[3 D A ^'^- Environmental Protection Agency Industrial Environmental Research
m*mt f\ Office of Research and Development Laboratory
Research Triangle Park, North Carolin.i 1^/711
EPA-600/7-77-006
_____
JSHUSr 1977
APPLICABILITY OF NOX
COMBUSTION
MODIFICATIONS TO CYCLONE
BOILERS (Furnaces)
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven 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
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal
Agencies, and approved for publication. Approval does not
signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names
or commercial products constitute endorsement or recommen-
dation for use.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-006
January 1977
APPLICABILITY OF
NO COMBUSTION MODIFICATIONS
TO CYCLONE BOILERS (FURNACES)
T.E. Ctvrtnicek andS.J. Rusek
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45407
Contract No. 68-02-1320, Task 20
Program Element No. EHE624a
EPA Task Officer: David G. Lachapelle
*
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
iCyclone furnaces are a significant source of stationary NOX emis-
sfions.! It was estimated that 0.76 x 106 tonnes of NOX (over 6% of
stationary source NOX) were emitted from all cyclone-coal-fired
utility boilers in 1973. This represents from 19% to 22% of the
total NOX emissions from all coal-fired utility boilers in the U.S.
[several techniques of combustion modifications were applied in the
past to cyclone boilers/furnaces in an attempt to lower their NOX
emissions. T,hese include boiler load reduction, low excess air
firing, two-stage firing, and switching fuels. This report summa.-
rizes available NOX emission data when applying these techniques j
to cyclone boilers/furnaces. Even though significant reductions
in NOX were achieved, none of the techniques was shown to reduce
NOX emissions to the level meeting the New Source Performance
Standard.
11
-------
CONTENTS
Abstract
Figures
Tables
Acknowledgment
1.
2.
3.
Introduction
Conclusions and Recommendations
Characterization of Cyclone Furnaces/Boiler Types,
Population, and Emissions
3.1 Cyclone Furnace/Boiler Types and Operation
3.1.1 Principles of Cyclone Combustion
3.1.2 Cyclone Furnace/Boiler Arrangements
3.1.3 Auxiliary Equipment
3.1.3.1 Coal Preparation and Feeding
3.1.3.2 Slag Handling
3.1.3.3 Fly Ash Control and Handling
3.1.3.4 Combustion Control
3.1.4 Examples of Cyclone-Fired Boiler
Installations
3.1.5 Fuel Requirements
3.1.5.1 Coal
3.1.5.2 Other Fuels
3.1.6 Corrosion
3.1.7 Advantages/Disadvantages of Cyclone
Furnaces/Boilers
3.2 Population
3.3 Baseline Emissions from Unmodified Cyclone Furnace
Installations
3.3.1 Emissions Data Summary
3.3.2 NAPCA Data (Boiler I. D. No. 1)
ii
vi
viii
xi
1
2
8
8
8
15
16
16
19
20
20
21
25
25
29
32
38
41
45
46
51
111
-------
CONTENTS (continued).
3.3.3 Boiler Manufacturer Data (Boiler I. D.
Nos. 2 to 5, 9 to 11, 15) 53
3.3.4 Exxon Data (Boiler I. D. Nos. 6, 12 to 14) 56
3.3.5 KVB Data (Boiler I. D. No. 7) 58
3.3.6 TVA Data (Boiler I. D. Nos. 8, 16) 60
3.3.7 Commonwealth Edison Data (Boiler I. D.
Nos. 17 through 28) 61
3.3.8 NEDS Data (Boiler I. D. No. 29) 63
3.4 Need for NO Control 64
X
4. Applicability of Combustion Modifications to Cyclone
Furnaces/Boilers 67
4.1 Combustion Modification Strategy in General 67
4.2 Combustion Modification Experiences with Cyclone-
Fired Boiler Units 72
4.2.1 Boiler Manufacturer Field Experience (B&W) 73
4.2.1.1 Low Excess Air Firing (LEA) 75
4.2.1.2 Combined LEA and Switched Fuel Test 75
4.2.1.3 Switched Fuel Test 76
4.2.1.4 Simulated Staged Firing 76
4.2.2 Exxon Field Experience 79
4.2.2.1 Boiler No. 6 79
4.2.2.2 Boiler No. 12 84
4.2.2.3 Boiler Nos. 13 and 14 85
4.2.3 KVB Field Experience (Industrial Boiler) 88
4.2.4 Load Reduction Field Test Data 90
4.3 Implications of Applying State of the Art NOX
Combustion Modifications to Existing Cyclone
Combustion Units 95
4.3.1 Switched Fuel Firing 97
4.3.2 Load Reduction 97
4.4 Recommendations for Further Work 103
References 106
Appendices
A. Cyclone-Fired Boiler Installations 109
IV
-------
CONTENTS (continued).
B. Load Factors and Fuel Consumption at Cyclone-Fired
Electric Power Stations in 1973 115
C. Proposed Cyclone Boiler Test Program (KVB) 119
v
-------
FIGURES
Number Page
1 Cyclone furnace side view. 10
2 The cyclone furnace. 12
3 Furnace schematic. 13
4 Interior rear view of cyclone furnace. 14
5 Cyclone furnace throat (right) and attached boiler
target (left). 14
6 Furnace arrangements for cyclone-type primary
furnaces. 15
7 Schematic of boiler-firing arrangements using
cyclone furnaces. 16
8 Bin, direct-firing, and direct-firing predrying
bypass systems for coal preparation and feeding
to the cyclone furnace (schematic). 17
9 Sizing of crushed coal fired in the cyclone
furnace. 18
10 Belt-type coal feeder for the cyclone furnace. 19
11 Batch-removal slag-handling system for cyclone-
furnace boiler. 20
12 Sectional side view of boiler unit No. 20-A,
Calumet Station, Commonwealth Edison Company,
Chicago, Illinois, with horizontal-cyclone-
burner firing. 22
13 Philo-6 once-through supercritical Benson boiler
installation (Ohio Power Co.). 26
14 Coal suitability factors. 30
15 Auxiliary power requirements of typical high-
capacity pressure-fired cyclone-furnace and
pulverized-coal units. 40
16 States with cyclone-fired boiler units showing
number of boilers and percent of total U.S.
primary cyclone steaming capacity (149 boilers
generating 2.6 x I0k kg/s steam (data courtesy
of the Babcock & Wilcox Co.). 42
vi
-------
FIGURES (continued).
Number
17 Boiler outline for cyclone-type unit showing
sampling positions. 52
18 Thermodynamic equilibrium data for natural gas
fuel 70
19 Kinetic data for natural gas fuel. 70
20 Two-staging concept. 77
21 Cyclone firing arrangement, TVA Unit No. 1,
Drakesboro, Kentucky (view facing front or rear
wall. 80
22 Recirculated flue gas entry points for boiler
No. 6. 83
23 Overall reduction of NOX emissions for six coal-
and oil-fired cyclone furnace boilers using load
reduction (stack % C>2 levels indicated adjacent
to data points). 91
24 NOX emissions as a function of boiler size and %
load (boiler loads indicated adjacent to data
points. 94
25 Steam-temperature control by use of increased
excess air. 99
26 Cyclone-fired boiler with gas tempering for gas-
temperature control and gas recirculation for
control of furnace absorption and reheat
temperature. 100
27 Flue gas recirculation. 101
-------
TABLES
Number Page
1 Description of Equipment; Unit No. 20-A, Calumet
Station 23
2 Typical Analyses of Central Illinois Coal and Ash 27
3 Typical Coals Suitable for Cyclone Firing 31
4 Range of Coal Analyses 32
5 Selected Samples of Natural Gas from United States
Fields 33
6 Typical Ultimate Analyses of Petroleum Fuels 34
7 Analyses of Wood and Wood Ash 35
8 Examples of Analyses of Coal Feeds and Resulting
Chars from Various Coal-Conversion Processes 36
9 Typical Analyses of Petroleum Cokes 37
10 State-by-State Population Distribution of Cyclone-
Fired Boilers 43
11 Summary of Baseline Emissions Test Methods 47
12 Summary of Baseline NOX Emissions Data for Cyclone
Boilers 48
13 Baseline CO, SC>2, SO$t and Particulate Emission
Data Ranges for Cyclone Boilers at Full Load
(>90% of Rated Capacity) 50
14 Emissions Summary for Unidentified Coal-Fired
Cyclone Boiler Unit 53
15 Average Coal, Boiler, and Flue Gas Data for
Unidentified Boiler Unit 54
16 Emissions from Cyclone-Fired Boilers 55
17 Summary of Exxon Emissions Data for Cyclone-Fired
Boilers 57
18 Summary of Boiler Operating Data Corresponding to
Exxon Emission Tests 58
19 Available Fuel Analysis Data for Boilers Tested
by Exxon 59
Vlll
-------
TABLES (continued).
Number
20 Emissions Data for New York Boiler (Boiler
I. D. No. 7) 60
21 Coal Analysis, A New York Boiler (Boiler
I. D. No. 7) 60
22 Baseline Flue Gas Concentrations and Emission
Rates of Nitrogen Oxides and Carbon Monoxide for
12 Commonwealth Edison-Owned Cyclone Furnace
Fired-Boiler Units 62
23 Annualized 1972 Emissions Data for a 235 MW
Cyclone-Fired Utility Boiler Unit (Boiler
I. D. No. 29) 63
24 Summary of Aerotherm NOX Emission Estimates for
All Cyclone-Fired Boilers in 1972 66
25 Potential Factors Controlling the Formation of
Thermal NOX in Cyclone Boilers 71
26 Cyclone Boiler Units Field-Tested for Combustion
Modification Applicability 74
27 Lignite-Fired Boiler (Boiler I. D. No. 9) 75
28 Effect of Fuel Switching on NOX Emissions from
Coal-Fired Boiler I. D. No. 5 76
29 Seven Fuel and Air Flow Combinations for Stacked-
Cyclone Firing (Pattern Firing) 79
30 Boiler Test Program Implemented by Exxon (Boiler
I. D. No. 6, TVA Paradise Unit No. 1) 81
31 Summary of Emission Data, Boiler I. D. No. 6,
704 MW Coal-Fired, TVA Paradise Unit No. 1 82
32 Summary of Emission Data from Boiler No. 12,
450 MW 84
33 Summary of Operating Conditions and Emissions
Data for Oil-Fired Boiler Nos. 13 and 14 (Ace,
B. L. England Units 1 and 2) 87
34 Baseline and Combustion Modification Emissions
Data (Boiler Located in New York, Boiler No. 7) 89
35 Load Reduction Test Data 93
36 NO , Reductions for a 20% Reduction in Boiler Load 94
IX
-------
TABLES (continued).
Number Pa9e
A-l Installations of Cyclone-Fired Boiler Units -
Utilities 110
A-2 Installation of Cyclone-Fired Boiler Units -
Industrial and Commercial 114
B-l Loa/i Factors and Fuel Consumption for Cyclone-
Fired Electric Power Plants in 1973 117
-------
ACKNOWLEDGMENT
The authors wish to acknowledge the significant contribution of
the Babcock & Wilcox Company and the data and information provided
by Commonwealth Edison and KVB Engineering, Inc.
XI
-------
SECTION 1
INTRODUCTION
Cyclone-fired boiler units are widely used for generating steam.
They are used primarily in large electric power plants and are
also used to a lesser extent by industry and by large institu-
tions to supply steam for power generation or other uses.
Cyclone-fired, primary steam-generating capacity in the U.S.
totals approximately 26,000 kg/s, about 9% of the total U.S.
steam-generating capacity.
Cyclone boilers have traditionally been labeled as high NO emit-
ters. Coal-fired cyclone boilers contribute nearly 20% ofxthe
NO emissions from all coal-fired utility boilers.
X
The first purpose of this study was to update available informa-
tion on cyclone furnace/boiler characteristics, population, sales
trends, and emissions. The second purpose was to develop judgmen-
tal information on the combustion modifications capable of reduc-
ing NO emissions from cyclone combustion.
X
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
1. A cyclone-fired boiler unit is a complex, integrated, combus-
tion system that operates under a well-defined and a rather
restricted range of conditions. It was developed to fire
troublesome coals high in ash content and having low ash fus-
ion temperatures. These coals are difficult to burn in both
stoker and pulverized coal combustion systems. The heart of
the cyclone boiler is the cyclone furnace. Successful opera-
tion of the cyclone furnace depends on maintaining a liquid
or wet slag within the furnace. To meet this condition, fur-
nace heat absorption rates as well as ash fusion temperatures
of coal fired in the furnace must be low. If these condi-
tions are not met, the advantages of cyclonic combustion are
lost, and other conventional methods of combustion become a
prerequisite.
2. Combustion of coal in a cyclone furnace results in heat
release rates 8 to 10 times higher than in pulverized coal
combustion. Cyclone furnace gas temperatures are about
1922K (3000°F). These temperatures are sufficient to melt
the ash into a liquid slag, a thin layer of which adheres to
the walls of the cyclone. Coal particles are thrown to the
walls by centrifugal force and are caught in the running slag
where they are quickly combusted. Slag is then tapped,
cooled, and sent to disposal.
3. The key to successful operation of the cyclone boiler is to
maintain a noncorrosive and fluid ash in the cyclone furnace
throughout the whole range of loads at which the boiler oper-
ates. The abilities of coal fuels to meet these requirements
can limit the use of certain coals. Lignite, oil, and gas
fuels are also routinely fired in cyclone boilers even though
the cyclone furnace was originally designed for bituminous
coal firing.
4. Cyclone boilers under normal operating conditions do not
exhibit any unusual or peculiar corrosion characteristics.
Operation at reduced excess air, however, aggravates forma-
tion of corrosive iron and iron sulfide, which has had cata-
strophic effects in several boilers.
-------
5. Cyclone firing has explicit operating advantages and disad-
vantages. Its primary advantage is the ability to burn slag-
ging coals economically, which cannot be done by other con-
ventional methods of combustion. Among other advantages are
simplicity and reliability, low excess-air requirement
(10 to 17%), low carbon loss when burning bituminous coals,
higher full-load boiler efficiency, lower total particulate
emissions, an ash more suitable for landfill, smaller size fur-
nace, low coal-preparation cost, and easy conversion to the
.firing of other types of fuel. Some disadvantages include
limited operating flexibility, high NOX emissions, high pres-
sure drop, high carbon loss when burning western coals of
nonwetting ash characteristics, and perhaps a respirable par-
ticulate emission problem from the fineness of the ash emit-
ted through the stack. All of these disadvantages are asso-
ciated with the principal method of firing utilized in
cyclonic combustion.
6. There are 149 cyclone boilers in the U.S. These boilers are
fired by a total of 736 cyclone furnaces and generate
26,000 kg/s of primary steam (about 9% of the total U.S.
steam generating capacity). Cyclone boilers are located in
26 states, with nearly half of the capacity and one-third of
the boilers located in Illinois, Missouri, and Indiana. A
significant portion of the steam raising capacity (94%) is
operated by electric utilities.
7. The Babcock & Wilcox Company (B&W) is the sole supplier and
manufacturer of cyclone boilers. Since 1973, B&W has not
sold a single cyclone unit. The decline of sales started
when New Source Performance Standards (NSPS) for SOX emis-
sion control were put into effect. The coals with low ash
fusion temperatures normally have high sulfur content. The
final event restricting the sale of bituminous-coal-fired
cyclone boilers was the implementation of the NSPS for NO
emissions. x
8. Due to the nature of cyclonic combustion requiring high com-
bustion temperatures, the cyclone boilers are the highest
NOX emitters among all presently available combustion methods.
At full load and when firing bituminous coals, they emit
nitrogen oxides in the concentration range of 960 to 1200 vppm.
This translates into 576 to 688 ng NO per joule of heat
input (1.4 to 1.6 lb/10 Btu). The present new source per-
formance standard (NSPS) applicable to cyclone boilers
burning bituminous coal is 301 ng NO per joule (0.7 lb/106
Btu). X
9. Bituminous coal firing produces the highest NOX emissions
among all fuel types. In general, the ranges of full-load
NOX emissions appear to decrease in the following order:
bituminous coal (960 vppm to 1,197 vppm [576 ng/J to 688 ng/J]),
-------
sub-bituminous coal (910 vppm [546 ng/J]), lignite (485 vppm
to 593 vppm [291 ng/J to 355 ng/J]), natural gas (415 vppm
to 650 vppm [207 ng/J to 325 ng/J]), and residual oil
(441 vppm to 530 vppm [254 ng/J to 310 ng/J]).
The corresponding NSPS for these fuels are:
ng/J
Bituminous coal 302
Sub-bituminous coal Not available
Lignite 258
Oil 129
Gas 86
10. It was estimated that in 1973, 62 x 109 kg (68.4 x 106 tons)
of coal were burned at power plants employing cyclone firing.
The overwhelming majority of this coal was bituminous with
an average heating value of 26 MJ/kg (11,200 Btu/lb).
11. Using the emission factor of 12.28 g NOX per kg of bitumi-
nous coal, it was estimated that 0.76 x 106 tonnes
(0.84 x 106 tons) of NOX were emitted from all cyclone-coal-
fired utility boilers in 1973. This represents from
19% to 22% of the total NOX emissions from all coal-fired
utility boilers. With respect to all (137) identified sta-
tionary sources of NOX, the cyclone-fired utility boilers
burning bituminous coal ranked third. It was estimated that
over 6% of the total NOX emitted by all 137 stationary
sources came from this equipment type.
12. The principles of NOX formation during cyclonic combustion
are not well understood. It is believed that NOX originates
from two sources. First, NOX is thermally formed by fixa-
tion of atmospheric nitrogen. This reaction is primarily
influenced by oxygen concentration, combustion temperature,
and reaction time at the combustion temperature. A second
source is fuel-bound nitrogen which is oxidized to NOX in
the combustion process. The most critical factor in fuel
NOX formation appears to be the local conditions under which
fuel volatilization takes place. Under reducing conditions,
fuel nitrogen may form N2 or other nitrogen intermediates
which can revert to N2, whereas in an oxidizing environment,
NO is formed.
13. Preferably, any combustion modification made to a cyclone
boiler to reduce NOX emissions should act on both the ther-
mal and fuel-bound NO formation mechanisms. Available
information reveals that four types of combustion modifica-
tions have been applied, singly or in combination, to
cyclone combustion units. These are (1) boiler load reduc-
tion, (2) low excess air (LEA) firing, (3) simulated staged
-------
firing, and (4) switched fuel firing. Most of the data are
available for the first two modifications. Despite the fact
that the cyclone furnaces are significant NOX emitters, only
a relatively small number of cyclone boilers were found to
have been examined and tested in some way to determine the
effect of combustion modifications on NOX emissions. One
reason for the lack of adequate field data on this combustion
equipment class is the relative inflexibility of the cyclone
boilers toward modification.
14. Boiler load reduction produced the highest and the most con-
sistent degree of NOX emission reduction. Reducing the
boiler load from 100% to 80% resulted in an average NOX reduc-
tion of 29% with coal fired units and an average NOX reduc-
tion of 19% with oil fired units. Even with this reduction
in NOX emissions, however, the cyclone boilers could not
meet New Source Performance Standards.
15. Low excess air (LEA) firing reduced NOX generally at the
cost of increased CO emissions. Reducing excess air in one
lignite boiler resulted in an NOX reduction of 50% although
supplemental oil was required to maintain ignition in the
furnace. Applying LEA firing to oil-fired units resulted in
less dramatic changes (10% to 16% reductions) at acceptable
CO levels. One bituminous coal-fired unit tested showed an
11% reduction in NOX with no change in CO. NOX reductions
achieved with LEA firing alone in cyclone boilers could not
meet the NOX NSPS.
16. Staged firing was simulated in several coal-, oil-, and gas-
fired boilers. A 28% to 36% NOX reduction was achieved by
firing eastern coal, and a 48% reduction was achieved by
firing natural gas using a two-stage concept. Pattern fir-
ing, the other simulated stage firing concept applicable to
boilers with a multitude of cyclones and consisting of
varied fuel and air flows through these cyclones, produced
mixed results for a variety of fuels. B&W does not recom-
mend the stage firing methods for application because of the
combination of the following reasons: lack of significant,
long-term testing experience; reluctance of boiler owners to
accept the method on a permanent basis; and risk of cata-
strophic tube corrosion.
17. Switching from bituminous coal to fuels such as natural gas,
residual oil, or lignite can cut NOX emission in half. How-
ever, different standards apply to different fuels, and none
of the fuels cyclonically fired can meet NOX emissions
standards.
18. The data in this report indicate that the NOX New Source Per-
formance Standard (NSPS) cannot be met at normal full-load
firing for any type of fuel fired in a cyclone boiler unit,
-------
with the possible exception of lignite. The proposed lig-
nite standard was met during one test conducted by B&W at
very low excess air levels and with supplemental oil fuel.
Both load reduction and switching fuels may be considered as
practical interim measures in reducing NOX emissions from
some cyclone boilers during serious episode conditions.
Both of these methods, however, are least desirable to the
boiler owner for operational and economic reasons.
19. Flue gas recirculation (FGR) was applied to only one boiler
unit and showed no change in NOX level of a coal-fired unit.
The potential of this method cannot be properly evaluated
because of limited data and application of FGR to the point
in the cyclone boiler where maximum NOX control effective-
ness could not be achieved.
20. Because detailed information was not available on conditions
under which the tests summarized in this report were imple-
mented, it is not certain if the tests were thorough, compar-
able, and representative of all permanent cyclone operations
and if all possibilities for cyclone boiler modification to
achieve reduced NOX emissions were thoroughly explored.
Many tests were cost- and/or scope-limited. It is therefore
recommended that a comprehensive test program be developed
that would concentrate on causes and locations of NOX forma-
tion in the cyclonic combustion process. The comprehensive
test program should investigate the following:
a. Variations in operation of individual cyclones in
multicyclone fired facilities and the influence of
these variations on the total boiler NO emissions.
x
b. The effect of boiler modifications on cyclone fur-
nace NOX emissions with a comprehensive evaluation
of variables and conditions having a strong influ-
ence on NO formation.
X
c. Long-duration tests of the successful modifica-
tions to develop reliability and operational data.
d. Applicability of the successful modifications to
all cyclone boiler facilities.
e. Cost of such modifications.
21. Whether any NOX emission control methods, even if effective
in meeting NOX emission standards, will be practical and
acceptable to boiler operators is not presently known. For
the last 6 years, no cyclone boilers have been sold in the
United States. A large majority of cyclone boilers (49%
steaming capacity) are 12 to 32 years old. The normal life
expectancy for boiler facilities is from 25 to 35 years.
-------
Thus, within the next 3 to 13 years, some 49% of cyclone
steaming capacity will be obsolete and will have to be
replaced. This process may be dramatically encouraged by a
strong enforcement of existing NOX emission standards. A
study is therefore recommended to determine the influence of
the NSPS applicable to NO emissions from cyclone boilers on
American industries. x
-------
SECTION 3
CHARACTERIZATION OF CYCLONE FURNACES/
BOILER TYPES, POPULATION, AND EMISSIONS
The following four sub-sections characterize cyclone furnaces/
boilers. All references to cyclone-furnace-fired boiler units in
this report are made solely to units originally patented, devel-
oped, and designed by the Babcock and Wilcox Company for the orig-
inal purpose of burning coal. At least one other manufacturer
markets a "cyclonic" method of firing coal. Fluor Utah, Inc., a
subsidiary of Fluor Corporation, markets the Lucas Cyclonic Fur-
nace System developed in England, which is primarily used for tire
disposal.
Section 3.1 gives a detailed account of the principles of opera-
tion, design, auxiliary equipment, and fuel requirements. Section
3.2 presents available population-related data pertaining to
installations. Section 3.3 summarizes the available emissions
data. Section 3.4 gives an updated estimate of the need for NO
control in this equipment class.
3.1 CYCLONE FURNACE/BOILER TYPES AND OPERATION
3.1.1 Principles of Cyclone Combustion
A cyclone-fired boiler unit is a complex, integrated combustion
system whose purpose is to generate steam. The heart of this sys-
tem is the cyclone furnace. The cyclone method of firing coal was
developed about 35 years ago by the Babcock and Wilcox Company.
At the time it was introduced, it represented a major breakthrough
in the art of firing troublesome coals high in ash content and
having low ash fusion temperatures. These coals proved difficult
to burn in both stoker and pulverized units. Cyclone firing still
is one of the better ways to burn this low-quality coal. Although
all existing cyclone furnaces were originally designed to burn
coal, many other types of fuels have been and are still being suc-
cessfully fired in them. These fuels include residual and distil-
late oils, solid waste (wood bark, coke), and natural gas.
-------
Figure 1 shows a side view of a typical cyclone furnace in opera-
tion.1 The diameters of existing furnaces range from 1.8 m to 3.1 m
(6 ft to 10 ft) with lengths normally 3.4 m (11 ft) or longer.2
These sizes correspond to heat input firing rates ranging from
44 MW to 123 MW (150 and 420 million Btu/hr).3 With a typical
low-rank coal of 24.4 MJ/kg (10,500 Btu/lb), this corresponds to
a coal feed rate between 1.8 and 5.0 kg/s (7 and 20 tons/hr) per
furnace. The full load heat release rate is independent of fur-
nace size and is approximately 2.52 MW/m2 (800,000 Btu/hr - ft2)
of wall surface. "* This rate is generally 8 to 10 times higher
than that for pulverized coal furnaces. Full-load heat release
rates on a furnace volume basis for a 3.1 m (10 ft) diameter
cyclone are 4.7 MW/m3 (450,000 Btu/hr - ft3) and increase to
7.8 MW/m3 (750,000 Btu/hr - ft3) for a 1.8 m (6 ft) diameter fur-
nace. A typical heat release rate on a furnace volume basis for
a pulverized-coal-fired unit is 0.2 MW/m3 (20,000 Btu/hr - ft3).
Because of the small amount of wall surface area and the insulat-
ing effect of the wall refractory and the liquid slag layer, the
heat absorption through the cyclone furnace walls is low. Heat
absorption ranges from 0.13 to 0.25 MW/m2 (40,000 to 80,000 Btu/
hr - ft2) with high temperature refractory (e.g., Super Hi-Bond
3000®) covering studded tubes lining the furnace walls.2 This
amounts to approximately 5% to 10% of the total heat release rate
at full load. Successful operation of the cyclone furnace depends
on maintaining a liquid or wet slag condition within the furnace.
This is a characteristic of a properly operated cyclone. Heat
absorption rates as well as ash fusion temperatures must be low to
provide this condition. Consequently, the cyclone furnaces oper-
ate under a well-defined and a rather restricted range of condi-
tions. If these conditions are not met, the advantages of
cyclonic combustion are lost, and other more conventional methods
of combustion (pulverized coal firing, stokers, etc.) become a
prerequisite.
S. T. Potterton of Babcock & Wilcox (B&W) has described the gen-
eral coal firing arrangement in cyclone furnaces.3 His descrip-
tion is paraphrased as follows:
*New Cyclone-Fired Boiler for E. H. Werner Station, Jersey Central
Power & Light Company. Bulletin G-81 by the Babcock and Wilcox
Company, New York, New York, 1953. 9 pp.
2The Babcock and Wilcox Company. Steam—Its Generation and Use.
38th Edition. New York, New York, 1972.
3Potterton, S. T. Combination Fuel Firing in Cyclone Furnaces.
The Babcock and Wilcox Company. (Presented to 1970 Industrial
Coal Conference. Lafayette, Indiana October 7, 1970). Barber-
ton, Ohio. 9 pp.
^Shields, Carl D. Boilers - Types, Characteristics, and Functions.
F. W. Dodge Corporation, New York, New York, 1961. 559 pp.
-------
CRUSHED COAL
4017-1
nan
Figure 1. Cyclone furnace side view.1
"The cyclone furnace is a water-cooled, horizontal cyl-
inder in which fuel is fired, heat is released, and
combustion is completed. Coal crushed so that 95%
will pass through a 4 mesh screen is introduced tan-
gentially through a primary burner at the front of the
cyclone. About 20% of the required combustion air
also enters the cyclone tangentially through the pri-
mary burner at about 349 K (250°F). This pre-heated
air enters tangentially and serves to distribute the
coal over the surface of the cyclone. The remaining
combustion air is also preheated and enters the
cyclone at a high velocity of about 91.4 m/s (18,000
FPM) and at about 672 K (750°F) through a secondary
air port near the top of the cyclone and extending
over almost its full length. All air is preheated by
using a heat exchanger operating off the waste heat
of the flue gas. The port is arranged for tangential
air entry."
The tangential admission of secondary high velocity air increases
the whirling or centrifugal action on the fuel. A small amount of
air (up to about 5%) is also admitted through the inlet at the cen-
ter of the burner behind the primary coal/air stream and is called
tertiary air. The tertiary air port is used to cool the burner
10
-------
and help maintain fuel ignition. Either a gas or oil lighter with
a capacity of 2.9 MW (10 million Btu/hr) is used to ignite the
fuel. The lighter (igniter) is normally lit prior to the introduc-
tion of the crushed coal.
Figure 2 shows the arrangement of cyclone components.3 Along with
the basic coal furnace components are shown the oil and gas
burners which can be used for multifuel firing or ignition of coal.
G. W. Kessler of B&W has described the cyclone combustion process
environment.5 The following paragraphs paraphrase his description;
"The fuel in the cyclone furnace is burned at heat
release rates exceeding 5.2 MW/m3 (500,000 Btu/hr - ft3)
and with gas temperatures of about 1922 K (3000°F).
These temperatures are sufficient to melt the ash into a
liquid slag, a thin layer of which adheres to the walls
of the cyclone. The incoming fuel particles, except
those fines burned in suspension, are thrown to the
walls by centrifugal force and are caught in the running
slag. The secondary air entering the cyclone tangen-
tially sweeps past the coal particles embedded in the
slag surface at high speed. Thus, the air required to
burn the coal is quickly supplied, and the products of
combustion are rapidly removed.
"The products of combustion are discharged through a
water-cooled reentrant throat at the rear of the cyclone
into the boiler furnace. The part of the molten slag
which adheres to the cyclone walls, or impacts on the
boiler target, flows toward the rear of the cyclone and
is discharged through a tap hole into the boiler furnace.
