EPA-600/9-89-062a
June 1989
PROCEEDINGS:
1989 JOINT SYMPOSIUM ON STATIONARY COMBUSTION NOx CONTROL
San Francisco, CA, March 6-9, 1989
Volume 1
Compiled by
Claudia Runge
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 93404
EPA Project Officer: EPRI Project Manager:
William P. Linak David Eskinazi
U.S. Environmental Protection Agency Electric Power Research Institute
Air and Energy Engineering Research Laboratory 3412 Hillview Avenue
Research Triangle Park, NC 27711 Palo Alto, CA 93404
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/9-89-062 a
3. RECIPIENT'S ACCESSIOf+NO.
PB89 22Q 52 9 IK
4. TITLE AND SUBTITLE
Proceedings: 1989 Joint Symposium on Stationary Com-
bustion NOx Control, San Francisco, CA, March 6-9,
1989, Volume 1
5. REPORT DATE
June 1989
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Claudia Runge, Compiler
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, California 93404
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
NA (Inhouse)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PEFIIOD COVERED
Proceedings; 3/87-3/89
14. SPONSORING AGENCY CODE
EPA/600/13
is. supplementary notes ^EERL project officer is William P. Linak, Mail Drop 65, 919/
541-5792.
16. abstract T^e procee(jings document presentations at the 1989 Joint Symposium on
Stationary Combustion NOx Control, held March 6-9, 1989, in San Francisco, C/.
The symposium, sponsored by the U. S. EPA and EPRI, was the fifth in a series de-
voted solely to the discussion of control of NOx emissions from stationary sources.
Topics discussed included low-NOx combustion developments such as burner design
modifications and reburning; coal-, oil-, and gas-fired boiler applications; flue gas
treatment processes; fundamental combustion studies; and industrial and commer-
cial applications. Also presented were manufacturers' updates of commercially
available technology and an overview of environmental issues involving NOx control.
The 4~day meeting was attended by persons from 14 nations. More than 50 papers
were presented by EPA and EPRI staff members, utility company representatives,
boiler and related equipment manufacturers, research and development groups, and
university representatives. The proceedings are in two volumes. Volume 1 includes
background, combustion NOx developments I and II, manufacturer's update, advan-
ced combustion technology, and incineration. Volume 2 includes SCR coal applica-
tions, fundamental combustion research, post combustion NOx control development,
fundamental combustion research, new developments, and oil and gas combustion.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Toxicity
Nitrogen Oxides Measurement
Combustion Electric Power Plants
Fossil Fuels Boilers
Catalysis Research
Incinerators Development
Wastes
Pollution Control
Stationary Sources
Hazardous Waste
Municipal Waste
Research and Develop-
ment
13B 06T
07B
2 IB 10B
21D 13 A
07D 14 F
14 G
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
55J
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
i
-------
ABSTRACT
The 1989 Joint Symposium on Stationary Combustion N0X Control was held in San
Francisco, California, March 6-9, 1989. This Symposium, jointly sponsored by EPRI
and EPA, had as its objective the exchange of information regarding recent
technological and regulatory developments pertaining to stationary combustion NOx
control in the United States and abroad. Topics covered during the Symposium
included: low-NOx combustion development; coal-, oil-, and gas-fired boiler
applications; flue gas treatment; fundamental combustion studies; and environmental
and regulatory issues. The Symposium Proceedings is published in two volumes.
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
i i
-------
PREFACE
The 1989 Joint Symposium on Stationary Combustion N0X Control was held March 6-9,
1989, in San Francisco, California. This symposium, jointly sponsored by EPRI and
EPA, was the fifth of its kind devoted solely to the discussion of control of N0X
emissions from stationary sources. Specific topics discussed included the low-NOx
combustion developments such as burner design modifications and reburning, coal-,
oil-, and gas-fired boilers applications, flue gas treatment processes, fundamental
combustion studies, and industrial and commercial applications. Also presented were
manufacturers' updates of commercially available technology and an overview of
environmental issues involving N0X control.
The four-day meeting was attended by persons from 14 nations. Over fifty papers
were presented by EPRI and EPA staff members, utility company representatives,
boiler and related equipment manufacturers, research and development groups, and
university representatives.
