrate
in small particles.
126
-------
3. The volatization-condensation process which deposits volatile
metals onto small fly ash particles is more effective in a plant
firing a dRDF-coal mixture than in a plant firing coal only. The
greater volatile metal deposits in the coal-dRDF firing were
probably due to the low combustion temperatures.
4. Except for Ni and P, the metals in the coal-dRDF fly ash increase
in solubility with increasing dRDF content.
5. Both trace and matrix metals have a significantly greater solubility
in small particles than in large particles.
In summary, the results from the trace compound emissions test results
indicate that the quantities of the trace organic emissions for the blends
studied were so small that the ground-level concentrations would probably not
exceed 1 percent of the threshold level limits. Hence, unless the data from
future tests indicate higher levels of trace organics, the emission levels
from coal-dRDF firings would be within acceptable limits. The quantities of
metals present in the fly ash and bottom ash suggest that further studies
need to be carried out to establish if there is a health hazard due to
(1) increased bulk solubility (as well as greater trace metal mobilization)
of bottom ash with increased dRDF substitution and (2) the adsorption of
volatilized metals during combustion onto the surface area of small (aerosol
size) particulates.
127
-------
REFERENCES
1. Winegartner, E. C. Coal Fouling and Slagging Parameters. American
Society of Mechanical Engineers, 1974.
2. Fontana, M. G., and M. D. Green. Corrosion Engineering.
McGraw-Hill, N.Y.C., 1967.
3. Watson, H. H. American Hygiene Association Quarterly. Volume 15,
1954. p. 21.
4. Murphy, T. D., Jr. Design and Analysis of Industrial Experiments.
Chemical Engineering, June 1977. pp. 169-182.
5. Natrella, M. G. Experimental Statistics. National Bureau of
Standard Handbook 91. U.S. Government Printing Office,
Washington D.C., 1963.
6. Kaakinen, J. W., et al. Trace Element Behavior in Coal Fired Power
Plants. Environmental Science and Technology, Volume 9. pp. 862-869.
128
-------
APPENDIX A
EMISSIONS, FUEL, AND ASH DATA SUMMARIES
TABLE A-l. FIELD TEST RESULTS FOR COAL ONLY 1:0 FIRING
NJ
VO
JATE X-
3/19
3/19
3/21
3/21
3/22
3/28
3/31
3/31
4/1
5/3
5/4
5/4
5/5
5/5
5/16
5/16
5/17
5/17 .
1/20
1/20
1/21
1/24
1/25
12/10
12/10
NOTES :
51
51
40
40
40
38
27
27
35
25
20
22
31
51
21
36
17
17
53
53
47
53
50
1.
2.
3.
4.
5.
6.
GR/FCF BOILER ' Cl F SO,.M
.311 1 72 8
.458 1 84 15
.267 2 1397
.325 2 75 7.2
.226 1 55 7.3
.197 1 38 9.3
.286 1 35 8.1 1000
.234 1 36 6.5
.281 1 44 7.5 1148
.295 2 25 5.4 1118
.339 2 19 7.9 1495
2 1080
.206 2 283
.199 2 19 6.8 367
.588 2 25 11.9 1302
.298 2 26 7. 3 1083
.448 2 1254
2 1118
.330 1 27 25 1734
.436 1 32 46 1734
.214 1 14 16 1155
.327 1 31 16 1265
1 1291
.229 1 98 4.6 1916
.240 1 115 23.6 1381
All values adjusted to 50% EA or 12* CO
GR/SCF is grains/standard ft3.
Lined out data are considered outliers.
PPM
soxwc
2312
1691
1397
1153
1040
1117
1324
1125
541
1341
1121
1149
4888
5946
2329
2217
2 .
NOXM NOXWC THC '
362
257
197
277
300
221
136
101
408
348
348
221
251
332
21
12
13
29
34
306
229
432 19
12
8
12
412 1
577 1
.9
.7
.7
.6
.1
.2
.0
.8
.6
.8
.6
%
EA
104
104
72
92
71
71
82
72
116
106
84
40
147
104
96
108
88
86
87
110
f
PPM g/hr g/hr fl-CM
OPACITY SO3 K4SJ.NJ1SUT COLLECT KBSIS.
38
38
32
44
68
68
46
50
70
70
54
54
70
70
91
91
39
42
37
41
39
59
59
<107
<107
110 <10'
110
76
76
75
82 103
102 140
20.7 152 90
70 106
99 123
252 (0.10) 592 (0.14)
235 (0.12) 579 (0.14)
491 (0.10) 666 (0.17)
oF
Tflame
2201
2236
2210
2197
2259
2209
2121
2020
2244
2440
2125
2302
2024
OF
Tflue
411
410C '/502C2
5100 2
509* 2
395*'
403
399
404
418
484
459
454
491
490
474
517
461
43Q
424
427
SO M is sulfur oxides measured with a meter (electrochemical transducer) .
SO*WC is sulfur oxides measured by wet chemistry.
In volumes headed by "g/hr," any second
fly ash density in g/cc.
value
(shown in
paranthei
36S)
represents
-------
TABLE A-2. FIELD TEST RESULTS FOR 1:1 BLEND FIRING
Uo
DATE
3/23
3/23
3/24
5/12
5/12
5/13
5/13
12/8
12/13
12/13
12/14
12/14
JL
.43
.43
.44
.34
.28
.39
. 30
.45
.45
.45
.43
.43
GR/SCF
. 138
.191
.196
.287
.303
.191
.172
.441
.224
.196
.177
BOILER
1
1
1
2
2
2
2
1
1
1
1
1
'Cl
311
321
265
195
172
146
281
437
182
58
309
F
9.1
7.0
13.7
8.4
7.4
11.8
27
22.3
33
SOXM
652
1003
1388
1503
1391
1054
1402
1809
1235
808
1234
PPM
SOXWC
794
755
1118
1502
1476
1268
992
NOXM
248
351
436
224
282
278
% PPM g/hr
NOXWC THC ^ EA OPACITY SO 3 REINJECT
8.3 74
18.4 74
13.4 102
387 109
109
415 20.0 101
101
9.0
337
225
27
27
22
59
59
42
42
70
73
73
74
74
141 (0
117 (0
131 (0
18.3 122 (0
87
87
82
.15)
.16
.14)
.14)
g/hr SJ-CM
COLLECT RESIS.
IxlO10
1x10 ' °
4xl09
384 (0.20) 7xl09
344 (0.18) 5xlOs
669 (0.16) 3x10'°
260 (0.17) 5x10'°
"F
T
Aflame
2226
2213
2275
2305
2264
2253
2210
OF
T
flue
423
429
430
506
493
505
497
NOTE:
See notes for Table A-l.
-------
TABLE A-3. FIELD TEST RESULTS FOR 1:2 BLEND FIRING
U)
3/29
3/29
3/30
5/10
5/10
5/11
5/11
. 36
. 36
.26
. 36
. 30
.36
. 34
. 20 i
. 176
. 180
. 282
.320
.248
1 4 i8
1 32 1
1 301
2 198
2 238
2 243
2
14.0
12.4
8.8
11.7
11 . 1
11.6
800
828
980
777
715
994
9 8 ft
46]
928
810
842
722
301:
" F
19.7
15.8
34.2
19.1
'14
'14
')]
I 38
116
1 32
114
31
31
42
45 123
45 133
45 240
45 10.3 138
82
82
98
(0.
(0.
(0.
(0.
.17)
10)
.20)
.17)
155
164
241
241
(0.
(0.
(0.
(0.
.21)
.23)
.27)
.26)
6x10' '
8x10'
2x10' ;
4x10' '
2287
2306
2313
2401
2322
2308
2299
415
409
391
515
512
518
506
NOTE: See notes for Table A-l.
-------
TABLE A-4. FIELD TEST RESULTS FOR dRDF 0:1 FIRING
1 — 1
CO
N3
DATE
5/14
5/14
JL
.27
.26
GR/SCF
.348
. 356
BOILER ' Cl F SO M
2 654 9.4 251
2 610 7.8 275
PPM
SO WC
X
303
268
NO M
X
486
456
NO WC
X
489
441
'1, %. PPM
THC ^ EA OPACITY SO,
56.3 133 48
106 48
q/hr
REINJECT
215
230
(0.
(0.
,85)
.84)
q/hr
COLLECT
273
271
(0.
(0.
.76)
.70)
ii-CM
RESIS.
1x10 ' J
IxlO1 '
"F °F
T T
flame flue
2326 470
2285 473
NOTE: See notes for Table A-l.
-------
TABLE A-5. CORRECTED FIELD TEST RESULTS FOR FOUR COAL:dRDF BLENDS
(Emission data were normalized to the March reference coal)
BLEND DATE
1:0 (coal only) 3/19
3/19
3/21
3/21
3/22
3/28
3/31
3/31
4/1
5/3
5/4
5/4
5/5
5/5
5/16
5/16
5/17
5/17
1/20
1/20
1/21
1/24
1/25
12/10
12/10
*0utliers were
1:1 3/23
3/23
3/24
5/12
5/12
5/13
5/13
12/8
12/13
12/13
12/14
12/14
1:2 3/29
3/29
3/30
5/10
5/10
5/11
5/11
0:1 (dRDF only) 5/14
5/14
BOILER
.51
.51
.40
.40
.40
.38
.27
.27
.35
.25
.20
.22
.31
.51
.21
.36
.17
.17
.53
.53
.47
.53
.50
omitted
.43
.43
.44
.34
.28
.39
.30
.45
.45
.45
.43
.43
.36
.36
.26
.36
.30
.36
.34
.27
.26
1
1
2
2
1
1
1
1
1
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
from
1
1
1
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
ASH-
GR/SCF
.311
.458
.267
.325
.226
.197
.286
.234
.281
.124
.142
.087
.084
.247
.125
.188
.238
.314
.154
.235
.165
.173
PPM
Cl
72
8-4
75
55
38
35
36
44
50
38
38
50
52
11
13
6
12
39
46
SOXM
1397
1000
1148
1375
1839
1328
348
451
1601
1332
1542
1375
1266
1266
843
923
942
1399
1008
soxwc
2312
1691
1397
1153
1040
1117
1324
1384
665
1649
1379
1413
-3^68*
4341
1700
1618
THC
21.9
12.7
13.7
29.6
34.1
17.3
13.1
9.6
13.7
2.0
1.7
NOXM
362
257
197
277
300
230
141
105
424
358
358
228
259
NOXWC
332
318
238
449
424
594
statistical analysis.
.137
.189
.194
.161
.170
.107
.196
.503
.255
.223
.202
.201
.174
.178
.175
.198
.154
.268
.274
280
289
239
234
206
175
202
315
131
42
222
407
300
280
216
259
265
654
610
652
1003
1693
1834
1697
1286
715
923
630
412
629
808
994
1176
932
858
251
275
794
755
1118
1832
1801
1547
1210
1004
996
466
1114
972
1010
866
303
268
8.1
18.0
13.1
18.8
8.1
19.3
15.5
33.5
18.3
55.7
255
326
405
302
381
375
232
223
265
272
255
360
386
455
304
192
238
274
247
133
-------
TABLE A-6. AS-RECEIVED COAL PROPERTIES
DATE
As Received
* Moisture
% Ash
i Volatile
% Fixed C.
Btu/lb
Dry Basis
I C
% H
% N.
% Cl
% s
% Ash
% 0^
Fusion
Initial
1st Softening
2nd Softening
Fluid
Mineral Analysis
Phos . Pent Ox.
Silica
Ferric Ox.
Alumina
Titania
Sodium Ox.
Potasium Ox.
Lime
Magnesia
Sulfur TriOx.
Undet.
DEC
2. 12
10.78
29.42
57.68
13,471
77. 15
4.77
1.26
.26
3.57
11.01
1.98
2060°F
2180°F
2280°F
2500°F
.48
35.43
34.94
22.39
. 56
.25
.99
1.63
.28
1.23
1.82
JAN
8. 00
11.78
18.33
61.89
12, 100
72.7
4.1
1.6
__
1.70
12.81
7.1
2460°F
2540°F
2500°F
2570°F
.92
43. 50
21.00
20.70
1.42
2.70
2. 37
.38
.58
.83
JAN
10.80
15.87
15.73
57.60
10, 910
69. 2
3.8
1.5
__
1.1
17.79
6.6
2540°
2610°
2660°
2720°
JAN
2.30
15.24
14.43
68.03
12,380
69.6
3. 5
1.6
__
2.6
15.60
7. 1
F 2280°F
F 2340°F
F 2390°F
F 2440°F
JAN
AVERAGE
7.03
14. 30
16.16
62.50
11, 797
70.5
3.8
1.57
1.80
15.40
6.93
2427°F
2497°F
2517°F
2577°F
.92
43.50
21.00
20.70
1.42
2.70
2.37
. 38
. 58
.83
MARCH
4.92
10.50
30.38
54. 20
12,675
75.21
4.90
1.34
.06
1.58
11.04
5.87
2330°F
2385°F
2430°F
2525°F
.41
52.05
12.74
25.64
.70
.47
1.87
2.18
. 36
1.66
1.95
MARCH
4.00
10.70
19. 87
65.43
12, 780
73. 1
4.2
1.6
.14
1.90
11.14
8.0
2340°F
2400°F
2450°F
2500°F
MARCH
2.20
10.50
18. 60
68. 60
13,210
73.9
4.1
1.7
.14
1.60
10.74
7.9
2330°F
2400°F
2460°F
2520°F
MARCH
4.00
9. 20
20.85
65.95
13,170
74.4
4. 3
1.7
.11
1.80
9. 58
8.2
2300°F
2360°F
2400°F
2450°F
MARCH*
AVERAGE
3.78
10.23
22.43
63. 55
12,959
74.15
4.38
1.59
.11
1.72
10.63
7.42
2325V
2386V
2435V
2499":"1
.41
52.02
12.74
25.64
.70
.47
1.87
2.18
.36
1.66
1.95
MAY
1.46
22.15
23.45
52.94
11,603
66.82
4.48
1.15
.08
1.01
22.48
3.98
2700+°F
2700+OF
2700+°F
2700+°F
MAY
1.00
28.27
16.91
53.82
10, 800
63.10
3.77
1.21
.03
1.09
28.56
2. 24
2700+°F
2700+°F
2700+°F
2700+°F
.30
62.23
3.83
26.83
.89
.27
2.52
.43
1.19
.02
1.49
MAY MAY
AVERAGE
1.35
15.43
27.29
55.93
12,715
72. 28
4.74
1.69
.05
1.56
15.64
4.04
2700+°F
2700+°F
2700+°F
2700+°F
.35
57.00
7.76
28.02
.91
.34
2.12
.64
1.10
. 62
1.14
1.27
21.95
22.55
54.23
11,706
67.4
4.33
1.35
.05
1.22
22.23
3.42
2700+°F
2700+°F
2700+°F
2700+°F
.33
59.62
5.80
27.43
.90
.31
2.32
.54
1.15
. 32
1.32
-------
TABLE A-7. MOISTURE AND ASH FREE COAL PROPERTIES
co
DEC
% Moisture
% Ash
Btu/lb
MAF Basis
% Vol
% Fixed C.
Btu/lb
% C
% H
% N
% Cl
% S
* 02
0
0
13,
33
66
15,
86
5
1
4
2
471
.78
.22
466
.70
.36
.42
.29
.01
.23
JAN
0
0
12,100
22.85
77.15
15,084
83. 38
4.70
1.83
-_
1.95
8.14
JAN
0
0
10,910
21.45
78.55
14, 878
84.18
4.62
1.82
--
1.34
8.03
JAN
0
0
12,
17
82
15,
82
4
1
3
8
380
. 50
.50
013
.46
.15
.90
.08
.41
JAN
AVERAGE
0
0
11.
20
79
14,
83
4
1
2
8
797
. f>
.4
996
.34
.49
.85
.12
.19
MARCH
0
0
12,
35
64
14,
84
5
1
1
6
675
.92
.08
986
. 55
. 51
.51
.07
.78
.60
MARCH
0
0
12,780
23.29
76.71
14, 982
82.27
4.73
1.80
.16
2.14
9. 00
MARCH
0
0
13,210
21.31
78. 58
15, 132
82.79
4. 59
1.90
.16
1.79
8.85
MARCH
0
0
13,170
24.02
75.98
15, 173
82.29
4. 76
1.88
.12
1.99
9.07
MARCH "
AVERAGE
0
0
12,
26
73
15,
82
4
1
1
8
959
.14
.84
069
.98
.90
.77
.13
.93
. 38
MAY
11,
30.
69.
15,
86.
5.
1.
1.
5.
0
0
603
70
30
189
20
78
48
10
30
13
MAY
0
0
10,
23
76
15,
88
5
1
1
3
800
.90
.10
270
. 32
.28
.69
.04
.53
.14
MAY
0
0
12,
32
67
15,
85
5
2
1
4
715
.80
.20
279
.68
.62
.00
.06
.45
.79
MAY
AVERAGE
0
0
11,706
29.13
70.87
13,246
86.73
5.56
1.72
.07
1.43
4.35
Note: Used as the reference fuel.
-------
TABLE A-8. AS-RECEIVED dRDF PROPERTIES
DEC DEC DEC
AVERAGE
As Received
'(, Moi sture
•i Ash
* Volatile
* Fixed C.
Btu/lb
Dry Basis
't C
'i II
* ^2
y, ci
% s
* Ash
I 02
Fusion
Initial
1st Softening
2nd Softeninq
Fluid
Mineral Analysis
Phos.Perrt.Ox
Silica
Ferric Ox.
Alumina
Ti tania
Sodium Ox.
Potasium Ox.
Lime
Magnesium
Sulfur TriOx.
Undet .
10.72 16.08
16.95 22.98
59.89 53.18
12.44 7.76
6667 6309
43.98
5.29
. 35
.40
.19 .60
18.99 27.38
30. 80
1955°F 2080T
2055°F 2120°F
2175°F
2290°F 2260°F
. 87
55. 52
2. 27
13.45
.66
6. 82
1. 30
10. 75
1 .14
6. 03
1.19
13.4
19.97
56. 54
10.1
6488
43.98
5.29
.35
.40
. 40
23. 19
30.80
2 0 1 8 ° F
2088°F
2175°F
2275°F
, -87
55. 52
2. 27
13.45
. 66
6.82
1.30
10. 75
1. 14
6.03
1.19
MARCH
9. 55
24. 55
57.84
8.06
6008
37. 04
5. 04
.47
.49
. 22
27. 14
29.60
1940°F
2 0 0 0 ° F
2060°F
2180°F
.67
64.29
2. 34
6.74
. 55
9.60
. 60
10. 08
1.77
1.49
1.87
MARCH
15. 69
24.28
50. 32
9.71
5059
41.3
3.9
. 30
.40
. 30
28. 80
25. 00
2 0 8 0 ° F
2160°F
2210°F
2 2 6 0 ° F
MARCH
2080°F
2130°F
2 1 7 0 ° F
2200°F
.71
77.45
1.91
2. 77
1.10
1.46
. 34
6.90
.81
1. 13
MARCH MARCH
AVERAGE
2060°F
2120°F
2180°F
2220°F
. 82
73.00
4.41
3.78
1.33
5.93
. 65
5. 54
. 77
1. 03
12.62
24.41
54.08
8.89
5534
39.17
4.47
. 39
.45
.26
27.97
27. 30
2040°F
2103°F
2 1 5 5 ° F
2215°F
.73
71. 58
2. 89
4.43
.99
5.66
. 53
7. 50
1.12
1.22
1.87
MAY MAY
7.00 13.15
23.51 30.04
58.42 47. 57
11.07 9.24
6015 5068
38.48 34.04
5.38 4.27
1.23 .58
.31 .32
.19 .27
25i28 34.59
29.13 25.93
1960°F 2050°
2060°F 2150°
2080°F 2170°
2190°F 2260°
.73 .58
63.06 63. 58
1.57 4.27
6.43 12.23
.56 .73
9.69 5.81
.65 1.22
10.54 7.82
1.71 1.54
3.60 .93
1.46 1.29
MAY
16
32
41
8
.51
.72
.'81
.96
4716
34
3
39
20
F
F
F
F
64
2
6
7
10
1
b
.38
.96
.73
.44
.37
.19
.93
. 64
. 31
.09
. 52
.79
.08
.86
.87
. 51
. 08
.25
MAY
AVERAGE
12.22
28. 75
49.27
9.76
5266
35. 63
4. 54
.85
.36
.28
33.02
25. 33
2005°F
2105°F
2125°F
2225°F
.65
63. 65
2.64
8.39
.69
7.53
.91
9.74
1. 59
3.20
1. 00
*GRAND
AVERAGE
12.75
24'. 38
53.30
9.58
5763
39.59
4.77
.53
.40
.31
28.06
27.81
Note :
The d-RDF"properties (grand average) were used as a basis for normalizing the emissions results for all test data.
-------
TABLE A-9. MOISTURE AND ASH FREE dRDF PROPERTIES
% Moisture
% Ash
Btu/lb
MAP Basis
% Vol
% Fixed C.