Slag is tapped into a slag tank, solidified, and disinte-
grated for disposal. The part of the molten slag which
does not adhere to the cyclone walls is discharged with
the combustion gases into the boiler furnace.
"The fundamental difference between cyclone furnaces and
pulverized coal-fired furnaces is the manner in which
combustion takes place. In pulverized coal-fired fur-
naces, the particles of coal move along with the gas
stream, and relatively large furnaces are required to
complete the combustion of the suspended fuel. With
cyclonic firing, the coal is held in the cyclone, and
the air is passed over the fuel. Thus, large quantities
of fuel can be fired and combustion completed in a rela-
tively small volume. The boiler furnace (boiler volume
outside of cyclone furnace proper) is used to cool the
5Kessler, G. W. Cyclone Furnace Boilers. The Babcock and Wilcox
Company. (Proceedings of the American Power Conference, 1954).
New York, New York. pp. 78-90.
11
-------
GAS BURNERS
EMERGENCY
STAND BY
OIL BURNER
CRUSHED
COAL INLET
RADIAL
BURNER
4017-2
OIL
BURNER
RE-ENTRANT
THROAT
Figure 2. The cyclone furnace.3
12
-------
products of combustion, and since the temperatures are
high (3000°F), high heat release rates are observed."
Figure 3 gives a better view of the primary and secondary air
inlets which cause the great degree of swirl turbulence within the
cyclone.3 The oil burner is normally retracted after startup when
burning coal.
SECONDARY
AIR
INLET
PRIMARY AIR
INLET
CENTER
OIL BURNER
Figure 3. Furnace schematic.3
Two views4'6 of existing cyclone furnaces are shown in Figures 4
and 5. Figure 4 shows an interior rear view of a cyclone furnace
with a normal accumulation of slag deposits (solidified). Fig-
ure 5 shows a view of the cyclone furnace discharge throat (at
right) and boiler target (at left).
6Grunert, A. E., L. Skog, and L. S. Wilcoxson. The Horizontal
Cyclone Burner. Transactions of the ASME, 69:613-634, August 1947
13
-------
Figure 4. Interior rear view
of cyclone furnace.6
Figure 5. Cyclone furnace throat
(right) and attached
boiler target (left).5
14
-------
3.1.2 Cyclone Furnace/Boiler Arrangements
The part of the steam-generating unit downstream of the primary
cyclone furnace but upstream of the main heat transfer surface is
called a secondary furnace or a boiler furnace. The cyclone fur-
nace is attached to the steam generating boiler in one of two
ways. Figure 6 shows the two possibilities.5 The configuration
on the left is called the screened furnace. It consists of a
slag screen of tubes dividing the boiler furnace into upper and
lower sections. The slag screen divides the boiler furnace into
the primary and secondary components. The second arrangement is
the open furnace, which has no dividing tubes. Either arrange-
ment may be found in existing installations. Space limitations
and heat transfer area required generally dictate which configura-
tion is used. In the screened furnace, the fly ash loading of
the flue gases will be about 10% of the total ash fired in con-
trast to 15% for the open furnace.14
SECONDARY
FURNACE
SCREEENED FURNACE
ARRANGEMENT
OPEN FURNACE
ARRANGEMENT
Figure 6. Furnace arrangements for cyclone-
type primary furnaces.7
One or more cyclone furnaces may be used to fire a single boiler
to achieve the heating capacity desired. Two general firing
arrangements are used, namely, one-wall firing and opposed firing,
A schematic of the firing arrangements is shown in Figure 7 using
open furnaces.2 The furnaces may also be stacked one on top of
another to obtain sufficient firing capacity.
15
-------
BOILER
CYCLONE
FURNACE
BOILER
CYCLONE
FURNACE
ONE-WALL FIR ING
OPPOSED FIRING
Figure 7. Schematic of boiler-firing arrange-
ments using cyclone furnaces.2
3.1.3 Auxiliary Equipment
A variety of auxiliary equipment supports the operation of cyclone-
fired boiler units. This support equipment includes:
• Coal preparation system and feeding arrangement
• Slag handling equipment
• Ash recovery and dust collectors
• Combustion controls
Each category is described in the following subsections.
3.1.3.1 Coal Preparation and Feeding—
B&W describes the types of coal preparation and feeding systems
used:2
"There are two general types of coal preparation and
feeding (see Figure 8), the bin or storage system and
the direct firing system. The former is preferred for
most bituminous coals when the plant layout permits.
The range of sizing of crushed coal required with
either system is given in Figure 9.
"With the bin system, coal is crushed in a central
preparation plant to a size suitable for firing, and
the crushed coal is delivered to the bunker. Because
the crushed coal is relatively large in particle size,
the hazards associated with pulverized coal systems do
not exist. The only precaution necessary is to pro-
vide adequate venting of the bunkers to assure removal
of the small amounts of combustible gases released
from freshly mined coal of certain types. With the
bin system, there is less equipment in the boiler room,
16
-------
CRUSHED COAL-
BUNKER
PRIMARY
AIR
CYCLONE
FURNAC
SLAG TANK
RAW COAL,
BUNKER/
DIRECT-FIR ING SYSTEM
DIRECT-FIRING PRE-DRYING
BYPASS SYSTEM
CONVEYOR
PRIMARY MECHANICAL SLAG TANK
AIR COLLECTOR
Figure 8. Bin, direct-firing, and direct-firing predrying
bypass systems for coal preparation and feeding
to the cyclone furnace (schematic).2
17
-------
SCREEN OPENING
. 03 .05 .1 .2.3.5 1.0 23
400
270 140 50 30 14 4
U.S. STANDARD SIEVE DESIGNATION
Figure 9.
Sizing of crushed coal fired
in the cyclone furnace.
and short crusher outages can be accommodated without
interrupting boiler operation.
"The second method of coal preparation is the direct
firing system, which has a separate crusher located
between the feeder and the burner of each cyclone fur-
nace. The crusher is swept by hot air, and the inti-
mate mixing of coal and hot air in the crusher helps to
dry the coal. This improves crusher performance and
ignition with high moisture coals. It is often easier
to accommodate the direct firing system in existing
plants, where the coal handling equipment cannot
readily be adapted to the bin system.
"The direct firing, predrying bypass system (Figure 8)
is a variation of the second method, incorporating a
mechanical dust collector between the crusher and the
cyclone furnace. The collector is vented to the
boiler furnace. This system is used when firing
extremely high moisture coals. Its advantage is that
moisture is removed from the coal during crushing and
then vented to the boiler furnace instead of the
cyclone furnace. This maintains maximum temperature
in the cyclone with improved performance and slag tap-
ping characteristics.
18
-------
"The coal feeders normally used are of the belt type,
illustrated in Figure 10.2 A rotating distributor is
provided at the coal discharge from the feeder to
assure a continuous and uniform rate of feed. This is
necessary because the coal is burned almost instanta-
neously when it reaches the cyclone furnace, and fluc-
tuations in feed are reflected in combustion conditions,
The rapidity of combustion makes the cyclone furnace
very responsive to load demands, and it has been demon-
strated that boiler output can be made to respond very
quickly to demand by changing coal-feeder speed. Con-
tinuous weighing devices can be applied to the belt
feeder so that it can serve the dual function of coal
scale and feeder.
CONVEYING
BELT
ROTATING
DISTRIBUTOR
COAL INLET
TO FEEDER
i
40,7.5 -COAL DISCHARGE FROM FEEDER
Figure 10. Belt-type coal feeder for the cyclone furnace.2
"Feeders of other types may also be used. Some are
equipped with an angled cutoff plate at the coal dis-
charge from the feeder to provide a uniform rate of
feed."
3.1.3.2 Slag Handling—
A slag handling system for cyclone-fired boiler units is shown in
Figure II.2 Newkirk describes the operation of the Dow Chemical
Company's slag taps and ash pits which is a typical facility:7
"The slag is discharged from the secondary furnace to a
slag chamber on each side of the furnace. Here it is
7Newkirk, M. Cyclone-Fired Pressurized Steam Generator.
actions of the ASME, Journal of Engineering for Power,
73:215-223, 1951.
Trans-
19
-------
BOILER
FURNACE"
TO SLAG
DISPOSAL
r
CYCLONE
FURNACE
SLAG
TANK
Figure 11.
Batch-removal slag-handling sys-
tem for cyclone-furnace boiler.
swept by a stream of hot gases, maintaining a liquefied
state, to the point where the ash is deposited in circu-
lated cool water through a water screen which expedites
the diffraction. The slag is distributed by propeller
agitation to three hopper bottoms. The ejection of the
ash is by oscillating hydrojets with upstream hydrojets
to help. The ash is ejected through a water-operated
slide gate directly to grinder—to sump—to ash pump—
to fill. In the new units, it is deposited through a
pressure door to a sluiceway—to an ash pit for deposi-
tion. "
Besides batch systems for removing slag from the furnaces, many
installations employ a continuous method of slag removal. After
the slag is quenched with water, it is removed immediately by
means of a conveyor belt to the disposal facility.
3.1.3.3 Fly Ash Control and Handling—
Cyclone firing produces a stack dust highly concentrated in fine
material (approximately 85% is less than 10 microns in diameter).
Electrostatic precipitators are most commonly used to collect this
dust efficiently. Disposal of the fines collected is difficult
because it is poor landfill material. As a result, it is rein-
jected into the cyclone furnace and converted into more easily dis-
posable slag.
3.1.3.4 Combustion Control—
A cyclone-fired boiler unit is very adaptable to automatic combus-
tion control. B&W describes the control criteria:2
20
-------
"Automatic combustion controls for cyclone-furnace
boilers are generally based on maintaining equal coal
weights and equal total air flows in the proper propor-
tion to each cyclone furnace. Where volumetric-type
feeders are used, equal coal weights are obtained by
maintaining equal feeder speeds. Where gravimetric-
type feeders are used, they measure and control the
coal weights to the cyclone furnaces.
"Combustion air flow is measured separately to each
cyclone. Where individual ducts supply combustion air
to individual cyclones, a venturi throat in each duct
measures the air to each cyclone. Where cyclones are
installed in a common windbox, secondary air flow is
measured at the bell-mouth section of the secondary air
port of each cyclone, then added to the primary and
tertiary air flows of that cyclone. These flows are
measured at orifices in the individual ducts.
"Using these measurements, the controls maintain equal
coal rates and air flows to each cyclone furnace. The
overall excess air is controlled in the usual manner
with a boiler meter based on steam flow and air flow.
Oxygen recorders are usually provided as operating
guides to monitor the controls."
3.1.4 Examples of Cyclone-Fired Boiler Installations
Figure 12 gives a sectional view of boiler unit No. 20-A at the
Calumet Station of Commonwealth Edison Company in Chicago,
Illinois.6 This unit, incidentally, was the first cyclone-fired
coal-burning installation and went onstream in September 1944.
The unit generates 18.9 to 22.7 kg/s of steam (150,000 to 180,000
Ib/hr). Design pressure of this boiler is 4.14 x 106 Pa (600 psi)
with temperatures reaching 755 K (900°F) in the superheater. It
was originally designed to burn low-grade Central Illinois coal.
Table 1 describes the equipment contained in unit No. 20-A.6
This installation utilizes one cyclone furnace of the one-wall
screened furnace arrangement. This boiler unit is operational
but is presently fired with natural gas rather than coal.
Another example of cyclone-fired installations is the Philo-6
unit owned by the Ohio Power Company. Unlike the Calumet instal-
lation, it is a supercritical steam unit. The supercritical
steam generator increases the thermal efficiency of the steam tur-
bine through the use of higher initial boiler pressure and multi-
ple reheat stages. The Philo-6 has throttle conditions of
31.0 x 106 Pa (4,500 psi) at 894 K (1150°F) with two reheats of
7.9 x 106 Pa (1143 psi) at 839 K (1050°F) with 1.3 x 106 Pa
(192 psi) at 811 K (1000°F).3 The boiler is fired with three
cyclone furnaces in a one-wall, screened furnace arrangement.
21
-------
STACK
INDUCED-DRAFT
FAN
8. F. BOOSTER PUMP
SECONDARY AIR
DUCT TO CYCLONE
TABLE
FEEDER
COAL
CONDITIONER
AIR HEATER
FORCED-DRAFT
FAN
FINAL SUPERHEATER
OUTLET
SECOND SECTION
INLET FROM
ATTEMPERATOR
OUTLET TO
ATTEMPERATOR
COAL
CONVEYOR
REDDER
CONVEYOR
COAL
BUNKER
COAL SCALE
PRIMARY AIR & COAL
INLET
TERTIARY AIR
INLET
Figure 12. Sectional side view of boiler unit No. 20-A, Calumet
Station, Commonwealth Edison Company, Chicago,
Illinois, with horizontal-cyclone-burner firing.6
22
-------
TABLE 1. DESCRIPTION OF EQUIPMENT; UNIT NO. 20-A, CALUMET STATION6
Boiler type B&W radiant water tube
Furnace
Dimensions: Plan. . . . 4.57 m x 2.74 m (15 ft x 9 ft) below 0.58 m (23 in)
drum, 3.58 m x 2.74 m (11 ft 9 in x 9 ft) above 0.58 m (23 in) drum
Volume 178.4 m3 (6,300 cu ft)
Secondary-furnace-wall construction. . .Rear wall tubes below intermediate
0.58 m (23 in) drum; side-wall and front-wall tubes are 8.3 cm
(3 1/4 in) centers
Floor Bailey block construction
Following sections are full-stud construction. . .Front wall below reflect-
ing arch over cyclone discharge; side walls under reflecting arch, and
also to approximately the top of the lower section of platen tubes, and
the rear wall from the furnace floor to approximately the top of the
lower section of platen tubes. The remaining wall sections have flat-
stud construction.
Reflecting arch construction. . . .Tubes 8.3 cm (3 1/4 in) OD, slag-screen
section fully studded and on 34.3 cm (13 1/2 in) centers. Baffle sec-
tion fully studded and arch section studded only on side facing floor
are spaced on 11.4 cm (4 1/2 in) centers. Remaining section above arch,
bare tubes on 34.3 cm (13 1/2 in) centers
Platen tube construction. . .Tubes 8.3 cm (3 1/4 in) OD, spaced on 34.3 cm
(13 1/2 in) centers. Lower section, fully studded. Remaining section,
bare tube
Cyclone-burner construction. . . . Burner walls 3.8 cm (1 1/2 in) OD fully
studded tubes on 5.7 cm (2 1/4 in) centers. Portion of tube section at
secondary air inlet bare with blocks welded between tubes for smooth air
entry. Each tube made in two semicircular sections of 2.44 m (8 ft)
diam. Rear wall and throat formed by fully studded front-wall tubes of
secondary furnace. Slag-tap opening in bottom rear wall. Inlet cone at
front of cyclone, 3.8 cm (1 1/2 in) OD tubes on 4.8 cm (1 7/8 in)
centers, bare tubes with blocks welded on tubes for wear surface.
Burner axis slopes 8.73 x 10~2 rad (5 deg) toward discharge end of fur-
nace.
Superheater type Convection, continuous-tube, pendant
Air heater type Vertical-tubular type enclosed in circular casing
Effective heating surface
Boiler 755.2 8,129
(continued)
23
-------
TABLE 1 (continued).
Effective heating surface m ft
Furnace side walls 182.0 1,959.5
Furnace front walls 95.6 1,029
Furnace rear walls 33.5 360.5
Furnace platens 132.7 1,428
Cyclone burner 29.4 316
Superheater, primary 349.6 3,763
Superheater, secondary 418.7 4,507
Attemperator 47.2 508
Desuperheater 47.2 508
Air heater 2,633 28,346
Induced-draft fan
Type American blower
Capacity 61.4 m3/s (130,000 cfm) at 450 K (350°F), 2.24 kPa (9 in)
static pressure
Drive Two-speed motor, 360 and 705 rpm, 261 kW (350 hp) , 2300 V
Volume and pressure control Damper on discharge duct
Forced-draft fan
Type B.F. Sturtevant Co. , No. 250, 365 compressor
Capacity 28.3 m3/s (60,000 cfm) at 300 k (80°F), 15.4 kPa (62 in)
static pressure at 3557 rpm
Drive Constant speed, 746 kW (1000 hp) , 2300 V
Volume and pressure control Adjustable inlet vane
Raw-coal scales
Type Richardson automatic
Capacity 136 kg (300 Ib) per dump—2.3 kg/s (9 tons/hr)
Boiler conveyor
Capacity 3.8 kg/s (15 tons/yr)
Feeder
Type B&W table
(continued)
24
-------
TABLE 1 (continued).
Capacity 3.1 kg/s (12 1/2 tons/yr)
Drive Variable-speed, d-c motor
Coal crusher (Original design, 24 in diam)
Type -.. BSW hammer
Capacity *• 3.1 kg/s (12 1/2 tons/hr)
Crusher speed 1140 rpm
Number of hammers 30
Classifier Perforated plate
Coal crusher (Second design, 48 in diam)
Type B&W hammer
Capacity:
Crusher speed Variable
Number of hammers 240
Classifier Grid
The three-cyclone boiler unit is designed to produce 85 kg/s of
steam (675,000 Ib/hr).
Figure 13 illustrates the general layout of the Philo-6 installa-
tion.4 The supercritical steam unit design is not typical of the
cyclone boiler population since only about seven supercritical
units are in operation. The Calumet installation is more repre-
sentative of the units in operation.
3.1.5 Fuel Requirements
3.1.5.1 Coal—
The cyclone furnace was developed to burn a particularly trouble-
some Illinois coal. Table 2 gives typical analysis data for this
coal.8 This is a high-ash content, low-ash fusion temperature
coal which is rather difficult to burn in stokers or dry-bottom
pulverized coal units.
8Stone, V. L. and I. L. Wade. Operating Experiences with Cyclone'
Fired Steam Generators. Mechanical Engineering. 74:359-368,
1972.
25
-------
10
Figure 13. Philo-6 once-through supercritical Benson
boiler installation (Ohio Power Co.).4
-------
TABLE 2. TYPICAL ANALYSES OF CENTRAL ILLINOIS COAL AND ASH8
Proximate, %
Moisture 14.0
Volatile matter 34.0
Fixed carbon 37.8
Ash 14.2
TOTAL 100.0
Ultimate, moisture and ash-free, %
Sulfur 6.48
Hydrogen 5.43
Carbon 77.16
Nitrogen 1.39
Oxygen 9.54
Chloride as NaCl, % on dry basis 0.37
Heating value, as fired, MJ/kg (Btu/lb) 23.4
(10,050)
Ash, % by weight
Silica, SiO2 40.43
Alumina, A1203 19.98
Iron oxide, Fe2O3 25.11
Calcium oxide, CaO 5.08
Magnesium oxide, MgO 1.09
Sulfuric anhydride, SO^ 5.19
Phosphorus pentoxide, P205 0.01
Alkalies, sodium and potassium oxides, Na2O and K20 3.04
Chloride, Cl 0.07
TOTAL 100.00
Water-soluble constituents in ash, % by weight
Ferrous sulfate, FeSOij 0.30
Sodium sulfate, NaSO^ 1.37
Calcium sulfate, CaSO^ 4.87
Magnesium sulfate, MgSO^ 1.43
Sodium chloride, NaCl 0.05
TOTAL 8.02
Coal ash-fusing temperatures (reducing atmosphere), K (°F)
Initial deformation
Softening
Fluid
1355
(1980)
1380
(2025)
1555
(2340)
27
-------
As mentioned previously, the cyclone furnace can also burn a wide
variety of other coals. There are, however, some restrictions to
the types of coals which can be successfully fired in the cyclone
furnace. The key to successful firing is to maintain a noncorro-
sive and fluid ash state in the furnace throughout the boiler
load range. The suitability of coals is thus dependent on the
moisture, ash, and volatile contents of the coal together with the
chemical composition of the ash.2
B&W states that the volatile matter should be higher than 15% (dry
basis) to obtain the required high combustion rate. Ash content
should be at least 6% to provide a proper slag coating in the fur-
nace and can be as high as 25% (dry basis). A wide range of mois-
ture contents is permissible depending on coal rank, secondary air
temperature range for drying, and fuel preparation equipment.2
The two most important factors which determine the suitability of
a coal for cyclone firing are slag viscosity or fluidity and tend-
encies of coal to form corrosive iron and iron sulfide. A fluid
slag layer in the furnace is desirable for proper combustion (i.e.,
bulk of coal burns in slag layer), ease of tapping, and minimal
ash accumulation. Formation of iron and iron sulfide in the fur-
nace can result in catastrophic tube failure. A further discus-
sion of these two factors is presented below.
Regarding slag viscosity, B&W indicates that it must be 25 Pa»s
(250 poises) or below at a temperature of 1700 K (2600°F). In
addition, the ash softening temperature must be 1600 K (2500°F) or
below when tested in a reducing atmosphere. The ash softening
point is that temperature at which an ash cone when heated has
fused down to a spherical lump (as per American Society for Test-
ing Materials—ASTM Standard Method D-1857).
Hot melt slag viscosity is often determined with a special viscome-
ter. When hot melt slag viscometer data are not available, a use-
ful correlation developed by Hay and Roberts can be used.9 This
correlation obtains a relative indication of slag viscosity at
1700 K (2600°F) based on ash composition.
The correlation states that if the Equivalent Silica Content
(often called Silica Ratio) is below a value of 72, the coal will
have a viscosity below 25 Pa«s (250 poises) at 1700 K (2600°F).
The Equivalent Silica Content (ESC) is defined as:
9Lowry, H. H. (ed). Chemistry of Coal Utilization. Supplementary
volume, prepared by the committee on Chemistry of Coal, Division
of Chemistry and Chemical Technology, National Academy of Sci-
ences - National Research Council. John Wiley and Sons, Inc.,
New York, New York, 1963. 1,142 pp.
28
-------
= Si02 x 100 (1)
SiO2 + Fe203 + MgO + CaO
where SiC>2 = weight percent of SiO2 in coal ash
Fe20s = weight percent of equivalent Fe2C>3 in coal ash
MgO = weight percent of MgO in coal ash
CaO = weight percent of CaO in coal ash
ESC = Equivalent Silica Content, %
Slagging tests performed by B&W on coals for cyclone furnaces indi-
cate that a wide range of coals meet the viscosity criteria. Fig-
ure 14a shows the results of these tests.5 However, to be truly
suitable, the coal must not have a marked tendency to form iron
and iron sulfide. Figure 14b indicates2 that the coal ash Fe2O3/
(CaO + MgO) ratio (termed CAR) and the percent sulfur in coal are
most important in determining this tendency. Corrosion is dis-
cussed in more detail in Section 3.1.6.
Table 3 lists analysis data10 for five coals typically suited for
firing in cyclone furnaces. These coals meet the two most impor-
tant criteria (low slag viscosity and minimal corrosion tendencies)
as well as thef other related suitability factors (ash, volatile
matter, and moisture content). Table 4 gives the range of proxi-
mate analyses of coals which have been successfully fired in
cyclone furnaces.2
In summary, criteria were presented which indicate the general
suitability of a coal for cyclone firing. A suitable analysis
alone is necessary but not sufficient to determine the combustion
characteristics of a specific coal. In practice, once a coal is
judged suitable, it then undergoes boiler firing tests which deter-
mine operating difficulties. Examples of difficulties that might
be encountered when changing to a suitable western coal include:
high combustible carbon carryover as a result of poor combustion;
need for increased coal feeding and drying capacity; and increased
crusher wear due to higher coal rates.
3.1.5.2 Other Fuels—
Natural gas and residual or crude oils are the most commonly used
alternate fuels in cyclone fired boilers. Wood bark, coal chars,
petroleum coke, and fuel oil may also be satisfactorily fired if
economics or other needs dictate. No actual instances of these
types of firings were found in the literature. Bark, chars, coke,
and certain coals may require supplemental fuel for successful fir-
ing. Firing fuels other than coal in cyclone furnaces is not nor-
mally competitive with other firing methods unless coal is the
principal fuel.2
10Selvig, W. A. and F. H. Gibson. Analyses of Ash from United
States Coals. Bull. No. 567. Bureau of Mines, U.S. Department
of the Interior. 1956. 33 pp.
29
-------
1538
2800
1700
2600
1589
2400
1463
2200
0- o
5 Q-
1366
2000
BORDERLINE
®
©
©
®
o
©
0
1366
2000
146S
2200
15S9
2400
1700
2600
ASH SOFTENING TEMPERATURE
(REDUCING ATMOSPHERE)
©©©
©00©
0©@
©
®@®@
COALS TESTED
CENTRAL PENNSYLVANIA
WESTERN PENNSYLVANIA
NORTHERN WEST VIRGINIA
OHIO
SOUTHERN WEST VIRGINIA
VIRGINIA
EASTERN KENTUCKY
WESTERN KENTUCKY
ILLINOIS
INDIANA
TENNESSEE
TEXAS (LIGNITE I
NORTH DAKOTA (LIGNITE I
(a) Results of slagging tests of coals for cyclone
furnaces5 (590 K combustion air temperature).
5?
30
D
t 25
c
o
en
J 20
o
ro
15
o
!• I0
5
(i
SI
\
JITABL
\
E
NOT
N
\
5UITAI
\
!LE
23456789
TOTAL SULFUR IN COAL, %'- DRY BASIS
10
(b) Coal suitability for cyclone furnaces based
on tendency to form iron and iron sulfide.2
Figure 14. Coal suitability factors.2/5
30
-------
TABLE 3. TYPICAL COALS SUITABLE FOR CYCLONE FIRING
10 '
State
County
Bed
Mine
Rank
Heating value, MJ/kg (Btu/lb)
Ash softening temperature, K (°F)
Equivalent silica content (ESC)a
Coal ash ratio (CAR)a
Sulfur, wt %
Illinois
Williamson
No. 6
No. 12
Bituminous
26.2
(11,260)
a 1,405
(2,070)
63.0
2.30
3.60
Montana
Carbon
No. 3
North Side
Subbituminous
25.7
(11,030)
1,400
(2,060)
47.0
1.30
1.70
Proximate analysis, wt % (as
Moisture
Volatile matter3
Fixed carbon
Ash
Ash (moisture free)3
Si02
A1203
Fe203
TiO2
P205
CaO
MgO
Na2O
K20
SO 3
9.60
33.4
44.4
12.6
13.9
Ash
41.2
15.9
23.1
0.80
0.12
9.40
0.40
0.60
1.90
7.40
10.1
34.5
46.7
8.70
9.70
analyses, wt
30.7
19.6
18.9
1.10
-
11.3
3.70
1.90
0.50
12.2
North Dakota
Mercer
Zap
Indian Head
Lignite
16.9
(7,250)
1,528
(2,290)
31.0
0.60
1.00
received basis)
33.6
28.4
31.1
6.90
10.3
%
17.5
9.90
15.0
-
-
17.8
5.60
-
-
28.3
Pennsylvania
Fayette
Pittsburgh
Banning No. 2
Bituminous
31.4
(13,520)
1,466
(2,180)
62.0
4.40
2.20
1.90
33.5
54.8
9.80
10.0
43.4
24.0
21.2
1.10
0.30
4.10
0.70
0.90
0.80
3.70
Kentucky
Muhlenberg
Nos. 9 and 11
Mixture from
two mines
Bituminous
26.6
(11,440)
1,389
(2,040)
60.0
3.10
3.90
6.10
36.1
43.6
14.2
15.2
44.0
18.3
21.7
0.90
0.34
6.20
0.80
-
-
6.4
Relates to coal suitability.
-------
TABLE 4. RANGE OF COAL ANALYSES2
ComponentPercent by weight
Moisture 2 to 40
Volatile matter (dry) 18 to 45
Fixed carbon (dry) 35 to 75
Ash (dry) 4 to 25
Typical compositions and heating values for natural gas and oils
are shown in Tables 5 and 6, respectively.2'11 Residual oil with
low sulfur and low vanadium contents is normally desired so as to
minimize boiler corrosion. Fuel oil is not commonly used due to
its high cost.
Wood bark may also be burned in the cyclone furnace. The analyses
and heating value of bark varies significantly. Table 7 gives
some typical analyses of wood bark and ash.2
Char, the nonvolatile portion of coal which results from certain
gasification processes, can be successfully burned in cyclone fur-
naces. Table 8 gives analyses of chars from coal conversion pro-
cesses. 11
Petroleum coke can also be burned in cyclone furnaces. Table 9
gives a typical range of analyses for the two most commonly occur-
ring petroleum cokes which originate from the delayed coke process
and the fluid coke process.12
3.1.6 Corrosion
The cyclone furnace boiler unit does not appear to exhibit any
unusual or peculiar corrosion characteristics. The axially fired
furnace design combined with the studded furnace wall sections are
the furnace characteristics to which low tube corrosion rates are
usually attributed. The integrity of the studded furnace wall
sections indicate when preventive furnace maintenance is to be
performed. However, furnace tube wastage as well as superheater
and reheater tube corrosion has occurred in several installations.
It is reasonable at this point to discuss corrosion characteris-
tics of cyclone furnaces because of the possible connection
between corrosion and combustion modifications for NO control.
X
lxPerry, R. H. and C. H. Chilton (eds). Chemical Engineers'
Handbook. 5th edition. McGraw-Hill Book Company, New York,
New York, 1973. 1,650 pp.
12Perry, R. H., C. A. Chilton, and S. O. Kirkpatrick (eds).
Chemical Engineers' Handbook, 4th edition.' McGraw-Hill Book
Company, New York, New York, 1963. 1,650 pp.
32
-------
TABLE 5. SELECTED SAMPLES OF NATURAL GAS FROM UNITED STATES FIELDS.2
U)
Analyses
Constituents, vol %
H2 Hydrogen
CH2 Methane
C2Hit Ethylene
C2H6 Ethane
CO Carbon monoxide
C02 Carbon dioxide
N2 Nitrogen
O2 Oxygen
H2S Hydrogen Sulfide
Ultimate, wt %
S Sulfur
H2 Hydrogen
C Carbon
N2 Nitrogen
O2 Oxygen
1
Pennsylvania
—
83.40
-
15.80
_
—
0.80
-
-
_
23.53
75.25
1.22
-
Sample
Source
2
Southern
California
-
84.00
- .