Symposium cochairpersons were David Eskinazi, Project Manager in the Air Quality
Control Program of EPRI's Generation and Storage Division; and Dr. William P. Linak,
Chemical Engineer in the Combustion Research Branch of the EPA's Air and Energy
Engineering Research Laboratory. Each of the cochairpersons made brief introductory
remarks. Mr. David Finnigan, Counsel for the House Energy and Commerce Committee
delivered the keynote address. Written manuscripts were not prepared for the
introductory remarks or keynote address; therefore, they are not published herein.
The Proceedings of the 1989 Joint Symposium have been compiled in two volumes.
Volume 1 contains papers from the following sessions:
• Session 1: Background
• Session 2: Combustion N0X Developments I
• Session 3: Combustion N0X Development II
• Session 4: Manufacturer's Update
• Session 5A: Advanced Combustion Technology
t Session 5B: Incineration
iii
-------
Papers from the following sessions are contained in Volume 2:
• Session 6A: SCR Coal Applications
• Session 6B: Fundamental Combustion Research
• Session 7A: Post Combustion N0X Control Development
• Session 7B: Fundamental Combustion Research
t Session 8: New Developments
• Session 9: Oil and Gas Combustion Applications
Also included in both volumes is an Appendix listing the Symposium attendees.
iv
-------
CONTENTS
VOLUME 1
Paper Page
SESSION 1: BACKGROUND
Chairman: I. Torrens, EPRI
"Innovative Clean Coal Technology N0X Control Technology,"J. Temchin
and H. Feibus 1-1
"Policies for N0X Control in Europe," A.-K. Hjalmarsson and J. Vernon 1-9
"Environmental Effects of Nitrogen Oxides," R. Perhac 1-29
"N2O Emissions from Fossil Fuel Combustion," W. Linak, J. McSorley,
R. Hall, J. Ryan, R. Srivastava, J. Wendt, and J. Mereb 1-37
"Measurement of N£0 from Combustion Sources," L. Muzio, M. Teague
T. Montgomery, G. Samuelsen, J. Kramlich, and R. Lyon 1-55
SESSION 2: COMBUSTION N0X DEVELOPMENTS I
Chairman: D. Eskinazi, EPRI
"The Application of Combustion Modifications for NOx-Reduction to Low-Rank
Coal-Fired Boilers," K. Hein 2-1
"Predicting Boiler and Emissions Performance by Comparative Turbulent/Low
N0X Burner Testing on a Large Test Facility," 2-23
J. Vatsky and C. Allen
"Reduction of NOx Emissions from a 500MW Front Wall Fired Boiler," P. Beard,
W. Brooks, K. Johnson, K. Matthews, P. Wells, and J. Vatsky 2-53
"NOx Emissions Results for a Low-NOx PM Burner Retrofit," R. Thompson,
G. Shiomoto, D. Shore, M. McDannel, and D. Eskinazi 2-67
"Retrofit and Boiler Performance Evaluation of the Low N0X PM Firing System
at Kansas Power & Light," R. Lewis, A. Kwasnik, R. LaFlesn, and D. Eskinazi 2-87
"ENEL's Ongoing and Planned N0X Control Activities," A. Benanti,D. Bonolis,
R. Tarli, A. Baldacci, A. Piantanida, and A. Zennaro 2-109
v
-------
Paper
SESSION 3: COMBUSTION NOx DEVELOPMENT II
Chairman: R. Hall, EPA
"Pilot Evaluation of Reburning for Cyclone Boiler NOx Control," H. Farzan,
L. Rodgers, G. Maringo, A. Kokkinos, and J. Pratapas
"Application of Reburning to a Cyclone Fired Boiler," R. Borio, A. Kwasnik,
D. Anderson, D. Kirchgessner, R. Lott, A. Kokkinos, and S. Durrani
"Design Methods for Low-NOx Retrofits of Pulverized Coal-Fired Utility
Boilers," S. Morita, K. Kiyama, T. Jimbo, K. Hodozuka, and K. Mine
"New Approach to NO^ Control Optimization of N0X and Unburnt Carbon
Losses," M. Kinoshita, T. Kawamura, S. Kaneko, and M. Sakai