Btu/lb
% C
% H
% N2
% Cl
% S
% 02
DEC
0
0
6667
82.80
17.20
9217
54.29
6.53
.43
.49
.23
38.02
DEC
0
0
6309
87.27
12.73
10353
.83
MARCH
0
0
6008
87.77
12.23
9117
50.84
6.92
.65
.67
.30
40.63
MARCH
0
0
5059
83.82
16.18
8427
58.00
5.48
.42
.56
.42
35.11
MAY
0
0
6015
84.07
15.93
8656
51.50
7.20
1.65
.41
.25
38.99
MAY
0
0
5068
83.73
16.27
8921
52.04
6.53
.89
.49
.41
39.64
MAY
0
0
4716
82.35
17.65
9290
56.54
6.51
1.20
.72
.61
34.42
AVERAGE
0
0
5692
84.54
15.46
9140
53.87
6.53
.87
.56
.44
37.80
S.D.
—
739
2.13
2.13
618
2.92
.58
.48
.12
.22
2.51
137
-------
APPENDIX B
SUMMARY SHEETS FOR ASME ABBREVIATED EFFICIENCY TESTS
AND BOILERS 1 AND 2 SPECIFICATIONS
SUMMARY SHEET
A.SME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
1:0
PTC 4.1-a(1964)
TEST NO BOILER NO.
DATE 5/4
O»NER OF PLANT LOCATION
TeST CONDUCTED BY
POLES MA* E & TYPE
DBJECTWE OF TEST
RATED CAPAC
DURA T (ON
TY
STOKER TYPE & SIZE
PULVERIZER. TYPE & SIZE BURNER, TYPE & SIZE
FUEL USED MINE COUNTY STATE
SIZE AS FIRED
PRESSURES & TEMPERATURES FUEL DATA
1
2
3
4
5
6
7
a
9
10
"
12
13
U
STEAM PRESSURE IN BOILER DRUM
STEAM PRESSURE AT S H. OUTLET
STEAM PRESSURE AT R. 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
(This is Reference Temperature) T
TEMPERATURE OF FUEL
GAS TEMP. LEAVING (Boiler) (Eton.) (Air Hlr.)
GAS TEMP. ENTERING AH (II conditions to be
psio
psio
psi a
psio
F
F
F
F
F
F
F
F
F
153
230
.95
68
457
UNIT 0 UANTITIES
15
16
17
ia
19
20
21
22
23
24
25
ENTHALPY OF SAT. LIQUID (TOTAL HEAT)
ENTHALPY OF (SATURATED) (SUPERHEAT ED)
STM
ENTHALPY OF SAT. FEED TO (BOILER)
(ECON.)
ENTHALPY OF REHEATED STEAM R.H. INLET
ENTHALPY OF REHEATED STEAM R. H.
OUTLET
HEAT ABS/LB OF STEAM (ITEM 16-ITEM 17)
HEAT ABS, LB R.H. STEAMOTEM 19-ITEM 18)
DRY REFUSE (ASH PIT » FLY ASH) PER LB
AS FIRED FUEL
Btu PER LB IN REFUSE (WEIGHTED AVERAGE)
CARBON BURNED PER LB AS FIRED FUEL
DRY GAS PER LB AS FIRED FUEL BURNED
Biu/lb
Btu/lb
Biu/lb
Btu/lb
Btu/lb
Btu'lb
Btu/lb
Ib/lb
Btu/lb
Ib/lb
Ib/lb
HOURLY QUANTITIES
26
27
28
29
30
31
ACTUAL WATER EVAPORATED
REHEAT STEAM FLOW
RATE OF FUEL FIRING (AS FIRED -t)
TOTAL HEAT INPUT (Item 28 X Item 41)
1000
HEAT OU'PUT IN BLOW-DOWN WATER
HEA*L "'"" J''llem 20) •(!(•", J7. Item 21) • Item 30
OUTPUT '000
b'hr
b,-hr
Ib/nr
kBAr
UB/hr
kB/n,
198.2
L195.8
198.2
997. ft
.37
5800
.517
21. b
10,200
1952
22,850
—
10,176
FLUE CAS ANAL. (BOILERMECON) (AIR MTR) OUTLET
32
33
34
35
36
CO,
°J
CO
Nj (BY DIFFERENCE)
EXCESS AIR
-. VOL
-, VOL
% VOL
r. VOL
%
5.9
11.3
—
82.7
104
COAL AS FIRED
PROX. ANALYSIS
37
38
39
40
MOISTURE
VOL MATTER
FIXED CARBON
ASH
TOTAL
41
42
Btu per Ib AS FIRED
ASH SOFT TEMP.'
ASTM METHOD
% wt
1.3
22.55
54.23
21.95
11,706
COAL OR OIL AS FIRED
ULTIMATE ANALYSIS
43
44
45
46
47
40
37
CARBON
HYDROGEN
OXYGEN
NITROGEN
SULPHUR
ASH
MOISTURE
TOTAL
66.52
4.27
3.38
1.33
1.20
21.95
1.3
COAL PULVERIZATION
48
49
50
64
GRINDABILITY
INDEX-
FINENESS %THRU
50 M-
FINENESS 1. THRU
200 M-
INPUT. OUTPUT
EFFICIENCY OF UNIT %
51
52
53
44
41
OIL
FLASH
Sp. Gro
POINT f
• ity Deg. API-
VISCOSITY AT SSU-
BURNER SSF
TOTAL
% wt
Btupe
HYDROGEN
Ib
CAS
54
55
56
57
58
59
60
t-61
CO
CH4 METHANE
CjH, ACETYLENE
C,H« ETHYLENE
C,H.
ETHANE
H,S
COj
H,
HYDROGEN
TOTAL
62
63
41
TOTAL
*B wt
HYDROGEN
% VOL
DENSITY 68 F
ATM. PRESS.
Btu PER CU FT
Btu PER LB
ITEM 31
• 100
ITEM 29
HEAT LOSS EFFICIENCY
65
66
67
68
69
70
71
72
HEAT LOSS DUE TO DRY GAS
HEAT LOSS DUE TO MOISTURE IN FUEL
HEAT LOSS DUE TO HjO FROM COMB.OFH,
HEAT LOSS DUE TO COMBUST- IN REFUSE
HEAT LOSS DUE TO RADIATION
UNMEASURED LOSSES
Btu'lb
A. F. FUEL
TOTAL
EFFICIENCY = (100 - t.m 71)
% ol A F
FUEL
17.9
.1
4.0
18.3
3.7
1.5
45.5
54.5
'Not Reouireo1 for Elfic «ic* Tesfng
> For Po.nt of Meosurem.nl See Por. 7.2.8 1. PTC 4.1-1964
138
-------
CALCULATION SHEET
ASME TEST FORM
FOR ABBREVIATED EFFICIENCY
1:0
PTC 4.1-b (1964)
TEST Revised September, 1965
CWNER OF PLANT TEST NO. BOILER NC. DATE
3C'
24
25
36
65
66
67
68
„
70
71
72
r ITEM 15 ITEM 171 kBlr
HE«T (TUTPUT N BOILEB "LC'.'iO'M »ATE p =' £ "i7 « • - = !; E> "• 'f"- r •= r f - , - 1
L 1300
/f impractical to we igb refuse, this
item can be estimated as follows 21.95
npv OFFIKF PFP inner .< r,PFCFUE' - ' ASH IN AS FIRED COAL f(OT[. |fr
,00 -'.COMB. IN REFUSE SAMPUE PITREFUSE
-UE DUST 1 ASH
DIFFER MATERIALLY
,__ r— —1 N COMBUSTIBLE CONTENT, THEY
££ PT«M ITEM«1 SHOULD BE ESTIMATED
CARBON BURNED 66.52 .37 x 5800 .517 SEPARATELY. SEE SECTION 7
FUEL 'DO [_ U.500 J I.UMKUIATIUNS.
DRY GAS PER LB llCOj « 80, • 7(Na « CO)
BURNED 3(CO> * C0) / v. 8
ITEM 32 ITEM 33 | ITEM 35 ITEM 34 ) ITEM 24 ITEM
11x5. 9 * « x jj.3 « 7^ 82.7' — / x .517 * 1.2
(ITEM 32 ITEM 34 \
.5.9 . « ) L
/
E,CE,t 0, - C° ,TVu,J, - 'TEM-3*
AIR* = 1 00 X 1 no x ^ =
.2682N, - (0 C0_ ) BZ./ 11.3 ITEM 34
* 3 3n83(ITFM3S) (ITFM13- "=•»•»)
1
HEAT LOSS EFFICIENCY
HEAT LOSS DUE LB DRY GAS .25 ITEM 25 ° . 25(| TEM ,31 _,|TEM , ,,
TO DRY GAS = PERLBAS xC x ('!., -•„;,)= _n"x«»4X ' Uit*ll) =
FIRED FUEL ' Uni, 21.6 457 68
MO*STURES|N1FUELI = AS F'RE"^! X 1 (ENTHALPY OF VAPOR AT 1 PSIA & T GAS LVG)
(ENTHA1 PY OF 1 IQUIDAT T AIR)] ~ ''^M 37 X[(FNTHAI PY OF VAPOR
1269 36 10° ,
AT 1 PSIA 1 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 GAS
4. 27 LVG) - (ENTHALPY OF LIQUID AT T AIR)]
- „ , !TEM 44 x [,ENTHALPr nf VAPOR AT 1 P^IA «. T ITEM 13) - (ENTHALPY OF LIQUID AT
HEAT LOSS DUE TO ITEM 22 ITEM 23
COMBUSTIBLE IN REFUSE = 37 x 5800 =
HCAT LOSS DUE TO TOTAL BTU RADIATION LOSS PER HR
KADiAtlGN' LB AS FISt D FUE L ITI^ZB
UNMEASURED LOSSES ••
TOTAL
EFFICIENCY = (100 - ITEM 71)
47~l
21.6
,67 J
104%
Blu/lb
AS FIRED
FUEL
2.1.Q1.
.16:0.
474
ZU6 .
LOSS „
HHV
100 =
41
41
67
__X100 =
41
— xlOO =
41
69
x 1 00 -
41
70 x 100 =
41
LOSS
X
17.9
.1
4.P
. ia,3
. . .3. -.7
1.5
45.5
54.5
II lo»ei ore noi meo.u.rd. u»e ABMA Slondo'd Rodrolion Lou Chorl. F,g. 8. PTC 4.1-1964
• Unm.oiuod loins li»>«d in PTC 4.1 but not lobulot.d obov.mo, by pro'idcd (or by o»>ijn,ng o •
og...d upon -o'u« fc>' I'"" 70.
139
-------
SUMMARY SHEET
A.SME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
1:1
PTC 4.1.a(1964)
TEST NO BOILER NO.
DATE 5/13
0»NER OF PLANT LOCATION
TEST CONDUCTED BY OBJECTIVE OF TEST
DURATION
BO LEO MAKE 8. TYPE RATE D CA PAD TY
STOKER TYPE & SIZE
PULVERIZER, TYPE & SIZE BURNER, TYPE & SIZE
FUEL USED MINE COUNTY STATE
SIZE AS FIRED
PRESSURES & TEMPERATURES FUEL DATA
1
2
3
4
5
6
7
8
9
10
11
12
13
14
STEAM PRESSURE IN BOILER DRUM
STEAM PRESSURE AT S. H. OUTLET
STEAM PRESSURE AT R. 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 MBOILER)
STEAM QUALITY?". MOISTURE OR P. P. M.
AIR TEMP. AROUND BOILER (AMBIENT)
TEMP AIR FOR COMBUSTION
TEMPERATURE OF FUEL
GAS TEMP. LEAVING (Boiler) (Econ.) (Air Htr.)
corrected to Quarontee)
piia
piia
pita
psio
F
f
F
f
f
F
F
F
F
155
221
.95
78
—
501
UNIT QUANTITIES
15
16
17
18
19
20
21
22
23
24
25
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/LB OF STEAM (ITEM 16-ITEM 17)
HEAT ABS/LB R.H. STEAM(ITEM 19-ITEM IB)
DRY REFUSE (ASH PIT « FLY ASH) PER LB
AS FIRED FUEL
Btu PER LB IN REFUSE (WEIGHTED AVERAGE)
CARBON BURNED PER LB AS FIRED FUEL
DRY GAS PER LB AS FIRED FUEL BURNED
3tu/lb
Btu/lb
Btu/lb
Btu/lb
Btu/lb
Blu'lb
Btu/lb
Ib/lb
Btu/lb
Ib/lb
Ib/lb
HOURLY QUANTITIES
26
27
28
29
30
31
ACTUAL WATER EVAPORATED
REHEAT STEAM FLOW
RATE OF FUEL FIRING (AS FIRED wt)
TOTAL HEAT INPUT (Item 28 x Item 41)
1000
HEAT OUTPUT IN SLOW-DOWN WATER
T|0!J*.L(ltem26.lt.m20).(ltem27»ltem2l).ltem30
OUTPUT 1000
b'hr
Ib/hr
Ib/hr
kB/h,
liB/hr
kB/n,
1196
189
1007
.36
5075
.41
12.62
20.018
—
3309
29.741
—
20,158
FLUE GAS ANAL. (BOILERHECON) (AIR HTR) OUTLET
32
33
34
35
36
CO,
o,
CO
N, (BY DIFFERENCE)
EXCESS AIR
r, VOL
••. VOL
% VOL
% VOL
*
8.1
9,9
82.0
82
COAL AS FIRED
PROX. ANALYSIS
37
38
39
40
MOISTURE
VOL MATTER
FIXED CARBON
ASH
TOTAL
41
42
Btu per Ib AS FIRED
ASH SOFT TEMP.-
ASTM METHOD
% wt
6.62
31.68
38.39
23.33
89.88
COAL OR OIL AS FIRED
ULTIMATE ANALYSIS
43
44
45
46
47
40
37
CARBON
HYDROGEN
OXYGEN
NITROGEN
SULPHUR
ASH
MOISTURE
TOTAL
54.06
4.10
9.83
1.06
L .86
23.33
6.62
COAL PULVERIZATION
48
49
50
64
GR1NDABIL1TY
INDEX-
FINENESS XTHRU
50 M*
FINENESS X THRU
200 M*
51
52
53
44
41
OIL
FLASH
Sp. Gro
POINT f
• ity Dej. API-
VISCOSITY AT SSU-
BURNER SSF
TOTAL HYDROGEN
Btu pe
Ib
GAS
54
55
56
57
58
j,
60
61
CO
CH4 METHANE
C,H, ACETYLENE
C,H« ETHYLENE
C.H.
ETHANE
H,S
CO,
H,
HYDROGEN
TOTAL
62
63
41
TOTAL HYDROGEN
Xwt
5 VOL
DENSITY 68 F
ATM. PRESS.
Btu PER CU FT
Btu PER LB
INPUT-OUTPUT ITEM 31 - 100
EFFICIENCY OF UNIT X ITEM 29
HEAT LOSS EFFICIENCY
65
66
67
68
69
70
71
72
HEAT LOSS DUE TO DRY GAS
HEAT LOSS DUE TO MOISTURE IN FUEL
HEAT LOSS DUE TO H,O FROM COMB. OF H
HEAT LOSS DUE TO COMBUST. IN REFUSE
HEAT LOSS DUE TO RADIATION
UNMEASURED LOSSES
Btu/lb
A. F. FUEL
TOTAL
EFFICIENCY = (100 - Item 71)
% of A F
FUEL
14.8
.9
5.1
20.3
1.8
1.5
44.4
55.6
'Not Required lor Efficiency Telling
t For Point =1 Me.iuremerit See Per. 7.2.8.1-PTC 4.1-1964
140
-------
CALCULATION SHEET
A.SME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
PTC4.1-b (1964)
Revised September, 1965
1:1
30
54
25
36
65
66
67
68
70
71
72
O. NEK OF PLANT T£ST NO. BOILER NO. PATE
I" ITEW 15 ITEV" '7~
HEAT CUT P. 7 IN E.OlLER Ell.0«-00' N »ATFR -L9 Oe »ATER &L.O»-"O»N PER H» x ' .... . . . - . _^
\ 1000
If impractical fa weigh refuse, ffiis
ifem con Se estimated as follows 23 33
100 - ?i COMB. IN REFUSE SAMPLE p|T REflJS€
WB n.
-UE DUST 4 ASH
DIFFER MATERIALLY
p- — , IN COMBUSTIBLE CONTENT, THEY
ITEM 43 F7TEM22 ITEM 23 1 SHOULD BE ESTIMATED
CARBON BURNED 5406 .36 „ 4205 .41 SEPARATELY. SEE SECTION 7.
PWVBMF.RED . 10() ^ U5M J , COMPUTATIONS.
DRY GAS PER LB HCOj + 80, + 7(N, + CO)
BURNED 3(C0' * C0) / \ _ .
ITEM 32 ITEM 33 / ITEM 35 ITEM 34) I ITEM J4
I1x .8.1 *8x..9.9. *^__82. » .—-./ x ,91.
+ 3 s)
ITEM 47 I
* ..:s.6.. 12.62
/ITEM 32 ITEM34\ L 267 J
3 x 1. .9:f. . * .."..1
EltrEI- °i C ITEM'S ITEM 34
EXCESS i • Q Q -
A l R t = i oo x — — -innv "••' * =
.2682N, - <0 . i5L | |TEM ,.
2 2<»7 (ITPM Tr) (ITFU n )
HEAT LOSS EFFICIENCY
HEAT LOSS DUE LB DRY GAS ITEM 25 (ITEMJ31 (ITFM11I
TODRYGAS = PERLBAS xC X Civ, -•«,)=," xMZ ' ~' " =
FIRED FUEL " Unit 12.62
MOmu^N^U^ '%*££&&* l™™^ Of VA^AT 1 PSI^^G^S LVG)
(EMTH4' PY "F I iiJU'PAT T AIR)] - " fc 7 » [(pMTHAl Pv OF VAPOB
100 46
AT 1 PSIA & T ITEM 13) -(ENTHALPY OF LIQUID AT T ITEM 11)] =
HEAT LOSS DUE TO H,0 FROM COMB. OF H, = 9H, x [(ENTHALPY OF VAPOR AT I PSIA & T GAS
4_1 LVG) - (ENTHALPY OF LIQUID AT T AIR)]
. , , ITEM 44 x [(ENTHALPY OF VAPOR AT 1 PSIA 1 T ITEM 11) - (ENTHALPY OF LIQUID AT
100 T ITEM 11)] =
HEAT LOSS DUE TO ITEM 22 ITEM 23
COMBUSTIBLE IN REFUSE = . 36 x 5075 ~
HEAT LOSS DUE TO TOTAL BTU RADIATION LOSS PER MR
RADIATION' LB AS FIRED FUEL ITtxze
UNMEASURED LOSSES ••
TOTAL
EFFICIENCY = (100 -ITEM 71)
82%
B.u/l.
AS FIRED
FUEL
1335
82
458
18Z7. .
LOSS x
HHV
100 =
65
X 100 =
41
** x 100 =
41
asur«d lo.^i li«»i( in PTC 4.1 but »oi iabulot«d above mo» b> provided lor by oinjn.ng o ^
O0,..d upon .olu. to, II.". 70.
141
-------
SUMMARY SHEET
A.SME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
1:2
PTC 4.1-o{1964)
TE;T NC BOILER NO. DATE 5/n
OiNER OF PLANT LOCATION
TS ST CONDUCTED BY OBJECTIVE OF TEST DURATION
B2:L;R MA-.E & TYPF RATE D CA PACi Tr
STOKER TYPE & SIZE
PULVERIZER, TYPE s. SIZE BURNER, TYPE & SIZE
FUEL USED MINE COUNTY STATE SIZE AS FIRED
PRESSURES & TEMPERATURES FUEL DATA
1
2
3
4
5
6
-i
e
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
STEAM PRESSURE IN BOILER DRUM
STEAM PRESSURE AT S. H. OUTLET
STEAM PRESSURE AT R. 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 )(BOILER)
STEAM QUALITY r. MOISTURE OR P. P.M.
AIR TEMP. AROUND BOILER (AMBIENT)
TEMP AIR FOR COMBUSTION
(Thij ii Reference Temp.rotw'e) T
TEMPERATURE OF FUEL
GAS TEMP. LEAVING (Boil.r) (Eton.) (Air Hlr.)
GAS TEMP. ENTERING AH (II conditions to b.
UNIT QUANTITIES
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/LB OF STEAM (ITEM 16-ITEM 17)
HEAT ABS/LB R.H. STEAM(ITEM 19-ITEM 18)
DRY REFUSE (ASH PIT » FLY ASH) PER LB
AS FIRED FUEL
Btu PER LB IN REFUSE (WEIGHTED AVERAGE)
CARBON BURNED PER LB AS FIRED FUEL
DRY GAS PER LB AS FIRED FUEL BURNED
psio
?SIO
psro
ps.a
F
F
F
F
F
F
F
F
F
Btu/lb
Btu/lb
Btu/lb
Blu/lb
Btu/lb
3tu'lb
Btu/lb
Ib/lb
Btu/lb
Ib/lb
Ib/lb
HOURLY QUANTITIES
ACTUAL »ATER EVAPORATED
REHEAT STEAM FLO*
RATE OF FUEL FIRING (AS FIRED -ly
TOTAL MEAT INPtJT (It*™ 28 X Item 41)
1000
HEAT OUTPUT IN BLOW-DOWN WATER
HEAT"" (l"m J«'ll«'" 20)>(it«™ 37.11..- 31) •it.m. 30
OUTPUT 1000
Ib K.
b/h.