14.80
-
0.70
0.50
-
—
—
23.30
74.72
0.76
1.22
number
of gas
3
Ohio
1.82
93.33
0.25
-
0.45
0.22
3.40
0.35
0.18
0.34
23.20
69.12
5.76
1.58
4
Louisiana
-
90.00
-
5.00
-
-
5.00
-
—
-
22.68
69.26
8.06
—
5
Oklahoma
-
84.10
—
6.70
—
-
8.40
-
—
-
20.85
64.84
12.90
1.41
Specific gravity
(relative to air)
Higher heat value
0.636
0.636
0.567
0.600
289 K, 101.3 kPa
J60°F, 30 in. Hg
0.630
MJ/m3 a
Btu/cu ft
MJ/kg of fuel
Btu/lb of fuel
42.1
1,129
53.9
23,170
41.6
1,116
53.3
22,904
35.9
964
51.4
22,077
37.4
1,002
50.8
21,824
36.3
974
46.9
20,160
-------
u>
TABLE 6. TYPICAL ULTIMATE ANALYSES OF PETROLEUM FUELS11
Composition,
%
Carbon
Hydrogen
Oxygen
Nitrogen
Sulfur
Ash
C/H ratio
No. 1
fuel oil
(41.5° A.P.I.)
86.4
13.6
0.01
0.003
0.09
<0.01
6.35
No. 2
fuel oil
(33° A.P.I.)
87.3
12.6
0.04
0.006
0.22
<0.01
6.93
No. 4
fuel oil
(23.2° A.P.I.)
86.47
11.65
0.27
0.24
1.35
0.02
7.42
Low sulfur ,
No. 6
fuel oil
(12.6° A.P.I.)
87.26
10.49
0.64
0.28
0.84
0.04
8.31
High sulfur,
No. 6
fuel oil
(15.5° A.P.I.)
84.67
11.02
0.38
0.18
3.97
0.02
7.62
-------
TABLE 7. ANALYSES OF WOOD AND WOOD ASH2
Wood analyses
(dry basis) , wt %
Proximate
Volatile matter
Fixed carbon
Ash
Ultimate
Hydrogen
Carbon
Sulfur
Nitrogen
Oxygen
Ash
Heating value, MJ/kg
(Btu/lb)
Ash analyses, wt %
Si02
Fe203
Ti02
A1203
yin30i+
CaO
MgO
Na20
K20
S03
Cl
Ash fusibility, K (°F)
Reducing
Initial deformation
Softening
Fluid
Oxidizing
Initial deformation
Softening
Fluid
Pine
bark
72.9
24.2
2.9
5.6
53.4
0.1
0.1
37.9
2.9
21.0
(9,030)
39.0
3.0
0.2
14.0
Trace
25.5
6.5
1.3
6.0
0.3
Trace
1466
(2180)
1500
(2240)
1539
(2310)
1483
(2210)
1522
(2280)
1561
(2350)
Oak
bark
76.0
18.7
5.3
5.4
49.7
0.1
0.2
39.3
5.3
19.5
(8,370)
11.1
3.3
0.1
0.1
Trace
61.5
1.2
8.9
0.2
2.0
Trace
1750
(2690)
1766
(2720)
1778
(2740)
1744
(2680)
1772
(2730)
1783
(2750)
Spruce
barka
69.6
26.6
3.8
5.7
51.8
0.1
0.2
38.4
3.8
20.3
(8,740)
32.0
6.4
0.8
11.0
1.5
25.3
4.1
8.0
2.4
2.1
Trace
Redwood
bark3
72.6
27.0
0.4
5.1
51.9
0.1
0.1
42.4
0.4
19.4
(8,350)
14.3
3.5
0.3
4.0
0.1
6.0
6.6
18.0
10.6
7.4
18.4
Salt-water stored.
35
-------
TABLE 8. EXAMPLES OF ANALYSES OF COAL FEEDS AND RESULTING CHARS FROM VARIOUS COAL-CONVERSION PROCESSES.11
ui
Process
Coal bed
Composition and properties
Analysis, wt %
Volatile matter
Fixed carbon
Ash
Sulfur
Heating value, MJ/kg
(Btu/lb)
FMC
Pittsburgh-Federal
Coal , dry
basis
36.8
57.0
6.2
2.9
33.6
(14,470)
Char , dry
basis
3.7
86.8
9.5
1.9
31.1
(13,400)
a
Illinois
Coal , dry
basis
38.6
50.0
11.4
3.8
29.3
(12,600)
No. 6
Char , dry
basis
3.5
76.4
20.1
3.1
27.6
(11,870)
IGTb
Pittsburgh
Coal, dry Char, dry
basis basis
32.7 1.2
52.3 77.5
14.1 21.3
4.3 1.7
30.7 28.4
(13,200) (12,200)
FMC process involves multistage fluidized-bed pyrolysis of coal to produce a liquid, residual char,
and some gas.
IGT process involves hydrogasification of coal to produce a gas of pipe-line quality (about 1,000
Btu/cu ft) and char.
-------
TABLE 9. TYPICAL ANALYSES OF PETROLEUM COKES12
Delayed-
process coke,
range
Fluid-
process coke,
range
Volatile, wt !
Ash, wt %
Bulk density,
True density,
Heating value
'•>
kg/m3
kg/m3
as received,
8 to 13
0.05 to 1.6
-
1.28 to 1.42
-
MJ/kg (Btu/lb)
Hydrogen, wt 5
Carbon, wt %
Sulfur, wt %
Ash-softening
Ash- softening
!>
temp , K
temp, °F
—
—
-
-
^
3.7
0.1
881
1,500
32.3
(13,900
1.6
88
1.5
1478
2200
to
to
to
to
to
to
to
to
to
to
to
7.0
2.8
1,041
1,600
33.5
14,400)
2.1
95
10.0
1811
2800
The only available corrosion information was for coal fuel. Cor-
rosion susceptibility of coal-fired units appears to stem not
from design or construction flaws but rather from improper fuel
selection or boiler operation (sometimes both).
During the course of this study, personnel at both B&W and Common-
wealth Edison Company (Chicago), an electric utility operating 20
cyclone boilers, were asked to describe any common corrosion prob-
lems. The consensus of opinion was that significant tube corro-
sion can occur when the furnaces are running at low excess air
levels. The localized reducing atmospheres within the furnace
aggravate the tendencies of a. coal to form iron and iron sulfide.
These species then attack the tubes and, if not corrected in time,
can cause catastrophic failure. Normally, careful control of com-
bustion conditions alleviates this problem. A bad batch of coal
that has inherent tendencies to form iron and iron sulfide can
also cause tube corrosion even more rapidly at a low excess air
level. Primarily for these reasons, B&W does not support or
recommend any N0y combustion modification that would result in
reducing conditions within the furnace.
Dow Chemical Company (Midland, Michigan) has experienced tube
failures in its cyclone furnaces, but from a different cause.
M. Newkirk of Dow summarizes his findings:7
"The major cause of tube failure in the cyclone has been
due to the up and down firing brought about by numerous
interruptions of coal feed. This intermittent operation
causes a wide variation in temperature which results in
expansion and contraction of the cyclone. Consequently,
the slag cracks and peels off leaving parts of the
throat and the cyclone tubes bare. Since the heat
37
-------
release in the cyclone burner itself is very great
(5.6 MW/m3 [545,000 Btu/hr - ft3]), extreme heat-
transfer is effected in this localized area. The high
temperature damages the protective film of iron oxide
which, under favorable conditions, covers the inside of
the tubes. Any condition which damages the protective
film permits corrosion to continue. It is this repeti-
tive process of forming an iron oxide coating that robs
the iron from the tubes. When this happens to a
limited area, the attack will be localized and eventu-
ally will result in tube failure."
Thus, a smooth, continuous firing operation with a minimal number
of shutdowns is desirable to avoid catastrophes.
Catastrophic failure of superheater and reheater tubes occurred
in both the Ridgeland and Will County Stations of Commonwealth
Edison (Chicago, Illinois).9'13 The high metal temperatures of
the superheaters (approximately 1478 K, [2200°F]) combined with
the high sulfur content (4.6%) and high alkali content of the
coal (0.63% as Na20) was determined to be the cause of the severe
corrosion. Reduction of gas temperatures and subsequent tube
metal temperature along with some minor mechanical modifications
reduced the corrosion significantly. Coals having both high ash
alkalinity and high sulfur should be avoided when operating at
high steam temperatures.
3.1.7 Advantages/Disadvantages of Cyclone Furnaces/Boilers
The cyclone firing method reduces the amount of ash passing
through the boiler and results in uniform and complete combustion
of the crushed coal. These and other operating advantages of
cyclone furnace firing with coal are summarized below:2'3'5'6
• Excess air requirement is low (10% to 17%), and carbon
loss is low. As a result, full-load boiler efficiency
is higher (88% to 90%) than that for stoker (66% to
80%) or pulverized coal-fired units (85% to 88%).
Some cyclone units, however, operate with excess air
levels as high as 42%. This is so mainly with older
boiler units to prevent local reducing conditions and
consequent boiler damange.
• The amount of ash going through the flue is about 10%
to 15% of that fired compared to 50% to 85% for pul-
verized coal firing. This ultimately results in
13Sedor, P., E. K. Diehl, and 0. H. Barnhart. External Corrosion
of Superheaters in Boilers Firing High-Alkali Coals. Transac-
tions of the ASME, Journal of Engineering for Power, 82:181-190,
1960.
38
-------
reduced slag formation on heat absorption surfaces
and reduced particulate air pollution.
• Fly ash collected in dust collecting equipment can
be reinjected so that all ash removed from the boiler
will be a granular, chemically inert slag (as a
result of molten slag quenching). This ash consist-
ency is easier to use for landfill.
• Furnace size is reduced because of reduced combustion
gas residence time requirements.
• There are savings in the cost of fuel preparation
since only crushing is required instead of pulveriza-
tion.
• The cyclone furnace can handle a wide variety of
coals and is easily adapted to firing gas and coal.
• Operation of the cyclone furnace is simple and
reliable.
• Stack dust from cyclone furnace boilers is much finer
than that from pulverized-coal-fired units. About
85% of the dust from cyclone-fired boilers is less
than 10 vim in size compared with about 30% for pulver-
ized units at the same stack loading. Thus, erosion
of boiler internals is greatly reduced using cyclone
firing.
Along with the decided advantages of cyclone-fired boiler units,
there are disadvantages to burning coal in cyclone furnaces
besides the high level of NOX emissions (see Section 3.3). The
pressure drop across the cyclone furnace is rather high, ranging
from 5.0 to 10.0 kPa (20 to 40 in. of water) depending on load.
This requires a rather powerful forced draft fan to maintain the
high volume and velocity of secondary air necessary for cyclone
combustion. The fan power accounts for over 90% of the total aux-
iliary power requirement. The auxiliary power requirements for
cyclonic firing are less than those of pulverized coal units fir-
ing low-grindability (difficult to pulverize), low-heating-value
coals, and greater than pulverized units firing high-grindability
(less difficult to pulverize), high-heating-value coals.5
Figure 15 compares auxiliary power requirements of typical
cyclone furnace boilers and pulverized dry bottom boiler units as
a function of grindability and heating value of coals fed to pul-
verized units. The decided advantage of cyclone furnace boilers
over pulverized dry bottom boilers burning coals with heating
values below 22.3 MJ/kg (9,600 Btu/lb) and grindabilities below
50 (Hardgrove Grindability Index) is apparent from Figure 15.
39
-------
PULVERI ZED-COAL-FIRED
(DRY ASH) BOILERS
All h BITUMINOUS -j SUB- LIGNITE
COALS BITUM
HEATING VALUE. BTU/LBIAF) 14,000 13,000 12,600 10,000 9,600 6,800
GRINDABILITY 100 65 58 55 50 50
^FDFAN CU PRIMARY AIR FAN
!Hi CRUSHER IB PULVERIZER
1 Btu / Ib = 4.299x ID J / kg
IBtu
1 kWhr = Q.278J
9.471x 10"4J
Figure 15.
Auxiliary power requirements of typical
high-capacity pressure-fired cyclone-
furnace and pulverized-coal units.2
Holyoak of Commonwealth Edison has described problems encountered
when burning western coals in cyclone-fired utility boilers not
specifically designed for them.1U One of the more serious prob-
lems encountered has been achieving proper combustion in cyclone-
fired boilers. In Holyoak's study, several low-sulfur Montana
coals were tested. Very high carbon carryover was experienced
when burning straight western coal. Adding 15% to 25% Illinois
coal to the western coals somewhat reduced carbon loss.
During initial tests at the Will County Station on Unit 2, Holyoak
indicated that only 80% to 90% of full load could be achieved
without excessive carbon carryover.14 Exceeding maximum load in
14Holyoak, R. H. Burning Western Coals in Northern Illinois.
Commonwealth Edison Company, Chicago, Illinois, ASME Paper
73-WA/FU-4, August 17, 1973. 8 pp.
40
-------
this unit resulted in carbon carryover so heavy that ash conveying
systems could not handle the quantity produced. Ash samples at
maximum loads had a heat content of 27.9 MJ/kg (12,000 Btu/lb).
Two serious air heater fires and extensive precipitator damage
resulted from the high carbon losses.
Holyoak also indicates that load cannot be regulated while burning
western coals without bursts of carbon carryover. During load
swings, the coal was not burning in the cyclone; it was carried
through the cyclone and boiler without burning.
The carbon carryover problems were postulated by Holyoak to be
caused by the "nonwetting" characteristics of western coal slag in
the furnaces. The cyclone combustion depends greatly on burning
the bulk of coal on the sticky slag layer. It is difficult and
impractical to increase wettability of western coal slag (Holy-
oak's conclusion). Equipment changes were made in the boilers to
provide more retention time in the cyclone along with better coal
fineness and higher temperature to shorten the combustion process.14
The 1973 Holyoak study summarized that there were many operating
problems and loss of capacity as well as the possibility of major
equipment damage from fire or from reducing atmospheres in combus-
tion spaces. Since then, many of the problems evidently have been
solved or at least temporarily abated because western coal is the
primary fuel in at least four Commonwealth Edison stations (Wauke-
gan, Fisk, State Line, and Joliet) .
The fineness of fly ash resulting from cyclone combustion may con-
stitute a serious health hazard. A large fraction of stack dust
may be in the respirable size range below 3 ym in diameter.
Information or quantitative data on this aspect of cyclone combus-
tion were not available. This subject requires further investiga-
tion.
3.2 POPULATION
The first full-scale cyclone-furnace-fired boiler unit was placed
on-stream in 1944 at the Calumet Station (Calumet, Illinois) of
the Commonwealth Edison Company based in Chicago, Illinois. Since
then, a total of 84 cyclone-fired installations have been built in
the United States. These installations are located in 26 states
and contain a total of 149 boiler units fired by a total of 736
cyclone furnaces generating approximately 26,000 kg/s of primary
steam (2 x 108 Ib/hr). Figure 16 shows the geographical distribu-
tion of boiler units and indicates that the bulk of boilers and
primary steaming capacity are in the states of Illinois, Missouri,
and Indiana. These three states account for nearly half of the
total cyclone steaming capacity and one-third of the boilers.
Table 10 gives a further breakdown of cyclone-fired boiler popula-
tion. It shows that over 94% of the total primary steaming
41
-------
to
Figure 16. States with cyclone-fired boiler units showing number of boilers and per-
cent of total U.S. primary cyclone steaming capacity (149 boilers gener-
ating 2.6 x 104 kg/s steam (data courtesy of the Babcock & Wilcox Co.).
-------
TABLE 10. STATE-BY-STATE POPULATION DISTRIBUTION OF CYCLONE-FIRED BOILERS
Electric utility units .
Number
of
State boilers
Alabama
Arkansas
Connecticut
Florida
Illinois
Indiana
Iowa
Kansas
Kentucky
Maryland
Michigan
Minnesota
Missouri
Nebraska
New Hampshire
New Jersey
New York
North Carolina
North Dakota
Ohio
Pennsylvania
South Carolina
South Dakota
Tennessee
West Virginia
Wisconsin
TOTALS
0
1
3
4
33
10
3
3
4
2
1
2
12
2
2
7
0
0
4
7
0
0
2
3
4
7
116
Number
of
cyclones
0
8
13
14
196
56
9
22
54
8
7
17
79
6
10
28
0
0 .
43
29
0
0
13
21
20
24
677
Primary steam
i flow, kg/s
0.0
289.8
425.3
539.4
6,454.2
2,551.4
290.5
906.4
2,375.1
342.6
264.6
673.5
3,058.7
195.2
392.5
931.6
0.0
0.0
1,003.5
943.9
0.0
0.0
436.0
756.0
726.9
696.2
24,253.3
(94.3% of total)
Percent
of
total
0.0
1.2
1.7
2.2
26.6
10.5
1.2
3.7
9.8
1.4
1.2
2.8
12.6
0.8
1.6
3.8
0.0
0.0
4.1
3.9
0.0
0.0
1.8
3.2
3.0
2.9
100.0
Industrial
Number
of
boilers
2
2
0
0
0
2
2
0
0
1
6
0
0
0
0
1
4
1
0
2
2
1
0
0
1
6
33
and commercial units
Number
of Primary steam
cyclones flow, kg/s
4
4
0
0
0
2
3
0
0
1
12
0
0
0
0
1
8
1
0
4
5
2
0
0
2
10
59
113.4
113.4
0.0
0.0
0.0
46.6
76.3
0.0
0.0
15.8
307.4
0.0
0.0
0.0
0.0
27.1
220.5
18.9
0.0
124.1
100.8
37.8
0.0
0.0
50.4
204.7
1,457.2
(5.7% of total)
Percent
of
total
7.8
7.8
0.0
0.0
0.0
3.2
5.2
0.0
0.0
1.1
21.1
0.0
0.0
0.0
0.0
1.8
15.1
1.3
0.0
8.5
6.9
2.6
0.0
0.0
3.4
14.2
100.0
Total
number of
boilers
2
3
3
4
33
12
5
3
4
3
7
2
12
2
2
8
4
1
4
9
2
1
2
3
5
13
149
Total
number of
cyclones
4
12
13
14
196
58
12
22
54
9
19
17
79
6
10
29
8
1
43
33
5
2
13
21
22
34
736
Total primary Percent
steam flow, of U.S.
kg/s total
113.4
403.2
425.3
539.4
6,454.2
2,598.0
366.8
906.4
2,375.1
358.4
572.0
673.5
3,058.7
195.2
392.5
958.7
220.5
18.9
1,003.5
1,068.0
100.8
37.8
436.0
756.0
777.3
900.9
25,710.5
0.4
1.6
1.6
2.2
25.1
10.3
1.4
3.5
9.2
1.4
2.2
2.6
11.9
0.8
1.5
3.7
0.8
0.1
3.9
4.2
0.4
0.2
1.7
2.9
3.0
3.5
100.0
Data courtesy of the Babcock and Wilcox Company.
-------
capacity is held by the electric utility sector (24,253 kg/s)
which operates 116 of the 149 boilers. These boilers are fired
by 677 furnaces. The 33 remaining boiler units are owned by
private industry and institutions. Industries employing cyclone-
fired units include pulp and paper manufacturers, chemical and
steel producers, and one glass manufacturer. Several large mid-
western universities employ cyclone firing to meet their utility
demands. Primary steam generating capacities of individual
boiler units built range from 16 to 70 kg/s (127,000 to 555,000
Ib/hr) for industrial and commercial units and from 23 to 1,160
kg/s (182,000 to 9,200,000 Ib/hr) for electric utility units.
Tables A-l and A-2, Appendix A, give a detailed listing of all 84
cyclone-fired installations organized by type, state, customer,
and size of installation.
Although the statistical population information indicates that
149 boiler units were erected since the inception of cyclone
firing, it is difficult to determine exactly the number of units
that are currently in operation. Some units may be at the end of
their useful life span and might well be in the process of being
replaced. This information problem was discussed with B&W, the
sole manufacturer of cyclone units in the United States. They
estimate that the majority of boilers are still in use, but some
may have been derated because of their age.
Since their inception in 1944, cyclone-fired boilers have sold
well. The technology was able to meet the demands of boiler
owners who wished to burn low-quality coals with low ash fusion
temperatures. In the 1950"s, 1960's, and early 1970"s, cyclone
boilers accounted for a major portion of B&Ws total sales. How-
ever, since about 1973, B&W has not sold a single cyclone unit.
The decline of sales started with the strict federal SOX regula-
tions imposed on new stationary sources [New Source Performance
Standards(NSPS)]. The low ash fusion coals burned in the cyclone
boiler normally have high sulfur content. Switching to low-
sulfur coals normally results in ash with a high fusion tempera-
ture. The higher ash fusion temperature coals cause slag tapping
and corrosion difficulties with cyclone boilers. Thus, a balance
could not, in general, be obtained between low SO- emissions and
adequate boiler operating characteristics. The final event which
restricted the sale of bituminous coal-fired cyclones was the
limitation of NO emissions as per the NSPS.15 These ITew Source
Performance Standards for NOX are given below:16
15Federal Register. 36 (247) :24879 , December 23, 1971.
16Shimizu, A. B., R. J. Schreiber, H. B. Mason, et al. NOX Combus-
tion Control Methods and Costs for Stationary Sources. Summary
Study. Aerotherm Division, Acurex Corporation, U.S. Environ-
mental Protection Agency, EPA 600/2-75-046, September 1975.
104 pp.
44
-------
Gas Oil Bituminous coal
86.1 ng/J 129.1 ng/J 301.2 ng/J
0.2 lb/106 Btu 0.3 lb/10g Btu 0.7 lb/10g Btu
^160 vppm ^225 vppm ^500 vppm
(3% O2 basis) (3% O2 basis) (3% O2 basis)
Cyclone firing results in the highest NOX production of any
coal-firing method and is the most difficult to control in this
regard. Presently, boilers burning lignite are exempt from the
NOX standard. EPA is expected to propose a standard of 257.9 ng
NOX/J (0.6 Ib NOX/106 Btu) for lignite-fired utility boilers dur-
ing the latter part of 1976.
3.3 BASELINE EMISSIONS FROM UNMODIFIED CYCLONE FURNACE INSTALLA-
TIONS
Baseline emissions are defined to be those NOX, SOX, CO, and par-
ticulate emissions reflecting normal or near-normal boiler opera-
tion at various loads. The available data on baseline emissions
from cyclone furnaces are included in this section. Some data
were simply "spot checks" of boiler operation and may or may not
be representative of the true emission levels.
Altogether, 29 cyclone-furnace-fired boiler units were found to
have been field sampled for NOX, SOX, CO, and particulates. The
types of data compiled include results of spot checks as well as
comprehensive boiler test programs. The data were gathered from
the open literature, from B&W (the boiler manufacturer), and from
Commonwealth Edison (an electric utility company). The testing
agencies whose data were found in the open literature were govern-
ment contractors (Exxon and KVB), Federal EPA, and the TVA. The
Natural Emission Data System (NEDS) was also thoroughly searched
for cyclone-furnace-fired boiler data.17 The potentially valuable
data source dated March 1, 1976 showed that only one of the 134
cyclone boilers listed had been field tested. Data for this
boiler unit are included here. The NEDS file indicated that the
remaining 133 boilers had been estimated via emission factors.
Because the emissions determined by means of an emission factor do
not lend themselves to observation of variations in emissions from
individual emission sources, the NEDS emission data for the other
133 boilers are not included in this report.
The data gathered reflect emissions arising from the cyclone com-
bustion of bituminous coal, sub-bituminous coal, lignite, residual
oil, and natural gas. The smallest unit tested at maximum contin-
uous full load produced 65 kg/s of steam. The largest unit pro-
duced 350 kg/s of steam.
17National Emissions Data System (NEDS). Computer File Listing of
Detailed Point Sources of Utility and Industrial Cyclone-Fired
Boilers, March 1, 1976. 134 pp.
45
-------
The bulk of the data was reported in units of volume or mass con-
centration (e.g., vppm, g/m^), with some organizations also report-
ing emission levels on a mass-per-heat basis (e.g., ng/J). One
set of emissions data was reported on an annualized mass basis
(NEDS data).
The data reported by the various testing organizations were -
obtained using a variety of sampling test methods. Table 11
gives a brief summary of the instruments and methods used in
obtaining the field data. Detailed descriptions of the boiler
sampling techniques are beyond the scope of this report but can be
obtained by referring to the references cited in Sections 3.3.2 to
3.3.7.
The accuracy of the emissions data from field sampling is unknown
since no error estimates were available from the pertinent data
sources. It is recommended that the baseline emission levels as
well as any other field-sampled emissions data in this report be
interpreted with proper caution because of the variance in test
methods and techniques used by different testing organizations and
because of the unknown data accuracy.
3.3.1 Emissions Data Summary
A summary of the comparable baseline NOX data compiled during the
course of this study is presented in Table 12. The data are orga-
nized by a boiler identification code number (boiler number), type
of fuel burned during the test, and organization performing the
testing. The rated boiler capacities are reported if they could
be determined from the open literature or the manufacturer's data.
With the exception of boiler number 7, which is an industrial unit,
all of the boilers are classified as utility units. NOX (NO + N02)
data in Table 12 are reported as a volume concentration on a dry
3% 02 corrected basis at the percent load at which the test was
made. Data on mass of emissions per heat input basis (ng/J) are
shown in parentheses.
Only 16 of the 29 boilers field sampled were included in Table 12.
The emissions data for the remaining 13 boilers could not be
expressed on a 3% O2 dry basis, and the test loads could be esti-
mated only qualitatively. Hence, the data on the last 13 boilers
would not be comparable with the data from the first 16 boiler
units.
As shown in Table 12, full-load NOX levels for 8 bituminous-coal-
fired units with the rated capacities between 167 and 618 kg/s of
steam ranged between 960 and 1,197 vppm (arithmetic average was
1,074 vppm). It appears that NOX volume concentrations for all
these units drop with decreasing load. This drop may be as high
as 30% for decreased loads of up to 30% (boiler numbers 1, 4, and
6). As suggested from the data for boiler number 7, a further
decrease of the load does not seem to have too much effect on NOX
46
-------
TABLE 11. SUMMARY OF BASELINE EMISSIONS TEST METHODS
Organization
NAPCA (EPA)
BSW
NO
X
Method 7 (phenoldisulfonic
acid colorimetric)
Dynascience Monitor NX330
SO
X
Methods 6 and 8 (barium-
thorin titrations)
Particulate
Method 5
(impingers)
CO
Method 10
(IR Spectrometer)
Exxon
KVB
TVA
NEDS (EPA)
Commonwealth Edison
(electrochemical cell,
0 to 5,000 ppm range)
Beckman nondispersive IR
and UV spectrometers
Thermo Electron Chemilumi-
nescent analyzer
UV photometric analyzer
Method 7
Wet methods and Beckman
spectrometers
Unknown NFS
NFSa NFS3
Titration with lead perchlo-
rate (Shell-Emeryville) Method 5
NFSa NFS3
Methods 6 and 8 Method 5
NFS NFS
Unknown
Beckman nondispersive
IR spectrometer
Beckman nondispersive
IR spectrometer
NFSa
NFS (emission factor
estimate)
Unknown
NFS = not field sampled.
-------
TABLE 12.
SUMMARY OF BASELINE NO EMISSIONS DATA FOR CYCLONE BOILERS
x
Boiler
I. D.
NO.
1
2
3
4
5
6
7
8
Ranges
Maximum unit
rated capacity
Fuel type
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
Testing
organization
NAPCA
B&W
BSW
BSW
BSW
Exxon
KVB
TVA
Electric,
MW
206a
NA
200
NA
240
704
NA
300
200 to 704
Steam,
kg/s
171
NA
167
NA
200a
618
64 7
250a
65 to 618
Arithmetic averages
5
Ranges
9
10
Ranges
Sub-bituminous coal
Lignite
Lignite
BSW
BSW
BSW
240
240
NA
NA
NA
2003
200
NA
NA
NA
Arithmetic averages
11
12
13
14
Ranges
Residual oil
Residual oil
Residual oil
Residual oil
BSW
Exxon
Exxon
Exxon
NA
450
136
168
136 to 450
NA
309
117
158
117 to 309
Arithmetic averages
15
16
Ranges
Natural gas
Natural gas
BSW
TVA
NA
300
300
NA
250
250
Arithmetic averages
NOX, vppm, dry
40 to 50
NA
NA
NA
NA
NA
NA
742 (447)
NA
742 (447)
742 (447)
NA
NA
NA
NA
NA
NA
NA
NA
261 (150)
NA
261 (150)
261 (150)
NA
NA
NA
NA
51 to 60
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
206 (124)
NA
NA
206 (124)
206 (124)
NA
NA
NA
NA
3% 02 basis
61 to 70
NA
NA
NA
NA
NA
NA
800 (482)
NA
800 (482)
800 (482)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
(ng/J) at % maximum boiler
71 to 80
784 (774)
1,020 (612)
NA
NA
NA
886 (532)
790 (473)
NA
784 to 1,020
(774 to 612)
870 (600)
NA
NA
NA
NA
NA
NA
NA
NA
404 (232)
NA
404 (232)
404 (232)
NA
NA
NA
NA
81 to 90
NA
NA
NA
730 (438)
NA
NA
NA
NA
730 (438)
730 (438)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
load
91 to 100
1,160 (946)
NA
1,020 (612)°
975 (585)°
960 (576)°
1,197 (688)
NA
1,130 (678)
960 to 1,197
(576 to 688)
1,074 (680)
910 (546) C
910 (546)
593 (355)k
h
485 (291)
485 to 593
(291 to 355)
539 (323)
h
460 (276)
530 (310)
441 (254)
361 (206)
441 to 530
(254 to 310)
448 (261)
415 (207)°
650 (325)°
415 to 650
(207 to 325)
532 (266)
Estimated value.
1 vppm % 0.6 ng/J was assumed.
Cl vppm % 0.5 ng/J was assumed.
NA = Not available.