"The XCL Burner - Latest Developments and Operating Experience," A. LaRue
"Application of Low N0X Combustion Technologies to a Low Volatile Coal
Firing Boiler," S. Miyamae, T. Kiga, and K. Makino, and K. Suzuki
SESSION 4: MANUFACTURER'S UPDATE
Chairman: G. Offen, EPRI
"NOx Control: The Foster Wheeler Approach," J. Vatsky
"N0X Control Update - 1989," A. LaRue, and P. Cioffi
"1989 Update on N0X Emission Control Technologies at Combustion Engineering,
R. Donais, M. Cohen, and M. McCartney
"Status of N0X Control Technology at Riley Stoker," R. Lisauskas,
E. Reicker, and T. Davis
SESSION 5A: ADVANCED COMBUSTION TECHNOLOGY
Chairman: N. Holt, EPRI
"The Conversion of Fuel-Nitrogen to N0X in Circulating Fluidized Bed
Combustion," Y. Lee, and M. Hiltunen
"N0„ Control in Coal Gasification Combined Cycle (IGCC) Systems," N. Holt,
E. Clark, and A. Cohn
vi
-------
Paper
Page
Development of N0X Control Technologies for Coal-Fueled Stationary Diesel
5A-29
SESSION 5B:
INCINERATION
Chairman: W. Linak, EPA
"Reduction of NOx Emissions from MSW Combustion Using Gas Reburning,"
C. Penterson, D. Itse, H. Abbasi, Y. Wakamura, and D. Linz 5B-1
"Application of Low N0„ Precombustor Technology to the Incineration of
Nitrogenated Wastes," R. Srivastava, J. Ryan, W. Linak, R. Hall,
J. McSorley, and J. Mulholland 5B-23
"The Effect of Fuel Nitrogen on N0X Emissions from a Rotary-Kiln
Incinerator," J. Lighty, D. Gordon, D. Pershing, W. Owens, V. Cundy, and
C. Leger 5B-45
"Reduction of Nitrogen Oxides from Coal-Fired Power Plants by Using the SCR
Process. Experiences in the Federal Republic of Germany with Pilot and
Commercial Scale Denox Plants," B. Schonbucher 6A-1
"Experience Gained by Neckarwerke from Operation of SCR DeNOx Units,"
P. Necker 6A-19
"Recent Developments in the SCR System and Its Operational Experiences,"
H. Kuroda, I. Morita, T. Murataka, F. Nakajima, Y. Kato, and A. Kato 6A-39
"The First De-N0„ Installation in the Netherlands, a Demonstration Project at
EPON - Nijmegen Power Station," J. Koppius-Odink, W. Weier, and W. Prins 6A-57
"Operating Experience of SCR Systems at EPDC's Coal-Fired Power Station,"
T. Mori, and N. Shimizu 6A-85
VOLUME 2
SESSION 6A:
SCR COAL APPLICATIONS
Chairman: E. Cichanowicz, EPRI
vii
-------
Paper
Page
"Technical Feasibility and Economics of SCR N0X Control in Utility
Applications," C. Robie, P. Ireland, and J. Cichanowicz 6A-105
SESSION 6B: FUNDAMENTAL COMBUSTION RESEARCH
Chairman: B. Martin, EPA
"Prediction of Fuel and Thermal NO in Advanced Combustion Systems,"
R. Boardman, and L. Smoot 6B-1
"N0X Reduction in Fuel-Rich Natural Gas and Methanol Turbulent Diffusion
Flames," M. Toqan, J. Teare, J. Beer, A. Weir, Jr., and L. Radak 6B-21
"Reduction of Fuel-NO by Increased Operating Pressure in a Laboratory-
Scale Coal Gasifier," K. Nichols, P. Hedman, and A. Blackham 6B-39
"Evolution and Reaction of Fuel Nitrogen During Rapid Coal Pyrolysis and
Combustion," G. Haussmann, and C. Kruger 6B-61
"The Effect of Process Variables on N0X and Nitrogen Species Reduction in
Coal Fuel Staging," K. Knill, and M. Morgan 6B-75
"N0X Emissions in a Pilot Scale Circulating Fluidized Bed Combustor,"
J. Zhao, J. Grace, C. Lim, C. Brereton, R. Legros, and E. Anthony 6B-93
SESSION 7A: POST COMBUSTION N0X CONTROL DEVELOPMENT
Chairman: C. Sedman, EPA
"Design and Operation of the SCR-Type NOx-Reduction Plants at the Durnrohr