1 b/h r
kB/hr
it B hr
kB/hr
154
230
.95
70
512
1195.8
198.2
997.6
.35
4205
.371
14.3
i7O3.a
3776
30,684
—
17,696
FLUE CAS ANAL. (BOILERMECON) (AIR HTR) OUTLET
32
33
34
35
36
CO,
Q
CO
N, (BY DIFFERENCE)
EXCESS AIR
". VOL
', VCL
% VOL
% VOL
%
6.4
11.0
—
3?. 6
qa.6
COAL AS FIRED
PROX. ANALYSIS
37
38
39
40
MOISTURE
VOL MATTER
FIXED CARBON
ASH
TOTAL
41
42
Btu per Ib AS FIRED
ASH SOFT TEMP.
ASTM METHOD
% .t
7.94
38.26
29.09
24.73
83.82
COAL OR OIL AS FIRED
ULTIMATE ANALYSIS
43
44
45
46
47
40
37
CARBON
HYDROGEN
OXYGEN
NITROGEN
SULPHUR
ASH
MOISTURE
TOTAL
47.29
4.07
14.19
.90
.66
24.73
7.94
COAL PULVERIZATION
48
49
50
64
GRINDABILITY
INDEX-
FINENESS %THRU
50 M*
FINENESS % THRU
200 M*
INPUT-OUTPUT
EFFICIENCY OF UNIT %
51
52
53
44
41
OIL
FLASH POINT f
Sp. Grovity Deg. AP •
VISCOSITY AT SSU'
BURNER SSF
TOTAL HYDROGEN
% wt
Btu per Ib
CAS
54
55
54
57
58
V»
60
61
CO
CH« METHANE
CjH, ACETYLENE
C,K, ETHYLENE
CjH, ETHANE
H,S
CO,
H, HYDROGEN
TOTAL
63
63
41
TOTAL HYDROGEN
X .t
%VOL
DENSITY M F
ATM. PRESS.
Btu PER CU FT
Btu PER LB
ITEM 31 . 100
ITEM 39
Blu/lb
HEAT LOSS EFFICIENCY A. F. FUEL
65
66
67
68
69
70
71
72
HEAT LOSS DUE TO DRY GAS
HEAT LOSS DUE TO MOISTURE IN FUEL
HEAT LOSS DUE TO H,0 FROM COMB OF H,
HEAT LOSS DUE TO COMBUST. IN REFUSE
HEAT LOSS DUE TO RADIATION
UNMEASURED LOSSES
TOTAL
EFFICIENCY = (100 - Item 71)
%•! A. F
FUEL
18.8
1.2
5.4
17.6
1 .8
1.5
46.3
53.7
'Not R.quir.d lo' Efflcl.nc, T.ltlng
t For Po.nt ol M.otur.m.nt S.. Pot. 7.2.8.1-PTC 4.1-1964
142
-------
CALCULATION SHEET
ASME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
PTC4.1-B (1964)1:2
Revised September, 1965
30
24
25
36
65
66
67
66
f-
70
71
72
OWNEK OF PLAHT TEST NO. BOILER NO. DATE
T ITEM 15 ITEM 1?"
^T •' ~ ;••_ - : ,'T II, B3ILES BLDW-DOWN ».'. TE " = _E OF » o -| p c . 3 , .- - . •. ;•£ R t-i r > < -
L
item con fae estimated as follows 24.73 .35
- = v prr.. x ITEM 44 x ((ENTHALPY OF VAPOR AT 1 PSIA 8. T ITEM 13) - (ENTHALPY OF LIQUIO AT
100 T ITEM ID] =
HE AT LOSS DUE TO ITEM 22 ITEM23
COMBUSTIBLE IN REFUSE ^5 * 42Q5
NT A T L OS^ DUE TO TOTAL RTURAPtATIONlOSSPr^HP
k /. ! 1 1 A 1 ! ON • I b l 1 ' . R I. : ^ J f > ' ' '-• . *
UNMEASURED LOSSES ••
TOTAL
EFFICIENCY = (100 -ITEM 71)
Btu/lb
AS FIRED
FUEL
15KQ
100
455
1472
LOSS
TiffT
100 :
41
^- x 100 s
41
67
XIOO =
41
68
— xlOO =
41
69
jl
70 x 100 -
41
LOSS
X
.18,8
. . 1,2.
5.4
17.6
1.8
1.5
46.3
53.7
If I0,»** or. nor mroiured, u.e ABMA Sfondo'd Rod.of.oo Lou Chort. F.g. 8, PTC 41-1964
" Unm«8«uf«d lo»»«I l>»fed rn PTC 4.1 but not lobut«l*d obov* moy by provided for by aitigmng o '
««r**d upon *otu« for Item 70.
143
-------
SUMMARY SHEET
A.SME TEST FORM
FOR ABBREVIATED EFFICIENCY TEST
0:1
PTC 4.l-a{1964)
TEST NO BOILER NO.
0*NER 0= PLANT LOCATION
*EST- CONDUCTED BY OBJECTIVE OF TEST
DATE 5/14/77
DURATION
BOLfR MAKE 4 TYPE RA T E D CAPACI TY
STOKER TYPE 8. SIZE
PULVERIZER, TYPE & SIZE BURNER, TYPE & SIZE
FUEL USED MINE COUNTY STATE
SIZE AS FIRED
PRESSURES & TEMPERATURES FUEL DATA
1
2
3
4
5
6
7
8
9
10
I 1
12
13
14
STEAM PRESSURE IN BOILER DRUM
STEAM PRESSURE AT S. H. OUTLET
STEAM PRESSURE AT R. 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 ((BOILER)
ST E AM QUA LI TYT. MOISTURE OR P. P.M.
AIR TEMP. AROUND BOILER (AMBIENT)
TEMP AIR FOR COMBUSTION
TEMPERATURE OF FUEL
GAS TEMP. LEAVING (Boil«r) (Econ.) (Air Htr.)
GAS TEMP. ENTERING AH (If condition, to be
psio
psio
psia
P!,0
F
F
F
F
F
F
F
F
F
152
226
.95
78
472
UNIT 0 UANTITIES
15
16
17
18
19
20
21
22
23
24
25
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 A8S/LB OF STEAM (ITEM I6-ITEM 17)
HEAT ABS/LB R.H. STEAM (ITEM 19-ITEM 18)
DRY REFUSE (ASH PIT * FLY ASH) PER LB
AS FIRED FUEL
Btu PER LB IN REFUSE (WEIGHTED AVERAGE)
CARBON BURNED PER LB AS FIRED FUEL
DRY GAS PER LB AS FIRED FUEL BURNED
Btu/lb
Btu/lt>
Btu/lb
Btu/lk
Btu/lb
Biu'lb
Btu/lb
Ib/lb
Btu/lb
Ib/lb
Ib/lb
HOURLY QUANTITIES
26
27
28
29
30
31
ACTUAL WATER EVAPORATED
REHEAT STEAM FLOW
RATE OF FUEL FIRING (AS FIRED wt)
TOTAL HFAT INPUT (Item 28 X liem 41)
1000
HEAT OUTPUT IN BLOW. DOWN WATER
H°ATL(l"m26""""20|'(l"m37'''""2"'''""30
OUTPUT 1000
Ib h,
Ib/hr
Ib/hr
fcBAr
VB/hr
kB/nr
1195.7
194.2
1001.5
.269
580
.298
6.404
tl,459
4991
5103
25^5J}Q
11.476
FLUE GAS ANAL. (BOILERMECON) (AIR HTR) OUTLET
32
33
34
35
36
co;
Gj
CO
N, (BY DIFFERENCE)
EXCESS AIR
% VOL
', VOL
% VOL
% VOL
%
7.3
11.5
83.2
112
COAL AS FIRED
PROX. ANALYSIS
37
38
39
40
MOISTURE
VOL MATTER
FIXED CARBON
ASH
TOTAL
41
42
Btu per Ib AS FIRED
ASH SOFT TEMP.'
ASTM METHOD
% wt
16.60
48.59
8.96
25.85_
5103
COAL OR OIL AS FIRED
ULTIMATE ANALYSIS
43
44
45
46
47
40
37
CARBON
HYDROGEN
OXYGEN
NITROGEN
SULPHUR
ASH
MOISTURE
TOTAL
30.90
1.7fi
?1 .78
.SS
.23
25.85
16.60
COAL PULVERIZATION
48
49
50
(4
GRINDABILITY
INDEX*
FINENESS %THRU
SO M-*
FINENESS % THRU
200 M-
INPUT-OUTPUT
EFFICIENCY OF UNIT *
51
52
53
44
41
OIL
FLASH
Sp. Gro
POINT F'
vity De9. API'
VISCOSITY AT SSU-
BURNER SSF
TOTAL HYDROGEN
X wt
Btu pe
Ib
GAS
54
55
VI
57
58
59
60
61
CO
CH. METHANE
C,H, ACETYLENE
C,H, ETHYLENE
C,H.
ETHANE
H,S
CO,
H,
HYDROGEN
TOTAL
62
63
41
TOTAL
X wt
HYDROGEN
SVOL
DENSITY 68 F
ATM. PRESS.
Btu PERCU FT
Btu PER LB
ITEM 31 > 100
ITEM 29
HEAT LOSS EFFICIENCY
65
66
67
68
69
70
71
72
HEAT LOSS DUE TO DRY GAS
HEAT LOSS DUE TO MOISTURE IN FUEL
HEAT LOSS DUE TO H,0 FROM COMB.OFH,
HEAT LOSS DUE TO COMBUST. IN REFUSE
HEAT LOSS DUE TO RADIATION
UNMEASURED LOSSES
B.u/lb
A. F. FUEL
TOTAL
EFFICIENCY = (100 - Item 71)
%of A. F
FUEL
12.4
T.7
7.5
3.1
3 7
1.5
31.9
68.1
'Not Required for Efficiency Totting
t For Point of Meolurement See POP. 7.2.I.I-PTC 4.1-1964
144
-------
CALCULATION SHEET
ASME TEST FORM
FOR ABBREVIATED EFFICIEKCY TEST
PTC4.1-b (1964) 0:1
Revised September, !96S
3C
24
25
36
65
66
67
68
>,
70
71
72
OWNER OF PLANT TEST NO. BOILER NO. DATE
T ITEM 15 ITEM 17"
HE,»- C'J-°J~ it, BOILER BLC»-DO'.'- » A - E <; =.5 OF WATER faLC».:iOWN PEB *=! » ', -
L 1C::
If impractical to weigh refuse, this
item con be estimated as follows 95 85 7fiQ
PPV pFr,,7 CTF" IN (1TFU T iicmj«j
81 . 2 11 . 5 2
HEAT LOSS EFFICIENCY
HEAT LOSS DUE LB DRY GAS ITEM 25 ^ITEMISI IITFMMI
TO DRY GAS = PERLBAS x C x ('!.,- 'oir) = ...XXB2J . , ' ,„ =
FIRED FUEL " Uni, 6.404 472 78
MO*STURES|N FUEL3 = AS nRE^FUEL X l (ENTHA1-PY OF VA116R. 6*T ' PS'*ft^*§ LVG)
(EN THAI PY OF I 'QUIP AT T AIR)] - ' ^M K [(ENTHAI PY OF VAPOR
100 46
AT 1 PSIA & T ITEM 13) -(ENTHALPY OF LIQUID AT T ITEM 11)] =
HEAT LOSS DUE TO H,O FROM COMB. OF H, = 9H, x [(ENTHALPY OF VAPOR AT 1 PSIA & T GAS
3.76 1178.9 LVG) - (ENTHALPY^F LIQUID AT T AIR)]
- 0 * ITEM 44 » [(ENTHALPY OF VAPOR AT 1 PSIA & T ITEM 131 _ (FNTHALPY OF 1 IQUIt) AT
HEAT LOSS DUE TO ITEM 22 ITEM 23
COMBUSTIBLE IN REFUSE = .269 x 580 =
HfAT LOSS DUF TO TOTAL BTU RAOIA1ION LOSS PF R HR
KA.TItT.ON- LB AS FIRED FUFL i - 1 M . I,
UNMEASURED LOSSES "
TOTAL
EFFICIENCY = (100 -ITEM 71)
* !7234! 6.404
Jw J=
..112%
B.u/lb
AS FIRED
FUEL
. "I .
188
383
156
LOU
TiHV
100 =
41
^1 X 100 =
41
67
X100 =
41
68
— X100 =
41
69
x 100 -
41
70 x 100 =
41
LOSS
.W,4.
3.7
7.5
..3,1.
.3.7
1.5
31.9
68.1
. 9.2 - PTC 4.1-1964
• IMo'i!.',".'. 'o"'meoLr«
-------
BOILER NO. 1
ERIE CITY IRON WORKS
Erie, Pennsylvania
FOR MARYLAND INSTITUTION FOR MEN
BREATHEDSVILLE, MARYLAND
Predicted Performance Only
Guaranteed efficiency at 60,000 Ib/steam/hour is 75.0%
October 1, 1963
G.O. 96/63
17
Percent of load
33 67
1.0
Peak
1. Evaporation, actual, pounds per hour
2. Pressure in the drum, psig
3. Feedwater temperature °F
4. Temperature of flue gases leaving the
furnace °F
5. Temperature of flue gases leaving the
boiler °F
6. Excess air at boiler outlet %
(See note below)
7. Temperature of air at windbox °F
8. Dust collector draft loss, in H20
(See below)
9. Gas dust and damper draft loss (See
note below)
10. Furnace and boiler draft loss
11. Total draft required inches H20
12. Induced draft design static
13. Stoker and windbox air resistance
(See note below)
14. Duct and Damper air resistance
(No orifice in duct)
15. Total air resistance inches H20
16. Forced draft fan static design
17. Flue gas leaving the boiler, Ib/hr
18. Air required for combustion, Ib/hr
19. Fuel burned, Ib/hr
20. Furnace heat liberation Btu/cu ft
21. Heat release Btu/sq ft active grate
(at 129 sq ft grate)
22. Moisture in steam leaving the boiler %
23. Unit efficiency %
24. Losses:
A. Dry gas %
B. Moisture in fuel and hydrogen %
C. Moisture in air %
D. Unburned combustible % (See note
below*)
E. Radiation %
F. Unaccounted for %
G. Total %
13,083
175
220
422
96.45
80
0.20
0.10
0.20
0.50
0.11
0.10
0.21
25,837
24,080
1,274
4,320
93,600
1/2 of
77.71
12.34
4.02
0.33
0.80
0.80
3.30
1.50
22.29
NOTE: Items 8 and 9 are per Aerotec Industries,
Items 6 13 and 24-D are per Hoffman Comb.
June 12, 1963.
*See Item 24-D above
26,167
175
220
450
57.11
80
0.40
0.10
0.33
0.83
0.32
0.10
0.42
40,121
36,748
2,447
8,300
181,000
1 1/2 of 1
80.76
10.73
4.07
0.29
1.00
1.00
1.65
1.50
19.24
52,333 78,500
175 175
220 220
1,885
513 580
40.5 34.0
80 80
1.03 2.55
0.20 0.45
0.53 0.95
1.76 3.95
See Fan Design Below
0.70 1.40
0.20 0.45
0.90 1.85
See Fan Design Below
72,084 104,693
65,349 94,464
4,885 7,420
16,550 25,800
358,000 543,000
1/2 of 1 1/2 of 1
80.43 78.91
11.27 12.43
4.17 4.28
0.30 0.33
1.50 2.00
1.50 2.00
0.83 0.55
1.50 1.50
19.57 21.09
85,000
175
220
597
31.3
80
3.00
0.55
1.10
4.65
1.60
0.55
2.15
111,535
100,435
8,052
28,000
589,000
1/2 of 1
78.53
12.61
4.31
0.34
2.20
2.20
0.51
1.50
21.47
Inc. performance dated July 22, 1963.
Engineering
Company performance dated
The unburned combustible loss is as given by Hoffman Comb.-Engr. Company based on
reinjection from boiler hoppers and decantation collector. Guaranteed overall boiler
and stoker efficiency is 75.0% at 60,000 pounds of steam per hour.
146
-------
BOILER NO. 2
ERIE CITY IRON WORKS
Er ie, Pennsylvania
FOR MARYLAND INSTITUTION FOR MEN
BREATHEDSVILLE, MARYLAND
Predicted Performance Only
Guaranteed efficiency at 60,000 lb/steam/hour is 75.0%
October 1, 1963
G.O. 97/63
17
Percent of load
33 67
1.0
Peak
10,000
175
220
432
96.45
80
0.20
0.10
0.20
0.50
0.11
0.10
0.21
19,854
18,504
979
4,620
100,000
1/2 of 1
77.07
12.41
4.02
0.34
0.80
3.66
1.50
22.93
20,000
175
220
454
57.11
80
0.40
0.10
0.35
0.85
0.32
0.10
0.42
30,792
28,203
1,878
8,850
192,000
1/2 of
80.46
10.85
4.07
0.29
1.00
1.83
1.50
19.54
40,000 60,000
175
220
519
40.5
80
1.00
0.20
0.56
1.76
See fan design
0.70
0.20
0.90
See fan design
55,277
50,112
3,746
17,700
175
220
1,870
590
34.0
80
2.40
0.45
1.05
3.90
below
1.40
0.45
1.85
below
80,340
72,490
5,694
26,900
383,000 583,000
1 1/2 of 1 1/2 of 1
80.15
11.45
4.18
0.31
1.50
0.91
1.50
19.85
78.58
12.67
4.30
0.34
2.00
0.61
1.50
21.42
66,000
175
220
611
31.3
80
3.00
0.55
1.25
4.80
1.60
0.55
2.15
87,073
78,407
6,286
29,600
642,000
1/2 of 1
78.09
12.97
4.34
0.35
2.20
0.55
1.50
21.91
1. Evaporation, actual pounds per hour
2. Pressure in the drum, psig
3. Feedwater temperature °F
4. Temperature of flue gases leaving the
furnace °F
5. Temperature of flue gases leaving the
boiler °F
6. Excess air at boiler outlet %
(See note below)
7. Temperature of air at windbox °F
8. Dust collector draft loss, in H20
(See below)
9. Gas dust and damper draft loss (See
note below)
10. Furnace and boiler draft loss
11. Total draft required inches H20
12. Induced draft design static
13. Stoker and windbox air resistance
(See note below)
14. Duct and Damper air resistance
(No orifice in duct)
15. Total air resistance inches H20
16. Forced draft fan static design
17. Flue gas leaving the boiler, Ib/hr
18. Air required for combustion, Ib/hr
19. Fuel burned, Ib/hr
20. Furnace heat liberation Btu/cu ft
21. Heat release Btu/sq ft active grate
(at 129 sq ft grate)
22. Moisture in steam leaving the boiler%
23. Unit efficiency %
24. Losses:
A. Dry gas %
B. Moisture in fuel and hydrogen%
C. Moisture in air %
D. Unburned combustible% (See note
below*)
E. Radiation %
F. Unaccounted for %
G. Total %
NOTE: Items 8 and 9 are per Aerotec Industries, Inc. performance dated July 22, 1963.
Items 6, 13 and 24-D are per Hoffman Comb. Engineering Company performance dated
June 12, 1963.
*See Item 24-D above
The unburned combustible loss is as given by Hoffman Comb. Engr. Company based on
reinjection from boiler hoppers and decantation collector. Guaranteed overall boiler
and stoker efficiency is 75.0% at 60,000 pounds of steam per hour.
147
-------
APPENDIX C
PROCEDURE FOR ESTIMATING STACK VELOCITY
The emissions were monitored at the stack using isokinetic sampling
techniques. Achieving isokinetic sampling at MCI was greatly complicated by
the fact that the stack is sized for all three boilers firing simultaneously
at full load. Only one boiler is operated at a time at approximately one-
half load. As a result, the average stack gas velocity is too low to be read
on an S-type pitot tube. Because of the dirty nature of the stack gas, use
of a hot wire anemometer or similar device capable of reading the low flow is
not practical. As an alternative approach, the stack gas velocity was com-
puted on the basis of the boiler load, the experimentally determined boiler
efficiency, the fuel characteristics, and the Orsat analysis at the stack.
This procedure is an extension of standard boiler monitoring techniques. The
calculation used to compute stack gas velocity is as follows:
(1) The fraction of nitrogen in the flue gas is determined from the Orsat
analysis (which condenses all the water prior to analysis) by recognizing
that:
N2 = 100 - C02 - 02 - CO (C-l)
(2) The weight of nitrogen and carbon per mole of dry flue gas is then
computed by Equations (C-2) and (C-3).
lb/mole DFG = T^T x 28 (C-2)
N2
lbc/mole DFG = 2 x 12 (C-3)
(3) The nitrogen-to-carbon ratio in the flue gas is computed as the
ratio of the nitrogen and carbon levels in the dry flue gas.
N2 _ Equation (C-2) (C-4)
lb,. Equation (C-3)
L.
148
-------
(4) The nitrogen-to-carbon ratio in the flue gas is then multiplied by
the carbon-to-fuel ratio computed by Equation (C-5) and the air-to-
nitrogen ratio to determine the air-to-fuel ratio.
Ib mp np
TT^ = T~ (%C ) + TZ— (%CJ (C-5)
Ib,. , mp +np, c mp +np, d
fuel ^c d c d
where m is the volume fraction coal and n is the volume fraction
dRDF; %C and p are the as-received carbon content and the bulk density
of the respective fuels c(coal) and d(dRDF).