-------
volume concentration. It is probably the initial decrease of the
boiler load that has the most significant influence on NOX emis-
sions. All of the emission data appear relatively independent of
boiler size at common loads (1,074 vppm + 11.5%, -10.6% at loads
between 91 and 100%, and 870 vppm + 17.2%, -9.9% at loads between
71 and 80%).
One boiler unit, number 5, was fired with two types of coal, bitu-
minous and sub-bituminous. Its NOX level at full load when fired
with sub-bituminous coal (910 vppm) was slightly lower than when
fired with bituminous coal (960 vppm). The rated size of boiler
number 5 is 200 kg/s of steam (primary flow).
Two lignite-fired boilers were field sampled. NOX levels at full
load ranged from 485 to 593 vppm and averaged 539 vppm. The sizes
of these boilers were not available.
Data from four residual-oil-fired units indicated full-load NOX
emission levels ranging between 441 and 530 vppm with an average
of 448 vppm. Rated sizes of these boilers ranged between 117 to
309 kg/s of steam, with one boiler size unknown. The partial load
data indicate that NOX emission concentration appears more signifi-
cant at higher boiler load reduction than the decrease observed
when firing bituminous coal.
The two natural-gas-fired units shown in Table 12 (units 15 and
16) emitted between 415 and 650 vppm (average 532 vppm) NOX at
full load. The size of boiler number 16 was known and is 250 kg/s
of steam. Partial load data for gas-fired units were not available,
At full load, none of the bituminous coal-, oil-, or gas-fired
cyclone units were able to meet the New Source Performance Stand-
ards for NOX with respect to each fuel (refer to Section 3.2 for
standards). In general, the full-load NOX emission data indicate
that the NOX concentrations decrease with a fuel type in the fol-
lowing order: bituminous coal firing %sub-bituminous coal firing
>:>lignite firing fy residual oil firing % natural gas firing.
The summary of full-load baseline emission levels for CO, S02, SO^,
SOX, and particulates is shown in Table 13. The summary is broken
down by fuel types fired at full boiler load (e.g., >90% of rated
capacity) and shows the identification numbers of units tested for
these emissions. No data were found for boilers firing natural
gas.
CO emissions were highest for units firing residual or heavy oils
(3 to 8.5 vppm dry) and were generally zero for bituminous coal and
lignite units. No information was found on CO emissions for sub-
bituminous-coal-fired cyclone boilers. Emissions of sulfur oxide
fluctuated greatly, and the highest levels occurred in bituminous
coal-fired units. The single unit tested for particulate
49
-------
TABLE 13. BASELINE CO, S02, S03, AND PARTICULATE EMISSION DATA RANGES
FOR CYCLONE BOILERS AT FULL LOAD (>90% OF RATED CAPACITY)
CO,
Fuel type vppm dry
Bituminous coal 0
(boiler I. D. Nos.) (1, 3)
Subbituminous coal , NA
(boiler I. D. Nos.) (NA)
en
o Lignite 0
(boiler I. D. Nos.) (9)
Residual oil 3 to 85
(boiler I. D. Nos.) (11, 12, 13, 14)
S02
vppm dry
1,360 to 2,140
(I, 5)
535
(5)
580 to 800
(9)
NA
(NA)
S03
vppm dry
14 to 31
(1, 5)
14
(5)
NA
(NA)
NA
(NA)
SOx (S0£ + SOs) , Particulates , g/md
vppm dry (12% CO2, dry basis
1,374 to 2,171
(1, 5)
549
(5)
NA
(NA)
NA
(NA)
0.89a
(1)
NA
(NA)
NA
(NA)
NA
(NA)
Downstream of electrostatic precipitator operating at 74.5% mass efficiency.
Refers to units field sampled for these emissions.
NA = not available.
-------
emissions showed a level of 0.89 g/m3 at 74.5% collection effi-
ciency firing bituminous coal.
The following sections present the available baseline emission
data in greater detail. The sections are organized by the organi-
zations which performed the boiler tests.
3.3.2 NAPCA Data (Boiler I. D. No. 1)
In 1967 the National Air Pollution Control Administration (now EPA)
published a report on emissions from coal-fired power plants.18
This report included results of testing performed on a cyclone-
fired boiler unit. The boiler is rated at 171.5 kg/s of steam
(1,360,000 Ib/hr) at 16.65 MPa (2,400 psig) and 839 K (1050°F).
It is assumed that four cyclone furnaces fired this unit although
this was not specified in the report. Two forced-draft fans with
a capacity of 174.6 m3/s (370,000 scfm) supply combustion air to
the furnace and maintain positive pressure throughout the boiler
system. Flue gas leaving the boiler passes through secondary and
primary superheater sections, an economizer, an air preheater, and
finally a fly ash collector. The fly ash collectors include two
parallel electrostatic precipitators. Collected fly ash is nor-
mally reinjected into the furnace. Figure 17 shows the general
equipment and sampling arrangements at this installation.18
The location of this boiler was not specifically identified. Dur-
ing emissions testing of the boiler, a single type of high-volatile
bituminous coal from Pennsylvania was burned. Three tests were
run at about full load, two of which included fly ash reinjection.
Two additional tests were run at 75% load, both with fly ash rein-
jection. Here, full load is defined on the basis of maximum con-
tinuous steaming capacity (171.5 kg/s). All tests were conducted
with normal amounts of excess air (42% to 46.2%). The results of
the emissions testing are presented in summary form in Table 14,
The data indicate the expected high levels of NOX (1,200 vppm)
emitted at full load from the cyclone-furnace-fired boiler when
burning high-volatile A or B bituminous coals. This high value is
in contrast to a full load NOX level of 221 vppm emitted by a ver-
tically fired dry-bottom unit burning pulverized coal that was
also tested in this study.18 Coal of the same rank and nitrogen
content (1.4% by weight) was burned in the pulverized coal unit.
Excess air level in the pulverized coal unit was slightly higher,
44% versus 42% for a cyclone-fired unit.
With the exception of the particulate data, which are corrected to
12% C02, all other data are listed at stack conditions. The data
represent averages from the five tests performed. No CO data were
available for this unit at full load. Sampling probes were placed
18Cuffe, S. T. and R. W. Gerstle, Emissions from Coal-Fired Power
Plants: A Comprehensive Summary. U.S. Department of Health,
Education, and Welfare, NAPCA, Durham, North Carolina, 1967.
26 pp.
51
-------
SECONDARY SUPERHEATER
AND REHEAT HEADERS
.xSAMPLING POINTS. A
ELECTROSTATIC
/"PRECIPITATOR
PRIMARY
UPERHEATER
o
COAL
BUNKER
PRECIPITATOR
COAL
''SAMPLING.
POINT|
FURNACE!
-BURNERS--^
SECONDARY
AIR DUCT
RECIRCULA- SLAG
TING FAN |_TANK | f
COAL SAMPLING
POINT
FEEDERS
COAL CRUSHERS
FLUE GAS TEMPERING DUCT
Figure 17. Boiler outline for cyclone-type
unit showing sampling positions.18
in two locations—before the dust collection device and after it.
A side project of this study was to determine if gaseous emission
levels were affected by passage through the particulate collection
device. No significant change in any of the levels is evident
from the data in Table 14. Table 15 lists fuel, boiler, and flue
gas data averages which represent the conditions under which the
testing occurred. The data are presented as they appear in the
literature (with the exception of their conversion to metric
equivalents).
52
-------
TABLE 14. EMISSIONS SUMMARY FOR UNIDENTIFIED
COAL-FIRED CYCLONE BOILER UNIT18
Boiler I. D. No. 1 Full load 75% Load
Particulates,a g/m3
Before*3 3.4 4.1
After 0.89 0.50
ESP collector efficiency, % 74.5 86.3
Nitrogen oxides,
vppm (dry 3% 02) , ng/J
Before" xxWxu ^u,
After0
Carbon monoxide, vppm (dry 3% O2)
Beforeb
After
Sulfur dioxide, vppm (dry 3% 02)
Beforeb
After0
Sulfur trioxide, vppm (dry 3% O2)
Beforeb
After
1,204
1,160
1,350
1,360
21
31
; 1,075
(2.5)
946
(2.2)
No data
No data
742;
784;
15
10
1,380
1,370
13
22
817
(1.9)
774
(1.8)
aCorrected to 12% CO2 dry basis standard conditions.
Before fly ash collector.
°After fly ash collector.
Reported as NO2.
3.3.3 Boiler Manufacturer Data (Boiler I. D. Nos. 2 to 5, 9 to 11,
15)
B&W, the manufacturer of cyclone-furnace-fired boiler units, has
tested several cyclone boilers for NOX and other emissions. The
emission data made available for this study by B&W are presented
in Table 16. The data presented reflect uncontrolled emissions
from boiler units and are given at stack conditions which are
unknown. There are some data gaps in defining the load the boiler
was under during testing. For example, it is not known whether
the full load is defined as the maximum continuous rating of the
53
-------
TABLE 15. AVERAGE COAL, BOILER, AND FLUE GAS
DATA FOR UNIDENTIFIED BOILER UNIT18
Boiler I. D. No. 1
Proximate analysis of coal (as fired) , %
Moisture
Volatile matter
Fixed carbon
Ash
Ultimate analysis of coal (as fired) , %
Hydrogen
Carbon
Nitrogen
Oxygen
Sulfur
Ash
Heating value, MJ/kg
Boiler conditions
Steam rate, kg/s
Coal feed rate, kg/s
Flue-gas volume, m^/s
Before
After0
Average flue-gas temperature, K (°F)
Before
After0
Flue moisture, %
Before
After0
C02, %
Before
After0
02, %
Before
After0
Excess air, %
Before
After0
Full load
1.1
37.0
54.5
7.4
5.2
77.4
1.4
6.1
2.4
7.7
32.4
168.0 (98% load)
16.2
263
237
410 (279)
397 (255)
6.3
5.9
12.8
12.7
6.4
6.3
42.6
42.0
75% Load
1.1
37.0
54.5
7.4
5.2
77.4
1.4
6.1
2.4
7.4
32.4
128.9
10.4
208
190
402 (264)
388 (239)
6.6
6.4
12.0
12.2
6.8
46.0
46.2
Expressed at standard conditions of 288 K (59°F) and 1.013 x 105
(1 atm).
b
Before fly ash collector.
f*
"After fly ash collector.
Measured at fly ash collector.
54
-------
TABLE 16. EMISSIONS FROM CYCLONE-FIRED BOILERS
Ui
Ul
Boiler
I. D. No.
2
3
4
4
5
5
9
9
9
9
9
9
10
10
11
15
Gross boiler
Fuel burned load, MW
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
Bituminous coal
(Illinois)
Subbituminous coal
(Montana)
Lignite
Lignite
Lignite
Lignite
Lignite
Lignite
Lignite
Lignite
Residual oil
Natural gas
NA
200
NA
NA
240
240
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Percent of-
full load
76
100
100
85
100
100
100
100
100
100
100
100
100
100
100
100
02 in flue
gas, %
4.5
3.6
5.4
5.2
2.6
3.5
7.0
6.4
4.9
4.6
5.6
5.1
5.6b
4.8
5.0
4.9b
3.2
2.9
NOX, vppm dry
(3% 02 basis)
1,020
1,020
975
730
960
910
NA
685
562
503
640
575
593b
480
490
485b
460
415
CO,
vppm dry
15
0
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
50
NA
S02,
vppm dry
NA
NA
NA
NA
2,140
535
765
NA
NA
NA
590
580
645b
NA
NA
NA
NA
S03,
vppm dry
NA
NA
NA
NA
14
14
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Data courtesy of the Babcock & Wilcox Company, Engineering Services Group, Barberton, Ohio.
Average of results obtained from a specific boiler.
NA = not available.
-------
boiler or some other basis such as normal full load (a percentage
of maximum load). Although the information contained in Table 16
is less complete than the previous data of Cuffe and Gerstle,18 it
gives some idea of the variation of NOX emissions with fuel type.
In general, at full load, these data indicate that NOX emission
levels decrease in the following order: bituminous coal, sub-
bituminous coal, lignite, residual oil, and natural gas.
3.3.4 Exxon Data (Boiler I. D. Nos. 6, 12 to 14)
Exxon Research and Engineering Company (Linden, New Jersey) has
performed an extensive series of emissions testing on utility
boilers.19'20 The primary objective of this work was to develop
NOX and other pollutant control technology through combustion modi-
fication (see Section 4.1). In the course of Exxon's work, a
total of four cyclone-fired steam generators were tested, and
their uncontrolled NOX and CO emissions were determined. Three of
these units were oil fired, and one was coal fired. The emission
data are presented in Table 17.
The unidentified boiler is an oil-fired unit with a maximum contin-
uous rating of 309 kg/s primary steam flow and a full-load gross
electrical rating of 450 MW. It is fired by eight cyclone fur-
naces. The next four sets of data were also obtained for oil-
fired units located at the B. L. England Station of Atlantic City
Electric (New Jersey).
Boiler unit No. 1 has a maximum continuous rating of 117 kg/s pri-
mary steam flow with a full-load gross electrical rating of 136 MW.
Three cyclones fire this unit. Boiler unit No. 2 has a maximum
continuous rating of 158 kg/s primary steam flow with a full-load
gross electrical rating of 168 MW. Four cyclones fire unit No. 2.
No fuels analyses were given for these two boiler units.
The last two sets of data were obtained from a large coal-fired
unit owned by the Tennessee Valley Authority located at the Para-
dise, Kentucky station, where the boiler is designated as Unit 1.
The maximum continuous rating of the boiler is 618 kg/s primary
steam flow with a full-load gross electrical rating of 704 MW.
19Bartok, A. R., Crawford, and G. J. Piegari. Systematic Field
Study of NOX Emission Control Methods for Utility Boilers. Esso
Research and Engineering Company (for: U.S. Environmental Pro-
tection Agency. Research Triangle Park, North Carolina, Con-
tract No. CPA 70-90). December 31, 1971. 215 pp.
20Crawford, A. R., E. H. Manny, and W. Bartok. Field Testing:
Application of Combustion Modifications to Control NOX Emissions
from Utility Boilers. Exxon Research and Engineering Company,
Government Research Laboratory. (for: U.S. Environmental Pro-
tection Agency, Washington, D. C. EPA-650/2-74-066). June 1974.
151 pp.
56
-------
TABLE 17. SUMMARY OF EXXON EMISSIONS DATA FOR CYCLONE-FIRED BOILERS19/20
Boiler
I. D.
No.
12
12
13
13
13
14
6
6
Boiler unit
identification
Unidentified
boiler
Unidentified
boiler
B. L. England
boiler unit 1
B. L. England
boiler unit 1
B. L. England
boiler unit 1
B. L. England
boiler unit 2
TVA Paradise
unit 1
TVA Paradise
unit 1
Fuel burned
Residual
Residual
Residual
Residual
Residual
Residual
oil
oil
oil
oil
oil
oil
Bituminous coal
Bituminous coal
Gross
boiler
load,
MW
415
258
133
62
105
167
665
545
Percent
of
fuel
load
91
57
98
45
77
100
95
77
Stack
temperature ,
K (°F)
611
589
679
603
659
645
601
585
(640)
(601)
(763)
(626)
(727)
(702)
(622)
(594)
02
% dry
4.0
4.6
1.5
4.2
2.7
2.2
5.3
5.3
vppm dry
3% O?
530a
206a
441
261
404
361
1,197
886
NO
X
ng/J
(lb/106 Btu)
310 (0.72)
NA (NA)
254 (0.59)
150 (0.35)
232 (0.54)
206 (0.48)
688 (1.60)
NA (NA)
CO , vppm
dry 3% 0?
6.5
3
57
54
59
85
NA
NA
Average of two runs.
NA = not available.
-------
The only actual excess air measurement found in the Exxon litera-
ture was for TVA Paradise Unit No. 1 boiler, which was operated at
20% excess air level. The other boilers were operated at excess
air levels reflecting their normal operation (15% to 30%).
Details concerning the test conditions and boiler characteristics
are given in Table 18.
TABLE 18. SUMMARY OF BOILER OPERATING DATA CORRESPONDING
TO EXXON EMISSION TESTS19'20
Boiler designation
B. L. England
Unidentified
boiler
Boiler
unit 1
Boiler
unit 2
TVA Paradise
unit 1
Boiler I. D. No.
Maximum continuous steam
rating, kg/s
Full -load rating, MW
Initial year of operation
Nominal heat rate, Btu/kW hr
Fuel burned
Furnace volume, m3
Furnace heating surface, m2
Number of cyclones
Main steam pressure, kPa
Main steam temperature, K (°F)
12
309
450
1964
NA
Oil
4,313
1,932
8
NA
NA
13
117
136
1957
NA
Oil
NA
NA
3
1,251
811
(1000)
14
158
168
1964
NA
Oil
NA
NA
4
1,251
811
(1000)
6
618
704
1963
8,777
Coal
9,646
3,818
14
1,654
840
(1053)
NA = not available.
Table 19 lists the available oil and coal analyses representative
of test conditions.
3.3.5 KVB Data (Boiler I. D. No. 7)
The first phase of an ongoing study to determine the effectiveness
of combustion modification techniques to control emissions of NOX
from industrial boilers has been completed by KVB Engineering,
Incorporated of Tustin, California (see Section 4.1 for details).21
As part of this study, one cyclone-fired boiler was tested for its
baseline emissions at various loads firing bituminous coal.
21Cato, G. A., H. J. Buening, C. C. DeVivo, B. G. Morton, and
J. M. Robinson. Field Testing: Application of Combustion Modi-
fications to Control Pollutant Emissions from Industrial Boilers-
Phase I. KVB Engineering, Inc. Tustin, California. (for EPA,
Office of Research and Development. EPA-650/2-74-078-a).
PB 238 920. October 1974. 196 pp.
58
-------
TABLE 19. AVAILABLE FUEL ANALYSIS DATA
FOR BOILERS TESTED BY EXXON19
Oil analyses
B. L. England
Unidentified
(Boiler I. D. 12)
Boiler unit 1
(Boiler I. D. 13)
Boiler unit 2
(Boiler I. D. 14)
Ash, wt %
C, wt %
H, wt %
N, wt %
S, wt %
Fe, ppm
Ni, ppm
V, ppm
High heating value,
MJ/kg (Btu/lb)
Kin. Vis. at 372 K
(210°F), m2/s
0.02
85.2
11.8
0.5
0.46
2.0
7.5
29
44.6 (19,185)
35.86 x ID"6
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Coal analyses for the TVA Paradise boiler
unit 1, boiler I. D. 6 (as received basis)
Proximate analysis, wt %
Moisture
Ash
Volatile matter
Ultimate analysis, wt %
6.8
15.6
31.0
Carbon
Hydrogen
Nitrogen
Sulfur
Oxygen
High heating value, MJ/kg
(Btu/lb)
61.6
5.1
1.3
3.6
15.0
25.8
(11,090)
NA = not available.
The cyclone-fired industrial boiler tested during the KVB program
is located in New York and was built in 1967. It has a rated maxi-
mum continuous primary steam capacity of 64.7 kg/s (513,000 Ib/hr).
Two cyclone furnaces each 3.66 m long and 3 m in diameter fire a
water tube secondary furnace which has a tube furnace area of
35 m2 and a volume of 32 m3. The burners are spaced 1 m apart and
59
-------
are located horizontally on a single wall. The full-load furnace
heat release is 3.9 MW (gross)/m2 (1,250,000 Btu/hr - ft2) on an
area basis and 4.7 MW (gross)/m3 (454,500 Btu/hr cu ft) on a vol-
ume basis. The primary air temperature when burning 100% coal
during the emissions testing was 559 K (546°F).
Table 20 summarizes the available emissions data for this boiler
extracted from the KVB study. Full load is defined in terms of
maximum steam capacity (64.7 kg/s). Table 21 gives a representa-
tive analysis of the coal burned during the emissions testing.
TABLE 20. EMISSIONS DATA FOR NEW YORK BOILER (BOILER I. D. NO.
Test Percent Flue vppra
load, of gas Stack dry
full 02 temperature, 3* 02
S02
Particulates,
»,/» iu.il «2 «..,«....„»„, 1
Fuel burned steaa load % dry K CD basis (lb/106 Btul vppn. (lb/106 Btu) vppm (lb/106 Etui vppm (lb/106 Btu) (lb/106 Btu)
Bituminous coal 40.3
Bituminous coal 30.3
78.4
62.4
46.8
3.2
3.4
3.2
416 (293)
418 (293)
418 (293)
790
800
742
473 (1.10)
482 (1.12)
447 (1.04)
0
0
0
0
0
0
NA
1,122
NA
NA
937 (2.18)
NA
NA
13
NA
NA
10.7 (0.025)
NA
NA
513 (1.19)
NA
HA * not available.
TABLE 21. COAL ANALYSIS, A NEW YORK
BOILER (BOILER I. D. NO. 7)21
Moisture, wt % 1.43
Heat of combustion gross, MJ/kg 30.7
Carbon, wt % 76.6
Hydrogen, wt % 5.5
Sulfur, wt % 2.9
Nitrogen, wt % 1.6
Ash, wt % 7.8
Carbon residue, wt % 52.7
3.3.6 TVA Data (Boiler I. D. Nos. 8, 16)
The Tennessee Valley Authority (TVA) owns and operates several
cyclone-fired utility boilers. NOX emission data were found for
three of these boiler units. One set of data for a TVA station
(Boiler I. D. No. 6, Paradise Station Unit No. 1) has already been
presented as part of Exxon's data (Section 3.3.4). Presented here
are data for two other units tested by TVA.22
22Hollinden, G. A., S. S. Ray, N. D. Moore, J. T. Reese, and
C. Gottschalk. NOX Control at TVA Coal-Fired Steam Plants.
Tennessee Valley Authority, Chattanooga, Tennessee. Paper Pre-
sented at National Symposium ASME Air Pollution Control Division.
30 pp.
60
-------
The two boilers tested are located in Memphis, Tennessee at the
T. H. Allen station. Unit No. 1 (boiler I. D. No. 16) has a full-
load rating of 300 MW, is fired by seven cyclones, and was fueled
with gas during its test. Unit No. 2 (boiler I. D. No. 8) has sim-
ilar characteristics, but was fueled with bituminous coal during
its test. The maximum continuous primary steaming rate of each
unit is 380 kg/s. Both tests occurred under full-load conditions
of 290 MW for each boiler (97% of maximum continuous rating). The
NOX concentration for gas-fired unit No. 1 was 650 vppm (3% 02
dry) or about 430 ng/J (1.0 lb/106 Btu). The NOX concentration
for coal-fired boiler unit No. 2 was 1,130 vppm (3% 02 dry) or
about 663 ng/J (1.54 lb/106 Btu). Boiler unit No. 2 emitted about
the same amount of NOX at 290 MW as did the Paradise unit No. 1
firing at 665 MW.
3.3.7 Commonwealth Edison Data (Boiler I. D. Nos. 17 through 28)
The largest number of cyclone boiler units is owned by the Common-
wealth Edison Company, a large electric utility company in Chicago,
Illinois. It is estimated that about 25% of the total U.S.
cyclone-fired steaming capacity is owned by this company. This
section includes baseline data for 12 boiler units spot-checked by
Commonwealth Edison's Operational Analysis Department over a
period of about 3 years. The remainder of the NOX emission data
on Commonwealth Edison facilities were generated by the equipment
supplier, B&W, and were not made available for this study.
The available emission data are presented in Table 22. The data
give NOX, NC>2, and CO emissions and are organized by type of fuel
burned during testing. At the time when these tests were per-
formed, in most cases the mass emission rates (ng/J) were not cal-
culated. According to Commonwealth Edison, it would be difficult
if not impossible to reconstruct these tests. Consequently, the
Commonwealth Edison data should be evaluated with caution and
should be viewed as spot-check results rather than the results rep-
resentative and typical of cyclone furnace operation.
The exact boiler loads for the data in Table 22 could not be deter-
mined; however, a qualitative indication of the loads during these
tests is provided as given from the data supplier. In addition,
none of the emissions data were corrected to a comparable "dry
3% 02 basis." For these reasons, it would be difficult to prop-
erly compare the Commonwealth Edison data with emission data pre-
sented elsewhere in this report. The Commonwealth Edison data
were also excluded from the Emissions Data Summary, Section 3.3.1,
and Load Reduction Field Test Data, Section 4.2.4.
Four fuel types were tested by Commonwealth Edison; i.e., bitumi-
nous coal, blend of bituminous and subbituminous coals, subbitumi-
nous coal, and residual oil. The full-load data (NFL in Column 6
of Table 22) were arranged by fuel type. In general, these data
indicate that at full load, NOX emission levels decrease in the
61
-------
TABLE 22. BASELINE FLUE GAS CONCENTRATIONS AND EMISSION RATES OF NITROGEN OXIDES AND CARBON
MONOXIDE FOR 12 COMMONWEALTH EDISON-OWNED CYCLONE FURNACE FIRED-BOILER UNITS3
Boiler
I. D.
No.
17
18
18
19
19
20
20
20
21
22
23
23
23
24
25
26
27
27
28
Station
Kincaid
Powerton
Powerton
Will County
Will County
Stateline
Stateline
Stateline
Fisk
Waukegan
Will County
Will county
Will County
Ridgeland
Ridgeland
Ridgeland
Ridgeland
Ridgeland
Ridgeland
Unit
No.
2
51
51
2
2
4
4
4
18
17
1
1
1
1
2
4
5
5
6
Fuel burned during spot test
Illinois bituminous coal 4.2% S
Illinois bituminous coal 3.6% S
Illinois bituminous coal 3.6% S
Illinois bituminous coal 3.6% S
Illinois bituminous coal 3.6% S
50/50 Blend western subbituminous
and Illinois bituminous coals
50/50 Blend western subbituminous
and Illinois bituminous coals
50/50 Blend western subbituminous
and Illinois bituminous coals
Western subbituminous coal 0.7% S
Western subbituminous coal 0.6% S
Western subbituminous coal 0.4% S
Western subbituminous coal 0.4% S
Western subbituminous coal 0.4% S
No. 6 residual oil with additive A
No. 6 residual oil 1% S
No. 6 residual oil with additive B
No. 6 residual oil 1% S
No. 6 residual oil 1% S
No. 6 residual oil with additive c
Net
MW
rating
616
425
42b
157
157
358
358
358
78
119
144
144
144
74
84
74
156
156
138
Boiler
load
NFL
NFL
RL
NFL
SRL
MFL
RL
MFL
NFL
NFL
NFL
ML
LL
NFL
NFL
NFL
NFL
NFL
NFL
N0x
vppm
467
913
437
525
518
674
645
582
292
442
484
241
360
286
268
419
185
215
187
(as N02>
ng/J
(lb/106 Btu)
NA
671
(1.56)
409
(0.952)
856
(1.99)
899
(2.09)
NA
NA
NA
NA
NA
NA
NA
NA
241
(0.561)
NA
379
(0.882)
NA
NA
125
(0.291)
vppm
NA
11
1
NA
NA
2
0.9
0.9
0.05
1.0
0.05
0.05
0.05
NA
NA
NA
NA
NA
NA
NO 2
ng/J
(lb/106 Btu)
NA
0.86
(0.002)
0.86
(0.002)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
vppm
200
22
4
1
1
NA
NA
NA
132
227
560
260
475
10
NA
NA
NA
NA
NA
CO
ng/J
(lb/106 Btu)
NA
9.9
(0.023)
2.1
(0.005)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
4.3
NA
NA
NA
NA
NA
Data courtesy of the Commonwealth Edison Company.
bAll data have been corrected to 29.92 in Hg and 70°F but not to a dry 3% O2 basis. Test methods may vary. Instrumental methods were used in
some instances. All concentrations and rates are approximate averages; in some cases, individual tests showed substantial variation. MW shown
are net ratings, not loads experienced.
°NFL = near full load, RL = reduced load, ML = medium load, LL = low load, SRL = slightly reduced load. No quantitative load data were available.
NA = not available.
-------
following order: bituminous coal = 50/50 blend bituminous/sub-
bituminous coals > subbituminous coal » residual oil. This trend
is identical to the one indicated by the data developed by B&W
(see Section 3.3.3). In addition to this trend, it appears that
CO emissions from cyclone boilers burning western subbituminous
coals are excessively high and may constitute an emissions problem.
The same is true for one of the bituminous coals from Illinois con-
taining 4.2% sulfur.
3.3.8 NEDS Data (Boiler I. D. No. 29)
As stated previously, only one boiler listed in the NEDS file17 as
of March 1976 was designated as being "source tested." Source
testing refers to actual stack sampling measurements made on that
boiler unit. The information on this utility boiler is presented
in Table 23. The unit is owned by Minnkota Power Cooperative, Inc.,
and constitutes the Milton R. Young generating station (235 MW)
located 5 miles southeast of Center, North Dakota. The boiler
burns lignite and is fired by a total of seven cyclone furnaces
generating 216 kg/s of primary steam at full load. The boiler is
relatively new (made operational in 1972), and the data indicate
that the boiler was operating at its maximum design capacity (load
factor = 1.04). A centrifugal collector operating at 70% mass
efficiency was used for removing particulates during this testing.
No other details are known concerning the assumptions made in
arriving at these annualized emissions.
TABLE 23. ANNUALIZED 1972 EMISSIONS DATA FOR A 235 MW CYCLONE-
FIRED UTILITY BOILER UNIT (Boiler I. D. No. 29)17
NEDS data
Estimated data
Fuel burned
Particulates
NO
cox
SO
v
Normal operation
Annual load factor
Stack temperature
Flue gas rate (actual)
Fuel consumption
Maximum fuel design rate
Fuel heat value
Boiler design heat release rate
Lignite
0.281 kg/s?
0.375 kg/s^
0.025 kg/s
0.500 kg/s
24 hours/day
336 days/year
1.04
439 K (331°F)
78.8 m3/s
45.87 kg/s
47.69 kg/s
15.1 MJ/kg
732.6 MJ/s
0.65 g/m3
683 vppm (410 ng/J)
75 vppm (45 ng/J)
645 vppm (392 ng/J)
430 m3/s
bSulfur content = 0.7%, ash content = 8.
Source test.
approved non-EPA emission factor.
Based on maximum fuel design rate.