Power Station in Austria," M. Novak, and H. Rych 7A-1
"Denox Catalytic Converters for Various Types of Furnaces and Fuels -
Development, Testing, Operation," L. Balling, and D. Hein 7A-27
"Assessment of Japanese SCR Technology for Oil-Fired Boilers and Its
Applicability in the U.S.A.," P. Lowe, W. Ellison, and L. Radak 7A-41
"N0X Control in a Brown Coal-Fired Utilty Boiler," J. Hoffmann,
J. von Bergmann, D. Bokenbrink, and K. Hein 7A-53
SESSION 7B: FUNDAMENTAL COMBUSTION RESEARCH
Chairman: W. Linak, EPA
"Formation of Nitrous Oxide from NO and SO2 During Solid Fuel Combustion,"
G. deSoete 7B-1
viii
-------
Paper Page
"Fuel Nitrogen Mechanisms Governing N0X Abatement for Low and High Bank
Coals," J. Wendt, A. Bose, and K. Hein 7B-14
"Methanol Injection - A New Method for Nox and SO3 Control," R. Lyon 7B-29
SESSION 8: NEW DEVELOPMENTS
Chairman: E. Plyler, EPA
"Catalyst Poisoning in the Selective Catalytic Reduction Reaction,"
R. Yang, J. Chen, M. Buzanowski, and J. Cichanowicz 8-1
"Catalytic Filter Bags," M. Kalinowski, and P. Aubourg 8-9
"Advanced In-Furnace N0X Control in New European Coal-Fired Power Plant,"
K. Bendixen, and J. Pedersen 8-15
"Two-Stage DeNOx Process Test Data from Switzerland's Largest Incineration
Plant," D. Jones, L. Muzio, E. Stocker, P. Niiesch, S. Negrea,
G. Lautenschlager, E. Wachter, and G. Rose 8-21
"Effects of Catalyst Developments of the Economics of Selective Catalytic
Reduction," T. Gouker, J. Solar, and C. Brundrett 8-27
"Full Scale Demonstration of Additive N02 Reduction^) with Dry Sodium
Desulfurization," V. Bland 8-33
"Shell Process for Low-Temperature N0X Control," F. Goudriaan, C. Mesters,
and R. Samson 8-39
"N0X Reduction of Waste Incineration Flue Gas," B. Herrlander 8-45
"Reburning and Repowering for N0X Control on Large Utility Boilers,"
S. Chen, E. Moller, D. Pershing, and A. Walters 8-51
SESSION 9: OIL AND GAS COMBUSTION APPLICATIONS
Chairman: A. Kokkinos, EPRI
"Retrofit of an Advanced Low-NOx Combustion System at Hawaiian Electric's
Oil -Fi red Kahe Generating Station," J. Yee, R. Freitas, D. Giovanni,
S. Kerho, and M. McElroy 9-1
"Gas Turbine Nitrogen Oxide (N0„) Control Current Technologies and
Operating Experiences," L. AngeTlo, and P. Lowe 9-19
"Demonstration of an Automated Urea Injection System at Encina Unit 2,"
J. Nylander, M. Mansour, and D. Brown 9-35
ix
-------
Paper Page
"Retrofit Implications of a Low N0X Burner System on a 230-MW Oil- and
Gas-Fired Boiler," J. Carnevale, and J. Klueger 9-57
"Retrofitting Low-NOx Burners for Gas and Oil Firing," J. Gerdes, Jr.,
R. Waibel, and L. Raaak 9-75
"Engineering Evaluation of Combined N0X/S02 Removal Processes: Interim
Report," W. DePriest, J. Jarvis, and J. Cichanowicz 9-89
APPENDIX A: LIST OF ATTENDEES
x
-------
Session 1
BACKGROUND
Chairman: I. Torrens, EPRI
1-i
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
INNOVATIVE CLEAN COAL TECHNOLOGY N0X CONTROL TECHNOLOGY
Jerome Temchin and Howard Feibus
U.S. Department of Energy Fossil Energy
Office of Coal Combustion, Coal Preparation and Control Systems
Washington, D.C. 20545
ABSTRACT
The recent selections of the Innovative Clean Coal Technology program were
made on September 27, 1988, and consisted of 16 projects. A number of the
projects involved techniques for controlling NO -i.e., from coal-fired power
plants. These projects will be described in detail.