Ib .
air
-LU ., ..
fuel
Note: — is the constant 1.30
(5) The wet flue gas is then determined by recognizing that the total
amount of flue gas must be equal to the fraction of the input fuel
burned plus the pounds of air added to the fuel.
Ib of WFG _ / lbfuel _ lbash \ lbair (C-7)
Ib of fuel \ Ib. , Ib, -, J Ib. ,
fuel fuel/ fuel
(6) The wet flue gas is then converted into a volumetric flux by multiplying
by the standard flue gas density corrected to stack conditions. Note
that the actual density could be computed, but the error introduced by
assuming standard combustion products is of a lower order than the
sampling error of the composition of the fuel submitted to the labora-
tory for ultimate analyses.
cu ft of WFG @ T .„„ 460+T
s = WFG s ( _.
Ib of fuel 0.071 560 v '
Note: .071 is assumed to be the density of flue gas at 560°R,
and T is the stack temperature in °F.
149
-------
This result is then converted into an average gas velocity by
recognizing that the gas flux will be the cubic feet of gas per pound of
fuel burned multiplied by the rate of fuel consumption in the boiler
with this product then divided by the cross-sectional area of the stack.
These relationships [expressed in Equation (C9)] assume that the velocity
flux is to be computed over the same heat balance period as the boiler
efficiency test. The ratio of pounds of blend divided by run time can
be replaced by the amount of fuel needed to produce the steam generation
rate (as read off the strip chart recorder) once the boiler efficiency
is known for a given excess air level.
U = (cu ft of WFG/lb fuel) (Ib fuel/sec) = ft/sec (c_9)
f °2
where D is the stack diameter, and U is the gas velocity.
(7) The calculated gas velocity can be readily converted into a velocity
head (feet of air) by employing a rearrangement of Bernoulli's equation:
Ah = . , (C-10)
2g (Pm-p)
where p is the gas density under stack conditions (Ib/cu ft), p is the
density of the manometer fluid (Ib/cu ft) , and g is the gravitational
constant (ft/sec2).
Equation (C-10) is the velocity head as would be measured by a
standard pitot tube. As a result, this Ah needs to be multiplied by a
correction factor («0.91) which relates standard pitot tube results to
the measurements of an S-type pitot probe and by a constant to convert
from feet of air head to inches of water. This calculated apparent
stack velocity is then used to calculate an isokenetic flue gas
sampling rate.
150
-------
APPENDIX D
CASCADE IMPACTOR DATA
TABLE D-l. DECEMBER CASCADE IMPACTOR RAW DATA
DKCF.MBF.K
Total Klapsed Impactor Assumed Part Impactor Stack
Flow Time Flow Rate Density Temp. Temp.
(CF) (Mill) (CFM) (t;/cc)' ('F) ("F)
1:0-1 14.47 30 2.6 1.0
Date: 12/6/77
1:0-2 A 1.02 60 2.6 1,0
Date: 12/6/77
1:0-3 15.91 30 .57 1.0
Date: 12/10/77
1:1-1 34.36 30 .58 1.0
Date: 12/8/76
1:1-2 28.26 30 .59 1.0
Date: 12/13/76
1:1-3 30.77 30 .59 1.0
Date: 12/13/76
345 345 D,0
mg
CUM 7,
345 345 D.,o
mg
CUM X
370 370 D,0
mg
CUM ?„
390 390 D50
mg
CUM 7,
390 390 D,0
mg
CUM 7.
380 380 D. o
mg
CUM ',',
1
J6
0
0
16
0
0
34
11.9
14.44
31.0
0
0
31.0
7.0
4.14
31
0
0
2
7.
0
0
7.
4.
7.
17
7.
24.
16.
A.
3.
16.
A.
6.
16
12.
9.
7
7
2
18
9
03
0
6
33
0
7
91
6
87
3
2.8
9.6
12.53
2.8
9.3
23.08
5.7
8.1
33.86
5.7
40.0
32.32
5.7
31.-.
25.47
5.7
34.8
37.15
4
1.
7,
21,
1.
7,
36,
2
12.
48.
2,
30.
54.
2.
32,
44.
2.
22.
54.
.3
.1
.80
3
.9
.58
7
.2
,67
,7
.1
13
7
3
56
7
5
78
5
48.
84.
12.
58.
1.
2.
52.
1.
0.
54.
1.
15.
53.
1.
16.
67.
7
0
46
7
8
46
5
9
18
9
3
35
4
7
84
4
2
A8
0
84
-
8
72
8
62
22
70
40
77
3
70
6
.46
-
.3
.1
.31
.6
.5
.50
.7
.2
.43
.7
.6
.84
.7
.3
.06
7
0
8A.46
...
O.A
72.99
j,
11.9
76.94
.4
13.5
80.22
.A
1.7
78. 8A
.A
22/6
87.77
Filter
11.9
100
15.8
100
19.0
100
27.3
100
35.8
15.6
100
-------
TABLE D-2. JANUARY CASCADE IMPACTOR RAW DATA
JANUARY
1:U-1
Dace: 1/20/77
1:0-2
Date: 1/21/77
1:0-3
Date: 1/21/77
1:0-5
Date: 1/24/77
Total Elapsed Tmpactor
Flow Time I1' low Rate
(CF) (Min) (CFM)
19.92 30 .763
19.1 30 .feW
22.53 30 .54
'
13.95 30 .441
Assumed Part Impactor Stack
Density Temp. Temp.
(K/ec) no CF)
1.0 '390 390 D.,tJ
me
CUM 7,
l.Q 375 375 D.,,,
mg
CUM 7,
1.0 375 375 D™
mg
CUM 7
1.0 400 400 0,j0
mg
CUM '/,
1
30
2.0
2.54
31
0
0
33
0
0
37
0
0
2
13.
6.
10.
15
4.
12.
16
3.
6.
18
0
0
3
5
4
69
9
37
7
48
4.
28.
4h.
5.
11.
40,
5,
15,
32,
6,
6
16
9
3
69
6
1
,40
.9
.1
.92
.5
.1
.53
4
2
33
89
2
19
90
2
9
48
3
15
58
.25
.7
.57
.55
.9
.66
.7
.0
.69
.0
.6
.81
5
1.
3,
93.
1.
3,
98,
1
13,
71,
1
5
72
25
.4
89
45
,2
,74
.53
.0
.45
.7
.2
.90
6
0
93
0
98
4
79
2
79
.56
.89
.65
.74
.69
* 5
.33
.77
.4
.40
0
93.
0
98,
0
79,
0
79
35
89
.42
,74
.44
.33
.51
.40
Filter
4.
100
0,
100
11
100
7
100
,8
.5
.8
.6
-------
TABLE D-3. MARCH CASCADE IMPACTOR RAW DATA
MARCH
Total Elapsed Impactor Assumed Part Impactor Stack
Flow Time Flow Rate Density Temp. Temp.
(CF) (Mln) (CFM) (g/cc) (°F) (°F)
1:0-1 11.33 30 .61 1.0
Date: 3/19/77
1:0-2 7.16 20 .58 1.0
Date: 3/19/77
1:0-3 10.44 30 .58 1.0
Date: 3/21/77
1:0-4 8.13 20 .67 1.0
Date: 3/21/77
1:0-5 7.62 20 .62 1.0
Date: 3/22/77
1:0-6 10.29 30 .54 1.0
Date: 3/28/77
1:0-7 7.01 20 .54 1.0
Date: 3/31/77
1:0-8 5.54 15 .60 1.0
Date: 4/1/77
1:0-9 5.22 15 .55 1.0
Date: 4/1/77
390 390 D 50
mg
CUM %
390 390 D,o
rag
CUM %
415 415 D,o
mg
CUM 7.
415 415 D50
mg
CUM 7.
370 370 D,o
mg
CUM %
370 370 D5o
mg
CUM %
360 360 D5o
mg
CUM %
390 390 D,o
mg
CUM X
390 390 D,o
mg
CUM
1
31
0.9
1.43
32
13.1
25.10
32
0.9
1.45
29.5
1.0
1.94
31
1.6
4.18
33
3.6
6.01
33
2.3
5.18
31
1.2
3.00
33
0.8
2.33
2
15
3.9
7.61
15.5
3.7
32.18
15.5
1.8
4.36
14.5
1.3
4.46
15
2.9
11.75
16.5
2.9
10.85
16.5
4.3
14.86
15
2.5
9.25
16
1.6
6.98
3
5.5
20.7
40.41
5.7
11.9
54.98
5.7
19.5
35.86
5,3
17.3
37.98
5.4
12.7
44.91
5.9
18.4
41.57
5.9
5.4
27,03
5.6
12.6
40.75
5.8
9.8
35.47
2
15
64
2
9
73
2
15
61
2
9
57
2
9
68
2
11
60
2
6
41
2
8
61
2
7
57
4
.5
.4
.82
.7
.9
.95
.7
.7
.23
.5
.9
.17
.4
.2
.93
.8
.2
.27
.8
.5
.67
.55
.1
.00
.75
.5
.27
5
1.
9.
79.
1.
5.
84.
1.
8.
74.
1.
7.
70.
1.
4.
81.
1.
8.
75.
1.
14.
73.
1.
6.
76.
1.
6.
74.
4
5
87
45
4
29
45
4
80
4
1
93
3
7
20
55
9
13
55
3
87
40
3
75
50
0
71
6
.64
3.8
85.90
.67
2.6
89.27
.67
2.2
78.35
.61
3.1
76.94
.64
1.9
86.16
.70
4.2
82.14
.70
2.2
78.83
.66
2.1
82.00
.69
2.6
82.27
7
.41
1.5
88.27
.34
0.7
90.61
.43
1.6
80.94
.40
5.8
88.18
.42
0.8
88.25
.45
2.6
86.48
.45
1.9
83.11
.42
1.1
84.75
.44
0.6
84.01
Filter
7.4
100
4.9
100
11.8
100
6.1
100
4.5
100
8.1
100
7.5
100
6.1
100
5.5
100
(continued)
-------
TABLE D-3. (continued)
MARCH
Total Elapsed Impactor Assumed Part. Impactor Stack
Flow Time Flow Rate Density
(CF) (Min) (CFM) (g/cc)
1:1-1 11.18 30 .62 1.0
Date: 3/23/77
1:1-2 8.76 30 .45 1.0
Date: 3/23/77
1:1-3 12.43 30 .68 1.0
Date: 3/24/77
1:1-1 8.54 25 .53 1.0
Date: 3/29/77
1:1-2 6.77 20 .51 1.0
Date: 3/29/77
1:2-3 6.3 20 .48 1.0
BLANK 10.71 30 .60 1.0
Date: 3/16/77
Temp . Temp
(°F) <°F)
385 385 D,0
mg
CUM %
400 400 D3o
mg
CUM %
390 390 D50
mg
CUM %
380 380 D50
mg
CUM %
390 390 D,o
mg
CUM Z
380 380 D,o
mg
CUM X
400 400 D50
mg
CUM X
1
31
0.
0.
37
0.
0.
31
4.
6.
33
0.
1.
34
0.
0.
35
3.
0.
3
68
2
44
7
89
9
87
2
48
39
6
2
15
1.5
4.09
18
1.1
2.86
16
2.2
10.12
16.5
0.9
3.74
17
0.7
2.18
17.5
7.5
0.1
3
5.4
15.4
39.09
6.7
8.9
22.47
5.6
13.9
30.50
5.9
8.9
22.25
6.0
10.1
26.48
6.4
28.72
0.5
4
2.
10.
62.
3.
9.
42.
2.
9.
44.
2.
7.
38.
2.
7.
44.
2.
48.
0.
5
4
73
0
1
51
5
4
28
75
7
25
75
5
58
7
30
3
5
1.30
4.4
72.73
1.70
4.6
52.64
1.30
3.5
49.41
1.55
2.6
43.66
1.60
3.6
53.28
1.60
62.92
0.7
6
.65
2.8
79.09
.77
2.4
57.93
.65
1.1
51.03
.70
3.9
51.77
.70
2.9
60.28
.74
66.32
0.7
7
.43
4.6
89.55
.50
4.4
67.62
.42
11.3
67.60
.45
7.2
66.74
.47
3.8
69.48
.49
70.23
0.3
Filter
4.6
100
14.7
100
22.1
100
16.0
100
12.7
100
11.4
100
3.1
-------
TABLE D-4. MAY CASCADE IMPACTOR RAW DATA
Ul
MAY
1:0-1
Date: 5/3/77
1:0-2
Date: 5/4/77
1:0-3
Date: 5/5/77
1:0-4
Date: 5/5/77
1:0-5
Date: 5/16/77
1:0-6
Date: 5/16/77
BLANK 1-0
Date: 5/4/77
1:1-1
Date: 5/12/77
1:1-2
Date: 5/12/77
1:1-3
Date: 5/13/77
Total
Flow
(CF)
10.84
7.18
5.3
6.84
3.6
7.03
4.46
8.96
7.95
6.36
Elapsed
Time
(Min)
30
20
15
20
10
15
15
20
20
15
Impact or
Flow Rate
(CFM)
.56
.57
.62
.56
.60
.80
.45
.77
.66
.70
Assumed Part. Impact or Stack
Density Temp. Temp.
(g/cc) (°F) (°F)
1.0 360 360
1.0 360 360
1.0 420 420
1.0 400 400
1.0 380 380
1.0 435 435
1.0 360 360
1.0 415 415
1.0 400 400
1.0 400 400
DBO
mg
CUM X
Dso
mg
CUM X
D,0
mg
CUM X
D,o
mg
CUM X
D,o
mg
CUM X
D50
mg
CUM X
D,o
mg
CUM X
D,0
mg
CUM X
D,o
mg
CUM X
D,o
mg
CUM X
1
32.0
0.7
1.17
32.0
0.7
1.78
31
0.8
1.47
32
0.9
2.07
31
0.2
1.27
27
0.0
0
0.3
28
0.6
1.19
30
0.2
0.47
29
0.0
0
2
15.5
1.8
1.17
15.5
1.1
4.58
15
1.2
3.68
15.5
0.8
3.91
15.5
0.6
5.10
13
1.7
2.94
0.4
14
1.8
4.77
10.4
1.3
3.55
14
2.0
6.13
3
5. BO
17.4
33.22
5.80
9.6
29.01
5.5
14.3
29.96
5.8
11.9
31.26
5.7
1.4
14.01
4.8
25.7
47.40
1.0
4.9
16.8
38.17
5.4
12.3
32.70
5.2
9.9
36.50
4
2.60
14.7
57.76
2.60
7.5
48.09
2.5
14.0
55.70
2.6
12.9
60.92
2.6
4.4
42.04
2.2
12.9
69.72
0.5
2.2
10.5
59.05
2.4
9.9
56.16
2.3
7.4
59.20
5
1.50
9.6
73.79
1.50
4.1
58.53
1.4
10.0
74.08
1.5
7.4
77.93
1.45
3.2
62.42
1.2
4.1
76.82
0.5
1.2
4.1
67.20
1.35
4.7
67.30
1.3
1.8
64.72
6
0.69
3.9
80.30
0.69
3.5
67.43
.62
6.4
85.84
.69
1.9
82.30
.65
3.5
84.71
.54
3.0
82.01
0
.55
2.7
72.56
.60
2.7
73.70
.50
1.3
68.71
7 Filter
0.47 0.47
2.9 8.9
85.14 100
0.47
3.1 9.7
75.32 100
.40
1.7 6.0
88.97 100
.47
1.5 6.2
85.75 100
.42
1.6 0.8
94.90 100
.35
3.3 7.1
87.72 100
0.2 0.3
.35
3.6 10.2
79.72 100
.39
2.0 9.1
78.44 100
.37
2.5 7.7
76.38 100
(continued)
-------
TABLE D-4. (continued)
Wl
MAY
BLANK
Date:
1:2-1
Date:
1:2-2
Total Elapsed Impactor Assumed Part. Impactor Stack
Flow Time Flow Rate Density
(CF) (Min) (CFM) (g/cc)
1-1 6.29 15 .69 1.0
5/13/77
7.33 20 .63 1.0
5/10/77
7.34 30 .63 1.0
Date 5/10/77
1:2-3
Date:
BLANK
Date:
0:1-1
Date:
0:1-2
Date:
8.68 20 .73 1.0
5/11/77
1-2 8.53 20 .72 1.0
5/11/77
8.0 20 .66 1.0
5/14/77
11.9 30 .65 1.0
5/14/77
Terap . Temp .
(°F) (°F>
400 400 D
mg
CUM %
420 420 D
mg
CUM %
425 425 D
mg
CUM %
410 410 D
mg
CUM %
420 420 D
mg
CUM %
380 380 D
mg
CUM %
390 390 D
mg
CUM %
1
0.4
30
0.8
1.63
30
1.2
2.51
28
1.2
2.01
0.5
29
0.0
0
29
0.4
0.45
2
0.1
15
1.6
4.89
15
1.2
5.02
14
1.7
4.87
0.3
14.5
0.5
1.17
14.5
1.8
2.49
3
0.6
5.4
11.6
28.51
5.4
2.4
10.04
5
20.1
38.59
1.1
5.2
0.0
1.17
5.2
18.9
23.87
4
0.
2.
8.
46.
2.
12.
35.
2.
13.
60.
1.
2.
7.
18.
2.
13.
39.
6
5
9
64
5
2
56
3
3
91
1
4
5
69
4
8
48
5
1.3
1.4
3.1
52.95
1.4
16.5
70.08
1.3
5.0
69.30
0.8
1.3
2.4
24.30
1.3
3,9
43.89
6
0.7
.62
2.8
58.66
.62
1.2
72.59
.57
2.4
73.32
0.7
.62
1.7
28.27
.62
4.2
48.64
7
0.6
.4
6.1
71.08
.4
2.1
76.99
.37
3.2
78.69
0.5
.39
7.1
44.86
.39
13.0
63.35
Filter
0.7
14.2
100
11.0
100
12.7
100
0
23.6
100
32.4
100
-------
APPENDIX E
DISCUSSION OF MONSANTO'S ESP TEST DATA
SYSTECH analyzed Monsanto's ESP test data to determine how the fly ash
from coal-only and 1:1 and 1:2 blend tests affected the precipitator
performance. Figure E-l plots the results of this analysis with inverse
penetration as a function of the specific collector area, the applied voltage,
and the square root of the current. These parameters were chosen as axes
because the Deutch Equation (E-l) indicates that these axes should yield a
straight line on semilog paper as demonstrated in the following.
- e - (E-l)
where ri = probability of particle capture
P = penetration
A = collector electrode area ft2
Q = gas glow CFM
W = migration velocity
Further reduction of Equation (E-l) yields the following:
In P = - ^ (E-2)
and
In ^ = ^ (E-3)
The migration velocity is a function of electric field strength (E ) and
particle charge (q).
W a q E (E-4)
For a uniform field, the field strength (E ) is the applied voltage (V)
to the collector electrodes divided by the gap between the electrodes. Also,
the charge per particle (q) is a function of the electron cloud density which
is proportional to the square root of the current flow (i) for collision
157
-------
100
90
80
70
60
50
40
30
20
10
9
8
7
6
5
BLEND SYMBOL
1:0 •
1:1 A
1:2 •
I
I
I
10,000 20,000 30,000 40,000
(SCA) kV VT
50,000
Figure E-l.
Precipitability of blend fly ash component to
coal-only conditions.
158
-------
charging. Therefore, substituting these relationships into Equation (E-4)
yields the following equation:
W a V y/i
(E-5)
Since A/Q is commonly referred to as SCA, Equation (E-3) can now be
expressed as:
1 / /
In ^ a SCA V,/i (E-6)
Table E-l summarizes all of the Monsanto data, and Figure E-l presents
a logarithmic plot of 1/P versus SCA for the various blends. If the precipi-
tator performs normally, regardless of the variation in these parameters, the
results should plot as a straight line.
However, the coal-only ash had resistivities too low for proper
precipitation. The 1:1 and the 1:2 ash were slightly less and slightly more
precipitable, respectively, than the coal-only ash.
The difference between the resistivity for the first 1:2 test data and
the resistivity for the rest of the data is probably real. After the
coal-only runs were completed, the precipitator was thoroughly cleaned.
Consequently, most of the power applied to the cells for this first 1:2 test
probably passed through the air gap. During the 1:1 test, cell after cell
was taken off line because of the short-circuiting over the dielectric
bridge. Consequently, while the applied voltage or field was correct, a
significant amount of current probably leaked through the insulating hangers
to the grounded electrodes. If it is assumed that half of the current
bypassed the air gap, then the aerosol data for the 1:1 test aligns with the
aerosol data for the 1:2 test. The collection data for the 1:1 test, there-
fore, has questionable validity.
Because the first 1:2 run had more fly ash precipitation than the coal-
only run, it is likely that substituting dRDF for coal will not seriously
degrade the precipitator performance. This conclusion, however, must be
verified by further testing.
159
-------
TABLE E-l. FIELD PORTABLE ELECTROSTATIC PRECIPITATOR DATA
DATE
5/5/77
1:0
coal
only
5/10/77
1:2
Blend
FLOW
(ACFM)
1500
1500
SCA
(Ft.2/KCFM)
320
320
Temp.
(°F)
480
470
AVERAGE
ELEC. COND.
CELL VOLT C.D.
(KV) (UA/Ft?)
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
27
34
32
32
32
32
36
34
32
36
15
30
22
16
30
4
12
15
10
42
INLET COND.