63
-------
One possible discrepancy was noted in these data from the NEDS
file. The flue gas rate of 78.8 m3/s appears low for a unit of
this size (235 MW). A report by W. S. Smith and C. W. Gruber23
indicates that for a 235 MW coal-fired plant, the stack effluent
rate should be approximately 283.2 m3/s at 289 K and 100 kPA.
Assuming an ideal gas law, the flue gas rate can be corrected to
439 K. This results in a flue gas rate of 430 m3/s, a rate more
than five times higher than that approximated by NEDS. This cor-
rected rate, rather than the NEDS figure, was used in arriving at
the estimates shown in Table 23. It was also assumed that the NO
was reported as NO2, and the SOX was reported as S02. Emission
rates (ng/J) were estimated by assuming that 1 vppm ^0.6 ng/J
(bituminous coal ratio). None of the estimates made could be cor-
rected to a consistent basis such as 3% 02 dry and 12% C02 because
of insufficient information.
3.4 NEED FOR NO CONTROL
An estimate of the total annual amount of NOX emitted from the pop-
ulation of cyclone-fired boiler units can be made using the emis-
sion factor method. Total NOX emitted is obtained by multiplying
the total quantity of fuel burned by the appropriate emission fac-
tor (weight of NOX emitted/unit of fuel consumed). The only cur-
rently accepted emission factor for cyclone firing is for large
boiler units (>29 x 106 J/s) burning bituminous coal. The value
of this emission factor is 12.28 g NOX emitted per kg of coal
burned.24
The electric utility industry was found to have the most complete
information on annual coal consumption. This industry is most
important in terms of NOX emissions because over 94% of the pri-
mary steam produced by cyclone furnaces is generated in the elec-
tric utility sector (see Section 3.2).
By examining a compilation of boiler records published by the
National Coal Association, an estimate of the total amount of coal
burned at power plants employing cyclone furnace firing was made.25
No actual data on fuel consumption in cyclone furnaces per se
could be found in the literature. Appendix B includes information
which was used to arrive at the estimates of major fuel types
23Smith, W. S., and C. W. Gruber. Atmospheric Emissions from Coal
Combustion - An Inventory Guide. U.S. HEW, Public Health Serv-
ice, Division of Air Pollution. Cincinnati, Ohio. April 1966.
112 pp.
24Anon. Compilation of Air Pollutant Emission Factors. 2nd edi-
tion, U.S. Environmental Protection Agency, April 1973.
pp. 1.1-1 to 1.4-3.
25Anon. Steam-Electric Plant Factors, 1974 Edition. National
Coal Association, 24th edition, 1974. 110 pp.
64
-------
(coal, oil, and gas) used for utility cyclone firing. There are
some limitations connected with these estimates. Also contained
in Appendix B are fuel consumption, electric load factor, and net
power generation data for all power plant installations in the
United States employing cyclone fuel firing (coal, oil, and gas).
It was estimated that in 1973, 62 x 109 kg (68.4 x 106 tons) of
coal were burned at power plants employing cyclone firing. The
overwhelming majority of this coal was bituminous with an average
heating value of 26 MJ/kg (11,200 Btu/lb). If the 62 x 109 kg of
coal burned is multiplied by the emission factor of 12.28 g NOX
emitted per kg coal burned, the result is 0.76 x 106 tonnes
(0.84 x 106 tons) of NOX emitted from all cyclone coal-fired util-
ity boilers in 1973.
A recent EPA-sponsored study by the Aerotherm Division of the
Acurex Corporation estimated that in 1972, 3.44 x 106 tonnes/yr
(3.79 x 10^ tons/yr) of NOX were emitted from all coal-fired util-
ity boilers.16 Using this 1972 estimate as the emission base and
a cyclone NOX estimate of 0.76 x 106 tonnes (0.84 x 106 tons/yr),
bituminous coal-fired cyclone furnace utility boilers contributed
22% of the total NOX emissions from all coal-fired utility boilers
in 1973. (The emissions base was assumed to be unchanged for
1973.) These statistics clearly indicate a need for NOX control
in this equipment class.
The same Aerotherm study ranked cyclone-fired utility boilers burn-
ing bituminous coal third out of a possible 137 ranked stationary
sources of NOX in 1972. It was estimated that over 6% of the
total NOX emitted by all 137 sources (10.58 x 106 tonnes,
11.66 x 106 tons total NOX in 1972) came from bituminous-coal-
fired cyclone furnace utility boilers. The first-ranked source
was gas-fired, spark ignition internal combustion engines, which
contributed 16% of total 1972 NOX. The second was tangentially
fired bituminous-coal-burning utility boilers, which contributed
12% of total 1972 NOX.15 Aerotherm estimated that 19% of the
total NOX from all coal-fired utility boilers in 1972 was contrib-
uted by bituminous-coal-fired cyclone furnace utility boilers. A
summary of pertinent statistics from the Aerotherm study is given
in Table 24. From this table, it is seen that within the cyclone-
fired boiler category, over 91% of the NOX emissions result from
utility boilers burning bituminous coal.
65
-------
TABLE 24. SUMMARY OF AEROTHERM NOX EMISSION ESTIMATES
FOR ALL CYCLONE-FIRED BOILERS IN 197216
Rank out
of 137 Boiler
NOx sources type
3
55
64
79
89
130
TOTAL
Utility
Industrial
Utility
Industrial
Utility
Utility
a N°x
Fuel tonnes/yr x 10°
Bituminous coal
Bituminous coal
Residual oil
Residual oil
Lignite
Distillate oil
0.65b
0.025
0.017
0.012
0.008
0.001
0.713
Distribution of
NOX within
cyclone boiler
category , %
91.2
3.5
2.4
1.7
1.1
0.1
100.0
3No data available for natural gas firing.
MRC estimate was 0.76 x 106 tonnes/yr in 1973 for these two categories based
on information in Appendix B.
66
-------
SECTION 4
APPLICABILITY OF COMBUSTION MODIFICATIONS
TO CYCLONE FURNACES/BOILERS
The term combustion modification as used in the context of this
report refers primarily to any modification or change in the major
combustion operating conditions or fuels of a boiler unit to sup-
press formation of NOX. Some equipment modifications related to
the basic combustion equipment design could also be considered as
combustion modifications. In the combustion process, oxides of
nitrogen are formed both from the nitrogen in the combustion air
(thermal NOX) and by conversion of chemically bound nitrogen in
the fuel (fuel NOX). Section 4.1 presents a review of combustion
modification strategy with reference to controlling NOX formed
both thermally and from the fuel. Section 4.2 cites instances of
field experience with cyclone-fired boiler units. Section 4.3
reviews the significance of the NOX combustion modification experi-
ence. Section 4.4 includes recommendations for further work.
4.1 COMBUSTION MODIFICATION STRATEGY IN GENERAL
Combustion modifications seek to suppress the formation of NOX
which results from two sources; namely, chemically bound nitrogen
in the fuel and atmospheric nitrogen. Atmospheric nitrogen reacts
with oxygen at high temperatures during the combustion process to
form NOX. Oxygen also reacts with nitrogen chemically bound in
the fuel during the combustion process. For natural gas and light
distillate oil firing, the bulk of NOX is formed via atmospheric
(nitrogen) fixation.26 With residual (crude) oil and coal, the
contribution from fuel-bound nitrogen can be significant, and
under certain operating conditions, it can be predominant.26 Pohl
states that U.S. coals contain 0.5% to 2.0% nitrogen by weight, of
which about 10% to 50% may be converted to nitric oxide in combus-
tion.27 The fate of the remaining 50% to 90% of the nitrogen in
coal is not well known. Fine, Slater, and Sarofin, et al.
26Brown, R. A., H. B. Mason, and R. J. Schreiber. Systems Analy-
sis Requirements for Nitrogen Oxide Control of Stationary
Sources. Aerotherm/Acurex Corporation (California).
U.S. Environmental Protection Agency, EPA-650/2-74-091, 1974.
27Pohl, J. H. and A. F. Sarofim. Fate of Coal Nitrogen During
Pyrolysis and Oxidation. Fuels Research Laboratory, Massachu-
setts Institute of Technology, Paper presented at "Symposium on
67
-------
postulate that the remainder of the bound nitrogen in U.S. coals
is converted into molecular nitrogen (N2).28
The major oxide of nitrogen formed during the combustion process
is nitric oxide, NO. Other oxides of nitrogen formed, such as
nitrogen dioxide, NO2, and its dimer, N20it, require prior forma-
tion of NO. The formation of thermal and fuel NO is discussed
below in relation to combustion modifications.
The kinetics of thermal NO formation are complex and coupled to
the kinetics of fuel oxidation. Both fuel and atmospheric nitro-
gen oxidation kinetics are influenced by effects of turbulent mix-
ing in the flame zone.26 It is generally accepted that the most
significant reactions which form thermal NO are those of the
Zeldovich chain mechanism involving formation of oxygen radicals:
O2 + M ^± O + 0 + M (1)
N2 + 0 ^± NO + N (2)
O2 + N ^±: NO + 0 (3)
Reaction 2 is rate controlling. M is any third-body molecule,
which can result in formation of oxygen radical.
Brown, et al. further describe thermal NO formation26 in combus-
tion equipment:
"Due principally to the high energy required to break
the N2 bond in Reaction 2, the activation energy for NO
formation via the Zeldovich mechanism is considerably
larger than for typical rate-controlling reactions in
hydrocarbon oxidation. This entails that thermal NO
formation is initiated well after initiation of fuel com-
bustion and is extremely temperature sensitive with vir-
tually all NO being formed in the high temperature
regions of the flame. For the time scales involved in
the flow through commercial combustors, the high tempera-
ture dependence of the NO system means that total NO emis-
sions are far below equilibrium levels. NO formation
is thus kinetically controlled with the emission level
dependent on time of exposure to the high temperature."
The amount of combustion air supplied for fuel oxidation also
influences the level of thermal NO. Increasing the level of air
Stationary Source Combustion" sponsored by Combustion Research
Section, U.S. Environmental Protection Agency, September 24-26,
1975 (Atlanta, Georgia), 22 pp.
28Fine, D. H., S. M. Slater, A. F. Sarofim, and G. C. Williams.
Nitrogen in Coal as a Source of Nitrogen Oxide Emission from
Furnaces. Fuel, 53 (4) :120-125, 1974.
68
-------
(02 level) above the amount required for complete theoretical com-
bustion generally increases the concentration of NO formed. Thus,
there are three factors that influence the extent to which thermal
NO is formed:
(1) Peak temperature
(2) Time of exposure to peak temperature
(3) Oxygen level at peak temperature
The current strategy behind combustion modifications is to act on
the three factors mentioned above. Figures 18 and 19 illustrate
the importance of each factor.29 The data presented in Figures 18
and 19 are for natural gas fuel where NO is formed largely due to
thermal effects with no or negligible fuel NO contribution. Data
presented are not actual field data but were theoretically derived
from fundamental kinetic and thermodynamic relationships.*9 Fig-
ure 18 shows the equilibrium concentration of NO in the combustion
products of natural gas versus percent theoretical air with temper-
ature as a parameter. Equilibrium NO levels are the highest
achievable at a given temperature and percent theoretical air. A
cyclone furnace operating at 1920 K (3000°F) and 20% excess air,
for example, should have an equilibrium NO concentration of approx-
imately 2,000 ppm.
Figure 19 gives theoretical kinetic information (assuming the
Zeldovich mechanism) for the reaction between nitrogen and oxygen
versus percent theoretical air with time and temperature as para-
meters. James of B&W concludes the following from Figures 18
and 19.28
• "As the oxygen concentration in the combustion prod-
ucts is increased, the amount and the rate of NO forma-
tion increases.
• "As the temperature of the combustion products is
raised, the amount and the rate at which thermal NO is
formed increases.
• "As the time available for reaction at high tempera-
ture increases, the amount and the NO concentration
also increase."
James also states that the converse of these conclusions may be
observed under specific boiler combustion conditions. Since the
majority of cyclone furnaces operate at temperatures above 1920 K
(3000°F), it is evident that the NO formation in these furnaces
never reaches equilibrium (e.g., about 2,000 ppm NO at 20% excess
29James, D. E. A Boiler Manufacturer's View on Nitric Oxide Forma-
tion. The Babcock & Wilcox Company, Presented to the Fifth Tech-
nical Meeting, West Coast Section of the Air Pollution Control
Association, San Francisco, California, October 8-9, 1970.
26 pp.
69
-------
REDUCING CONDITIONS
OXIDIZING CONDITIONS
10, 000
n-
Q_
rr:
' 1
'.I
g
CJ
1,000
REDUCING CONDITIONS
OXIDIZING CONDITIONS
80 90 100 110 120
PERCENT OF THEORECTICAL COMBUSTION AIR
Figure 18. Thermodynamic equilibrium
data for natural gas fuel.29
10, 000
RESIDENCETIMEIN
SECONDS AT THE
TEMPERATURE
SHOWN
80 90 100 110 120
PERCENT OF THEORECTICAL COMBUSTION AIR
Figure 19. Kinetic data for
natural gas fuel.29
-------
air, Figure 18), based on this theory. The highest NOX emissions
measured on a cyclone furnace were just over 1,100 ppm.
A heirarchy of effects leading to thermal NOX formation in cyclone
boilers is shown in Table 25. The boiler's primary equipment and
fuel parameters must be dealt with if thermal NOX is to be reduced
through a combustion modification approach. The causal relation-
ships between the primary, secondary, and fundamental parameters
are not known for cyclone boilers.
TABLE 25. POTENTIAL FACTORS CONTROLLING THE FORMATION OF THERMAL NO IN CYCIONE BOILERS
Primary equipment and fuel parameters
Secondary combustion parameters
Fundamental
thermal NOx
parameters
Degree to which primary parameters
have been explored in existing
cyclone boilers for NO,, control
1. Combustion air temperatures
(primary, secondary, tertiary)
2. Combustion air velocity
(primary, secondary, tertiary)
3. Cyclone furnace aerodynamics
4. Fuel type (switching)
1. Not explored
2. Not explored
3. Not explored
4. Fairly well explored
5.
6.
7,
a.
9.
10.
11.
Fuel composition within same fuel
type or rank
Injection pattern of fuel and air
(staging)
Size of fuel particles or droplets
Excess air
Monitoring individual cyclone NO
behav ior
Load reduction
Turbulence with n the furnace
Heat removal ra e in the furnace
Mixing of combu tion products into
flame
Turbulent disto tion of flame zone
Oxygen level 5 .
Peak temperature Thermal 6.
NO
Exposure tune at x
peak temperature
7.
_, 8.
9.
10.
11.
Not explored
Some staging patterns have been
applied to boiler furnaces but
not to individual cyclones
where bulk of NO is formed
Not explored
Fairly well explored for oil-
fired units only
Hot explored
Not explored
Well explored
Also shown in Table 25 is the degree to which primary parameters
have been explored in existing cyclone boilers to effect NOX reduc-
tion. This information summarizes qualitatively the state of the
art of combustion modifications for cyclone boilers as determined
from the literature and the boiler manufacturer (B&W). Of the 11
primary parameters shown, only 4 have been explored in some way
for NOX reduction potential (4, 6, 8, 11). A description of field
testing performed to determine the significance of primary para-
meters 4, 6, 8, and 11 is given in Section 4.2.
Research on the role of fuel-bound nitrogen in forming NOX is in
its preliminary stage. As mentioned previously, conversion of
nitrogen in the fuel could account for 10% to 50% of the NO pre-
sent in the flue gas. Fuel nitrogen conversion is generally
regarded to be relatively insensitive to temperature. Brown
states that26
"The most critical factor in fuel NOX conversion
appears to be the local conditions in which volatiliza-
tion and formation of nitrogen intermediate compounds
occur. In a reducing atmosphere, it is suspected that
the intermediate compounds go to form H2r or other more
complex unknown nitrogen compounds with little
71
-------
subsequent conversion to NO. In an oxidizing atmos-
phere, conversion of the intermediates to NO is thermo-
dynamically favored over conversion to N2- Although
basic understanding of these phenomena is only in the
preliminary stage, a promising strategy for fuel NOX
reduction appears to be modification of the burner or
combustion conditions to allow volatilization to occur
prior to massive entrainment of oxygen in the flame
zone.
"The fate of fuel-bound nitrogen which does not go to NO
under oxidation conditions is uncertain. There are indi-
cations that other pollutants, such as HCN, may result
when NO formation is suppressed. This possibility may,
indeed, constitute a limitation to fuel NOX reduction
strategies and requires further investigation."
Brown's observation should be explored further in light of poten-
tial application to cyclone boilers.
4.2 COMBUSTION MODIFICATION EXPERIENCES WITH CYCLONE-FIRED BOILER
UNITS
Section 3.4 of this report states that out of 137 stationary
sources of NOX in 1972, the cyclone-fired, bituminous-coal-burning
utility boilers ranked third. They were the source of 6% of sta-
tionary source NOX emissions in the U.S. despite the fact that the
cyclone furnaces are significant NOX emitters. Only a relatively
small number of cyclone boilers were found to have been examined
and tested in some way to determine the effect of combustion modi-
fications on NOX emissions. One reason for lack of field data on
this combustion equipment class is the relative inflexibility of
the cyclone boilers toward modification. Robert Lundberg, an
expert on cyclone furnaces for over 30 years at Commonwealth
Edison of Chicago, describes the cyclone as being ". . .more like
a digital device that really functions in about one mode." The
rigid operating specifications of the cyclone furnace are dic-
tated by proper furnace temperature and high heat release rates to
maintain furnace slagging. Operating experiences suggest that
these parameters cannot be altered to the degree required for ade-
quate NOX control without ending with a furnace which is no longer
a "cyclone."
The literature as well as the boiler manufacturer (B&W) reveal
that four types of combustion modifications have been applied
either singly or in combination to reduce NOX emissions from
cyclone furnaces. These modifications are: load reduction, low
excess air firing, simulated staged firing, and switched fuel fir-
ing. No boiler unit has ever been tested under sustained condi-
tions with any of these four modifications. The modification tech-
niques applied most often have been load reduction and low excess
air firing because they require no modification or changes of
72
-------
existing cyclone units. To date, NOX reduction for all boilers
tested was achieved by load reduction.
The prognosis for long-term application of any of the modifica-
tions previously mentioned to existing boiler units is dim. The
reasons for this prognosis are further discussed in Section 4.3.
Table 26 lists the boiler units tested and shows the type of com-
bustion modification applied in each test. The boilers are iden-
tified by numbers corresponding to those used in Section 3.3.
This enables the reader to identify the boiler with its design and
operation characteristics and available emission data.
All forms of modifications utilize one or more of the three fac-
tors that influence thermal NO formation; i.e., peak flame tempera-
ture, residence time of gas at peak temperature, and oxygen level.
One of the methods, staged firing also acts on the variables that
control fuel NO formation. The concept of staged firing provides
a localized reducing atmosphere which favors chemical conversion
of nitrogen containing intermediates in the fuel (coal and heavy
oils) to N2 or unknown nitrogen compounds with little subsequent
conversion to NO.26 Basically, the method works on the principle
of lowering the oxygen supply to the burner zone where fuel nitro-
gen is volatilized.
The following subsections present the combustion modification
field data extracted from the open literature and the cyclone
boiler manufacturer (B&W). Except for the tests performed with
load reduction, all other data are presented in alphabetical order
by the organization performing the tests. All available data on
load reduction are combined and presented last in Section 4.2
since load reduction is perhaps the least desirable NOX control
alternative from an operational standpoint. Section 4.3 summa-
rizes and discusses all the field data in light of the suitability
of combustion modifications as NO control alternatives and pre-
sents recommendations for furtherxwork.
4.2.1 Boiler Manufacturer Field Experience (B&W)
B&W tested at least six cyclone boilers to determine the effects
of nonload-reduction combustion modifications on NOX emissions.
No written reports concerning these tests were available in the
open literature. In general, the data that were obtained from B&W
during the course of this work were sketchy and incomplete because
of the age of the data and because of confidentiality agreements
between B&W and the boiler owners. None of the boiler units could
be specifically identified. However, all the boilers were of the
utility type. Modifications of cyclone furnaces investigated by
B&W include low-excess-air firing (LEA), fuel switching, and simu-
lated staged firing.
73
-------
TABLE 26. CYCLONE BOILER UNITS FIELD-TESTED FOR COMBUSTION MODIFICATION APPLICABILITY18"21
Testing
agency
B&W
B&W
B&W
B&W
B&W
B&W
Exxon
Exxon
Exxon
Exxon
KVB
NAPCA (EPA)
Boiler
I. D.
number
3
4
5
9
11
15
6
12
13
14
7
1
Boiler identification
Unidentified utility boiler
Unidentified utility boiler
Unidentified utility boiler
Unidentified
Unidentified
Unidentified
TVA Paradise Station, Unit
No. 1 utility boiler
Unidentified utility boiler
Atlantic City Electric —
England Station, Unit
No. 1 utility boiler
Atlantic City Electric —
England Station, Unit
No. 2 utility boiler
Unidentified industrial
boiler located in New York
Unidentified utility boiler
Size
(maximum
continuous
steam) ,
kg/s
167b
(200 MW)
NA
200b
(240 MW)
NAa
NA
NA
618
309
117
158
64.7
171.5
Combustion modifications applied
Fuels used Load
during tests reduction
Bituminous coal,
coal and gas
Bituminous coal X
Bituminous, sub-
bituminous, and
50/50 coal blend
Lignite
Residual oil
Natural gas
Bituminous coal X
Residual oil X
Residual oil X
Residual oil
Bituminous coal,
blends of coal
and oil X
Bituminous coal X
Low Switched-
excess Staged fuel
air firing firing
X X
X
X
X X
X
X
X
X X
X
X X
NA = not available.
Estimated.
-------
4.2.1.1 Low Excess Air Firing (LEA)—
One boiler (I. D. No. 9) firing lignite and one boiler (I. D. No. 11)
firing residual oil were tested under LEA conditions to determine
NOX emission reductions. Boiler No. 9 was tested at its full
(unknown) load. The results of this test are shown in Table 27.
Reducing excess air by 75% (6.4% 02 reduced to 1.6% 02 in the
flue gas) reduced the NOX level by 47%. However, CO emissions
increased as the excess air was being reduced.
TABLE 27. LIGNITE-FIRED BOILER
(Boiler I. D. No. 9)a
02 in
flue
gas, %
7.0
6.4
5.6
5.1
4.9
4.6
4.3
4.0
2 9
f. , y,
1.6b
NOx , vppm
dry 3%
02 basis
_
685
640
575
562
503
600
640
540
360
CO,
vppm
_
-
-
-
-
-
10
12
12
17
S02
vppm
765
—
590
580
-
-
735
660
845
800
Data courtesy of the Babcock &
Wilcox Co.
Low excess air required supple-
mental oil to maintain ignition.
In addition, the lowest excess air setting (1.6% in flue gas)
required supplemental oil to maintain ignition. At present there
is no lignite NOX NSPS which could be used to compare the NOX
reduction achieved during this test. The proposed standard is
258 ng/J (0.6 lb/106 Btu) or approximately 430 vppm. Table 27
indicates that this proposed standard can be met for boiler No. 9
but only at the 1.6% 02 level with supplemental oil fuel.
Boiler No. 11 was fired with residual fuel oil. Normal firing of
this cyclone with oil at 2.6% O2 in the flue gas yielded about
360 vppm NOX and 1,000+ vppm CO. When the 02 was increased to
3.2% in the flue gas, the NOX increased by 28% to 460 vppm, and
CO decreased to about 50 vppm. The reduction in NOX could not be
justified by the high levels of CO which occurred during this test.
4.2.1.2 Combined LEA and Switched Fuel Test—
Boiler No. 3 was a 200-MW (about 167 kg/s primary steam) coal-
fired unit equipped with flue gas recirculation for steam
75
-------
temperature control. At normal full load, it operated at 3.6% O2
in the flue gas, yielding 1,020 ppm NOX with negligible CO. This
unit was operated at 1.6% O2 in the flue gas, firing one cyclone
with coal and the remainder with natural gas. It is not known
how many cyclones the boiler had. At full load, the test yielded
an NOx level of 700 ppm. At the same time, the CO level increased
to 2,100 ppm. Flue gas recirculation was used during this test
but not for NOX control. From the available data, it was not
possible to distinguish the individual effects of the switched
fuel and LEA on NOX emissions. As with boiler No. 11, the NOX
reduction achieved could not be justified by the high CO levels.
4.2.1.3 Switched Fuel Test—
Boiler No. 5 was a 240-MW (about 200 kg/s primary steam) coal-
fired unit equipped with flue gas recirculation and gas tempering
for steam temperature control. This unit was tested to determine
the effect of coal type on boiler emissions. Table 28 summarizes
the data. Sulfur oxide emissions were significantly reduced by
switching to the western coal, which apparently had low sulfur
content. At the same time, NOX emissions were reduced only
slightly. Blending the eastern and western coals resulted in NO
emissions slightly higher than when these coals were individually
fired. No other test details were available. For the purposes
of NOX control, at least, switching to low-sulfur western coal
did not appear to significantly affect the high NOX level which
is characteristic of coal firing. Sulfur oxide emissions were
reduced, however, by 75%.
TABLE 28. EFFECT OF FUEL SWITCHING ON NOX EMISSIONS
FROM COAL-FIRED BOILER I. D. NO. 5a
Coal type
Illinois (bituminous)
Montana (subbituminous)
O2 in
flue
gas, %
2.6
3.5
Concentration ,
NOX
960
910
S02
2,140
535
ppm
SO 3
33
14
50/50 Blend of
Illinois and Montana 3.4 1,020 1,365 NA
aData courtesy of the Babcock & Wilcox Co.
4.2.1.4 Simulated Staged Firing—
Babcock & Wilcox, the originators of the staged-firing concept,
have applied simulated staged firing to several cyclone boilers
with limited success. Staged firing consists of sustaining part
of the combustion in a reducing atmosphere zone. Combustion is
then completed in an oxidizing atmosphere. The concept of staged
firing is particularly attractive in that it can simultaneously
lower both fuel and thermal NO contributions. Several forms of
76
-------
staged firing have been applied to specific boiler units. The
forms specifically applied by B&W to cyclone boilers include two-
staging and pattern firing.
Two-staging consists of operating the cyclone(s) slightly fuel-
rich. The amount of combustion air fed into the cyclone is reduced.
Consequently, the bulk of combustion with the cyclone occurs at
slightly fuel-rich or reducing atmosphere conditions. This
reduces fuel NO formation. The remainder of the required combus-
tion air is fed into the boiler at a point near the exit of the
cyclone furnace proper, usually through the flue gas recirculation
(FGR) ductwork. This means that the overall combustion is
extended over a longer time and furnace space resulting in lower
cyclone furnace heat release rates and temperatures. Lower fur-
nace temperatures then result in lower thermal NO formation. Fig-
ure 20 shows the schematic of the two-staging concept. Two-staging
requires some additional ductwork and controls and therefore slight
boiler modification.
' CYCLONE
FURNACE (S)
FUEL OPERATING
\SLIGHTLY
FUEL RICH'
85 - 99 %
THEORECTICAL
AIR
SUPERHEATER
TUBE
SECTION
SECONDARY
FURNACE
GAS
MIXING ZONE
ECONOMIZER
SECTION
TO AIR
PREHEATER
AND STACK
PRIMARY
FURNACE
SLAG
TANK
EXISTING FGR
BLOWER SYSTEM
COMBUSTION
MAKEUP AIR FROM
AIR PREHEATER
DAMPERS
KADDED DUCT
WORK AND
CONTROLS
Figure 20. Two-staging concept,
77
-------
B&W evaluated an eastern-coal-fired boiler and a gas-fired boiler,
both modified for two-staging tests. The first unit (unidenti-
fied) firing eastern coal showed a 28% to 36% reduction in NOX
emissions. NOX emissions at full load are normally about
1,110 ppm. With two-staging, NOX levels were between 700 ppm and
800 ppm. This does not appear sufficient to meet the 500 ppm
stated in the NSPS. B&W also found that oil supplement may be
required to maintain ignition and flame stability in the cyclone
furnace depending on the primary fuel burned. According to B&W,
more oil may be required for lignitic coals than for bituminous
coals.
The gas-fired cyclone boiler unit (boiler No. 15) showed a 48%
reduction in NOx emissions using the two-staging concept. Specif-
ically, the NOX emissions were reduced from 500 vppm to 260 vppm.
This is 100 vppm above the NSPS of 160 vppm. B&W indicated a
strong reluctance to recommend two-staging as a viable NOX control
alternative. The reasons given include a combination of the fol-
lowing factors: (1) lack of significant long-term testing experi-
ence, (2) reluctance of boiler owners to accept two-staging on a
permanent basis, and (3) risk of catastrophic tube corrosion in
the primary furnace (refer to Section 3.1.6).
Another form of staged firing applied to cyclone boilers is called
pattern firing. Pattern firing is possible only in multiple
cyclone units in stacked configuration. The idea behind pattern
firing is to operate the upper row of cyclones and the bottom row
of cyclones in such a way as to produce a staged effect. Normally,
in stacked multiple cyclone firing at full load, each cyclone (in
both the upper and lower rows) is fed equal amounts of fuel at
identical air/fuel ratios. With pattern firing, the amount of
fuel fed in each cyclone as well as the air-to-fuel ratios are
adjusted in one of the combination modes shown in Table 29. The
effect of all of the pattern combinations shown is to produce a
fuel-lean condition in the upper row(s) of cyclones. Using the
experience from two-staging, the most effective pattern-firing com-
bination for reducing NOX is to run the lower cyclones under reduc-
ing conditions and the upper row of cyclones under oxidizing condi-
tions. This practice, however, is prohibited by B&W for safety
reasons. Sustained operation of cyclones in reducing atmospheres
can cause catastrophic tube failure due to iron sulfide and iron
formation (see Section 3.1.6).
A residual oil unit was tested by B&W to determine the degree of
NOX control achievable with pattern firing. This unit (No. 11)
was tested at full load, both in normal operation and in the
pattern-firing mode. Under normal full-load conditions with
1.3% 02 in the stack (about 6.3% excess air) the baseline NOX
level was about 380 vppm on a 3% 02 dry basis. Patterned firing
capable of full load showed NOX levels between 290 vppm and
300 vppm (3% O2, dry) at 1.6% O2 in stack (about 7.8% excess air).