Three of the projects represent innovative combustion control techniques and
one involves catalytic reduction. In addition, five of the selected projects
involve a combination of SO- and NO control -- two employ flue gas cleanup
techniques and three employ repowering techniques.
INNOVATIVE CLEAN COAL TECHNOLOGY N0X TECHNOLOGIES
As a result of the recent Innovative Clean Coal Technology Solicitation, 16
projects, valued at more than $1.3.billion, were selected to demonstrate
advanced technologies for controlling SO- and NO emissions in existing
coal-fired plants. The solicitation is part of the commitment made by
President Reagan in March 1987 for a $5.0 billion program of which the Federal
Government would contribute $2.5 billion to demonstrate improved techniques
that could be used in the 1990's to reduce pollutants from existing coal-fired
plants that are commonly associated with acid rain. The projects comprise a
wide variety of SO- and NO control techniques. Ten of the 16 projects
involve NO controf techniques. Of these, seven will demonstrate retrofit
technologies which modify existing facilities to reduce emissions. The
remaining three projects will demonstrate repowering technologies which
replace a significant portion of the original facility and, in addition to
achieving significant emission reductions, often increase capacity, extend the
life of the plant, and improve efficiency of the system.
Brief descriptions of the ten projects involving N0X control are provided.
Retrofit Technologies:
Three of the seven NO control projects selected for demonstration of retrofit
technologies were proposed by Southern Company Services (SCS), Inc., of
Birmingham, Alabama. They comprise advanced wall-fired combustion techniques,
advanced tangentially-fired combustion techniques, and selective catalytic
reduction (SCR) technology.
1-1
-------
SCS Low NOx-Wall Fired
SCS plans to demonstrate the long-term optimal cost/performance potential of
several NO control techniques used both alone and in combination in a single
boiler. Tne NO control technologies used in these projects will be
characterized by comparing long-term pre-retrofit data with long-term
post-retrofit data. Sufficient data is to be generated to allow extrapolation
to the general boiler population through the use of statistical techniques.
The SCS project is intended to supplement previous short-term evaluations by
burner manufacturers to show the NO reduction potential of their equipment.
However, these short-term evaluations have not generated sufficient data to
extrapolate to the general population of boilers. The dynamic requirements of
utility boiler operations usually diminish the actual NO reduction achieved
in short-term evaluation since some additional margin is required to avoid
periodic upsets (slagging, high convective pass temperatures, etc.).
The majority of electric utility coal-fired boilers in the United States are
either wall-fired or tangentially (corner)-fired. Each class of boiler has a
characteristic range of NO emissions that derives from these different
configurations.
NO control technologies to be evaluated for wall-fired boilers will consist
of advanced overfire air (AOFA) and Low-NO burners both singularly and in
combination. In conventional Over Fired Atr (OFA) configurations, some of the
combustion air is diverted away from the burners driving the stoichiometry of
the burner toward a fuel-rich condition; the diverted air is supplied above
the combustion zone to complete combustion before the combustion products
reach the convective pass of the furnace. However, standard OFA designs do
not obtain complete mixing of the partially combusted gases above the burners.
AOFA incorporates injection port and duct configurations (sometimes in
combination with booster fans) of aerodynamic design to provide superior
mixing of the partially combusted fuel and air to achieve more complete
combustion and enhance NO reduction performance. AOFA is expected to achieve
about 35 percent N0X reduction.
The concept of overfire air was conceived by Babcock & Wilcox (B&W) in the
1950's for application on gas- and oil-fired boilers in Southern California to
meet stringent NO regulations, and has been used on utility boilers in
response to the 1971 NSPS standards. Although OFA techniques have been
effective in NO reduction, problems have been associated with their use in
coal-fired boilers. These problems have been associated with poor overfire
air mixing, which caused reducing environments within the furnace and flame
carryover into the convective pass (both conditions responsible for corrosion
problems with the boiler tubes). To counteract this problem some of the
combustion air can be diverted along the walls creating an oxidizing boundary
layer across the boiler tubes. Although most manufacturers have recognized
this shortcoming and are redesigning their OFA ports and ducts, such AOFA
techniques have not been applied to a full-scale boiler.
The new Low-NO burner systems for coal employ a technique of separating the
fuel and air streams in the primary combustion zone. The complete systems
"also incorporate the standard OFA configuration.