(gr/SCF) MEAS.
Avg.
Avg.
0.16
0.19
0.175
0.251
0.
0.201
0.226
OUTLET COND,
MEASURED
(gr/SCF)
0.0012
0.0082
0.0047
0.0040
0.00047
0.00226
EFFICIENCY
(%)
99.25
95.7
97.3
98.4
99.8
99.0
(continued)
-------
TABLE E-l. (continued)
DATE
5/11
1« o
• f.
Blend
5/12
1:1
Blend
FLOW
(ACFM)
3000
1500
SCA
(Ft.2/KCFM)
160
320
Temp.
(°F)
530
470
AVERAGE
ELEC. COND.
CELL VOLT C.D.
(KV) (UA/Ft.2)
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
32
37
30
38
37
23
25
24
30
32
5
7
0.8
12
17
10
15
'9
8
15
INLET COND.
(gr/SCF) MEAS.
AVG.
AVG.
0.15
0.098
0.124
0.20
0.18
0.19
OUTLET COND.
MEASURED
(gr/SCF)
.0178
.0114
0.0146
0.059
0.030
0.0445
EFFICIENCY
(%)
88.1
88.3
88.2
70.6
83.3
76.6
(continued)
-------
TABLE E-l. (continued)
DATE
5/13
1:1
Blend
5/16
(Note:
1:0 coal
only
5/16
(Note:
FLOW
(ACFM)
3000
1200
Operating
1200 ^
Operating
SCA
(Ft.2/KCFM)
160
320
4 cells)
240
3 cells)
Temp.
(°F)
500
450
450
AVERAGE
ELEC. COND.
CELL VOLT C.D.
(KV) (UA/Ft.2)
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
24
25
22
23
42
27
33
33
28
.»
26
32
32
—
12
10
8
8
20
10
20
28
18
— —
11
14
14
__
INLET COND.
(gr/SCF) MEAS.
AVG.
0.12
0.23
0.175
0.164
0.127
OUTLET COND.
MEASURED
(gr/SCF)
0.0425
0.0331
0.0378
0.00605
0.0103
EFFICIENCY
(%)
64.6
85.6
78.4
96.3
91.9
-------
APPENDIX F
HEAVY METALS EMISSIONS DATA SUMMARIES
TABLE F-l. AA ANALYSIS OF AEROSOL CAPTURE FOR COAL-ONLY (1:0) FIRING
OJ
Part
4-34
4-35
4-37
4-3H
4-39
4-40
4-41
4-42
3-42
3-43
3-44
3-45
3-46
3-47
3-48
Hg/m3
Filter
3/19
3/19
3/21
3/22
3/28
3/31
3/31
4/1
5/3
5/4
5/5
5/5
5/16
5/16
5/17
Probe Wash
4-51 3/19
4-53
4-54
4-55
4-56
4-57
4-58
3-38
3/22
3/22
3/28
3/31
3/31
4/1
Composite
May
Pb
156
254
147
758
169
190
190
200.
184
165
222
369
144
517
116
28.1
13.8
18.7
32.3
27.4
17.8
16.1
Cd
2.60
4.54
2.62
2.43
<3.38
7.15
2.38
3.73
2.30
2.35
3.13
5.76
2.41
8.01
2.00
4.02
1.32
2.07
1.40
1.59
.697
1.02
2.15
As
168
439
237.
115.
143
65.6
87.3
187.
196.
162
195.
199
76.8
151
71.2
184.5
44.2
45.3
38.5
23.7
13.1
11.2
16.1
Hg Cr
<0.69 29.8
27.3
36.7
36.4
<0.56 22.5
<1.03 7.95
<0.99 11.9
58.6
69.0
20.6
287.
60.2
12.0
32.8
46.0
44.7
25.7
82.5
409.
74.7
38.3
44.1
134.2
Ni
24.2
18.4
34.0
30.3
28.2
15.9
23.8
40.0
65.2
35.3
36.5
60.2
11.4
34.3
35.6
5.43
31.2
88.9
5253.
67.9
44.8
37.1
183.
Mn
26.0
46.3
59.7
41.9
37.2
21.5
22.2
59.2
29.9
24.7
20.4
31.9
27.5
28.8
10.4
17.5
9.99
62.1
13.6
8.07
10.2
—
Zn
474.
613.
445.
531
507
397
337
546
115
<58.9
679.
1047.
337
713 .
286.
101.
53.8
346
105
70.8
47.3
42.4
56.4
Cu
33.5
45.4
41.9
42.5
<56.3
<39.7
<39.7
<37.3
38.3
35.3
52.3
68.0
24.1
51.2
24.9
12.1
7.05
7.90
21.0
5.66
<4.98
3,96
6.4
Sn
1.86
1.82
.524
1.21
1.13
<.79
<.79
1.33
.767
.883
1.57
3.14
.722
1.19
<.498
.277
.296
.467
.509
.300
.107
Sb
<37.2
<90.9
<32.4
<60.7
<113.
< 79.5
< 79.3
< 53.3
< 76.7
< 58.9
< 52.2
< 52.3
< 48.2
< 32.0
< 49.5
< 10.1
< 6.94
< 9.86
< 23.3
< 11.3
< 9.96
< 5.66
< 1.08
Ag
<3.72
<9.09
<5.24
<6.07
<11.3
<7.9
<7.93
<5.33
<7.67
<5.89
<5.22
<5.23
<4.82
<3.20
<5.0
<13.7
<9.71
< .988
<2.33
<1.13
< .996
<1.13
< .537
Vn
37.2
90.9
<52.4
<60.6
<113
<79.5
<79.3
<53.3
<76.7
<58.9
52.3
52.3
<48.2
37.0
<50.0
10.1
<7.05
<9.9
<23.3
<11.3
<9.96
<5.66
<.5.37
-------
TABLE F-2. AA ANALYSIS OF AEROSOL CAPTURE OF 1:1 BLEND FIRING
ug/m
1'b
Cd
As
Part.
4-43
4-44
4-45
3-49
3-50
3-51
Probe
4-59
4-60
4-61
j-39
Filter
3/23
3/23
3/24
5/12
5/12
5/13
Wash
3/23
3/23
3/24
Composite
May
2705.
4638
3492
5609
2847
2438
48.1
112.
1Z2.
108
69.
84.
62.
74.
64.
47.
2.
1.
2
2
1
3
8
5
0
6
,09
.68
.34
.16
55.1
46.7
23.4
159.
165.
64.5
Hg
<0.65
<0.54
<0.68
<0.50
Cr
25.0
37.1
29.1
33.5
28.9
29.5
34.5
11.2
13.2
13.5
Ni
27.6
34.5
25.2
37.7
31.1
23.6
35.6
11.2
22.0
116.
Mi
43
78
54
62
64
33
18
6
11
10
i
.6
.0
.6
.8
.0
.0
.4
.14
.0
.2
Zn
6011
5994
4341
5441
4892
4325
207
196
173
132.
Cu
75.1
101.
85.3
83.7
66.7
62.9
5.23
<27.9
<14.6
7.73
Sa Sb
2.50 <30.1
4.77 <53.0
1.94 <38.8
3.35 <41.9
2.67 <44.5
.786 <39.3
.418 <10.5
<.559 <55.9
<29.3
.232 <7.73
Ag
10.
12.
9.
8.
<4.
<3.
2.
<5.
<2.
<.
5
7
30
37
45
93
09
59
93
773
<50
<53
<38
<41
<44
<39
<10
<55
<29
<7
Vn
.1
.0
.8
.8
.5
.3
.5
.9
.3
.73
-------
TABLE F-3. AA ANALYSIS OF AEROSOL CAPTURE FOR 1:2 BLEND FIRING
(J\
>J
Part.
4-46
4-47
4-48
3-52
3-53
3.54
Probe
4-62
4-63
4-64
3-40
.«/.'
Filter
3/29
3/29
3/30
5/10
5/10
5/11
Wash
3/29
3/29.
3/30
Comp
May
Pb
10,140.
5,151.
3,621.
6,345.
5,595.
7,893.
337
218
249
257
Cd
269.
216.
94.0
162
275
90.5
6.59
5.41
4.04
5.82
As
60.6
23.3
36.5
103
105
95.5
12.1
12.2
4.7
20.4
Hg Cr
38.6
<0.65 43.7
<0.90 34.8
61.3
<0.80 39.6
35.2
161
79.6
49.4
68.0
Ni
35.8
14.6
48.7
78.3
29.7
20.1
130.
78.1
67.4
113.
Mn
72.7
76.3
98.2
96.4
107.
75.4
31.2
60.4
29.7
20.6
Zn
11020
3686
6264
7191
6435
6539
295.
263.
371.
318.
Cu
110
36.9
55.7
140.
94.1
95.5
<17.4
<15.0
<22.5
12.1
Sn
2.2
2.13
7.66
2.33
4.46
1.51
.347
2.11
1.35
.605
Sb
<55.1
<89.9
<69.6
<42.3
<49.5
<50.3
<34.7
<30.1
<44.9
<7.56
Ag
23.7
11.7
6.96
19.5
13.9
14.1
< 3.47
<3.01
<4.49
<.756
Vn
<55.1
<19.4
<69.6
<42.3
<49.5
<50.3
<34.7
<30.1
<44.9
<7.56
-------
TABLE F-A. AA ANALYSIS OF AEROSOL CAPTURE FOR 0:1 BLEND FIRING
Pb Cd As Hg Cr Ni Mn Zn Cu Sn Sb Ag Vn
Part. Filter
3-55 5/14
3-56 5/14
Probe Wash
Composite
May
9557 240 39.5 <1.06 60.8 26.1 200. >821 165. 3.48 <86.9 24.3 =86.9
6351 187 39.6 66.9 20.9 241 5016 159. 6.27 <83.6 23.4 <83.6
326 15.8 11.8 34.0 31.3 — 605 21.8 .816 <27.2 <2.7 ( <27.2
-------
TABLE F-5. AA ANALYSIS OF ASH SAMPLES FROM 1:0 FIRING
B.A.
4-66
3-111
3-105
F.A.
3-107
3-84
3-99
F.A.
4-70
4-68
3-83
3/28
5/3
5/17
Coll #1
5/4
5/5
5/17
Relnj «
IfD 3/28
3/31
5/3
Pb
mg/kg
<12.5
<12.5
<12.5
30.0
17.5
15.0
134.2 -
16.3
20.0
Cd
mg/kg
<.75
<.75
<.75
<.75
<.75
<.75
2.87
<.75
<.75
As
mg/kg
11.0
17.9
14.3
102.4
77.0
49.5
- 625.
34.4
48.1
Hg
mg/kg
<0.4
<0.4
<0.4
<0.4
<0.4
<0.4
'0 '<
<0.4
<0.4
Cr
mg/kg
22.5
10.0
7.5
20.0
15.0
18.1
25 0
8.75
21.3
Hi
mg/kg
20.0
23.7
16.2
37.5
25.0
25.0
25 6
15.0
26.2
Mn
mg/kg
51.0
40.5
40.5
69.7
42.0
37.5
300
105
45.0
Zn
mg/kg
31.2
37.5
100.
50.0
81.2
50.0
inc.
31.2
50.0
Cu Sn
mg/kg mg/kg
15.0 .50
25.0
<12.5 .50
20.0 2.38
12.5 2.13
12.5 1.75
27 5 2 37
<12.5 1.88
15.0 1.0
Sb
mg/kg
<25
<25
<25
<25
<25
<25
^25
<25
<25
Ag
mg/kg
<2.5
5.75
120.
<2.5
<2.5
<2.5
<2.5
<2.5
Vn
mg/kg
<25
<25
<25
25.0
25.0
25.0
<25
25.0
Notes: B.A. denotes bottom ash and F.A. denotes fly ash.
-------
TABLE F-6. AA ANALYSIS OF ASH SAMPLES FROM 1:1 FIRING
oo
B.A.
3-102
3-103
4-67
F.A.
3-85
3-91
3-92
3-93
F.A.
4-72
4-71
3-94
5/12
5/13
3/23
Coll #1
5/9
5/12
5/12
5/13
Reinj. #2
3/23
3/24
5/13
Pb
mg/kg
36.3
56.3
26.3
148.
250.
300.
168.
97.5
97.5
92.5
Cd
rag /kg
.75
<.75
<.80
1.75
4.7
5.37
2.5
1.25
1.0
1.0
As Hg
mg/kg mg/kg
40.0 <0.4
30.0 <0.4
39.3 :0.4
50.0 <0.4
100. <0.75
124. <0.5
47.5 0.53
38.5 0.68
39.0 *r6*
49.5 0.58
Cr
mg/kg
25.0
22.5
25.6
20.6
22.5
23.1
12.5
13.8
11.3
21.3
Ni
mg/kg
40.0
26.2
27.4
25.0
35.0
35.0
21.2
17.5
16.2
23.7
Mn
mg/kg
185.
155.
138.
101
197.
196.
86.3
161
225
120
Zn
mg/kg
150.
75.0
73.3
197
431
512
231
125.
144.
156.
Cu
mg/kg
218.
225.
152.
-15.0
40.0
36.3
20.0
15.0
15.0
22.5
Sn
mg/kg
2.63
4.0
3.58
2.13
4.38
5.3
2.5
2.0
3.0
3.0
Sb
mg/kg
<25
<25
<25
<25
<25
<25
<25
<25
<25
<25
Ag
mg/kg
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
Vn
mg/kg
<25
<25
<25.5
-25
35
30
<25
25.0
<25
25.0
Notes: B.A. denotes bottom ash, F.A. denotes fly ash, and £ represents that the number reported is an
average of two in which one was below the detection limit and the other was just above.
-------
TABLE F-7. AA ANALYSIS OF ASH SAMPLES FROM 1:2 FIRING
BA
4-65
3-100
3-101
F.A.
3-86
3-89
3-90
F.A.
4-69
3-87
3-88
3/30
5/10
5/11
Coll #1
5/10
5/11
5/11
Relnj «2
3/29
5/10
5/11
Pb
mg/kg
128
47.5
82.5
275
246
300
109
132
198
Cd
mg/kg
.75
.75
.75
5.75
6.25
6.0
1.5
1.3
2.5
As
rag /kg
36.7
30.2
27.5
49.5
37.5
66
32.4
37.5
Hg
mg/kg
<0.4
<0.4
<0.4
<0.4
<0.4
2.3
<0.4
<_0.43
<0.4
Cr
mg/kg
42.5
18.8
22.5
26.1
38.7
40.
15.0
28.4
18.7
Ni
mg/kg
139
16.3
27.8
22.0
37.5
36.2
18.7
20.2
26.2
Mn
mg/kg
250
135
135
317.
379.
288.
300
146.
250
Zn
mg/kg
188.
81.2
65.6
598
619.
606
194
203
381.
Cu
mg/kg
200
186
29.7
50.0
40.
17.5
24.3
60
Sn
mg/kg
5.0
3.0
4.25
1.08
2.69
1.13
2.25
3.0
Sb
mg/kg
<25
<25
<25
<27.5
<25
<25
<25
<27
<25
Ag
mg/kg
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
<2.5
<2.7
<2.5
Vn
mg/kg
<25
<25
<25
27
30
25
25.
<27
25.
Noces: B.A. denotes bottom ash, F.A. denotes fly ash, and £ represents that the number reported is an
average of two in which one was below the detection limit and the other was just above.
-------
TABLE F-8. AA ANALYSIS OF ASH SAMPLES FROM 0:1 FIRING
B.A.
3-104 5/14
Pb
mg/kg
169.
Cd
mg/kg
2.125
As
rag /kg
56.3
Hg
mg/kg
£0.4
Cr
mg/kg
50.0
Ni
mg/kg
35.0
Mn
mg/kg
43.5
Zn
mg/kg
539.
Cu
mg/kg
205
Sn
mg/kg
3.6
Sb
rag /kg
<25
Ag
mg/kg
88.8
Vn
mg/kg
<25
F.A. Coll #1
3-96 5/14
3-97 5/14
925.
1100.
26.0
24.8
122.
85.3
£0.5
0.61
176.
193.
81.9
68.8
130.
117.
176.
127.
11.3
4.86
<25.
<27.5
12.7
14.8
40.0
43.2
F.A. Reinj #2
3-95 5/14
3-98 5/14
350 8,50
375 7.25
52.5
40.0
£0.48
<0.4
65.0
73.7
45.0
36.9
82.3
80.3
1138.
1097.
95.0
81.7
4.38
5.3
<25
<25
6.5
5.5
25.0
<25.
Notes: B.A. denotes bottom ash, F.A. denotes fly ash, and < represents that the number reported is an
average of two In which one was below the detection limit and the other was just above.
-------
TABLE F-9. TOTAL METAL EMISSION RATE IN THE FLUE GAS
March
4-34
4-35
4-37
4-38
4-39
4-40
4-41
4-42
May I
3-42
3-43
3-44
3-45
3-46
3-47
3-48
1:0
3/19
3/19
3/21
3/22
3/28
3/31
3/31
4/1
X
:0
5/3
5/4
5/5
5/5
5/10
5/16
5/17
X
Pb
202
327
171
•96i
220
233
218
222
228
295
210
242.
333
171
-5W
129
230
Cd
3.37
5.85
3.05
3.08
<4.40
8.76
2.73
4.14
<4.43
3.69
2.99
3.42
6.21
2.86
8.98
2.23
4.33
As
218
***,
276.
146.
186.
80.4
100.
208.
173
315
206.
213.
215.
91.0
169.
79.4
Hg
11.1
9.4
10.4
8.0
<5.3
3.3
<8.1
7.2
<7.85
<8.1
6.0
2.8
4.8
<5.1
7.3
<4.9
£5.57
Cr
38.7
35.2
42.7
46.2
29.3
9.74
13.6
65.1
35.1
111.
26.2
•3*3-.
64.9
14.2
36.8
51.3
50.7
TOTAL PARTICULATE
/ig/m3
Nl Mn Zn
31.4
23.7
39.6
38.4
36.7
19.5
27.3
44.4
32.6
105.
45.0
39.9
64.9
13.5
38.4
39.7
49.5
33.7
59.6
69.5
53.1
48.4
26.3
25.4
65.7
47.7
48.0
31.5
22.3
34.4
32.6
32.3
11.6
30.4
615
790.
518
673
660
486
386
606
592
185.
*V-r
742
1129
399.
799.
319.
596
P.P.
Cu
43.5
58.5
48.7
53.9
<73.3
<48.6
<45.5
41.4
-51.7*
61.5
45.0
57.1
73.3
28.6
57.4
27.8
50.1
/ig/m3
Sn
2.41
2.34
0.610
1.53
1.47
<.968
<.905
1.48
<1.46
1.23
1.10
1.71
3.39
.855
1.33
<.555
£1.45
Tot Part g
P.F.g
Sb
<48.3
<117.
<61.0
<77.0
<147
<97.4
<90.9
<59.2
<87.2
123.
<75.
<57.0
<56.3
<57.1
<35.9
<55.2
<65.6
Ag
<4.83
<11. 7
<6.10
<7.70
<14.7
<9.74
<9.09
<5.92
<8.72
<12.3
<7.5
<5.70
<5.63
<5.71
<3.59
<5,57
<6.56
Vn
48.
117
<61
<77.
<147
<97.
<90.
<59.
<87.
<12.
< 7.
<57.
<56.
<57.
<35.
<55.
<65.
3
0
4
9
i
2
3
5
0
3
1
9
7
6
Total
Part.
.8215
.5618
.6814
.4946
.2203
.3747
.3851
.5850
.4090
.5716
.6955
.6578
.9075
1.0586
.6837
Part.
Ftlter
.6329
.4361
.5857
.3901
.1692
.3059
.3361
.5268
.2548
.4488
.6368
.6101
.7660
.9447
.6131
(continued)
-------
TABLE F-9. (continued)
TOTAL PARTICULATE
Ug/m3
March
4-43
4-44
4-45
May
3-49
3-50
3-51
I-1
— J
NJ
March
4-46
4-47
4-48
May
3-52
3-53
3-54
May
3-55
3-56
1:1
3/23
3/23
3/24
X
1:1
5/12
5/12
5/13
X
1:2
3/29
3/29
3/30
X
1:2
5/10
5/10
5/11
X
0:1
5/14
5/14
X
Pb
3035
4917
3973
3975
6656
3512
2543
4237
12289
6054
4636
7660.
7116.
7338
10196
8217
12046
7859
9953.
Cd
77.5
89.4
71.4
79.4
88.4
79.0
49.7
72.4
326.
254
120
233
182
361
117
220
303
231
267
As
61.8
49.5
26.6
45.9
189.
204
673
153
73.4
27.4
33.9
44.9
116.
138
123
126
49.8
49.0
49.4
Hg
21.3
11.1
26.5
19.6
20.0
15.0
12.2
15.7
15.4
14.7
6.9
12.3
13.7
9.1
11.5
11.4
90.8
98.6
94.7
Cr
28.0
39.3
33.1
33.5
39.8
35.7
30.8
35.4
46.8
51.4
44.6
47.6
68.8
51.9
45.5
55.4
76.6
82.8
79.7
Ni
31.0
36.6
28.7
32.1
44.7
38.4
24.6
35.9
43.4
17.2
62.3
41.0
87.8
38.9
26.0
50.9
32.9
25.9
29.4
Mn
48.9
82.7
62.1
64.6
74.5
79.0
34.4
62.6
88.1
89.7
126.