Thus, NOX reductions achieved ranged between 21% and 24%. At the
78
-------
TABLE 29. SEVEN FUEL AND AIR FLOW COMBINATIONS FOR
STACKED-CYCLONE FIRING (PATTERN FIRING)
Firing mode
Change in fuel feed Change in excess air feed
1. Normal full load
2. Normal partial load
3. Normal partial load
4. Pattern full load3
5. Pattern partial loadc
6. Pattern partial loadc
7. Pattern partial loadc
Zero
Zero
One or more burners
out of service
Zero
- Upper row
+ Lower row
Upper row air only
+ Lower row
- Upper row
+ Lower row
Zero
Zero
Zero (cut air flow to
inoperative burners)
+ Upper row
- Lower row
+ Upper row
- Lower row
- Upper row
- Or normal lower row
Zero
Lower cyclone row should not be run in highly reducing atmosphere because
of corrosion risks.
same time, the level of excess air increased by 22%. No other
details or implications were made available concerning this test.
Also, the degree of boiler operating difficulties, efficiency pen-
alties, pattern firing combination, and corrosion potential were
not defined.
4.2.2 Exxon Field Experience
The Government Research Laboratory of Exxon Research and Engineer-
ing Company, Linden, New Jersey, has tested four cyclone boilers
(Nos. 6, 12, 13, 14). These boilers were field tested during the
course of EPA-sponsored programs to determine application of com-
bustion modification to control NOX from utility boilers.20/21
Altogether, three residual-oil-fired units and one bituminous-coal-
fired unit were tested. The combustion modifications applied were
LEA, FGR, pattern firing, load reduction, and combinations thereof,
4.2.2.1 Boiler No. 6—
Boiler No. 6 constitutes Unit 1 of the TVA Paradise Station
located in Drakesboro, Kentucky. The maximum rated capacity of
the unit is 704 MW at 618 kg/s (4.9 x 106 Ib/hr) of primary steam
flow. High-sulfur bituminous coal is used as fuel. The main
steam pressure is 16,546 kPa (2,400 psi) at 840 K (1053°F). The
unit is equipped with a steam reheat capacity of 435 kg/s
(3.45 x 106 Ib/hr) at 2,103 kPa (305 psi) and 813 K (1003°F).
79
-------
Fourteen B&W cyclones fire the unit and are arranged in two facing,
front and rear, walls containing seven cyclones each. The
cyclones are arranged in the stacked cyclone configuration shown
in Figure 21. Each cyclone is 3 ra in diameter and 3.6 m long.
The cyclones are located on approximately 5 m centers.
BOILER WALL
UPPER ROW
LOWER ROW
Figure 21.
Cyclone firing arrangement, TVA
Unit No. 1, Drakesboro, Kentucky
(view facing front or rear wall).
Exxon formulated the boiler test program based upon recommenda-
tions provided by B&W, a subcontractor for the program. B&W stipu-
lated that during Exxon's testing, the total air feed to the
cyclones would not be reduced below 122% to prevent fireside cor-
rosion. In addition, B&W required that cyclone temperatures not
be reduced below a point where slag chilling could result in slag
tapping problems.
The entire test program was of short duration (2 days) and did not
involve any major hardware changes. The baseline full and partial
load NOX data generated during this program were presented in Sec-
tion 3.3.4 of this report and are also part of the load reduction
data presented in Section 4.2.4 (Runs 1 and 4).
Exxon's boiler test program is presented in Table 30. All burners
were in service during the six test runs. Runs 1 through 3 were
performed at near full boiler load (94%). Run 1 established
boiler baseline NOX emissions. In run 2 the ratio of gas
80
-------
TABLE 30. BOILER TEST PROGRAM IMPLEMENTED BY EXXQN
(BOILER I. D. NO. 6, TVA PARADISE UNIT NO. 1)
Run
No.
1
2
3
4
5
6
Load,
(MW)
100
(704)
100
(704)
100
(704)
78
(550)
78
(550)
78
(550)
Excess
FGR variations Cyclone
level, Gas
% recirculation
120 Minimum
120 Increase
120 Minimum
120+ Minimum
120+ Minimum
120+ Minimum
Change in
Tempering fuel feed
Maximum Zero
Decrease Zero
Maximum + Bottom
Maximum Zero
Maximum Bottom normal
- top (50%)
Maximum Bottom normal
- top (50%)
Change in
air feed
Zero
Zero
Zero
Zero
Zero
Zero - bottom
+ - top
recirculation and gas tempering was varied to determine the
effects of FGR. If applied properly, FGR can lower the peak
flame zone temperature and also reduce the amount of 02 available
for NO formation. Section 4.3.3 defines and illustrates FGR,
which is normally used for steam temperature control in boiler
units so equipped. For Run 3, more coal was fed to the bottom
cyclone rows than to the top rows at no air flow change and full
load. Runs 4 through 6 were performed at about 3/4 load. Run 4
is a baseline NOX run at this partial load. During Run 5, the
normal full load amounts of coal were fed to the bottom cyclones,
and about 50% less than normal, full-load amounts were fed to the
top row of cyclones at zero air flow change. The same coal flows
were used during Run 6, but the secondary air was increased to
the upper row of cyclones until boiler excess 62 increased from
3.9% to 4.9%. The overall attempt in Runs 3, 5, and 6 was to
produce a staqinq effect of the top and bottom cyclone combina-
tion. The upper burners in Runs 3, 5, and 6 were operated under
highly lean conditions. B&W prohibited the lower cyclones to be
run under reducing conditions for any of these tests.
Table 31 presents the results of the test program. Exxon and B&W
concluded that no significant change in boiler efficiency was
experienced during the tests, and nitric oxide production was sig-
nificantly decreased by reducing the boiler load. Flue gas recir-
culation (FGR) applied to the maximum extent reduced NOX by 7%
(1,197 vppm to 1,112 vppm, Runs 1 and 2). Feeding more coal to
the bottom cyclones at full load at zero air flow change in Run 3
81
-------
TABLE 31. SUMMARY OF EMISSION DATA, BOILER I. D. NO. b, 704 MW COAL-FIRED, TVA PARADISE UNIT NO. I1
00
N)
Run
No.
1
2
3
4
5
6
Type of test
Full load baseline
FGR effects at full load
Changed coal feed between
upper and lower cyclones
at full load
Partial load baseline
Changed coal feed between upper
and lower cyclones at 3/4 load
Changed coal feed and excess air
flow between upper and lower
cyclones at 3/4 load
Gross
boiler
load,
MW
665
668
660
545
545
548
Operating data
Boiler
Boiler excess
efficiency, air
% level
91.1 Normal
91.5 Normal
90.8 Normal
91 . 6 Normal
92.0 Normal
91.7 Normal
Flue gas components
Distribution of NOX
Total coal between Change Oj CO2 3* Q£
coal upper and in air dry dry dry
flow, lower rows, flow basis, basis, basis,
FGR
Minimum
Maximum
Minimum
Minimum
Minimum
Minimum
kg/s
70.5
70.1
75.0
56.2
59.8
60.6
% upper/lower feed % % vppm
51/49
51/49
44/56
51/49
38/62
38/62
Zero 5.3 13.1 1,197
Zero 5.3 13.2 1,]12
Zero 5.4 13.0 1,203
Zero 5.3 13.3 886
Zero 5.1 13.6 915
Increased0 5.6 13.0 846
Flue gas
temperature,
K
604
597
602
585
594
595
(°F)
(627)
(615)
(624)
(593)
(609)
(611)
3Average of 16 data points per run. Each data point from a composite of three gas sample streams. CO emissions were not measured. Hydrocarbons
emissions measured <1 ppm.
"Staged firing" simulated by operating top cyclone burners under highly fuel-lean conditions.
CBoiler excess O2 increased 25% by increasing secondary air to upper rows.
-------
had no effect on NOX but reduced boiler efficiency slightly. At
partial load, NOX increased by 3% when coal feed was changed and
at no air flow change (Runs 4 and 5). Operating the boiler with
both fuel and air changed lowered the NOX level by 4% (886 vppm to
846 vppm, Runs 4 and 6) .
The relatively insignificant changes in NOX level witnessed during
these runs perhaps exemplify the improper application of FGR and
staging to cyclone boilers. As regards FGR, Figure 22 shows the
entry points of recirculated flue gas in boiler No. 6. It should
be noted that the recirculated flue gas enters the secondary
boiler furnace rather than the cyclone furnace. FGR was origi-
nally built into boiler units for purposes of steam temperature
control (refer to Section 4.3.2). The entry points of recircu-
lated flue gas in boiler No. 6 provide for optional steam tempera-
ture control but are not effective for NOX control. To be truly
effective, FGR must be applied directly to or very near the combus-
tion flame zone within each cylone. In an existing unit such as
boiler No. 6, this is difficult to do without extensive equipment
modifications. Modifying boiler No. 6 in such a fashion was
beyond the scope of Exxon's program.
FLU EG AS
FLU EG AS
SECONDARY
FURNACE
.ENTRY,
POINT
FLU EG AS
7 CYCLONES
7 CYCLONES
Figure 22. Recirculated flue gas entry points for boiler No. 6
83
-------
The staging tests performed by Exxon were very limited in that
none of the boiler combustion occurred under reducing atmosphere
conditions (B&Ws lower limit of 122% total air), which are neces-
sary for NOX reduction.
4.2.2.2 Boiler No. 12—
Residual-oil-fired boiler No. 12 was tested for NOX emissions at
LEA conditions, simulated staged combustion (pattern firing), and
at reduced load. Baseline emissions and load reduction data for
this unit are contained in Sections 3.3.4 and 4.2.4 of this report
and are also included here for comparison.
A summary of emissions data for boiler No. 12 is given is Table 32,
The emissions from the two ducts sampled are given separately in
Table 32 due to wide differences in the gas composition in these
sampling locations. The significance of the data variability was
not explained in the literature. Evidently some difficulties in
determining excess 62 in the boiler were experienced during this
test, as indicated by the flue gas 02 measurements made by Exxon
and the boiler recording instrument. The point of minimum excess
air was not defined. In any case, it can be assumed that Runs 1
and 3 were baseline runs at full and partial load, respectively,
and that Runs 2 and 4 were LEA runs as suggested by the NOX emis-
sion data. The suggested classification of the runs is also pre-
sented in Table 32 (second column).
TABLE 32. SUMMARY OF EMISSION DATA FROM BOILER NO. 12, 450 MW21
Flue gas compositions and temperatures
No.
1
2
3
4
•>
Gross
boiler
Type of test MW
Full load
baseline 421
LEA 410
Partial load
baseline 255
LEA 262
Pattern
firing 275
Duct No. 1
._ Dry basis, pptn, 3% Oj _
No. of *, \_ • Temp,
n % dry basis
firing O2 CO2 NOX CO ("F)
594
8 4.1 12.2 548 8 (610)
8 4.9 11.4 505 6 597
(615)
584
8 4.6 12.0 214 NA (592)
8 2.5 13.3 211 3 578
(581)
600
6C 5.1 11.2 315 1 (621)
Duct No. 2
Dry basis, ppm, 3% 0?
. Temp,
% dry basis *'
02 C02 NOx CO (°F)
625
2.7 12.8 572 6 (666)
4.3 11.8 497 6 628
(671)
618
6.5 7.4 200 NA (653)
4.9 11.3 200 3 578
(581)
591
6.9 9.5 306 2 (604)
O2,
%b
2.1
2.7
2.2
4.2
4.5
3Average of four data points. Each data point from compositie of three gas sample streams.
Boiler O2 recorder data.
6 cyclones firing oil and 2 cyclones on air only to simulate staged combustion.
NA = not available.
84
-------
At full load (about 93%) , the NOX level for this cyclone unit was
reduced with LEA firing by only 11% (based on average values from
ducts No. 1 and 2; 560 reduced to 501 vppm). The 3% reduction in
boiler load between test Runs 1 and 2 may account for some of the
NOX lowering. At partial load (57%), LEA did not reduce baseline
NOX although the boiler O2 data indicate that excess air was
reduced by one-third (Exxon Q£ data averages). As has been the
usual case with cyclone boilers, these data also show that load
reduction results in the most dramatic NOX level decrease (over
60% NOX reduction for 38% load reduction).
Run 5 was made to simulate staged combustion (within the flexibil-
ity of this boiler). Two upper-level cyclones were fired on air
only, while the other six cyclones were fired at increased rates
to maintain load. This change resulted in an increase of NOX emis-
sions by about 50% (206 ppm to 310 ppm), presumably because of the
higher intensity firing of the operating six cyclones.
4.2.2.3 Boiler Nos. 13 and 14—
Boiler No. 13 and boiler No. 14 constitute steam generating Units
1 and 2 of the B. L. England Station owned by Atlantic City Elec-
tric (New Jersey). Boiler No. 13 (117 kg/s steam, 136 MW) is
fired by single mechanical atomizing oil burners,in each of the
three cyclones. The cyclones are arranged in a triangular fashion
with two cyclones on one level and the third cyclone elevated.
All three cyclones are in the front wall of the furnace.
Boiler No. 14 (157 kg/s steam, 168 MW) is also fired by single
mechanical atomizing oil burners. This boiler has four cyclones
arranged in a square pattern, two cyclones at each elevation. All
four cyclones are installed in the front wall of the boiler. Both
boilers burn crude oil.
During the testing performed on boiler No. 13, the influence of
excess air, load reductions, and combinations thereof were studied.
Boiler No. 14 was tested at normal and LEA conditions at full load
only. Baseline emissions and load reduction data for these two
boilers are contained in Sections 3.3.4 and 4.2.4 of this report
and also in this section for comparison.
For boiler No. 13, LEA at full load was defined as 1.1% on the
boiler O2 meter (0.5% avg O2 measured by Exxon). At these oxygen
concentrations, the smoke density on ACE's smoke meter was normal,
and no visible emissions were apparent from the stack. However,
carbon monoxide emissions increased (>1,500 ppm) and continued
operation at this low level of excess air could not be recommended,
At partial load (103 MW), the minimum excess air was defined as
that which produced only a slightly visible stack plume, no appre-
ciable increase in smoke density, and reasonable CO emissions
(about 200 ppm max.).
85
-------
For boiler No. 14, LEA was defined similarly to that for boiler
No. 13 at partial load.
The operating and emissions data for these two boilers are pre-
sented in Table 33. With the exception of Run 3 for boiler No. 13,
normal cyclone firing patterns were used. Data for boiler No. 13,
Run 3 were taken at low load with the middle (upper) cyclone taken
out of service.
The comments of A. R. Crawford, E. H. Manny, and W. Bartok best
summarize the conclusions made during this boiler testing program.20
Boiler No. 13
"Baseline operations (test run No. 1) conducted with all
three cyclones operated normally, produced average flue
gas concentrations of 441 ppm NOX (3% O2, dry basis) at
1.5% oxygen. Reducing the excess air level to 1.1% and
0.5% oxygen in the flue gas resulted in a reduction in
average emission levels at this load to 396 ppm and 313
ppm NOX (3% O2, dry basis), respectively. Baseline oper-
ation at 105 MW output produced 404 ppm NOX (3% O2, dry
basis) at the level of 2.7% 02 in the flue gas. Reduc-
ing excess air to 2.4% and 1.0% 02 in the flue gas
reduced NOX emissions to 364 ppm and 241 ppm, respec-
tively, at the intermediate load. At the minimum load
of 62 MW, a baseline emission level of 261 ppm NOX (3%
O2, dry basis) was measured at 4.2% oxygen. This level
is about the same as the emissions at the intermediate
load level of 105 MW at low excess air conditions, indi-
cating the particularly significant contribution of fuel
nitrogen oxidation to NOX emission at intermediate to
low load levels; i.e., at lowered combustion intensity
conditions.
"Decreasing excess air levels at both full and inter-
mediate loads had a substantial effect on reducing NOX
emission levels. With cyclone operation, at least at
present, staged firing patterns which might effect fur-
ther reductions are not possible.
"To sum up, this boiler has baseline NOX emissions of
441 ppm which are higher than the original recommended
new source emission standards of about 225 ppm for oil-
fired boilers. (The maximum NOX reduction achievable at
full load with an acceptable level of CO was 10%.) Low
excess air operation at full and intermediate loads
resulted in significant lowering of NOX emissions as
shown in Table 33. However, decreases in load and reduc-
tions in excess air levels could not reduce emissions
below the recommended standards for new boilers which
are subject to reassessment at present by EPA."
86
-------
TABLE 33. SUMMARY OF OPERATING CONDITIONS AND EMISSIONS DATA FOR OIL-FIRED
BOILER NOS. 13 AND 14 (ACE, B. L. ENGLAND UNITS 1 AND 2)20
00
Average
Boiler operating
Test
run
No.
1
2
3
4
5
6
7
1
2
Gross
load,
MW
133
133
132
62
105
105
103
167
167
Excess air
level
Normal
Intermediate
Lowb
Normal
Normal
Intermediate
Low
Normal
Low
conditions
flue gas measurements3
NOX
3% Oo
Firing pattern
All cyclones
All cyclones
All cyclones
on
on
on
Middle cyclone off
All cyclones
All cyclones
All cyclones
All cyclones
All cyclones
on
on
on
on
on
Smoke
density
Boiler
30
30
30
24
26
26
25
Boiler
24
24
dry
O2 , basis,
% vppm
No. 13
1.5
1.1
0.5
4.2
2.7
2.4
1.0
No. 14
2.2
1.6
441
396
313
261
404
364
241
361
303
NOX/ ng/J
(lb/106 Btu)
254
228
181
151
232
206
138
206
181
(0.59)
(0.53)
(0.42)
(0.35)
(0.54)
(0.48)
(0.32)
(0.48)
(0.42)
CO
3% 02
basis,
vppm
57
74
1,523
54
59
53
68
85
231
C02,
13.1
13.1
13.2
11.9
12.7
12.9
13.8
13.5
13.5
Temperature,
K (°F)
679 (762)
678 (760)
671 (748)
603 (625)
695 (726)
653 (715)
643 (697)
645 (701)
643 (697)
Flue gas measurements made on composite gas samples from three individual sampling tubes. Measurements
shown are averages of three analyses from three sampling tubes (short, medium, and long) for each of four
probes.
^
Excessively high CO emissions at this condition.
-------
Boiler No. 14
"Baseline operations (test run No. 1) conducted at full
load with all four cyclones firing crude oil, produced
average flue gas NOX concentrations of 361 ppm (3% 02 /
dry basis) at 2.2% oxygen. Reducing the excess air
level to that corresponding to 1.6% oxygen in the flue
gas resulted in an average level of 303 ppm NOX (3% ©2/
dry basis), or a decrease of 16% from baseline conditions.
"To sum up, this boiler has baseline NOX emissions of
361 ppm NOX which are higher than the original EPA rec-
ommended standards of about 225 ppm for new oil-fired
boilers. Low excess air operation resulted in a 16%
reduction in NOX emissions, but could not reduce them
below recommended standard levels. This reduction in
NOX emissions was achieved without any adverse effects,
such as significantly increased smoke, unburned combus-
tible emissions, or reduced operability."
4.2.3 KVB Field Experience (Industrial Boiler)
During the first phase of an EPA-sponsored study to determine
application of combustion modifications to control NOX from indus-
trial boilers, KVB Engineering, Inc., Tustin, California, field
sampled 47 representative boilers of various firing types ranging
from 1.3 to 65 kg/s of steam capacity.22 One of these units was a
large industrial boiler of 64.7 kg/s steam capacity fired by two
cyclone furnaces; it is identified in this report as boiler No. 7.
Details on this boiler unit, which is located in New York state,
are given in Section 3.3.5. Boiler No. 7 normally fires coal at a
baseline load of 40 kg/s (62% of max. load).
The combustion modifications testing performed on boiler No. 7
were limited in scope. Altogether eight test runs were made. One
LEA run and four switched fuel runs were performed in addition to
three baseline runs. Table 34 presents all of the cyclone boiler
data compiled in KVB's Phase I final report.22
Run 4 was conducted at about 9% lower excess air than that used
for normal firing at the same load (62% of full load). Comparing
Run 2 (baseline) with Run 4 shows that NOX emissions were reduced
by 11%, from 800 vppm to 755 vppm. At the same time, carbon mon-
oxide concentration did not increase.
Mixtures of coal and No. 5 fuel oil were used to determine their
effect on boiler NOX emissions. KVB's conclusions for the
switched fuel tests follow:
"A 9% increase of NOX was shown when 30% of No. 5 fuel
oil was fired along with the coal at normal excess air
levels (run No. 6). But when the excess oxygen was
88
-------
TABLE 34. BASELINE AND COMBUSTION MODIFICATION EMISSIONS DATA (BOILER LOCATED IN NEW YORK, BOILER NO. 7)20
OD
Run
NO.
1
2
3
4
5
6
7
8
Type
of
test
Baseline
Baseline
Baseline
LEA
Switched
fuel
Switched
fuel
Switched
fuel
Switched
fuel
Fuel
burned
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituirtinous
coal
70/303
70/30a
SO/503
50/50a
Test
load,
kg/s
steam
40.3
40.3
30.3
40.5
40.3
51.4
40.3
50.4
Flue
gas
Full 02,
load , %
% dry
78.4 3.2
62.4 3.4
46.8 3.2
62.6 3.1
62.4 3.4
79.5 3.6
62.4 3.5
77.9 3.7
Stack
temp,
K (°F)
418
(293)
418
(293)
418
(293)
418
(293)
423
(302)
423
(302)
421
(298)
421
(298)
Dry
3% 02
basis ,
vppm
790
800
742
755
710
860
716
797
NOX
CO S02 SO 3
ng/J ng/J ng/J ng/J
(lb/106 Btu) vppm (lb/106 Btu) vppm (lb/106 Btu) vppm (lb/106 Btu)
473
482
447
455
408
494
408
456
(1
(1
(1
(1
(0
(1
(0
(1
.10) 0
.12) 0
.04) 0
.06) 0
.95) 0
,15) 0
.95) 0
.06) 0
0 NA NA NA NA
0 1,122 937 (2.18) 13 10.7 (0.025)
0 NA NA NA NA
0 NA NA NA NA
0 NA NA NA NA
0 NA NA NA NA
0 NA NA NA NA
0 NA NA NA NA
Particulates,
ng/J
(lb/106 Btu)
NA
513 (1.19)
NA
NA
184 (0.43)
NA
NA
NA
The ratio of bituminous coal and No. 5 fuel oil fired during testings.
NA = not available-
-------
returned to the baseline level of 3.4% (as measured in
the flue gas) and the load was reduced to baseline level
in run No. 5, the NOX decreased by 11% (compare run
Nos. 2 and 5). Apparently, at this mixture the nitrogen
oxides formation was very sensitive to the amount of
excess air being fired. A 50-50 mixture of coal and oil
showed no change in NOX; however, there was insufficient
time available to investigate completely whether or not
it was possible in this latter case to lower the excess
oxygen and thereby lower the NOX, as had been done in
run No. 5."
As was true with Exxon's tests, KVB observed that the cyclone-
fired boiler had the highest NOX emission levels of any other
boiler firing method that they field sampled. It was noted during
KVB's tests that the molten slag acts as thermal insulation, help-
ing to produce more of an adiabatic combustion zone. In addition,
of all boilers tested during KVB's study, boiler No. 7, the only
cyclone boiler, had the smallest furnace volume per unit heat
release rate [0.19 m3/MW (gross)] and the highest burner heat
release rate [75 MW (gross)/burner]. These two factors account
for high combustion temperatures (>1920 K) and associated high NOX
emissions.
4.2.4 Load Reduction Field Test Data
Under reduced load conditions, a boiler operates at a fraction of
its maximum steaming rate because of lowered fuel and air settings.
Load reduction is commonly applied to boilers at off-peak loads
to save fuel. For cyclone-fired boilers, the maximum load reduc-
tion is about 40% of the maximum continuous steam rating. Reduc-
tion below this point can result in flame instability with pos-
sible loss of ignition, lack of adequate steam temperature control,
and excessive slagging on cyclone walls when firing coal due to
decreased cyclone furnace temperature.
Load reduction of a boiler unit results in decreased combustion
turbulence and volumetric heat release rate. The net effect of
load reduction is to produce lower effective peak temperatures for
NO formation in the primary and secondary furnace sections of the
boiler unit. This ultimately results in lower stack NOX emissions.
Load reduction in cyclone-fired boilers results in consistently
lower NOX emissions when compared to the same boilers at full or
normal loads. Load reduction primarily affects thermal NO forma-
tion and cannot effectively reduce the conversion of fuel-bound
nitrogen to NO. In addition, it is usually considered an econom-
ically unattractive method for reducing NOX emissions because of
the penalties incurred and because of reduced thermal efficiency
and reduced boiler flexibility.
90
-------
Figure 23 summarizes the available data on load reduction for six
cyclone-fired boilers. (These data also appear in Section 3.3 of
this report.) NO emissions levels were determined for these
boilers at both full and partial loads, making reduction compari-
sons possible. The only data available were for bituminous-coal-
fired and oil-fired units. None of the Commonwealth Edison data
presented previously in Section 3.3.7 are presented or discussed
here because of insufficient data on the exact boiler loads and
uncertainties as to whether the data were corrected to a compar-
able dry 3% O2 basis.
CO
I7i
CD
cv
O
ac.
o
1300
1250
1200
1150
1100
1050
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
SYMBOL
BOILER
I.D. No.
MAX. N0x
REDUCTION, %
38
25
23
6
61
41
EXTRAPOLATION
BITUMINOUS
COAL FUEL
DATA
.NEW SOURCE PERFORMANCE
STANDARD APPLICABLE TO COAL FIRING
RESIDUAL
OIL FUEL
DATA
•NEW SOURCE PERFORMANCE STANDARD APPLICABLE TO OIL FIRING
, s , « . . . .
4.6^ i
40 50 60 70 80 90 100
% OF MAXIMUM CONTINUOUS RATED BOILER LOAD
Figure 23.
Overall reduction of NOX emissions for six
coal- and oil-fired cyclone furnace boilers
using load reduction (stack % O2 levels
indicated adjacent to data points).
91
-------
The purpose of Figure 23 is to show the degree of NOX reduction
achievable when load reduction was the only combustion modifica-
tion applied to six specific boiler units. Five of the six units
tested showed an overall reduction in NOX emissions as load was
reduced. One bituminous-coal-fired unit showed a 10 vppm gain in
NOX as the load was reduced by 16%. The highest overall reduction
in NOX—61%—was achieved in the two residual-oil-fired units.
Bituminous-coal-fired units showed overall NOX reductions ranging
from 7% to 38% compared to full or normal load NOX levels. Avail-
able data for the six boiler units shown in Figure 23 are pre-
sented in summary form in Table 35. Figure 23 indicates that the
amount of excess air fed to the furnaces at reduced loads gener-
ally increases or remains constant in comparison to full-load con-
ditions. Increasing excess air at low loads serves to negate any
dramatic gains in NOX reduction by furnishing more N2 and Q£ for
thermal and fuel NO formation. In oil-fired boiler No. 13, for
example, the percent 02 in the stack rose from 1.5% at full load
to 4.2% at reduced load (see Figure 23 and Table 35). This cor-
responds to an excess air level of 7.3% at full load and 24% at
reduced load. It is partially for this reason that the NOX
reduction in boiler unit No. 10 (41%) is not as dramatic as that
of boiler No. 12 (61%), which operated at near constant excess air
as load was reduced. Why operators increase the excess air at low
loads is discussed further in Section 4.3.3.
Several other trends in NOX control by load reduction become appar-
ent by further data analysis. In this analysis, the boiler load
span between 80% and 100% is investigated because this is the span
where the bulk of available data exists (boiler Nos. 1, 4, 6, 12,
and 13, Figure 23).
None of the data covered the entire span. Hence, they were extrap-
olated over the load range between 80% and 100%. For a 20% reduc-
tion in load, an average NOX reduction of 29% was calculated for
the three coal-fired boilers (boiler Nos. 1, 4, 6), and a 19%
reduction was estimated for the two oil-fired boilers (boiler
Nos. 12 and 13). Thus, it appears that slightly better NOX reduc-
tions with reduced boiler loads could be obtained with coal-fired
units in the load range between 80% and 100%. A summary of the
calculations is given in Table 36.
Even with the significant decreases in NOX emissions achieved by
load reduction, Figure 23 shows that only boiler No. 12 was able
to meet the NSPS for NOX, and this occurred at a load reduction of
over 40%. The approximate NSPS limits in units of vppm NOX for
both coal and oil are also shown in Figure 23 (dotted horizontal
lines).
The data for all six boiler units, whether fired with oil or coal,
indicate that NOX emissions levels remain fairly constant with
widely varying boiler unit size. For example, at full load the
92
-------
TABLE 35. LOAD REDUCTION TEST DATA
Boiler
No.
1
1
4
4
6
6
7
7
7
12
12
13
13
13
Boiler identification
Unidentified utility boiler
Unidentified utility boiler
Unidentified utility boiler
Unidentified utility boiler
TVA Paradise Station, unit
No. 1, utility boiler
TVA Paradise Station, unit
No. 1, utility boiler
Unidentified industrial
boiler located in New York
Unidentified industrial
boiler located in New York
Unidentified industrial
boiler located in New York
Unidentified utility boiler
Unidentified utility boiler
Atlantic City Electric
England Station, unit
No. 1 utility boiler
Atlantic City Electric
England Station, unit
No. 1 utility boiler
Atlantic City Electric
England Station, unit
No. 1 utility boiler
Testing
agency
NAPCA
NAPCA
BSW
B&W
Exxon
Exxon
KVB
KVB
KVB
Exxon
Exxon
Exxon
Exxon
Exxon
Exxon
Fuel
burned
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituminous
coal
Bituminous
coal
Residual
oil
Residual
oil
Residual
oil
Residual
oil
Residual
oil
Heating
value,
MJ/kg
32.4
32.4
NAa
cl
NA
25.8
25.8
30.7
30.7
30.7
44.6
44.6
NAa
NA3
NAa
Nitrogen
in fuel ,
wt %
1.4
1.4
NAa
NA
1.3
1.3
1.6
1.6
1.6
0.5
0.5
^
NAa
NA3
NA
Maximum
continuous
steaming
rate,
kg/s
171.5
171.5
NAa
NAa
618
618
64.7
64.7
64.7
309
309
117
117
117
Boiler
load,
98
75
100
85
95
77
78
62
47
92
57
98
77
45
NOx,
dry
3% 02
basis,
vppm
1,204
742
975
730
1,155
386
790
800
742
530
206
441
404
261
reduction,
(gain) , %
38
]-j
25
b
23
(1)
6
61
8
41
02 in
flue
gas.
6.4
6.8
5.4
5.2
5.3
5,3
3.2
3.4
3.2
4.0
4.6
1.5
2.7
4.2
Symbol on
Figure 23
0
m
m
0
A
A
V
V
V
aNA = not available.
Not applicable.
-------
TABLE 36. NO REDUCTIONS FOR A 20% REDUCTION IN BOILER LOAD
Boiler
No.
1
4
6
Average
12
13
Average
Fuel
Coal
Coal
Coal
Oil
Oil
NOX at
100%
load ,
vppma
1,240
975
1,220
1,145
600
445
522
NOX at
80%
load ,
vppma
840
660
930
810
420
410
415
NOX reduction,
vppm
400
315
290
335
180
35
107
%
32
32
24
29
30
8
19
Extrapolations from Figure 23, vppm dry 3% O2
basis.
coal-fired units with boiler steaming rates between 100 kg/s and
600 kg/s showed emission levels ranging between 1,000 vppm and
1,200 vppm. At three-quarter loads, the boilers with steaming
rates between 100 kg/s and 500 kg/s emitted between 750 vppm and
850 vppm. The trend is similar for oil-fired boilers, but NOX
emission levels are substantially lower, normally one-half or less
of coal emissions for similar loads and sizes. These trends are
illustrated in Figure 24.
1550
1450
1350
1250
1150
1050
950
850
750
650
550
450
350
250
isn
SYMBOL BOILER I.D. No.
-
o 1
0 4
0 6
• 7
* 12
f98 T 13 95
/ P
/ /
/ /
-62 / ^
-/*78 I7r BITUMINOUS COAL
"7 FUa DATA
/92 RESIDUAL OIL
FUa DATA
..
100 200 300 400 500 600
BOILER STEAMING RATE, kg/S
700 800
Figure 24.
NOX emissions as a function of boiler
size and % load (boiler loads
indicated adjacent to data points).
94
-------
4.3 IMPLICATIONS OF APPLYING STATE OF THE ART NOX COMBUSTION
MODIFICATIONS TO EXISTING CYCLONE COMBUSTION UNITS
This section examines the information presented in Section 4.2,
including operating experience with cyclone boiler and furnace
modifications for potential application of these modifications to
existing cyclone boilers.
In summary, the most consistent NOX emission reduction in any of
the boilers tested was achieved with load reduction. Reducing
boiler load from 100% to 80% resulted in average NOX reductions of
19% with oil-fired units and 29% with coal-fired units. However,
even with this reduction in NOX emissions, the cyclone boilers
could not meet the New Source Performance Standard.
The fuel type and characteristics generally have some influence on
the NOX level emitted by cyclone boilers. From rather limited
information, it was observed that natural gas-, residual oil-, or
lignite-fired boilers had NOX emissions significantly lower (up to
50%) than units burning the following fuels:
• bituminous coal
• subbituminous coal
• blends of bituminous/subbituminous coals
• blends of fuel oil and bituminous coals
The data in this report indicate that the NOX NSPS cannot be met
at normal full-load firing for any type of fuel in a cyclone
boiler unit, with the possible exception of lignite. The proposed
lignite standard was met during one test conducted by B&W at very
low excess air levels and with supplemental oil fuel (refer to
Section 4.2.1.1, boiler No. 9). Thus fuel switching does not
appear to be a viable method of NOX control for purposes of meet-
ing the appropriate fuel NOX NSPS.
Both load reduction and fuel switching may, however, be practical
interim measures in reducing NOX emissions from some units during
serious episode conditions. But both of these methods are the
least desirable to the boiler owner for operational and economic
reasons. Implications of applying these combustion modifications
to cyclone furnaces/boilers are further discussed in Sections 4.3.1
and 4.3.2.
A form of flue gas recirculation (FGR) was applied to only one
unit as witnessed by Exxon's data for boiler No. 6. This short
duration test, with no hardware modifications, produced no NOX
reduction. FGR for this particular boiler NOX emission test
implied that the recirculated gas was reinjected downstream of the
cyclone furnace combustion zone rather than directly into the
cyclone where it could possibly be more effective.
95
-------
Low excess air (LEA) firing gave mixed results for different
boilers and fuel types. Of the boilers tested under LEA firing
conditions, a lignite boiler showed the highest NOX reduction;
residual oil units showed lesser reductions; and a coal-fired unit
showed the least reduction. The general trend observed indicates
that NOX can be reduced by this method at the cost of higher CO
emissions. Reducing excess air in one lignite boiler (boiler No. 9,
Section 4.2.1.1) resulted in NOX reductions of 50% with a negligi-
ble CO emission increase. Supplemental oil fuel was required at
the lower air levels to maintain ignition. Applying LEA firing to
residual-oil-fired units (Section 4.2.2) reduced NOX by 10% to 16%
at full load and yielded acceptable CO concentrations (<200 vppm).
Reducing excess air further to achieve higher NOX reductions
resulted in excessive CO, with resultant boiler carbon loss and
loss of boiler efficiency. A bituminous-coal-fired boiler unit
running under LEA conditions showed an NOX reduction of 11% with
no change in CO emissions (0 vppm). Thus NOX reductions achieved
with LEA firing alone in cyclone boilers have not been spectacular,
and the resultant NOX emission levels are well above NSPS.
Staged firing was simulated in several coal-, oil-, and gas-fired
boilers. The first method, two-staging, showed a 28% to 36% reduc-
tion in NOX emissions when eastern coal was used. Two-staging
with gas firing showed a 48% reduction in NOX. No test data or
details were available to support these figures given by B&W, the
originators of the concept.
Pattern firing, the other simulated staged firing concept applied
to cyclone boilers, gave mixed results with different boiler units.
While B&W claims that 21% to 24% NOX emission reductions are
achievable using this method when burning residual oil, Exxon
noticed an increase in NOX (about 34%) with pattern firing due to
increased combustion turbulence in the operating cyclones. Exxon
also noticed no change in NOX level when a bituminous-coal-fired
unit was subjected to pattern firing. In either case, B&W does
not recommend these staged firing methods for application to exist-
ing units for reasons stated in Section 4.2.1.4.
According to B&W, the cyclone boiler manufacturer, all practical
possibilities of NOX reduction have been exhausted for existing
cyclone boilers. Even though data on NOX emissions from a variety
of cyclone combustion units are available, it does not appear that
the data are sufficient to fully evaluate the effectiveness of dif-
ferent combustion modifications to reduce NOX emissions. It is
true that most combustion modifications have a strong influence on
NOX formation within a cyclone boiler. Despite this influence,
NOX emission reduction to the level that would satisfy existing
environmental regulations was rarely achieved. Cyclone furnaces
are rather unique and complex systems that were developed prima-
rily to burn low quality fuels. The combustion process in these
furnaces is very efficient due to the highly turbulent conditions.
Moreover, cyclonic furnaces must operate within a specific range
96
-------
of operating conditions which, if not met, defeat the purpose for
which the cyclonic furnace was originally developed, and the
cyclonic method of combustion loses its advantages compared to
other conventional combustion methods.
The variables that influence cyclonic combustion are not fully
understood. This is especially true for variables that can influ-
ence formation of NOX in cyclones. Conversations with boiler
owners and the manufacturer (B&W) alike stressed that cyclonic
furnaces do not all operate alike. Even in cases of the same
boiler fired by multiple but identical cyclones burning identical
fuels, it is believed that not all cyclones operate in the same
way. This variability requires further definition. When firing
at LEA conditions, for example, although the total boiler flue gas
O2 levels could be readily determined, the 02 levels of the
exhaust gases from individual cyclones are not known. As a conse-
quence, NOX contributions from different cyclones might vary sig-
nificantly. Therefore, if the combustion characteristics of each
cyclone are individually monitored, it is believed that better NOX
reductions might be achievable. It is not possible to state in
advance, however, that this reduction would be adequate to meet
existing NOX emission regulations. Additional restrictions and
controls applied to the cyclone boilers would further reduce their
flexibility, increase their costs, and worsen their acceptance by
the users.
4.3.1 Switched Fuel Firing
Bituminous-coal-fired units produce the most NOX. Switching from
coal to another fuel such as natural gas can reduce NOX by as much
as 50%. The data contained in this report also indicate that
(1) there is no NOX reduction advantage when bituminous coal is
blended with subbituminous coal or No. 5 fuel oil, (2) there is
only a slight NOX reduction advantage when subbituminous coal is
fired in place of bituminous coal, (3) switching from bituminous
coal to lignite can reduce NOX as much as using natural gas,
(4) switching from bituminous coal to residual oil fuel also can
reduce NOX by as much as 58%.
The implications of fuel switching to achieve NOX control may be
complex. Fuel availability, economics, operating difficulties,
and possible derating of boilers must be considered on an individ-
ual boiler basis. Even if fuel switching were implemented for NOX
control of cyclone boilers, it is unlikely that the NSPS for the
substitute fuel will be met, but the absolute contribution of NOX
from the cyclone-fired units could be significantly reduced.
4.3.2 Load Reduction
Load reduction effectively reduces the quantity of NOX emitted
from the stack. The strategy of load reduction as an NOX control
alternative implies that boilers be operated at sustained reduced
97
-------
loads. The concept of operating utility or industrial boiler
units at sustained reduced loads is not attractive to the boiler
operator. The major factors to consider under low-load conditions
are steam temperature control to maintain thermal efficiency, abil-
ity to meet load fluctuations, and fouling. Full-load steam tem-
peratures (i.e., temperatures as high as at full load), are desira-
ble at lower loads to maintain prime mover (steam turbine) effi-
ciency. There are two widely used methods for steam temperature
control at partial load. These will be discussed in terms of sus-
tained boiler operation such as would be required to implement NOX
control via load reduction.
The bulk of the heat absorbed in the superheater, reheater (if so
equipped), and economizer is transferred by convection.2 In a
boiler unit, the heat transfer to these surfaces is controlled by
two variables which both control convective heat transfer; namely,
gas temperature and gas mass velocity. The extent to which these
variables are controlled largely determines the degree of super-
heat recoverable at low or partial loads.
The first widely used steam temperature control method is excess
air control. This method primarily affects the gas mass velocity
rather than the gas temperature. When excess air steam tempera-
ture control is used in a boiler unit, the boiler operators will
slightly increase the excess air level at partial load. This
increased gas flow produces an increase in recoverable superheat.5
At the same time, the gas temperature within the boiler unit and
the furnace heat absorption decrease, affecting boiler efficiency.
The resulting greater gas weight increases the heat loss through
the stack, which also lowers the overall boiler efficiency. The
boiler operators, however, under low boiler load conditions, are
willing to sacrifice boiler efficiency to some extent for the
amount of recoverable superheat, which is important to maintaining
high steam temperatures at the steam turbine inlet. The concept
of steam temperature control by excess air control is illustrated
in Figure 25.
Steam temperature control by the excess air method was probably
used in oil-fired boiler Nos. 12 and 13 and coal-fired boiler No. 2
whose NOX data are shown in Figure 23 and Table 35. The excess
air levels were increased by 6% in boiler No. 1, 15% in boiler
No. 12, and 35% in boiler No. 13 during the load reduction tests.
It is not known, however, if the excess air was increased in these
units solely for steam temperature control. There is also the
problem of reduced cyclone furnace combustion turbulence at lower
loads, which may require correction with more excess air to
achieve adequate fuel combustion.
Another common method of steam temperature control in cyclone-
fired boilers is flue gas recirculation (FGR). This method works
on similar principles as excess air control but results in higher
boiler efficiencies than excess air control. FGR influences both
98
-------
GREATER GAS WEIGHT
UP STACK INCREASES
STACK LOSS
MORE HEAT
CONTENT IN
GASES ENTERING
SUPERHEATER
INCREASED
COMBUSTION
AIR
NORMAL
COMBUSTION
AIR
ADDITIONAL \f HEAT AVAILABLE FOR
SUPERHEATER AND REHEATER ABSORPTION
RESULTING FROM INCREASED
EXCESS AIR /TOTAL HEAT AVAILABLE
HEAT IN GAS
ENTER ING SUPERHEATER
ABSORPTION IN
CONVECTION SURFACES
25 50 75
LOAD, %
HEAT IN GAS
LEAVING
ECONOMIZER
Figure 25.
Steam-temperature control by
use of increased excess air.
gas mass velocity and gas temperature, the major convective heat
transfer variables.
The flue gas used for recirculation is normally withdrawn from the
economizer outlet. The hot gas is then distributed by the FGR sys-
tem to a point above the combustion zone (termed gas recirculation)
and/or to a point near the furnace exit (termed gas tempering).
The points of reintroduction are dictated by the control desired.2
Gas recirculation flow controls furnace absorption and superheat
and reheat temperatures, while gas tempering conditions and con-
trols the temperatures of the flue gas entering the superheater
sections only. Proportioning the amounts of flue gas recirculated
to each point in the boiler results in control of steam temperature
Figure 26 depicts a cyclone-fired boiler unit equipped with gas
tempering and gas recirculation control. Boiler No. 6 is equipped
with an FGR system of this sort. The effects of FGR on boiler
heat absorption patterns are shown in Figure 27. Both FGR and
excess air control may be used simultaneously in a specific boiler
unit for steam temperature control.
FGR control of steam temperature at low loads has several advan-
tages. Sensible heat loss up the stack is negligible compared to
excess air control and is comparable to the loss for full-load fir-
ing. In addition, boiler slagging and fouling are minimized
because the gas recirculation temperature can help to keep the
99
-------
GAS
TEMPERING
CONTROL
DAMPER
\J
[SECONDARY
SUPERHEATER IEATI
GAS
TEMPERING
PRIMARYB
SUPERHEATER
.ECONOMIZER:
GAS
RECIRCULATION
STEAM-
TEMPERATURE
CONTROL DAMPER
GAS
RECIRCULATING
FAN
Figure 26.
Cyclone-fired boiler with gas tempering
for gas-temperature control and gas
recirculation for control of furnace
absorption and reheat temperature.2
100
-------
1316
I 2400 )
20 30 40
GAS TEMPERING,
(a) Effect of gas tempering on
heat-absorption pattern at
a constant firing rate.
i o
20 30
GAS RECIRCULATION,
-10
60
(b) Effect of gas recirculation
on heat-absorption pattern
at a constant firing rate.
Figure 27. Flue gas recirculation.2
-------
boiler temperatures below the ash fusion point.^ However, sus-
tained operation of the FGR system at low loads can become expen-
sive. This is primarily due to the higher operating cost of its
operation (increased FGR fan horsepower) when compared to excess
air control. Exact horsepower cost figures were not available.
Besides being a steam temperature control method at low boiler
loads, FGR is considered as a combustion modification for NOX con-
trol in large utility and industrial boilers of various firing
types (if FGR equipped) at full loads. In pulverized-coal-fired
units, for example, FGR serves to lower the peak flame zone temper-
ature and reduces the amount of 02 available for NO formation.
Only one instance of its application to cyclone firing was in the
literature. A limited test of FGR on NOX emissions was performed
during Exxon's study, as discussed in Section 4.2.2. NOX emis-
sions for this one bituminous-coal-fired boiler were not affected
by varying flue gas recirculation or gas tempering flows This is
probably due to the fact that most of the combustion occurs within
the cyclone and FGR has very little if any effect on NOX formation
after the combustion gases leave the cyclone.
Either of the two steam temperature control methods discussed
above (i.e., excess air and FGR) were originally developed to keep
the boiler unit efficiency fairly constant throughout load fluctua-
tion. Neither control method was intended, nor was it suitable,
for sustained operation such as might be required for NOX control
implementation. Loss of boiler efficiency and higher boiler oper-
ating costs are the associated penalties.
In addition to these drawbacks, a derated boiler will not be able
to meet the load demands for which it was originally designed.
Implementation of load reduction might in some cases require peak-
ing power supplied by prime movers such as reciprocating engines
and gas turbines whose NOX emissions are also quite high. Such
strategy could defeat the purpose of cyclone boiler load reduction
for NOX control. Further study is recommended in this area.
When coal-fired cyclone boilers are operated at reduced loads, a
serious problem with fouling can occur. The lower combustion tur-
bulence, heat release rate, and temperatures in the furnace at par-
tial loads can cause slag buildup with certain high-ash-fusion-tem-
perature coals which are marginally suitable for cyclone firing.
Sustained burner operation at low loads can result in plugging and
resultant shutdown. Areas that require further investigation in
this regard before implementing load reduction for NOX control
include (1) changing to coals having desired slagging characteris-
tics at partial loads and (2) use of fluxing agents to lower slag-
ging viscosity of marginal coals.
The load reduction NOX data analysis performed in Section 4.2.4
suggests that the NOX NSPS cannot be met at practical reduced
loads. Implementing load reduction to reduce NOX emissions from
102
-------
cyclone boilers during severe urban pollution episodes may be a
practical control alternative; but, clearly, sustained operation
of cyclone-fired boiler units at partial loads is not likely to be
readily accepted until all implications are resolved.
4.4 RECOMMENDATIONS FOR FURTHER WORK
Previous sections of this report have summarized available data on
NOX emissions from cyclone furnaces/boilers. Even though large
amounts of data were available in most cases, it was not possible
to obtain information on all the details and conditions under
which the data were generated. Thus, it is not possible to be cer-
tain that all the tests were thorough, comparable, and representa-
tive of all permanent cyclone operations and that all possibili-
ties for cyclone boiler modification for achieving reduced NOX
emissions were thoroughly explored.
Because the boilers tested were fully operational, there was too
much at stake for the boiler owner as well as for the boiler manu-
facturer to expose the boiler to conditions that would signifi-
cantly deviate from the conditions at which the boiler was origi-
nally designed to operate. Examples of factors that significantly
influenced many of the tests include the cost of the test program
and the necessary boiler modifications, safety, serious damage to
an existing facility, and maintaining boiler efficiency during the
test program. Because of these reasons, some furnace/boiler modi-
fications could not be properly tested for their effect on NOX
formation.
Also, the penalties associated with these modifications could not
be properly assessed. This is especially true for the modifica-
tions involving flue gas recirculation (FGR) and forms of staged
firing. To be effective, these modifications should be applied as
near the flame front as possible. The bulk of combustion and asso-
ciated NO formation in cyclone boilers occurs within the cyclone
furnace. Any FGR or staged-firing concept applied to places where
the combustion is nearly complete could not be too effective in
influencing NOX formation.
The cyclone furnace has special operational characteristics. It
operates at high temperatures (3000°F). Coal particles are caught
and burned in the molten slag. This is a well balanced system
that has demonstrated low corrosion rates. A change of the condi-
tions under which the cyclone furnace operates can upset this deli-
cate balance and cause serious corrosion, equipment damage, and
accidents (refer to Section 3.1.6). Operation under changed condi-
tions permanently would require re-evaluation of the balance and
development of materials (e.g., for the boiler tubes and/or for
the protection of the tubes) with adequate corrosion resistance in
the new environment. The cyclone furnace would have to be
redesigned and rebuilt using these materials. Then the rebuilt
furnace would be able to operate under fuel-rich conditions (low
excess air).
103
-------
The low-excess-air (LEA) combustion modification as applied to
cyclone furnaces could not be fully evaluated during test programs
mentioned in this report due to the apparent danger of corrosion
and equipment damage. Therefore, the potential success of the LEA
combustion modification for reducing NOX emissions from cyclone-
fired boilers will depend on the ability to operate the cyclone
furnace safely under fuel-rich combustion conditions. In the opin-
ion of B&W, there are presently no materials that could be used in
this application. Nevertheless, a study is recommended to EPA for
their consideration which would evaluate the potential for rede-
signing cyclone furnaces for permanent operation in a reducing
atmosphere. The study should also determine the costs of furnace
redesigning if applied to the existing cyclone-fired units.
It is not certain whether any of the possible cyclone boiler modi-
fications could reduce NOX emissions to the level adequate for
meeting existing NOX emission regulations and at the same time
maintaining cyclone boilers as a competitive combustion technique.
In order to determine this, a comprehensive boiler test program
should be developed. This program should concentrate on determin-
ing the causes for NOX formation and their location in the
cyclonic combustion process. Better understanding of cyclonic com-
bustion would surely identify new opportunities as well as proper
conditions for application of available combustion modifications
to cyclone boilers to attain NOX emission control. Sampling a
large number of cyclone combustion facilities for NOX emissions
would not be considered an adequate substitute for such a test pro-
gram, which should evaluate:
• Variations in operation of individual cyclones in
multicyclone-fired facilities and the influence of
these variations on the total boiler NOX emissions;
• The effect of boiler modifications on cyclone furnace
NOX emissions, with a comprehensive evaluation of vari-
ables and conditions having a strong influence on NOX
formation;
• Long duration operation of the successful modifica-
tions to develop reliability and operational data;
• The applicability of the successful modifications to
all cyclone boiler facilities; and
• The cost of such modifications.
An example of a test program for further evaluation of cyclone fur-
nace combustion conditions and their influence on NOX formation
has been prepared by KVB Engineering, Inc., Tustin, California.
This test program is offered to EPA for consideration as an ini-
tial step in the assessment of cyclone boilers and is presented in
Appendix C.
104
-------
Whether any NOX emission control methods, even if effective in
meeting NOX emission standards, will be practical and acceptable
to boiler operators is not presently known. For the last 6 years,
no cyclone boilers have been sold in the United States. The large
majority of cyclone boilers (49% steaming capacity) are 12 to 32
years old. The normal life expectancy for boiler facilities is
from 25 years to 35 years. Thus within the next 3 years to 13
years, some 49% of the present cyclone steaming capacity will be
obsolete and will have to be replaced.
This process may be dramatically encouraged by a strong enforce-
ment of existing NOX emission standards. A study is therefore rec-
ommended to determine the influence of the NSPS applicable to NOX
emissions from cyclone boilers on American industries.
105
-------
REFERENCES
1. New Cyclone-Fired Boiler for E. H. Werner Station, Jersey
Central Power & Light Company. Bulletin G-81 by the Babcock
& Wilcox Company/ New York, New York, 1953. 9 pp.
2. The Babcock & Wilcox Company. Steam—Its Generation and
Use. 38th Edition. New York, New York, 1972.
3. Potterton, S. T. Combination Fuel Firing in Cyclone Furnaces.
The Babcock & Wilcox Company. (Presented to 1970 Indus-
trial Coal Conference. Lafayette, Indiana. October 7, 1970).
Barberton, Ohio. 9 pp.
4. Shields, Carl D. Boilers - Types, Characteristics, and Func-
tions. F. W. Dodge Corporation, New York, New York, 1961.
559 pp.
5. Kessler, G. W. Cyclone Furnace Boilers. The Babcock &
Wilcox Company. (Proceedings of the American Power Confer-
ence, 1954). New York, New York. pp. 78-90.
6. Grunert, A. E., L. Skog, and L. S. Wilcoxson. The Horizontal
Cyclone Burner. Transactions of the ASME, 69:613-634, August
1947.
7. Newkirk, M. Cyclone-Fired Pressurized Steam Generator.
Transactions of the ASME, Journal of Engineering for Power,
73:215-223, 1951.
8. Stone, V. L. and I. L. Wade. Operating Experiences with
Cyclone-Fired Steam Generators. Mechanical Engineering.
74:359-368, 1952.
9. Lowry, H. H. (ed). Chemistry of Coal Utilization. Supple-
mentary volume, prepared by the committee on Chemistry of
Coal, Division of Chemistry and Chemical Technology, National
Academy of Sciences - National Research Council. John Wiley
and Sons, Inc., New York, New York, 1963. 1,142 pp.
10. Selvig, W. A. and F. H. Gibson. Analyses of Ash from United
States Coals. Bull. No. 567. Bureau of Mines, U.S. Depart-
ment of the Interior. 1956. 33 pp.
106
-------
11. Perry, R. H. and C. H. Chilton (eds). Chemical Engineers'
Handbook. 5th edition. McGraw-Hill Book Company, New York,
New York, 1973. 1,650 pp.
12. Perry, R. H., C. A. Chilton, and S. O. Kirkpatrick (eds).
Chemical Engineers' Handbook, 4th edition. McGraw-Hill Book
Company, New York, New York, 1963. 1,650 pp.
13. Sedor, P., E. K. Diehl, and 0. H. Barnhart. External Corro-
sion of Superheaters in Boilers Firing High-Alkali Coals.
Transactions of the ASME, Journal of Engineering for Power,
82:181-190, ^1960.
14. Holyoak, R. H. Burning Western Coals in Northern Illinois.
Commonwealth Edison Company, Chicago, Illinois, ASME Paper
73-WA/FU-4, August 17, 1973. 8 pp.
15. Federal Register. 36 (247):24879, December 23, 1971.
16. Shimi2u, A. B., R. J. Schreiber, H. B. Mason, et al. NOX
Combustion Control Methods and Costs for Stationary Sources.
Summary Study. Aerotherm Division, Acurex Corporation, U.S.
Environmental Protection Agency, EPA 600/2-75-046, September
1975. 104 pp.
17. National Emissions Data System (NEDS). Computer File List-
ing of Detailed Point Sources of Utility and Industrial
Cyclone-Fired Boilers, March 1, 1976. 134 pp.
18. Cuffe, S. T. and R. W. Gerstle. Emissions from Coal-Fired
Power Plants: A Comprehensive Summary. U.S. Department of
Health, Education, and Welfare, NAPCA, Durham, North Carolina,
1967. 26 pp.
19. Bartok, A. R. , Crawford, and G. J. Piegari. Systematic Field
Study of NOX Emission Control Methods for Utility Boilers.
Esso Research and Engineering Company (for: U.S. Environ-
mental Protection Agency. Research Triangle Park, North
Carolina, Contract No. CPA 70-90). December 31, 1971.
215 pp.
20. Crawford, A. R., E. H. Manny, and W. Bartok. Field Testing:
Application of Combustion Modifications to Control NOX Emis-
sions from Utility Boilers. Exxon Research and Engineering
Company, Government Research Laboratory. (for: U.S. Envi-
ronmental Protection Agency, Washington, D. C.
EPA-650/2-74-066). June 1974. 151 pp.
107
-------
21. Cato, G. A., H. J. Buening, C. C. DeVivo, B. G. Morton, and
J. M. Robinson. Field Testing: Application of Combustion
Modifications to Control Pollutant Emissions from Industrial
Boilers—Phase I. KVB Engineering, Inc. Tustin, California.
(for EPA, Office of Research and Development. EPA-650/2-74-
078-a). PB 238 920. October 1974. 196 pp.
22. Hollinden, G. A., S. S. Ray, N. D. Moore, J. T. Reese, and
C. Gottschalk. NOX Control at TVA Coal-Fired Steam Plants.
Tennessee Valley Authority, Chattanooga, Tennessee. Paper
Presented at National Symposium ASME Air Pollution Control
Division. 30 pp.
23. Smith, W. S., and C. W. Gruber. Atmospheric Emissions from
Coal Combustion - An Inventory Guide. U.S. HEW, Public
Health Service, Division of Air Pollution. Cincinnati, Ohio.
April 1966. 112 pp.
24. Anon. Compilation of Air Pollutant Emission Factors. 2nd
edition, U.S. Environmental Protection Agency, April 1973.
pp. 1.1-1 to 1.4-3.
25. Anon. Steam-Electric Plant Factors, 1974 Edition. National
Coal Association, 24th edition, 1974. 110 pp.
26. Brown, R. A., H. B. Mason, and R. J. Schreiber. Systems
Analysis Requirements for Nitrogen Oxide Control of Station-
ary Sources. Aerotherm/Acurex Corporation (California).
U.S. Environmental Protection Agency, EPA-650/2-74-091, 1974.
27. Pohl, J. H. and A. F. Sarofim. Fate of Coal Nitrogen During
Pyrolysis and Oxidation. Fuels Research Laboratory, Massa-
chusetts Institute of Technology, Paper presented at "Sympo-
sium on Stationary Source Combustion" sponsored by Combustion
Research Section, U.S. Environmental Protection Agency,
September 24-26, 1975 (Atlanta, Georgia), 22 pp.
28. Fine, D. H., S. M. Slater, A. F. Sarofim, and G. C. Williams.
Nitrogen in Coal as a Source of Nitrogen Oxide Emission from
Furnaces. Fuel, 53(4):120-125, 1974.
29. James, D. E. A Boiler Manufacturer's View on Nitric Oxide
Formation. The Babcock & Wilcox Company, Presented to the
Fifth Technical Meeting, West Coast Section of the Air Pollu-
tion Control Association, San Francisco, California,
October 8-9, 1970. 26 pp.
108
-------
APPENDIX A
CYCLONE-FIRED BOILER INSTALLATIONS
Tables A-l and A-2 list specific information related to each of
the 84 known cyclone-fired boiler installations. Information sup-
plied in these tables includes all known plant expansions to date
and was derived from a list of contract agreements made and fur-
nished to us by Babcock & Wilcox. Boiler steam flows listed are
capacities at maximum continuous rating. Table A-l pertains to
electric utility installations while Table A-2 presents informa-
tion on industrial and commercial boiler units.