101
108.
140
97.4
115.
252.
298
275
Zn
6744
6354
4939
6012
6457
6035
4512
5668
13355
4332
8020
8569
8065
8440
8447
8317
9858
6207
8033
Cu
84.3
107
97
96.1
99.3
82.3
65.6
82.4
133
43.4
71.3
82.5
157.
123
123
134
208
197
203
PF ug/tn1
Sn
2.80
5.06
2.21
3.36
3.98
3.29
0.82
2.70
2.67
2.50
9.81
4.99
2.61
5.85
1.95
3.47
4.38
7.76
6.07
3 Tot
Sb
<56.2
<56.2
<44.1
<52.2
<49.7
<54.9
<47.0
<48.5
<66.8
106.
<89.1
<87.3
<47.4
<64.9
<65.0
<59.1
<109.
<103
Part &
PFg
Ag
11.8
13.5
10.6
12.0
9.93
<5.49
<4.10
£6.51
28.7
13.8
8.91
17.1
21.9
18.2
18.2
30.6
28.9
29.7
Vn
<56.2
<56.2
<44.1
<52.2
<49.7
<54.9
<41.0
<48.5
<66.8
<22.8
<89.1
<59.6
<47.4
<64.9
<65.0
<59.1
<109
<103
Total
Part.
.4104
.4923
.5222
.7092
.7000
.4035
.4518
.4694
.2845
.8542
.7227
.5889
.3518
.3246
Part.
Filter
.3658
.4644
.4590
.5976
.5674
.3868
.3728
.3994
.2222
.7616
.5510
.4559
.2791
.2623
Notes: B.A. denotes bottom ash, F.A. denotes fly ash, and < represents that the number reported is an
average of two in which one was below the detection limit and the other was just above.
-------
APPENDIX G
PHYSICAL AND CHEMICAL CHARACTERIZATION OF RDF/COAL FLY ASH
INTRODUCTION
As part of the effort to identify the potential for environmental
impact by fly ash generated by combustion of a coalrdRDF blend, fly ash
samples were sent to Colorado State University for analysis. The aims of
this testing were to identify any differences in the composition of fly ash
generated by different coal:dRDF fuels, to ascertain whether concentrations
of certain materials existed at the surface of the particles, and to deter-
mine if fly ash composition was a function of particle size. D. F. S. Natusch
of Colorado State University was selected to perform the analysis because of
his extensive background in this area. The tentative conclusions put forth
in this section are the result of Dr. Natusch's analysis.
Specifically, the investigation was designed to answer the following
questions:
1. What are the morphological and compositional characteristics of fly
ash generated by burning dRDFrcoal fuel blends?
2. What is the elemental composition of fly ash generated by burning
dRDFrcoal fuel blends with respect to (a) the size of the particles
and (b) the ratio of dRDF to coal?
3. What factors appear to be responsible for the partitioning of
elements present in the fly ash as a function of size and RDF
content?
4. To what extent can individual elements present in the fly ash be
mobilized into solution as a result of an aqueous leaching process?
In the following sections we present a brief description of the analyti-
cal methodology employed, a list of the results obtained, and a short dis-
cussion of the meaning of these results.
PHYSICAL CHARACTERISTICS OF RDF/COAL FLY ASH
The samples of dRDFrcoal fly ash received consisted of four sets of
seven samples. Each sample set was obtained from burning a different ratio
of dRDF to coal ranging from pure coal to pure dRDF, and the seven samples
corresponding to each set were obtained from the seven impaction stages of an
173
-------
MRI stack sampler. Each of the seven impactor stages for each blend ratio
consisted of composited scrapings from two to six stack samples. The dis-
tribution of particle mass within this 28-sample set is presented elsewhere
in this report.
Each of the 28 subsamples was investigated using a scanning electron
microscope (Hitachi, Model HHS-2R). Each sample was mounted on double-sided
Scotch tape and coated with carbon and gold for the purpose of observation.
The results showed the fly ash to consist of four different structural
types. The first type consisted of a "spongy-looking" material that may have
been formed as a result of condensation or agglomeration of extremely small
particles onto the surface of large particles (Figure G-l). The second type
was in the form of extremely thin sheets, some of which were found to roll or
unroll under the influence of the SEM electron beam (Figure G-2). The third
type was spherical in form and similar to conventional fly ash particles
obtained from the combustion of coal. It is interesting to note, however,
that the size distribution, as observed qualitatively under the microscope,
indicated that these spherical particles are significantly smaller than those
normally found in a conventional coal-fired power plant. In addition, it was
noted that even when pure coal was employed as fuel, the geometry of the so-
called spherical fly ash particles was significantly different from those
obtained in a full-size power plant, indicating that combustion conditions
employed in this particular experimental unit were, in fact, somewhat unique
(Figure G-3). The fourth type of particle was in the form of flaky material
(Figure G-4) with somewhat rounded edges which indicate the possibility of
some melting during the combustion process.
It was observed that each of these morphological types was present in
all of the subsamples investigated. There were, however, some variations in
the relative amounts of each structure obtained in particular subsamples.
For example, the spherical particles were found to occur in much greater
profusion in the small size fractions of the samples from the 1:1 and 1:2
RDF-to-coal blends.
It is apparent from the foregoing results that fly ash generated from
burning blends of dRDF:coal is significantly different in form from that
which results from the combustion of pure coal. In particular, the absence
of a large population of spherical particles indicates that the combustion
temperature employed in this particular system was insufficient to generate
molten fly ash material as occurs in the combustion zone of conventional
coal-fired units. This phenomenon provides relatively high specific surface
areas due to the predominance of the spongy-type material illustrated in
Figure G-l, and consequently the operation of strong surface-associative
effects in the distribution of potentially toxic species might be expected.
The second type of physical characterization performed was a determi-
nation of the distribution of the specific surface area of the fly ash
particles as a function of both particle size and the dRDF:coal ratio. This
was achieved using a Quantachrome Quantasorb Model QS-7 which enables deter-
mination of a specific BET surface area based on the adsorption of nitrogen.
Under normal operating conditions, samples are initially outgassed at 300°C
174
-------
Figure G-l. 0% RDF sample, S4, magnification 450 X.
(Note spherical particle at about 4 o'clock)
Figure G-2. 67% RDF sample, S4, magnification 3500 X.
175
-------
Figure G-3. 100% RDF sample, S4, magnification 10,000 X.
Figure G-4. 50% RDF sample, S4, magnification 2000 X.
176
-------
to remove any absorbed material prior to the determination of the surface
area. Due to lack of material, however, outgassing at 300°C was not employed
in the present instance for fear of losing potentially volatile elements such
as arsenic and cadmium. For this reason, outgassing at room temperature was
employed with the result that the samples were not, in our opinion, com-
pletely outgassed. Nevertheless, the specific surface area values obtained
are considered to be relatively consistent within the individual sample set
investigated. They should not, however, be taken as absolute values.
The results obtained are illustrated in Table G-l from which it will be
seen that insufficient material was available for determination of all size
fractions and dRDFrcoal ratios. Nevertheless, the results do establish
several points of interest. First, it is apparent that the stack sampler
employed for particle collection in this study is, in fact, producing a good
differentiation of the fly ash on the basis of particle size. In other
words, there is a clear dependence of the specific surface area of these
particles on aerodynamic particle size. Secondly, there appears to be a
general trend of increasing specific surface area with increasing RDF-to-coal
content. In short, the higher the RDF content of the fuel the greater the
specific surface area of the fly ash particles which result.
TABLE G-l. SURFACE AREAS (m2/g) OF dRDF:COAL FLY ASH FOR THE
SEVEN SIZE FRACTIONS COLLECTED USING AN MRI STACK SAMPLER
dRDF by Size fraction
volume SI S2 S3 S4 S5 S6 S7
0%
50%
67%
100% 2.7
4.96
7.26
5.5 15.3
5.73
9.02
9.61
17.9
8.38
11.4
20.8
10.2
9.90
23.0
10.1
30.0
-Increasing size-
177
-------
Table G-2 presents the size fraction cut-off for each of the impactor
plates (SI through S7). For example, Stage S2 is 50 percent efficient in
collecting particles in the size range of 15 to 30 microns.
TABLE G-2. NOMINAL CUT DIAMETERS FOR IMPACTOR TESTS
Stage Partical Size at 50 Percent Efficiency
(microns)
SI
S2
S3
S4
S5
S6
S7
Filter
30
15
6
2.4
1.5
0.65
0.37
Less than
0.37
NOTE: For further information see Meteorology Research, Inc., Instruction
Manual, Inertial Cascade Impactor, Model 1502 and 1503, 1976.
ELEMENTAL COMPOSITION
Elemental analyses of the various subsamples presented were carried out
in several ways. Semiquantitative analysis of a number of the samples was
undertaken using direct current arc emission spectrometry in which the
spectra were recorded photographically on a Baird-Atomic 3-meter grating
spectrometer, Model 6X-1. Samples were mixed with "Spex Mix" spectroscopic
graphite using a "wig-L-bug" and were completely vaporized using a direct
current arc. The integrated spectra obtained were analyzed using a manual
densitometer. The results obtained for the S3 size fraction are presented in
Table G-3.
Samples were digested in an acid mixture consisting of 3.5 mis of aqua
regia, 2.5 mis of 48 percent hydrofluoric acid, and 0.5 mis of water. The
resulting digest was neutralized using approximately 2 grams of boric acid to
remove excess hydrofluoric in the form of boron trifluoride. These samples
were then analyzed for 18 elements using an automated Spectrometrics plasma
emission spectrometer utilizing an Echelle monochromator. The results
reported by element are listed in Table G-3.
Specific analyses for arsenic were performed by generating the arsenic
hydride and identifying this concentration using conventional flame atomic
absorption spectrometry. The results obtained for arsenic are also included
in Table G-3.
178
-------
TABLE G-3. ELEMENTAL ANALYSIS OF INDIVIDUAL RDF COAL FLY ASH SAMPLES
CONCENTRATIONS IN yg/g
dRDF Analysis
Element by Volume Method SI
Size Fraction From Sampler
S2 S3 S4 S5
S6
S7
Aluminum
Antimony
0%
50%
67%
100Z
OZ
50Z
67Z
100Z
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
84,000
57,900 9,970
101,000 55,900
44,300 46,400 42,700 41,200
22
(1)
(5)
(10)
24 31
(20)
58
200
341
3.2* 5.5* 5.8* 111
(continued)
179
-------
TABLE G-3. (continued)
dRDF Analysis
Element by Volume Method SI
Size Fraction from Sampler
S2 S3 S4 S5
S6
S7
Arsenic
Barium
0%
50%
67%
100%
0%
50%
67%
100%
AAS
INAA
DCAES
AAS
INAA
DCAES
AAS
INAA
DCAES
AAS
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
312
548
(400)
(300)
(350)
500
888
(200)
1630
<105
<334
992
1100
1150
1211
1220
762
1230
488
1530
959
1050
1042
461
1392
1700
1571
1000
691
1220
1390
881
679
1050
252
(continued)
180
-------
TABLE G-3. (continued)
dRDF Analysis
Element by Volume Method
SI
Size Fraction from Sampler
S2 S3 S4 S5 S6 S7
Beryllium
Bromine
50Z
67Z
100Z
OZ
50Z
67Z
100Z
PES
1NAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
(9)
(9)
<53.4
(7)
13.0
(10)
21.7
9.15
<29.0
3.27
42.6
18.9 <36.2
973
85.
229
357
256 107 146 95 140 110
(continued)
181
-------
TABLE G-3. (continued)
dRDF Analysis
Element by Volume Method
SI
Size Fraction from Sampler
S2 S3 S4 S5
S6
S7
Cadmium
Calcium
0%
50%
67%
100%
0%
50%
67%
100%
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
PCAES
(3)
<931
(20)
(20)
<92.1
(3)
16,700
41,000 14,100
139
<108
<505 394
<167 <631
32,400 27,000
12,000 14,000 9,750 10,500
(continued)
182
-------
TABLE G-3. (continued)
dRDF Analysis
Element by Volume Method SI
Chromium
0% PES
INAA
DCAES
50% PES
INAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA
DCAES
Cobalt
0% PES
INAA
DCAES
50% PES
INAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA
DCAES
Size Fraction from Sampler
S2 S3 S4 S5 S6 S7
(150)
(300)
(300)
(70)
(60)
(50)
(30)
(60)
(continued)
183
-------
TABLE G-3. (continued)
Element
Copper
Dysprosium
dRDF Analysis Size Fraction from Sampler
by Volume Method SI S2 S3 SA S3 S6 S7
OZ PES 91.3
INAA
DCAES (100)
50Z PES <3,630 <283
INAA
DCAES (60)
67Z PES <134 <4,320
INAA
DCAES (50)
100% PES <747 <1,780 <168
INAA
DCAES (100)
OZ PES
INAA 18.2
DCAES
50Z PES
INAA 10.9
DCAES
67Z PES
INAA 7.5
DCAES
100Z PES
INAA <0.039 5.0 9.7 11.1 10.5 10.0 3.9
DCAES
(continued)
184
-------
TABLE G-3. (continued)
dRDF
Element by Volume
Europium
0%
50%
67%
100%
Gallium
0%
50%
67%
100%
Analysis Size Fraction from Sampler
Method SI S2 S3 S4 S5 S6 S7
PES
INAA 3.13
DCAES
PES
INAA 2.2
DCAES
PES
INAA 1.55
DCAES
PES
INAA <0.37 <0.74 1.79 1.93 1.90 2.08 0.73
DCAES
PES
INAA 120
DCAES
PES
INAA 80
DCAES
PES
INAA 49
DCAES
PES
INAA 32 22 69 80 114 173 165
DCAES
(continued)
185
-------
TABLE G-3. (continued)
dKDF Analysis
Element by Volume Method
SI
Size Fraction from Sampler
S2 S3 S4 S5
S6
S7
Lanthanum
Lead
0%
50%
67%
100%
50%
67%
100%
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
9.3
19 46
(800)
(1,000)
68
60
37
<10,400 757
(300)
49 48
189
<7,830 <7,745
25
<1,030 <1,290 <1,860 <7,042
(500)
(continued)
186
-------
TABLE G-3. (continued)
dRDF Analysis
Element by Volume Method SI
Magnesium
0% .PES
INAA
DCAES
50% PES
INAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA
DCAES
Maganese
0% PES
INAA
DCAES
50% PES
INAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA 947
DCAES
Size Fraction from Sampler
S2 S3 S4 S5 S6 S7
4,000
7,260 4,550
6,810 6,630
3,680 4,060 2,970 2,910
<391
378
(300)
<4,530 <548
555
(300)
<257
741
(400)
<3,260 <58.5 <321
666 206 189 183 167 79
(200)
(continued)
187
-------
TABLE G-3. (continued)
Element
Molybdenum
Nickel
dRDF Analysis Size Fraction from Sampler
by Volume Method SI S2 S3 S4 S5 S6 S7
0% PES <13.5
INAA
DCAES
50% PES <18.9
INAA
DCAES
67% PES <88.6
INAA
DCAES
100% PES <20.3 <111
INAA
DCAES
0% PES
INAA
DCAES (1,000)
50% PES 7,260
INAA
DCAES (600)
67% PES
INAA
DCAES (600)
100% PES 4,610 <37,200 214
INAA
DCAES (300)
(continued)
188
-------
TABLE G-3. (continued)
dRDF
Element by Volume
Phosphorous
0%
50%
67%
100Z
Potassium
0%
50%
67%
100%
Analysis
Method SI
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA 2,500
DCAES
Size Fraction from Sampler
S2 S3 S4 S5 S6 S7
3,710
<11,900 2,340
3,260 3,450
1,360 1,770 887 1,360
18,400
12,900
20,800 15,900
18,000
17,000 26,200
15,900
13,300 15,600 11,400 10,000
3,000 1,010 11,600 11,500 11,800 5,400
(continued)
189
-------
TABLE G-3. (continued)
dRDF Analysis
Element by Volume Method SI
Silicon
0% PES
INAA
DCAES
50% PES
IXAA
DCAES
67% PES
INAA
DCAES
100% PES
ISAA
DCAES
Sodium
0% PES
INAA
DCAES
50% PES
ISAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA 1,990
DCAES
Size Fraction from Sampler
S2 S3 S4 S5 S6 S7
74,600
33,000
19,000 88,900
24,300
3,300
1,960
28,200
45,600 4,000
28,200
2,920 23,500
1,470 2,680 2,930 3,000 3,350 1,720
(continued)
190
-------
TABLE G-3. (continued)
Element
dRDF Analysis
by Volume Method
SI
Size Fraction from Sampler
S2 S3 S4 S5 S6
S7
Strontium
50%
67%
100%
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
<193
1,180
1,270
<981 881
1,518
943
1,237
740 879
989 1,107
479
410 661
638 1,010
420
Titanium
0%
50%
67%
100%
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
PES
INAA
DCAES
(6,000)
(12,000)
(10,000)
(continued)
191
-------
TABLE G-3. (continued)
Element
Thalium
Vanadium
dRDF Analysis
by Volume Method SI
0% PES
INAA
DCAES
50% PES
INAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA
DCAES
0% PES
INAA
DCAES
50% PES
INAA
DCAES
67% PES
INAA
DCAES
100% PES
INAA
DCAES
Size Fraction from Sampler
S2 S3 S4 55 S6 S7
(0.5)
(2.0)
(0.6)
(2.0)
(1,500)
(500)
(300)
(800)
NOTE: PES - Echelle Plasma Emission Spectrometry
INAA - Instrumental Neutron Activation Analysis
DCAES - DC Arc Emission Spectroscopy
AAS - Atomic Absorption Spectrometry
* These values may be low due to arsenic Interference.
192
-------
Selected fly ash samples were subjected to instrumental neutron acti-
vation analysis utilizing a Triga II reactor having a neutron flux at the
sample of approximately 1012 neutrons/second/cm2. Since the primary objective
of this analysis was to obtain some relative measure of the accuracy of the
other analytical techniques and, specifically, to determine the arsenic and
antimony present, only a short irradiation was employed. Following removal
from the reactor, the samples were analyzed using a Ge/Li detector in
conjunction with a 4096 channel multichannel analyzer. The results were
transferred to magnetic tape and analyzed using the PIDAQ program. These
results are presented in Table G-3.
Finally, several subsamples were subjected to individual particle
analysis using a Kevex Model 5000A X-ray energy dispersive spectrometer (XES)
associated with the scanning electron microscope.
A total of 27 elements were determined using the techniques listed
above. It will be noted that the analyses were not performed by X-ray
fluorescence spectrometry as originally envisaged due to the small amounts of
sample provided. This decision was reached after initial results indicated
unacceptably low counting statistics for most elements. As a result, a
semiquantitative screening analysis was performed using DC arc' emission
spectroscopy (DCAES). This technique was applied only to the size fraction
S3 of each RDF:coal blend sample. The results are considered to have only an
order of magnitude precision so are presented in brackets in Table G-3.
The size fractions S3 to S6 were analyzed by plasma emission spectro-
metry. The results were generally disappointing due to the high blank
levels encountered. Consequently, only about half of the analyses performed
are considered to be meaningful, and only those are included in Table G-4.
(It will be noted that good analytical data were obtained for the aqueous
leachates presented later since extensive acid digestion was not required.)
The precision of the PES results is generally about 10 percent although in
this complex analytical matrix accuracy for the trace elements Be, Cd, Cu,
Pb, Mn, Mo, and Ni is likely to be considerably poorer.
Analyses performed by Instrumental Neutron Activation Analysis all
have a precision (based on counting statistics) of less than 10 percent. The
accuracy appears to be comparable. Precisions of 1 to 2 percent are
associated with the results obtained for AS, Mn, and Na.
It will be noted that quite good agreement is obtained between the
techniques employed for the elements As, Ba, K, and Sr. Agreement is poor
for the other elements. It must be strongly stressed, however, that in
situations such as this where the original samples are composites of grab
samples, sampling statistics are normally extremely poor. Since different
analytical techniques require different amounts of sample, this means that
the precision associated with each procedure will vary greatly even though
the intrinsic analytical precision is good. In brief, the spread of results
depicted in Table G—4 is fairly typical. Nevertheless,, the relative pre-
cision and accuracies do enable trends to be observed.
193
-------
The actual amounts of the various elements listed in Table G-3 are
hardly remarkable except in the case of As, which is fairly high. However,
some interesting trends are observed. Most notable is the apparent variation
of several elements with variations in the dRDF:coal ratio. Thus, Br, Mn,
Pb, and Sb show a tendency to increase in concentration with increasing dRDF
levels whereas As, Ni, and V show an opposite tendency. The trends are
hardly definitive but are probably real as indicated by similar trends
observed in these materials in the fly ash leachates.
Variations of concentration with particle size are also not well defined.
There is, however, a general tendency for the elements As, Ga, Na, and Sb to
increase in specific concentration with decreasing particle size. A pro-
nounced increase in the concentration of Br and Mn in large particles is
also observed.
The trends indicated above are not unexpected. Thus, one would expect
elements such as Br, Mn, Pb, and Sb to be present at higher levels in dRDF
than in coal, and the preferential association of As, Ni, and V with coal is
acceptable. Similarly, the preference of As, Ga, Na, and Sb for small
particles is to be expected since these elements are capable of being vola-
tilized during combustion and then preferentially adsorbed onto small
particles. The behavior of Br and Mn is not understood, however.
To investigate the association of different elements with individual
particles, the size fraction S4 was subjected to energy dispersive X-ray
emission analysis under a scanning electron microscope. A number of indi-
vidual particles were analyzed and found to contain Al, Ca, Fe, K, Si, Ti,
and Zn as consistent matrix elements (note: these are the only analyses for
Fe and Zn). The presence of As, Na, P, and S was indicated in some particles,
but signal intensities were too weak for absolute identification.
MECHANISM OF TRACE ELEMENT DISTRIBUTION IN FLY ASH FROM FUEL BLENDS
It has now been reasonably established that certain of the more
volatile trace elements (or the compounds in which they are present) are
volatilized at temperatures encountered during many combustion processes.
The resulting vapor-phase metallic species then either condense or adsorb
(probably the former) onto the surface of co-entrained fly ash particles as
both vapor and particles move away from the high temperature combustion
zone. This process results in the preferential redistribution of volati-
lizable species into small particles due to the fact that small particles
have a larger specific surface area than large particles.
This volatilization-condensation phenomenon has several undesirable
environmental consequences. First, it results in many of the more toxic
elements becoming preferentially associated with small particles which are
most readily emitted from most combustion operations, which can have long
atmospheric lifetimes, and which are preferentially deposited in the pulmonary
region of the human lung when inhaled. Secondly, the condensation phenomenon
results in the presentation of toxic species on the surface of particulate
matter, thereby making it most readily available to the external environment
194
-------
(for example, extracting body fluids). Because of these effects, it is of
considerable interest to determine the extent to which such a volatilization
condensation mechanism may operate. Consequently, it is appropriate to
determine trends in elemental concentrations with particle size.
The dependence of elemental concentration on particle size is indicated
in Table G-3. While there is some evidence that certain elements (e.g., As,
Ga, Na, and Sb) increase in specific concentration with decreasing particle
size (i.e., increasing particle specific surface area), these trends are not
sufficiently convincing to establish the existence of a volatilization-
condensation process.
To obtain direct information on surface concentrations, a number of
particles were examined using auger electron spectroscopy. Since the
analytical volume for this technique extends only about 10 to 20 A below
the particle surface, the spectra obtained are derived entirely from surface
associated material. In order to obtain depth profiles, inner layers of the
particles are exposed by etching into the particles using a stream of
positively charged argon ions.
A representative scan from the auger spectrometer is presented in
Figure G-5 for the S4 size fraction derived from 100 percent dRDF. This
shows that the elements Al, C, Ca, Fe, P, S, Si, and Ti are the predominant
surface constituents of these particles. Depth profiles for the most
readily identifiable elements associated with samples derived from 0, 50,
and 100 percent dRDF:coal blends are presented in Figures G-6 through G-8.
Precise depth scales were not established; however, a sputtering rate of
approximately 30 A/minute was employed.
These depth profiles do not show any pronounced surface predominance
for the elements observed. There is an indication that S and Si may be
surface enriched and C surface depleted in some samples; however, these
results could well be artifactual in the case of S and Si. Although not
shown in Figures G-5 through G-8, there appears to be an increase in the
weak Fe and Ti signals with depth. Chlorine, though observed initially, is
rapidly removed by the electron beam.
It can be concluded that these surface studies provide no evidence for
the occurrence of a volatilization-condensation mechanism. This does not
mean that such a mechanism does not exist but simply that it does not apply
for the elements observed by auger electron spectroscopy.
MOBILIZATION OF TRACE ELEMENTS IN SOLUTION
While the elemental composition of fly ash from dRDFrcoal fuel blends
was the primary measurement required in the present study, it is important
to establish the extent to which the species present can be mobilized in
solution. This is because the toxic trace elements exert an adverse
environmental impact only if they can be transferred from the solid material
to a liquid solution.
195
-------
(original data supplied by
the University of Colorado)
Figure G-5. Auger electron spectrum of 100 percent dRDF fly ash.
JR6
r-0-j-iH
Co
E. -fo«r" S
'SPECIMEN
; 1 ; j Element [ Etching
i ' v {*•/ : 1 Gun 2
?%> KDP Fly AS^ 3 c jy ' ' *V|P
- •— j •_.!-. A i£_: ^i^i"
j — — f>— R»le:
; . . | • i • •
(original data supplied by the University of Colorado)
TIME
MINUTES/DIVISION
ip- 3S v,
1 1 -•—— •— • / i
Figure G-6. Auger depth profile for 0 percent dRDF fly ash.
196
-------
SO*
SPECIMEN
)M-i4
*L_
Etching
T
s !
L-::1-::::!;:i: ::.:.:t-^
(original data supplied by the University of Colorado)
-f
TIME. 2__ MINUTES/DIVISION
[Ep- r
ir ;Vnxxi= z RC=
IOX NEUT. DATE: 2-'
Figure G-7. Auger depth profile for 50 percent dRDF fly ash.
I SPECIMEN
0='- S*
8
F«
Etehmf
(original data supplied by the University of Colorado)
TIME. 2. MINUTES/DIVISION
SEMS- IPX
o«TE: 3.-l'i-
Figure G-8.
Auger depth profile for 100 percent dRDF fly ash.
197
-------
In order to obtain information about the solubility characteristics of
the fly ash, the size fractions S3 through S7 were ultrasonically agitated
(Heat Systems Model W 200R Sonicator Cell Disrupter) for 2 hours with
15 mis of triply distilled water. Sample masses varied from 0.0019 to
0.0230 g. (Previous studies have established that water soluble material
can be quantitatively extracted under these conditions.) Following sonica-
tion, the samples were filtered through a 0.45-um millipore filter. Elementa
analyses of the filtrate were then performed using plasma emission spectro-
metry, and anion analyses were performed by ion chromatography using a
Dionex Model 10 ion chromatograph. A representative ion chromatogram showing
the presence of fluoride, chloride, nitrate, and sulfate in the leachate
from the fly ash is presented in Figure G-9.
The results obtained from these analyses are presented in Table G-4.
Correction has been made for all blank levels and precision is generally
j<10 percent.
These solubility studies provide considerable information about both
the chemical characteristics and the potential environmental impact of the
fly ash. Thus, with reference to the data in Table G-4, a number of trends
can be observed.
First, it is apparent that for most species the amount of material
which is soluble increases with increasing RDF levels in the original fuel
blend. This trend can be seen clearly for Ca, Cu, K, Mg, Mo, Na, Pb, Si,
Cl , N03 , and SO*,2 and may occur for B, Ba, Cd, and F~. A reverse trend
is observed for Ni and P and, possibly, for Cr and Sr. The fact that many
of the major matrix elements exhibit increasing solubility with increasing
dRDF percentage indicates that the addition of dRDF to coal will result in
greater bulk solubility (as well as greater trace element mobilization) as
compared with pure coal fly ash.
The second obvious trend is towards increasing mobilization of species
with decreasing particle size. This is apparent for Cd, Cr, Cu, K, Mn, Mo,
Na, Ni, Pb, and Cl~ and may occur for Ba, Be, P, and F~. This trend may be
due to condensation of these species from the vapor phase as discussed
earlier, or it may result from the more efficient formation of soluble
oxides (i.e., calcining) in small particles. In the former case one would
expect to see similar size dependenices for both the bulk (Table G-3) and
the separated soluble sample (Table G-4). This would not be the case if
solubility is the direct result of chemical reaction at a particle surface.
Unfortunately, the available data are not sufficient to rule out either
mechanism.
Further consideration of the data in Tables G-3 and G-4 provides some
interesting insights into the fractional solubility of fly ash from dRDF:coal
blends. As an initial general statement, it can be said that matrix elements
such as Al, Ba, Mg, P, Si, and Sr exhibit quite low solubility (<10 percent)
whereas minor and trace elements such as Be, Cd, K, Mn, and Na are fairly
soluble (approximately 20 to 80 percent). Calcium is a notable exception to
this rule insofar as it exhibits high solubility. In considering this
198
-------
iijljii'ijiiiip!;1^:^;!1;-!!;^'!
i'lii!;!!! ;i'i;^J: .•''•' iiii! i!:-:!
.; (original data supplied by
j the University of Colorado)
; •!•!.•;.i
:i:il-::. :..
;:.+!. ...I..!.; i -iJ *•»!,.! ••.;•
.ii!::!i:&fl!j!i;Si;.!:!
!',' i: i i:! • • i i
:•. 1...I. , 1 ;.M
:.'' "! i •' • J!' • I ! •
I;!:: i
':;:>;.; /;;i;i!:; • liljliP 'Hi'!"•!'*
' ll V1/1!^''!'!]!!:;;!;!'!!!!!^!''!!
(2:10 dilution, 0-3 micromhos full scale deflection)
Figure G-9. A representative Ion Chromatograph of Sample: 1-0-S7
199
-------
TABLE G-4. CONCENTRATIONS OF SEVERAL METALLIC ELEMENTS AND ANIONS
IN AQUEOUS LEACHATES FROM dRDF:COAL FLY ASH
(yg/g of fly ash leachate)
Element
Aluminum
Barium
Beryllium
Boron
Size
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
dRDF Fuel
0% dRDF
by volume
18.8
138
148
24.6
3.03
3.02
17.2
42.3
13.9
<12.3
0.123
<2.61
2.01
2.46
9.09
<1.62
20.4
22.1
12.3
<199
Blend Composition
50% dRDF 67% dRDF
by volume by volume
2.12
12
3.90
263
158
2.74
25.8
44.2
68.4
77.5
<0.342
<5.00
<1.31
<5.32
2.50
<1.79
25.5
77.9
<346
<165
6.15
8.40
—
—
—
3.93
11.0
—
—
—
<0.427
<4.70
—
—
—
<2.24
32.2
—
—
100% dRDF
by volume
31.3
112
194
181
<10.1
3.57
61.0
169
171
35.9
<0.714
<4.86
<2.81
4.76
1.71
<3.74
56.5
22.2
<313
<56.2
(continued)
200
-------
TABLE G-4 (continued)
Element
Cadmium
Calcium
Chloride
Chromium
Size
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
dRDF Fuel
0% dRDF
by Volume
1.17
<7.20
<12.8
<15.6
<167
343
750
1,140
862
830
<560
<137
<346
28.7
417
0.926
<1.16
6.04
5.74
33.3
Blend Composition
50% dRDF 67% dRDF
by Volume by Volume
<0.189
<13.9
<24.8
<100
120
620
2,300 2,
3,020
2,550
548
223
56.9
<669
<2,710
<1,290
0.822
<2.24
5.19
<21.3
10.0
<0.236
<13.0
769
370
—
—
—
72.6
792
—
—
—
0.684
<2.09
—
—
100% dRDF
by Volume
<0.395
<13.4
<53.0
252
527
1,680
7,750
16,700
8,490
899
2,310
6,310
13,000
33,700
54,900
1.71
<2.16
<11.2
28.6
6.84
(continued)
201
-------
TABLE G- 4 (continued)
Element
Copper
Fluoride
Lead
Magnesium
Size
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
dRDF Fuel
0% dRDF
by Volume
<0.926
2.33
<3.39
<4.14
27.3
<443
12.6
<72.8
28.5
107
<5.04
<1.04
18.1
13.9
60.6
33.1
178
310
281
248
Blend Composition
50% dRDF 67% dRDF
by Volume by Volume
0.274 0.684
<5.50 <5.15
<6.56
5.26
228
<491 72.6
27.8 792
<141
<1,140
<271
3.70 4.79
15.5 6.10
22.1
1,740
7,890
18.1 25.1
308 350
474
353
70.0
100% dRDF
by Volume
<3.14
<5.35
<14.0
76.2
292
275
221
49.3
2,130
<92.7
7.71
12.0
283
2,420
4,720
94.6
328
1,240
1,300
100
(continued)
202
-------
TABLE G-4 (continued)
Element
Manganese
Moybydenum
Nickel
Nitrate
Size
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
S3
S4
S5
S6
S7
dRDF Fuel
0% dRDF
by Volume
2.22
10.6
26.8
25.4
57.6
0.864
2.15
14.8
24.6
60.6
0.556
5.40
17.4
17.2
24.2
60,000
1,550
1.580
6,090
5,380
Blend Composition
50% dRDF 67% dRDF
by Volume by Volume
4.79 5.47
29.0 " 25.2
36.4
26.3
30.0
1.03 1.11
<2.14 2.90
9.09
<37.5
2.50
0.822 3.08
5.15 2.26
6.49
10.5
70.0
142,000 199,000
2,160 1,980
7 , 690
10,400
<19,400
100% dRDF
by Volume
2.43
4.00
74.0
119
36.8
4.14
9.00
27.8
42.9
8.55
1.57
<0.665
<33.8
9.52
1.71
207,000
2,620
<21,500
29,800
8,890
(continued)
203
-------
TABLE G-4 (continued)
Element Size
Phosphorous
S3
S4
S5
S6
S7
Potassium
S3
S4
S5
S6
S7
Silicon
S3
S4
S5
S6
S7
Sodium
S3
S4
S5
S6
S7
dRDF Fuel Blend Composition
0% dRDF 50% dRDF 67% dRDF
by Volume by Volume by Volume
<31.0
46.6
205
101
870
257
870
1,460
1,480
1,280
25.2
105
170
161
288
444
545
584
295
<3,280
<34.4 <11.7
40.7 34.2
18.2
<1,450
27.5
525 160
1,200 835
2,710
7,950
15,400
15.0 38.6
129 186
184
21.1
<200
514 1,050
3,340 3,120
7,900
18,100
32,600
100% dRDF
by Volume
18.9
4.66
142
890
200
630
2,260
11,100
24,000
38,000
163
325
1,120
948
212
1,920
6,700
29,200
63,600
82,400
(continued)
204
-------
TABLE G-4 (concluded)
dRDF Fuel Blend Composition
0% dRDF 50% dRDF 67% dRDF 100% dRDF
Element Size by volume by Volume by Volume by Volume
Strontium
Sulfate
S3 4.75 7.53 5.47 10.1
S4 34.8 49.6 25.8 40.0
S5 72.5 70.1 — 80.6
S6 32.8 42.1 — 19.0
S7 130 <25.3 — 3.42
S3 43,100 82,500 80,800 110,000
S4 10,800 17,900 17,700 25,400
S5 11,600 23,800 — 40,800
S6 11,900 56,800 — 66,200
S7 8,640 107,500 — 75,000
205
-------
fractional solubility, it is noteworthy that there is a trend of increasing
fractional solubility with decreasing particle size for several elements,
the most notable of which is Mn. Since Mn exhibits no dependence of concen-
tration on particle size in Table G-3, one can tentatively conclude that,
for this element at least, its solubility increase with decreasing particle
size is simply due to the more efficient calcining, and thus the greater
bulk solubility, of small particles.
CONCLUSIONS
The following tentative conclusions can be drawn:
1. The specific concentrations of trace elements present in the fly
ash from blend combustion are quite similar to those found in pure
coal fly ash. The dRDF seems to be the main contributor of Br,
Mn, Pb, and Sb in this particular case while the coal is the
primary source of As, Ni, and V.
2. There is a tendency for several elements to be preferentially
concentrated in small particles, notably As, Ga, Na, and Sb.
3. It is apparent that the volatilization-condensation mechanism
which is responsible for partitioning volatile elements into small
fly ash particles in a coal-fired power plant is less effective in
the plant used to burn the dRDF:coal blend. This is probably due
to the lower temperatures achieved in the blend combustion since
the most volatile elements still exhibit volatilization-condensation
partitioning.
4. The solubility of the fly ash increases with increasing dRDF
content.
5. The solubility of both trace and matrix species present in small
particles is significantly greater than in large particles.
The overall conclusion to be drawn is that utilization of dRDF supple-
ments to coal for energy generation will generally increase the amounts,
mobility, and toxic potential of inorganic species associated with the
emitted fly ash as compared with that associated with pure coal fly ash.
206
-------
APPENDIX H
PRECEDING COAL:dRDF BLEND STUDIES
INTRODUCTION
In recent years European and American industries and government agencies
have become increasingly interested in the commercial conversion of urban
solid waste into a stoker coal substitute. Since little of the European
research data are available, this appendix is limited to the American
developments.
The following synopses of American tests conducted between 1972 and 1976
represent all information that could be gathered on these tests. While the
data in these synopses are incomplete, they give a qualitative and quantita-
tive insight into the recent developments.
FORT WAYNE MUNICIPAL ELECTRIC STATION TESTS1
The municipal electric station in Fort Wayne, Indiana, conducted these
tests with RDF cubettes supplied by the National Recycling Corporation (NRC)
of the same city.
Fuel Preparation System
As shown in the flow diagram of Figure H-l, NRC prepared the cubettes
from paper and paper board scraps collected from local paper mills. The NRC
plant includes a modified John Deere stationary alfalfa cubetter which has a
nominal capacity of 4.6 to 9.1 Mg/hr (5 to 10 TPH). This capacity varies
with the relative density of the scraps being densified. Up to 20 TPH of
waste can be injected into the cubetter and processed into fuel. White
metals, yellow metals, glass, and ceramics were excluded from the cubettes
since their low softening temperatures would likely cause clinkering and make
the furnace ash handling difficult. The cubettes were approximately
1-1/2 x 1-1/2 x 2 in. and were free of metals and glass. Occasionally
moisture was added to improve the binding qualities of the scrap waste.
Hollander, H. I. and N. F. Cunningham, "Beneficated Solid Waste
Cubettes as Salvage Fuel for Steam Generation," Proceedings 1968 National
Incinerator Conference, ASME, pp. 75-86.
207
-------
NATIONAL RECYCLING CORPORATION
FORT WAYNE PLANT
STORAGE
FOR
RECYCLE
4
METALS
3 TON/HR
.^ STEEL
MILLS
•^REFINERS
LANDFILL
4 r
K
STORAGE
OR RECYCLE
•REFINERS
•GLASS PLANTS
MUNICIPAL
WASTE
30 TON/HR
INDUSTRIAL
WASTE
SIZE REDUCER
1000 HP
BULKY
WASTE
N3
O
oo
CLASSIFIER
a
SEPARATOR
MAGNETIC
SEPARATOR
i
L^
i
4*
\^
NON FERROUS
METALS, GLASS, CERAMICS,
STONE, DIRT, ETC.
t
7 TON/HR
6 TON/HR
PLASTICS a
LIGHT PAPER
SEPARATOR
a
SCREENER
»i
ROTARY
SCREEN
1
4 TON/HR 10 TON/HR
4 TON/HR
REJECTS
PAPER, WOOD, RUBBER, PLASTICS, ETC.
10 TON/HR
RECOVERED LOW GRADE PAPER
CUBETTER
BALER
T
(PRINCIPALLY MUNICIPAL WASTE)
RECOVERED FIBER STOCK
^>^^ «^ i ••«•
( PRINCIPALLY INDUSTRIAL WASTE )
10 TON/HR
CUBETTES
is TON/HR - MUNICIPAL
5 TON/HR - INDUSTRIAL
10 TON/HR
I
BALED FIBER
STOCK
_STEAM
PLANTS
PAPERMILLS
BOARDMILLS
Figure H-l.
Flow chart, and mass balance for cubette production at the
National Recycling Corporation.
-------
Furnace and Boiler Facility
The tests were performed in one of the four furnace-boiler units. Each
unit includes a multiple retort, an underfed retort stoker-fired furnace with
a Sterling boiler, an economizer, and an air preheater. The boilers produce
2654 kPag (385 psig), 371°C (700°F) steam to drive turbine generators which
have a combined capacity of 40 megawatts. None of the plant equipment had to
be modified to accommodate the tests.
Test Program
Approximately 36 Mg of cubettes were burned during two tests. A fuel
analysis revealed that the Btu content of the as-received cubettes ranged
from 15.9 to 19.8 MJ/kg (6850 to 8530 Btu/lb). The average fuel properties
are shown in Table H-l. The tests were run at a 3:1 (coal:cubette) ratio.
While the cubette firing improved the appearance of the fuel bed, the smoke
opacity remained the same as when firing coal only. Although the tests were
successful, no further cubette firing has been reported.
TABLE H-l. ANALYSIS OF FUEL BURNED IN FORT WAYNE TESTS
Characteristic
Content
Moisture
Volatile Matter
Fixed Carbon
Ash
HHV as Fired
Sulfur
Chlorine
Hemispheric Reducing
Atmosphere Temperature
15%
65%
14%
6%
6800 Btu/lb (15.8 MJ/kg)
0.25%
0.20%
1148°C
209
-------
SUNBURY STEAM ELECTRIC STATION TESTS2
The Sunbury Steam Electric Station of the Pennsylvania Power and Light
Company conducted these tests with pellets prepared by Elo & Rhodes in
Easton, Pennsylvania.
Fuel Preparation System
Elo & Rhodes had ground mixed municipal solid waste to less than 2.5 cm,
pelletized the ground particles, and then delivered the pellets to a storage
bin. According to the fuel property summaries in Tables H-2 and H-3, the
waste benefication reduced the ash level to 20.6 percent and the moisture to
10.3 percent. The high heating value of the pellets is suspect because on a
moisture and ash-free basis, the heating value for the Elo & Rhodes pellets
is 22.6 MJ/kg (9700 Btu/lb) while the value for the NCRR pellets is 21.2 MJ/kg
(9100 Btu/lb).