109
-------
TABLE A-l. INSTALLATIONS OF CYCLONE-FIRED BOILER UNITS - UTILITIES
Number of units
State
Arkansas
Connecticut
Connecticut
Connecticut
Florida
Florida
Florida
Florida
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
I— i
j_, Illinois
O Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Plant or city
Helana
Middletown
Bridgeport
Bridgeport
Gannon
Gannon
Gannon
Gannon
coffeen
Cof feen
Calumet
Fisk
Joliet
Joliet
Kincaid
Pekin
Pekin
Ridgeland
Ridgeland
Ridgeland
Ridgeland '
Waukegan
Waukegan
Will County
Baldwin
Baldwin (Randolph
County)
Marion
Marion
Dallman Plant
Dallman Plant
Lakeside
Lakeside
Customer
Arkansas Power s Light Company
Hartford Electric
United Illuminating Company
United Illuminating Company
Tampa Electric Company
Tampa Electric Company
Tampa Electric Company
Tampa Electric Company
Central Illinois Power Company
Central Illinois Power Company
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Illinois Power Company
Illinois Power Company
Southern Illinois Power Company
Southern Illinois Power Company
Springfield Water, Light, and Power Department
Springfield Water, Light, and Power Department
Springfield Water, Light, and Power Department
Springfield Water, Light, and Power Department
Boilers
1
1
1
1
1
1
1
1
1
1
1
2
2
1
2
2
1
2
2
1
1
1
1
2
1
1
3
1
1
1
1
1
Cyclones
3
5
5
3
4
4
3
3
14
8
1
8
6
9
28
20
10
8
8
6
6
4
2
10
14
14
6
4
3
3
2
2
Cyclones
per boiler
8
5
5
3
4
4
3
3
14
8
1
4
3
9
14
10
10
4
4
6
6
4
2
5
14
14
2
4
3
3
2
2
Boiler steam flow,
kg/s
Primary
289.8
207.9
144.9
72.5
158.8
146.2
119.7
114.7
529.2
315.0
22.7
94.5
75.6
277.2
529.2
382.7
NAa
88.2
88.2
138.6
138.6
104.6
39.8
151.2
1,159.2
529.1
42.2
NA
86.9
86.9
40.3
40.3
Reheat
246.6
161.9
122.1
63.0
143.6
128.5
103.3
99.2
469.1
233.1
0.0
0.0
0.0
250.1
474.3
438.0
NA
0.0
. 0.0
126.9
126.9
92.4
0.0
136.1
477.2
477.2
0.0
NA
0.0
0.0
0.0
0.0
Total primary
steam flow,
kg/s
289.8
207.9
144.9
72.5
158.8
146.2
119.7
114.7
529.2
315.0
22.7
189.0
151.2
277.2
1,058.4
765.4
NA
176.4
176.4
138.6
138.6
104.6
39.8
302.4
1,159.2
529.1
126.6
NA
86.9
86.9
40.3
40.3
(continued)
-------
TABLE A-l (continued).
Number of units
State
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Indiana
Iowa
Iowa
Iowa
Kansas
Kansas
Kansas
Kentucky
Kentucky
Kentucky
Kentucky
Maryland
Maryland
Michigan
Minnesota
Minnesota
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Plant or city
State Line
Tanners Creek
Breed
Michigan City
Michigan City
Baileytown
Baileytown
Michigan City
Southerland
Missouri River
Sioux City
Muscatine
Quindaro Station
(No. 2)
Kaw
La Cygne
E. Smith
Paradise
Paradise
Paradise
Crane
Crane
St. Clair
Riverside
Stillwater
(A. S. King)
New Madrid
New Madrid
Thomas Hill
Thomas Hill
Allen S. King
Chamois
Asbury
Sibley
Customer
Commonwealth Edison
Indiana S Michigan Electric Company
Indiana S Michigan Electric Company
Northern Indiana Public Service Company
Northern Indiana Public Service Company
Northern Indiana Public Service Company
Northern Indiana Public Service Company
Northern Indiana Public Service Company
Iowa Electric Light S Power Company
Iowa Public Service Company
Muscatine Municipal Electric
Kansas City Board of Public Utilities
Kansas City Board of Public Utilities
Kansas City Power & Light
City of Owensboro
Tennessee Valley Authority
Tennessee Valley Authority
Tennessee Valley Authority
Baltimore Gas S Electric
Baltimore Gas & Electric
Detroit Edison Company
Northern States Power
Northern States Power
Associated Electric Co-op
Associated Electric Co-op
Associated Electric Co-op
Associated Electric Co-op
Central Electric Power Co-op
Empire District Electric Company
Missouri Public Service Company
Boilers
1
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Cyclones
9
11
8
6
10
8
4
NA
3
3
3
2
2
18
3
23
14
14
4
4
7
5
12
14
14
6
4
2
5
8
Cyclones
per boiler
9
11
8
2
10
8
4
NA
3
3
3
2
2
18
3
23
14
14
4
4
7
5
12
14
14
6
4
2
5
8
Boiler steam flow,
kg/s
Primary
277.2
483.8
368.3
47.3
407.0
315.0
151.2
407.0
72.5
132.3
85.7
72.5
53.6
780.3
132.3
1,008.0
617.4
617.4
171.0
171.6
264.6
189.0
484.5
548.7
548.1
241.4
157.5
52.4
163.8
325.6
Reheat
250.1
401.4
335.7
285.8
0.0
0.0
233.1
130.5
NA
64.6
115.5
0.0
64.9
46.0
632.5
114.7
700.6
444.2
444.2
131.4
130.3
239.5
166.3
417.4
503.4
510.3
210.9
138.6
0.0
143.0
279.6
Total primary
steam flow,
kg/s
277.2
433.8
368.3
141.9
407.0
315.0
151.2
407.0
72.5
132.3
85.7
72.5
53.6
780.3
132.3
1,008.0
617.4
617.4
171.0
171.6
264.6
189.0
484.5
548.7
548.1
241.4
157.5
52.4
163.8
325.6
(continued)
-------
TABLE A-l (continued).
t\J
Number of units
State
Missouri
Missouri
Missouri
Missouri
Missouri
Nebraska
Nebraska
New Hampshire
New Hampshire
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
North Dakota
North Dakota
North Dakota
North Dakota
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
South Dakota
South Dakota
Tennessee
West Virginia
West Virginia
West Virginia
Plant or city
Sibley
Sibley
Lake Road Pt.
Sioux Plant
St. Charles Co.
Sheldon
(Hallam)
Sheldon
(Hallaro)
Merrimack Pt.
Merrimack Pt.
Beesley's Pt.
Beesley's Pt.
Deepwater
Sayreville
Sayreville
South Amboy
Marion
Leland Olds
Center
Center
Beulah
Conesville
Conesville
Miles
Muskingum
Muskingum
Philo
Ben French
Big Stone
Thomas H. Allen
Willow Island
Rammer
Kammer
Customer
Missouri Public Service Company
Missouri Public Service Company
St. Joseph Power s Light
Union Electric
Union Electric
Consumers Public Power Company
Consumers Public Power Company
Public Service of New Hampshire
Public Service of New Hampshire
Atlantic City Electric Company
Atlantic City Electric Company
Atlantic City Electric Company
Jersey Central Power s Light
Jersey Central Power & Light
Jersey Central Power S Light
Public Service Company of New Jersey
Basin Electric
Minnkota Power Co-op, Inc.
Minnkota Power Co-op, Inc.
Ottertail Power
Columbus & Southern Ohio Electric
Columbus & Southern Ohio Electric
Ohio Edison Company
Ohio Power Company
Ohio Power Company
Ohio Power Company
Black Hills Power & Light Company
Ottertail Power
Tennessee Valley Authority
Monongahela Power Company
Ohio Power Company
Ohio Power Company
Boilers
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
3
1
2
1
Cyclones
2
2
2
10
10
3
3
7
3
4
3
3
4
4
2
8
12
12
7
12
4
4
8
5
5
3
1
12
21
5
10
5
Cyclones
per boiler
2
2
2
10
10
3
3
7
3
4
3
3
4
4
2
S
12
12
7
12
4
4
4
5
5
3
1
12
7
S
5
5
Boiler steam flow,
kg/s
Primary
56.7
56.7
78.8
414.5
414.5
99.4
95.8
289.8
102.7
141.8
116.8
77.6
113.4
113.4
59.9
308.7
378.0
NA
216.0
409.5
126.0
126.0
111.5
191.9
191.9
85.1
26.5
409.5
252.0
151.2
191.9
191.9
Reheat
0.0
0.0
69.9
373.0
373.0
0.0
86.9
252.0
90.5
124.1
102.3
68.0
89.5
89.5
47.3
248.2
344.0
NA
193.3
364.1
106.5
106.5
98.9
148.4
148.4
65.5
0.0
364.1
204.1
135.5
148.4
148.4
325.3
Total primary
steam flow,
kg/s
56.7
56.7
78.8
414.5
414.5
99.4
95.8
289.8
102.7
141. 8
116.8
77.6
113.4
113.4
59.9
308.7
378.0
NA
216.0
409.5
126.0
126.0
223.0
191.9
191.9
85.1
26.5
409.5
756.0
151.2
383.8
191.9
(continued)
-------
TABLE A-l (continued).
Number of units
State
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
U.S. TOTAL
Plant or city
Bay Front
Nelson Dewey
Nelson Dewey
Rock River
Rock River
Sheboygan
Sheboygan
Customer
Lake Superior District Power Company
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Power
Power
Power
Power
Power
Power
S
S
&
&
&
s,
Light
Light
Light
Light
Light
Light
Company
Company
Company
Company
Company
Company
Boilers
1
1
1
1
1
1
1
a
Cyclones
2
3
3
3
3
7
3
a
Cyclones
per boiler
2
3
3
3
3
7
3
Boiler steam flow,
kg/s
Primary
40.3
88.2
88.2
66.2
66.2
271.5
76.6
Reheat
0.0
78.6
78.0
57.3
57.3
238.1
0.0
Total primary
steam flow,
kg/a
40
88
88
66
66
271
75
24,253
.3
.2
.2
.2
.2
.5
.6
.3
Total of available data excluding information not available.
NA = not available.
-------
TABLE A-2. INSTALLATION OF CYCLONE-FIRED BOILER UNITS - INDUSTRIAL AND COMMERCIAL
Number of units
State
Alabama
Arkansas
Indiana
Indiana
Iowa
Iowa
Maryland
Michigan
Michigan
Michigan
Michigan
New Jersey
New York
New York
New York
New York
North Carolina
Ohio
Ohio
Pennsylvania
Pennsylvania
South Carolina
West Virginia
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Wisconsin
U.S. TOTAL
Plant or city
Mobile
Pine Bluff
Terre Haute
Notre Dame
Clinton
Clinton
Woodland
Midland
Midland
Midland
Midland
Bound Brook
Kodak Park
Kodak Park
Kodak Park
Kodak Park
Enka
Barberton
Barberton
Erie
Clarton
Greenwood
Luke Md.
Biron
Green Bay
Tomahawk
Rhinelander
Kaukauna
Kaukana
Customer
Southern Kraft Company
Southern Kraft Company
Indiana State University
University of Notre Dame
Clinton Corn Products
Clinton Corn Products
St. Croix Paper Company
Dow chemical Company
Dow Chemical Company
Dow Chemical Company
Dow Chemical Company
American Cyanamid Company
Eastman Kodak Company
Eastman Kodak Company
Eastman Kodak Company
Eastman Kodak Company
American Enka Corporation
Columbia Southern Chemical
Columbia Southern Chemical
General Electric Company
U.S. Steel Corporation
Greenwood Mills
West Virginia Pulp S Paper Company
Consolidated Water Power S, Paper
Fort Howard Paper Company
Owens-Illinois
Rhinelander Paper Company
Thilmany Pulp s Paper
Thilmany Pulp s Paper
Boilers
2
2
1
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
33
Cyclones
4
4
1
1
2
1
1
4
4
2
2
1
2
2
2
2
1
2
2
2
3
2
2
2
2
1
2
2
1
59
Cyclones
per boiler
2
2
1
1
2
1
1
2
2
2
2
1
2
2
2
2
1
2
2
2
3
2
2
2
2
1
2
2
1
-
Boiler steam flow,
kg/s
Primary^
56.7
56.7
25.2
21.4
41.6
34.7
15.8
50.4
50.4
55.4
50.4
27.1
69.3
50.4
50.4
50.4
18.9
75.6
48.5
37.8
63.0
37.8
50.4
31.5
63.0
18.9
31.5
40.3
19.5
-
Reheat
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0,0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-
Total primary
steam flow,
kg/s
113.4
113.4
25.2
21.4
41.6
34.7
15.8
100.8
100. 8
55.4
50.4
27.1
69.3
50.4
50.4
50.4
18.9
75.6
48.5
37.8
63.0
37.8
50.4
31.5
63.0
18.9
31.5
40.3
19.5
1,457.2
-------
APPENDIX B
LOAD FACTORS AND FUEL CONSUMPTION AT CYCLONE-
FIRED ELECTRIC POWER STATIONS IN 1973
Table B-l lists the net power generated, electric load factor,
and fuel consumption for individual cyclone-fired electric
power stations in 1973. Information contained in this table
was obtained by matching the list of boiler contracts furnished
by Babcock & Wilcox (B&W) against the station operating records
published by the National Coal Association.25 The list of
cyclone-fired boiler contracts furnished by B&W was presented in
Tables A-l and A-2 of Appendix A.
Table B-l should be interpreted with caution. The data contained
therein do not represent the exact amounts of gas, oil, and coal
burned solely in cyclone-fired boiler units in 1973. Rather,
Table B-l reflects the total amounts of gas, oil, and coal burned
at power plants whose facilities contain cyclone firing as a mode
of power generation.
For the NOX emission estimate presented in Section 3.4, the only
annual fuel consumption which could be estimated with reasonable
accuracy was for coal fuel. The total coal consumption at sta-
tions known to have cyclone-fired boiler units was 62 x 109 kg
(68,386,000 tons) in 1973. This figure includes all ranks of
coal since no breakdown by coal rank was available. For the
emission estimate in Section 3.4, it was assumed that all of
thd s coal was bimuminous in order to use the NOX emission factor
for bituminous coal (12.28 g NOX per kg coal). The error of
this assumption is expected to be less than 5% since only small
amounts of lignite and no anthracite coals are burned in cyclone-
fired units.
Another possible source of error in the coal- fuel consumption
estimate concerns multiple-mode firing. At a particular power
station, coal may be burned in cyclone furnaces as well as some
other mode such as pulverized coal firing. It was difficult to
determine the extent of this error from the available data.
Mr. Robert Lundberg of Commonwealth Edison (Chicago) was con-
tacted concerning this. He estimated that nearly all of
Commonwealth Edison's cyclone-fired stations were 100% cyclone
fired and did not have multiple modes. Commonwealth Edison owns
nearly 25% of the cyclone-fired boiler capacity in the United
States. There are exceptions to single-mode cyclone firing, of
115
-------
course, such as the Leland Olds Station owned by Basin Electric
in North Dakota. This station operates a horizontally opposed
pulverized coal unit and will soon be operating a cyclone-fired
boiler unit. Our estimate of coal fuel consumption may include
some coal which is not necessarily cyclone-fired. However, this
amount of coal should not, in our estimation, exceed 5% of the
total coal fired in cyclone boilers.
The influence of the two sources of error just mentioned on our
coal fuel consumption estimate cannot be accurately determined.
However, both error sources suggest that our coal fuel consump-
tion estimate may be slightly inflated. As a result of these
possible error sources, we estimate that the total coal consump-
tion of 62 x 109 kg in 1973 is perhaps 5% to 10% high. Multiplying
the bituminous coal NOX emission factor of 12.28 g NOx/kg coal
burned by 62 x 109 kg coal burned in 1973 results in an annual
emission rate of 0.76 x 106 tonnes of NOX per year for 1973.
Because of the above-mentioned reasons, this NOX estimate may
also be 5% to 10% high.
The total amounts of gas and oil burned in cyclone-furnace-fired
boiler units could not be estimated from the available data.
Hence, NOX emissions from these fuel types burned in cyclone-
furnace-fired boiler units could not be accurately estimated.
The gas and oil fuel consumption data presented in Table B-l
include the amounts of oil and gas burned in cyclone furnaces
as well as the amounts burned in other gas- and oil-fired units
which may be present at a particular power staation. The types
of units present could include gas turbines, reciprocating
engines, and oil and gas boilers fired by other methods (spud,
ring burners, etc.).
A summary of the fuel data presented in Table B-l is given below.
Heat released as a result of burning gas, oil, and coal fuels at
cyclone-furnace-fired installations was 1.95 x 109 GJ in 1973.
As indicated from Table B-l, on a heat basis 82.4% of all fuel
burned at these plants in 1973 was bituminous coal and lignite.
The remaining 17.6% of heat was provided by oil and natural gas.
Residual and distillate oils provided 12.5% of the total heat
released, while gas provided 5.1%.
Using the data in B-l, the fuel average heating values were also
determined. Average heating value of coal was 26 MJ/kg (11,200
Btu/lb). Oil had an average heating value of 40 GJ/m3 (144,000
Btu/gal), and natural gas had a heat value of 37 MJ/m3 (1,000
Btu/ft3).
In 1973, a total of 62 x 109 kg (68,386,000 tons) of coal,
6.1 x 106 m3 (38,258,700 barrels @ 42 gallons each) of oil, and
2.7 x 109 m3 (94,348 million cubic feel) of natural gas were
burned at all utility installations possessing cyclone furnaces.
116
-------
TABLE B-l. LOAD FACTORS AND FUEL CONSUMPTION FOR CYCLONE-FIRED ELECTRIC POWER PLANTS IN 1973
State
Arkansas
Connecticut
Connecticut
Florida
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Illinois
Indiana
Indiana
Indiana
Indiana
Indiana
Iowa
Iowa
Iowa
Kansas
Kansas
Kansas
Kentucky
Kentucky
Maryland
Michigan
Plant or city
Helena
Middletown
Bridgeport
Gannon
Coffeen
Calumet
Fisk
Joliet
Kincaid
Pekin
Ridgeland
Waukegan
Will County
Baldwin
Marion
Dallman Plant
Lakeside
State Line
Breed
Tanners Creek
Baileytown
Michigan City
Southerland
Sioux City
Muscatine
Kaw.
Quindaro Station
(No. 3)
La Cygne
E. Smith
Paradise
Crane
St. Clair
Customers
Arkansas Power & Light Company
Hartford Electric
United Illuminating
Tampa Electric Company
Central Illinois Power Company
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Commonwealth Edison
Illinois Power Company
Southern Illinois Power Company
Springfield Water, Light, & Power Department
Springfield Water, Light, & Power Department
Commonwealth Edison
Indiana £ Michigan Electric
Indiana & Michigan Electric
Northern Indiana Public Service Company
Northern Indiana Public Service Company
Iowa Electric Light
Iowa Public Service Company
Muscatine Municipal Electric
Kansas City Board of Public Utilities
Kansas City Board of Public Utilities
Kansas City Power and Light
City of Owensboro
Tennessee Valley Authority
Baltimore Gas & Electric
Detroit Edison Company
Net power
generation,
10s KW/hr
3,500
3,590
3,720
4,880
2,980
295
1,940
8,150
4,900
4,020
2,830
4,190
4,940
6,270
607
815
343
5,030
2,720
6,390
3,000
803
1,070
65.1
604
668
938
938
993
14,500
2,480
10,500
Fuel consumed
Load
factor
0
0
.44
.49
Coal
106 kg
_b
18.1
0.65 -b
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.44
.34
.31
.41
.52
.42
.38
.76
.51
.44
.57
.65
.58
.27
.59
.63
.66
.56
.43
.78
.18
.59
.46
.45
.09
.76
.65
.71
.63
2,040
1,570
b
1,030
3,550
2,400
1,860
b
1,920
2,470
2,760
312
403
201
2,460
1,080
2,410
1,220
282
178
1.91
197
29.0
93.4
474
420
6,050
-
3,230
(MJ/kg)
_D
27.3
b
21.1
21.8
b
21.7
23.5
22.6
24.6
b
23.5
22.1
23.9
24.8
25.1
24.6
22.8
25.2
25.9
25.9
24.6
24.2
25.0
27.6
26.5
21.5
25.7
27.1
b
27.4
Oil
103 m3
532
963
987
b
3.97
b
b
b
b
b
802
b
_b
.795
.159
_b
.635
b
.636
3.97
_b
b
b
b
b
b
16.8
.063
b
665
321
(MJ/m3)
42,700
40,200
40,400
b
38,600
b
b
b
b
b
41,600
b
b
39,000
38,700
_b
37,900
b
38,400
37,800
_b
-b
b
b
b
b
38,300
37,900
b
40,500
41,700
Gas
106 m3
404
_b
_b
_b
b
107
21.4
121
b
b
b
b
b
b
b
b
b
_b
b
_b
5.38
94.9
288
36.9
56.3
201
216
b
b
b
b
5.66
(MJ/m3)
37.7
_b
_b
_b
b
37.2
38.8
38.4
b
b
b
b
b
b
b
b
b
_b
b
-b
37.2
37.2
37.1
36.7
37.2
37.0
35.0
b
b
b
b
37.7
(continued)
-------
TABLE B-l (continued).
H
H
00
Fuel consumed
State
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Missouri
Nebraska
New Hampshire
New Jersey
New Jersey
Mew Jersey
New Jersey
New Jersey
North Dakota
North Dakota
North Dakota
North Dakota
Ohio
Ohio
Ohio
Ohio
South Dakota
South Dakota
Tennessee
West Virginia
Plant or city
New Madrid
Thomas Hill
Allen S. King
Chamois
Asbury
Sibley
Riverside
Stillwater
Lake Road Pt.
St. Charles Co.
Sioux Plant
Sheldon
Merrimack Pt.
Beesley's Point
Deepwater
Sayreville
South Amboy
Marion
Leland Olds
Center
Beulah
Center
Conesville
Muskingham
Niles
Philo
Ben French
Big Stone
Thomas H. Allen
Willow Island
West Virginia Krammer
Wisconsin
Wisconsin
Wisconsin
Wisconsin
Bay Front
Nelson Dewey
Rock River
Sheboygan
Customers
Associated Electric Co-op
Associated Electric Co-op
Central Electric
Empire District Electric Company
Missouri Public Service Company
Northern States Power
Northern States Power
St. Joseph Power & Light
Union Electric
Union Electric
Consumers Public Power Company
Public Service of New Hampshire
Atlantic City Electric Company
Atlantic City Electric Company
Jersey Central Power and Light
Jersey Central Power and Light
Public Service Company of New Jersey
Basin Electric
MinnXota Power Co-op
Ottertail Power
Square Butte
Columbus & Southern Ohio Electric
Ohio Power Company
Ohio Power Company
Ohio Power Company
Black Hills Power & Light Company
c
Tennessee Valley Authority
Monongahela Power Company
Ohio Power Company
Lake Superior District Power
Wisconsin Power s Light Company
Wisconsin Power & Light Company
Wisconsin Power & Light Company
U.S. TOTALS AND AVERAGES
aData obtained by matching B
S W contract list to data in Reference 25.
Net power
106 KW/nr'
2,790
2,680
245
1,260
1,800
2,140
3,000
936.3
-
3,360
1,340
2,750
1,950
1,390
2,000
454
443
1,440
1,720
c
c
3,780
858
1,460
1,360
146
c
—
4,790
1,440
3,780
338
1,370
862
2,800
b
Load
factor
0.
0.
49
67
0.47
0.68
0.40
0.54
0.57
0.71
c
0.39
0.67
0.68
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.75
.SI
.66
.45
.40
.76
.84
c
c
.34
.64
.67
.31
.76
c
-
.55
.76
.61
.48
.69
.66
.67
_b
Coal
Id6 kg
1 ,240
1,210
116
600
746
875
1,230
58.1
c
1,420
236
945
33.6
12.7
b
b
b
1,080
1,358
c
c
—
1,560
3,530
631
729
103.4
c
"
1,410
621
1,430
72.6
548
376
1,200
62,000
(MJ/kg)
25.4
25.4
26.1
23.8
27.7
21.6
24.6
24.0
c
25.4
27.5
31.4
26.9
26.7
b
b
b
15.5
15.2
c
-
c
'
24.4
24.3
26.1
24.4
18.6
c
25.9
26.1
27.6
30.5
25.2
25.8
25,J.
26.0
Oil
103 m3
b
b
b
.635
b
4.45
b
4.93
c
1.58
.318
508
325
563
152
172
.479
.477
c
~
c
"
6.04
5.88
.159
1.27
.159
c
b
b
.635
35.6
.159
.317
.529
6,100
(MJ/m3)
b
_b
b
40,600
b
38,900
b
41,800
c
~
38,200
40,400
38,700
40,100
40,200
40,100
40,000
39,000
39,000
c
~
c
38,200
39,000
37,600 .
38,100
38,700
c
b
b
38,700
36,800
39,300
39,300
40,299
40,000
Gas
106 m5
_b
b
b
-
b
b
177
b
-
323
c
"
222
b
b
80.8
27.4
b
b
b
c
c
b
b
b
b
b
c
311
b
b
32.0
b
b
b
2,700
i
(MJ/m")
_b
b
b
-
b
b
37.7
b
-
35.8
c
37.1
b
b
38.3
37.7
b
b
b
c
c
b
b
b
b
b
c
40.0
b
b
37.2
b
b
b
37
Not applicable.
CNot available.
-------
APPENDIX C
PROPOSED CYCLONE BOILER TEST PROGRAM
(KVB)
INTRODUCTION
Cyclone furnaces constitute the major class of coal-fired utility
boilers for which means to reduce NOX emissions have not been
developed. They also have the highest NOX emissions of any
utility boiler coal firing system (in excess of 1,000 ppm at 3%
O2, dry basis). When the NOX generation has been as high on
other type boilers, and the mechanism controlling the rate of
NOX formation have been understood, then relatively simple
operating adjustments have led to 40% to 80% nitric oxide emis-
sion reductions.
Published literature on NOX emissions from coal-fired cyclone
units do not indicate that any substantial program has been under-
taken to explore the operating variables which could influence
NOX formation, these include:
1. Primary air flow
2. Tertiary air flow
3. Coal fineness
4, Excess air
5. Combustion air temperature
6. Staged combustion air
Load reduction of 20% has been shown to reduce NOX emissions by
25%, which suggests that NOX emissions might be reduced by
reducing the peak temperature achieved in the cyclone. The
temperature must be sufficiently high to maintain the ash in the
cyclone furnace in a molten state, but excessive flue gas tempera-
ture would only serve to generate additional NOX. The addition
of fluxing agents to the coal to reduce ash fusion temperature
might permit further lowering of flue gas temperature by such
means as water injection, gas recirculation, or lower air pre-
heat.
Staged firing has been attempted by varying fuel supply to indi-
vidual cyclone furnaces but has not been effective. It is sus-
pected that the increased heat release in the majority of the
cyclones offset any reduction in NOX generation produced by
excess air variation in individual cyclones.
119
-------
PROPOSED PROGRAM
It is proposed that a boiler with four cyclone furnaces in one
wall be selected for testing to develop an understanding of the
influence of operating variables on NOX emissions. The program
would be conducted in two phases.
In the first phase, the unit would be tested as normally operated
to ascertain variations in NOX emissions from individual cyclones
by furnace probing at the outlet of each cyclone. Measurements
at the economizer outlet would establish the NOX generated in the
overall bulk gas. It is expected that variations in NOX from
individual cyclones of several hundred ppm may occur. Signifi-
cant differences would be examined in terms of air distribution,
fuel distribution, excess air, slagging, or damper settings. Fuel
and air metering to individual cyclones would be used to verify
excess air measurements and heat release. This phase of the work
will first establish if the NOX varies with cyclone operation
and the reasons why it varies; and then secondly it would establish
the gains, if any, which could be made through individual cyclone
excess oxygen monitoring and adjustment.
In the second phase, the following variables would be investi-
gated with uniform air and fuel to each cyclone furnace.
1. Excess air
2. Primary air
3. Tertiary air
4. Combustion air temperature
5. Coal fineness
6. Boiler load
It has been observed that oil-fired cyclones exhibit the charac-
teristic of premixed flames in that high 02 results in a decrease
in NOX. If coal-fired cyclones exhibit similar behavior or if a
substantial amount of NOX is generated in the bulk gas, staged
combustion air with uniform fuel flow to each cyclone furnace
would also be investigated.
Fuel and ash properties will have a considerable bearing on the
ability to modify operation without interfering with normal slag-
ging conditions in the cyclone. The range of fuels available
would be a consideration in selecting a test site. The base fuel
should be one permitting maximum flexibility in cyclone operation.
Thus the possible reductions and the controlling mechanisms will
be established through this probing and adjustment program. A
second, additional fuel type will be tested to explore and
demonstrate fuel difference problem areas.
120
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-77-006
2.
3. RECIPIENT'S ACCESSION NO.
4.T.TLEANDSUBT.TLE Applicability of NOx Combustion
Modifications to Cyclone Boilers (Furnaces)
5. REPORT DATE
January 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
T.E. Ctvrtnicek and S. J. Rusek
8. PERFORMING ORGANIZATION REPORT NO.
MRC-DA-610
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45407
10. PROGRAM ELEMENT NO.
EHE624a
11. CONTRACT/GRANT NO.
68-02-1320, Task 20
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 3-12/76
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES JERL-RTP task officer for thjs report is D.G. Lachapelle, Mail
Drop 65, 919/549-8411 Ext 2236.
16. ABSTRACT
Cyclone furnaces are a significant source of stationary NOX emissions. It was
estimated that 0.76 x 10° tonnes of NOX (over 6% of stationary source NOX) were
emitted from all cyclone-coal-fired utility boilers in 1973. This represents from
19% to 22% of the total NOX emissions from all coal-fired utility boilers in the U.S.
Several techniques of combustion modifications were applied in the past to cyclone
boilers/furnaces in an attempt to lower their NOX emissions. These include boiler
load reduction, low excess air firing, two-stage firing, and switching fuels. This
report summarizes available NOX emission data when applying these techniques to
cyclone boilers/furnaces. Even though significant reductions in NOX were achieved,
none of the techniques was shown to reduce NOX emissions to the level meeting the
New Source Performance Standard.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Nitrogen Oxides
Boilers
Furnaces
Combustion
Air Pollution Control
Stationary Sources
Cyclone Boilers
Combustion Modification
Emission Factors
NOx Reduction
13B
07B
13A
21B
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
131
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
121
------- |