Furnace and Boiler Facility
The test boiler at the Sunbury plant was the No. 4 boiler which is
equipped with three ball-in-tube mills that are normally used for pulverizing
coal. While a Raymond Bowlmill was acquired for the pellet pulverizing, the
ball-in-tube mills had to be used since the bowl mill did not perform satis-
factorily. The maximum generator capacity under normal conditions was 140 Mw,
but only 10 of the 12 burners were operable, and the boiler ratings could not
be attained. The pellets were nominally 5/8 in. in diameter and 1 in. long.
Test Program
Forty tons of pellets were burned during the test. It was difficult to
unload the pellets from the hopper cars because they had packed and bridged
over the bottom of the sliding gates on the cars. The pellet handling
produced excessive dust in the plant.
As the firing of the coal-pellet blend stabilized, the boiler output
decreased from the 120 Mw with coal-only firing to 104 Mw with the blend
firing. This drop was due to the lesser pulverizer capacity that reduced the
fuel input to the boiler. The pellets were fired for 6 1/2 hours. Although
the exact mixture of pellets and coal was not determined, it was estimated
that the pellets accounted for 45 percent of the total heat input. Table H-4
summarizes the monitored gaseous stack emissions. While the S03 concentration
doubled, the total S02 concentration remained the same. The chlorides
increased as expected. In general, the tests were successful. However, the
Sunbury staff has not indicated any interest in continuing the pellet firing.
2Author-unknown, "Final Report on Burning of Processed Refuse Pellets in
No. 4 Steam Generator on May 29 and 30, 1975" Sunbury Steam Electric Station
report, undated, 9 pages.
210
-------
TABLE H-2. ANALYSIS OF FUEL BURNED IN SUNBURY TESTS
As Received
Processed Refuse* Bituminous to Mills
Proximate Analysis
Moisture
Ash
Volatile Matter
Sulfur
Heating Value
Ultimate Analysis
Btu/lb
Mj/kg
10.3
20.6
51.8
0.4
6,680
15.5
4.7
16.0
26.2
3.1
12,011
27.9
Nitrogen
Hydrogen
Carbon
Oxygen
Ash
Sulfur
% 0.74
% 5.16
% 39.42
% 31.18
% 23.00
% 0.50
0
4.16
65.81
11.10
14.90
4.03
* Average of two samples
TABLE H-3. PARTIAL ANALYSIS OF ASH IN SUNBURY TESTS
Refuse
Bituminous
Silicon Dioxide %
Aluminum %
Iron and Titanium %
Calcium %
Magnesium %
Sodium %
Potassium %
Chlorides %
Remainder %
41.94
6.77
4.89
6.95
1.86
3.24
2.14
0.04
32.17
36.96
8.19
23.48
0.66
0.61
0.39
1.50
trace
28.21
100.00
100.00
211
-------
TABLE H-4. FLUE GAS ANALYSIS AFTER THE INDUCED DRAFT FAN AT SUNBURY
Date
5/29/75
5/29/75
5/29/75
Time
8:00 AM
10:00 AM
11:20 AM
1:00 PM
2:00 PM
4:00 PM
SO 3
ppm
Baseline 9.3
1 18.7
2 18.5
S02
ppm
935
994
1008
CL
0
6 & 6
8 & 11
PIQUA, OHIO, BLACK-CLAWSON TESTS3
These tests with coal-pellet firings were conducted by Black-Clawson
Fibreclaim, Inc., at the Piqua Municipal Power Plant.
Fuel Preparation System
The Black-Clawson Fibreclaim Company in Franklin, Ohio, prepared 22 tons
of nominal 3/8-in.-diameter pellets for the Piqua plant tests. They trucked
wet fiber that had been produced in their solid waste processing plant to the
Toledo Alfalfa Company, Middletown, Ohio, where the fiber was dried in a Heil
Model 125 triple-pass rotary dryer. Upon the return of the fiber to Franklin,
Black-Clawson produced the pellets in a California Century pelletizer mill.
A composite sample of the pellets had the fuel properties listed in Table H-5.
Boiler and Furnace Facility
Boiler No. 4 in the Piqua Municipal Power Plant was used for the test.
Manufactured in 1947 by the Combustion Engineering Company, this boiler has a
rating of 18.9 kg/sec (150,000 Ib/hr). Normally, the maximum steam pressure
is 3130 kPag (454 psig) at 440°C (750°F). The stoker is a Lloyd/Combustion
Engineeering chain grate. The economizer was designed and built by Combustion
Engineering. The air preheater was rated at a 2.7 GJ/hr (2.6 MMBtu/hr) input.
Emission control devices were not installed in this facility.
3Marsh, Paul, Black-Clawson Fibreclaim, Inc., "Preliminary Test Report
on Handling and Combustion Characteristics of Franklin Pelletized Fuel and
Coal Mixes," November, 1975, 17 pages.
212
-------
TABLE H-5. CHARACTERISTICS OF THE PIQUA dRDF
Moisture %
Ash %
Volatile %
Fixed Carbon %
Sulfur %
Heating Value Btu/lb
MJ/kg
16.5
9.02
63.7
10.38
0.22
6382
14.8
Test Program
The 20 Mg (22 tons) of pellets were mixed with an equal volume of coal
by a bulldozer in the coal yard. The mixture was then pushed with a front-
end loader into the bucket elevator and then transported by a drag chain
conveyor to the overhead bunkers from which it was metered to the grate
through a weigh lorry. No modifications were made to the existing coal
handling system. There were no mechanical problems during the test. Except
for the normal airflow and bed-depth adjustments for a particular fuel, the
boiler operated as for coal-only firing.
The blend was approximately 1:1 by volume, and the pellet substitution
provided 20 to 24 percent of the heat generated. The steam pressure and
temperature were maintained during a 7-hr test. The plant's normal coal
analysis is shown in Table H-6.
Conclusions
The 1:1 tests demonstrated the feasibility of using dRDF as a coal
supplement. Additional tests with a 2:1 mix, along with a detailed combus-
tion and emission assessment, were judged desirable.
213
-------
TABLE H-6. ANALYSIS OF THE COAL CO-FIRED IN PIQUA, OHIO
Parameter As Received Dry Basis
Moisture Percent
Volatile Matter Percent
Fixed Carbon Percent
Ash Percent
Sulfur Percent
Btu/lb 11,
MJ/kg
Ash from the coal
4
37.51
46.70
11.79
3.36
680
27.2
Initial Deformation Temperature
Second Softening Temperature
Fluid Temperature
12
1263°C, 2306°F
1318°C, 2405°F
1471°C, 2680°F
39.07
48.65
12.28
3.50
,170
28.3
WRIGHT-PATTERSON AFB BLACK-CLAWSON TESTS4
These tests were conducted by Black-Clawson Fibreclaim, Inc., at the
Wright-Patterson Air Force Base central heating plant, Building 770.
Fuel Preparation System
As for the Piqua Tests, Black-Clawson similarly prepared pellets for the
Wright Patterson AFB tests. Typical properties of the latter pellets are
listed in Table H-7.
Furnace and Boiler Facility
The central heating plant contains two Edgemoor Ironworks (36,300 kg/hr)
(80,000 Ib/hr) boilers that produce 862 kPag (125 psig) of saturated steam.
Installed in 1956, these boilers are fired with Detroit Rotograte spreader
stokers. The emission control equipment includes a cyclone separator for
reinjected fly ash, multiclones for coarse particulate control, and an
installed, but inoperative, electrostatic precipitator.
Test Program
The pellets trucked to Wright-Patterson AFB were placed in hopper cars
for delivery to the rail car dumper and then into the power plant hoppers.
^Jackson, J. W., "A Bioengineering Study of Emissions from RDF,'
UASFEHL, McClellan AFB, ADA024661, 1976.
214
-------
TABLE H-7. PROPERTIES OF THE dRDF AND STOCK USED IN THE WPAFB TEST
Feed Stock dRDF
As Received As Received
Energy - Btu/lb 3300 5800
MJ/kg 7.6 13.5
Moisture Percent 55.5 21.5
Ash Percent 6.2 11.2
Chloride Percent 0.07
Sodium 0.03
Softening Temperature 1266°C (2310°F)
Density Kg/m3 46 58.2
Sulfur Percent 0.12
Volatile Percent 60.2
Fixed Carbon 10.46
Two methods of mixing the pellets with coal were used to investigate the
mixing behavior. In the first method, a pile of blended fuel was prepared in
the coal yard by having a car unloading crane first alternately scoop and
deposit dRDF and coal and then lift and drop the deposit for further mixing.
The mixture was then loaded into a hopper car and delivered to the coal
bunker. In the second method, alternate scoops of the two fuels were simply
loaded into the hopper car with the fuels being mixed as they were unloaded
from the car and passed through the materials handling system to the coal
bunker. Both methods produced a visibly well-mixed fuel. After the blends
were loaded onto the coal cars, they were covered to prevent rain from
reducing them to their original pulped form. Calculated blend properties are
presented in Table H-8.
Two blend ratios were burned: 1:1 (coal:dRDF) by volume for 34 hours
and 1:2 by volume for 6 hours. Although the emissions were monitored, they
were uncontrolled since the installed electrostatic precipitator had been
inoperative for several years.
As indicated by the summary of the stack emissions in Table H-9, the
results were encouraging.
In contrast to the Sunbury tests, S02 was reduced in the Wright-Patterson
AFB tests. Moreover, the unburned hydrocarbons were drastically reduced in
the latter tests. Halides and heavy metal emission test data, however, could
not be interpreted conclusively, although increases were detected.
215
-------
TABLE H-8. AVERAGE PROPERTIES OF WPAFB dRDF
Properties Coal 1:1 Mix 2:1 Mix
Heating Value MJ/kg
(As Fired) Btu/lb
Moisture Percent
Bulk Density Ib/CF
kg/m3
Ash Percent
Sulfur Percent
Chlorine Percent
Fixed Carbon Percent
Volatiles Percent
Hydrogens Percent
Carbon Percent
30.0
12,900
4.88
53.1
851
7.5
0.67
0.09
55.6
31.9
4.9
73.3
26.3
11,327
7.73
44.2
708
9.09
0.53
0.14
43.8
39.3
5.09
63.5
22.1
9,518
14.01
42.4
679
8.86
0.43
0.14
36.5
40.6
4.91
56.5
TABLE H-9. STACK EMISSIONS (COMPARED TO COAL)
Emissions 1:1 Mix 1:2 Mix
Particulates Unchanged Highly variable, on
the average, unchanged
S02 Reduced by Reduced by
approximately 50% approximately 60%
NO Drastically reduced Drastically reduced
X by approximately 80% by 95+ %
216
-------
The boiler operation was normal during the 1:1 mix test after a few
minor control adjustments. However, during the 1:2 mix test, a loss of
control over the fuel distribution in the boiler caused the fire to concen-
trate near the rear of the firebox. In the automatic scale operation, some
fuel segregation produced a higher concentration of pellets in the middle
spreaders. Clinkers formed near the heavy fire concentration at the rear of
the boiler. The bottom ashes produced by the 1:1 blend firing differed
little from the bottom ashes produced by the coal-only firing. With the 2:1
mix, however, the ashes were frequently fused into clinkers.
A furnace inspection after the test revealed that some slag had
deposited on the rear wall refractory and on the rear portions of the side
walls. These deposits may have been caused by the poor fuel distribution
during the 1:2 blend tests. Small slag cones which had formed at the base of
the slag deposit indicated that ash had melted. Another evidence of fouling
was the formation of flake-like deposits on the fire side of the wall tubes.
These deposits generally sloughed off the tubes when the boiler cooled.
After the grates were swept clean, black stains were found with those close
to the grate air holes being the most prominent.
Conclusions
While the 1:1 mixture firing was generally satisfactory, the 2:1 mixture
firing caused poor fuel distributions which likely could have been solved by
minor facility modifications. A larger and more dense pellet should be
tested since its handling characteristics might solve some of the fuel
distribution deficiencies at the higher coal:KDF ratio. The potential for
scaling and waste should be evaluated further. The stack emission changes
were generally acceptable, and the decrease in hydrocarbons was especially
noteworthy. Lead emissions, which increased significantly during the tests,
were generally submicron aerosols that indicated the deposition of lead
vapors. The NOX and chloride-fluoride emissions in both the 1:1 and the 2:1
blend tests were significantly greater than those in the coal-only firings.
SANDWELL INTERNATIONAL TESTS5
These tests with fluff and densified RDF firings were conducted by
Sandwell International, Inc., at a facility owned by the Eugene Water &
Electric Board, Eugene, Oregon.
Fuel Preparation System
The Vista Fiber and Chemical Company in Los Gatos, California, produced
21 tons of nominal 1-in.-diameter dRDF pellets for these tests. The material
density of the individual pellets was 881 kg/m3 (55 lb/ft3), and the bulk
density was approximately 593 kg/m3 (34 lb/ft3). The average calorific
value of the fuel was 12.0 MJ/kg (5156 Btu/lb) as received and 15.0 MJ/kg
(6436 Btu/lb) on a dry basis.
5Sandwell International, Inc., "Eugene Water & Electric Board, Eugene,
Oregon, Solid Waste Fuel Modifications, Second Series Burn Tests—Final
Report," Report W3508/2, December 23, 1974.
217
-------
Boiler and Furnace Facility
The facility consists of three wood waste and bark boilers of unreported
design or manufacture. Testing was done in the No. 3 boiler. This boiler
has a nominal capacity of 19.5 kg/sec (155,000 Ib/hr) of steam and produces
427°C (800°F) superheated steam. The boiler has an air preheater and is
fired with a spreader stoker/traveling grate combination.
Test Program
The 21 tons of pellets were fed into the furnace during a 90-minute
period or at a feed rate equivalent to 304 Mg/day (335 TPD). As the furnace
exit gas temperatures increased from 593°C to 760°C (1100°F to 1400°F), the
speed of the feeders was reduced. The steam output reached 18.9 kg/sec
(150,000 Ib/hr) at 1169 kPag (300 psig) and could have attained a higher
output if additional fuel had been available. Although the emissions during
the dRDF tests were not analyzed, the fluff RDF tests indicated that the
particulate emissions, especially the fine particulate content, were generally
greater than those from coal only.
WISCONSIN SOLID WASTE RECYCLING AUTHORITY TESTS6
Of the several dRDF tests sponsored by the Wisconsin Solid Waste
Recycling Authority, three are synopsized as follows:
University of Wisconsin Tests
These tests with dRDF pellets were conducted at the University of
Wisconsin heating plant. For these tests, Gruman Eco Systems in St. Louis
prepared the pellets by densifying in a Sprout Waldren pelletizer the RDF
produced at the municipal RDF pilot plant. The pellets were 3/4 in. in
diameter and up to 3 in. long with a bulk density of 62.4 kg/m3 (39 lb/ft3)
and a heating value of 14 MJ/kg (6000 Btu/lb). The test furnace-boiler was a
Wickes waterwall furnace equipped with a Detroit Vibragrate stoker and a
water tube boiler rated at 5.7 kg/sec (45,000 Ib/hr) of 1862 kPag (125 psig)
saturated steam. The underfire air was delivered to the grate through a
5-compartment wind box. A baffle chamber provided the means for some removal
of particulate emissions.
Three volumetric coal:dRDF blends were burned during the test: 1:1,
1:3, and 0:1 (100 percent pellets). In each test, the furnace-boiler per-
formed satisfactorily under automatic control with the facility operating
between one quarter and one half of the design load. The burnout was
excellent, no emissions were visible; and no clinkers were formed. Although
the overfire air was not adjusted, the underfire air was throttled so that it
was introduced primarily through the first zone of the 5-compartment wind
box. No emissions data were collected.
6Private Correspondence between Warren Porter of Wisconsin Solid Waste
Recycling Authority and H. G. Rigo, 1976.
218
-------
Appleton Division Tests
These tests were conducted in the boiler house of the Appleton Division.
A blend of dRDF and bark was fed through a gravity feed chute into a Dutch
oven-fired, refractory-walled bark burner. With the dRDF fed at a rate of
2.73 kg/hr (3 tons/hr), the furnace heat release rate rose above the design
limit after 2 hours of firing, and consequently the blend feeding had to be
stopped. A pile of unburned material in the middle of the furnace was high
enough to block the overfire air ports and to cause poor combustion, smoking,
and clinkering. However, when the test was repeated at a slightly lower feed
rate, the combustion was satisfactory.
During the second test, a single stack test had particulate emission
rate of 94 mg/MJ (0.219 Ib/MBtu). Most of the ash in the dRDF remained in
the boiler as bottom ash. There was no visible plume.
Menasha Paperboard Mill Tests
These tests were conducted in the boiler house of the Menasha Paperboard
Mill. For these tests, 11.6 MJ/kg (5000 Btu/lb) dRDF prepared by the Vista
Fiber and Chemical Company was blended with 31.3 MJ/kg (13,444 Btu/lb) coal
in the existing fuel handling system. The coalrdRDF blend, which had a
15 percent dRDF substitution rate, was fired in a spreader stoker boiler.
The plant evaporation rate decreased from 9.75 to 9.05 kg steam/kg fuel as
the blend entered the boiler.
The plant power chief stated that although the blend firing appeared
feasible since it required no feed equipment changes, its particulate
emission was high.
CHANUTE AIR FORCE BASE TESTS
The U.S. Army Construction Engineering Research Laboratory conducted
these tests with pellets at Chanute Air Force Base for the Naval Facilities
Engineering Command. These tests provided a unique experience since they
were conducted in a chain grate stocker boiler and most of the pellets had
deteriorated into fines.
Fuel Preparation System
For these tests, the Vista Fiber and Chemical Company of Los Gatos,
California, prepared 181 Mg (200 tons) of dRDF pellets from mixed municipal
solid waste. In this preparation, the waste was course shredded, magneti-
cally separated, fine shredded, reshredded, air classified, screened, and
pelletized in a California Pellet Mill pelletizer.
7Letter to Warren Porter from C. Eaton on December 27, 1976, concerning
the test.
219
-------
After the 181 Mg (200 tons) of pellets had aggregated at the California
plant, they were loaded into box cars which were routed to Charleston, South
Carolina. Then, after a delay with the pellets unprotected, the box cars
proceeded to Rantoul, Illinois, where the pellets were unloaded and delivered
to Chanute Air Force Base. Upon their arrival at the boiler site, the pellets
had so deteriorated that the RDF was mostly fines, and the remaining pellets
had rough sides rather than the smooth sides characteristic of newly formed
pellets.
When the 181 Mg (200 tons) of deteriorated dRDF were placed in bunkers
to a depth of 7.9 meter (26 feet), the fuel bridged and rat holed. Since the
sides of the rat hole were stable, bins were unloaded by flushing the bunkers
with a fire hose. After 3 weeks of storage, the dRDF ignited by spontaneous
combustion. While bunker flooding with water extinguished the fire, much of
the remaining dRDF had deteriorated further.
Boiler and Furnace Facility
The Chanute Air Force Base heating plant houses several low-pressure
saturated steam boilers fired with chain grate stokers. The wind box is
unsegmented, and the front and rear overfire air jets are modulated as a
battery. The overhead parabolic bunkers feed a weigh lorry.
Test Program
After the bunkers were unloaded, some pellets were salvaged for short-
duration tests. During these tests (fired at 100 percent pellets), the chain
grate could not be fed fast enough to maintain load. Also, since the fire
filled the front of the furnace, the pellets burned too rapidly. No further
information on these tests was acquired.
220
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-80-095
3. RECIPIENT'S ACCESSIO(*NO.
4. TITLE ANDSUBTITLE
A FIELD TEST USING COAL:dRDF BLENDS IN SPREADER
STOKER-FIRED BOILERS
5. REPORT DATE
August 1980 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Gerald H. Degler, H. Gregory Rigo, and
Boyd T. Riley, Jr.
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Systems Technology Corporation
245 North Valley Road
Xenia, Ohio 45385
10. PROGRAM ELEMENT NO.
C73D1C
11. CONTRACT/GRANT NO.
68-03-2426
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Cin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final (6/76 - 7/78)
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Carlton C. Wiles 513/684-7871
16. ABSTRACT
This program was conducted to characterize and demonstrate the technical, economic,
and environmental feasibility of combustion densified forms of refuse derived fuel
(dRDF) blended with coal in spreader stoker-fired boilers. A total of 258.5 Mg
(285 tons) of pelletized 1/2-inch-diameter x 3/4-inch-long dRDF was co-fired with
coal in 2.7 x 7.5 kg/sec (60,000 Ib/hr) and 3.6 x 10 kg/sec (75,000 Ib/hr) of
1.03 MPa (150 psig) saturated steam. The results indicate that coal:dRDF blends up
to 1:2 can be handled and burned in conventional spreader stoker-fired boilers with-
out major equipment modification. As more dRDF was substituted for coal, the flame
volume increased, the opacity decreased, the fly ash carbon burnout improved, and
the turndown ratio of boiler operation increased. The emissions from the blend
firing decreased slightly in mass flux, dropped significantly in particulate size
and stack opacity, and had satisfactory particulate resistivities.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Refuse
Energy
Combustion
Air pollution
Evaluation
b.lDENTIFIERS/OPEN ENDED TERMS
Municipal solid wastes
Densified refuse derived
fuels
Resource recovery
c. COSATI Field/Group
13B
3. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
235
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
221
•it U.S. GOVERNMENT PRINTING OFFICE: 1980--657-165/0071
------- |