Older Low-NO burner designs utilized turbulent burners to combust the
pulverized coal causing a significant portion of the fuel-bound nitrogen to be
converted to NO . The new Low-NO burner systems incorporate burners that
inject the fuel and air in a manner that prevents turbulent mixing during the
initial combustion. This localized staged combustion allows for
1-2
-------
devolatization of the fuel-bound nitrogen in an initial reducing atmosphere
converting a significant portion of the fuel-bound nitrogen to N2 with no
further oxidation to NO . In properly designed burners, short-term reductions
in the range of 40 to 5u percent have been achieved. The SCS demonstration
will evaluate three commercial prototype LNB designs--B&W DRB (Dual Register
Burner)-type XCL burner with impeller, Foster Wheeler CFSF (Controlled
Flow/Split Flame) burner, and Riley Stoker CCV (Controlled Combustion Venturi)
burner--to determine which one is most suitable for this application.
The combination of AOFA and Low NO Burner (LNB) is expected to achieve about
60 percent reduction in N0X emissions.
The demonstration of wall-fired combustion technologies is planned at the 500
MW Unit 4 of Georgia Power Company's Plant Hammond near Rome, Georgia. This
unit is representative of a large class of wall-fired boilers accounting for
35 percent of the 1985 capacity of all pre-1979 NSPS boilers (boilers without
FGD units and mostly without NO control devices) for 411 out of a total of
929 boilers in this group. The 411 boilers produced about 35 percent of the
total 1985 N0X emissions of 4.9 million tons for the 929 boilers.
SCS Low N0x--Tanaential Fired
The SCS program to demonstrate Low-NO technologies for tangentially-fired
boilers will encompass three techniques: AOFA, Low-NO Concentric Firing
System (LNCFS), and an Advanced Tangential Firing System (ATFS). The AOFA
configuration is similar to that described for the wall-fired boiler. The
application of LNCFS encompasses a modification to the conventional tangential
firing configuration which results in stabilizing the flame front and allows
the initial devolatization period to occur within a fuel-rich atmosphere. The
ATFS is a proprietary configuration and cannot be described here. Based on
short-term parametric testing the AOFA is expected to achieve 25 - 35 percent
NO reduction while LNCFS and ATFS are each expected to achieve 35 - 45
percent reduction. The combination of ATFS and AOFA is expected to achieve 50
- 60 percent N0X reduction based on the same short-term data.
The demonstration of tangentially-fired combustion technologies is planned at
the 180 MW Unit 2 of Gulf Power Company's Plant Smith in Lynn Haven, Florida
near Panama City. This unit is a Combustion Engineering corner-fired steam
generator. Tangentially-fired boilers account for 42 percent of the 1985
capacity of all pre-1979 NSPS boilers for 345 out of a total of 929 boilers in
this group. The 345 boilers produced 34 percent of the total 1985 NO
emissions of 4.9 million tons for the 929 boilers.
Because of the effect that Low-NO techniques have on furnace stoichiometry,
demonstration plans for both wall- and tangentially-fired combustion
technologies call for careful monitoring of particulate, sulfur oxide,
hydrocarbon emissions, and furnace wall corrosion. Particulate emissions can
be significantly affected by the use of Low-NO techniques especially with ESP
units because of the increase in carbonaceous particles due to incomplete
combustion products, increased particle size, and change in particle
resistivity.
SCS-SCR Technology
SCS recently conducted an extensive evaluation of commercial or near-
commercial post-combustion NO control processes using information supplied by
more than forty organizations worldwide. Selective Catalytic Reduction (SCR)
was selected for further evaluation because it is the most technically mature
of the available processes. Nevertheless, significant uncertainties remain
1-3
-------
about SCR design, performance, and cost when applied to high-sulfur
United States coal and the more variable load demand under which U.S. electric
power plants operate compared to those in Japan where there has been wide-
spread commercial use of SCR technology. SCR has the advantage, compared to
combustion modification NO reduction approaches as described above, of being
applicable to every type of boiler, and can achieve up to 90 percent NO
emission reduction.
The SCR process consists of reacting ammonia with the NO in the flue gas
prior to the preheater in the presence of a catalyst to form nitrogen and
water. Because it is necessary to ensure a minimum temperature to the SCR
unit, it is necessary to provide a bypass around the economizer at rear of the
furnace which may have a negative effect on overall thermal efficiency of the
boiler. Also, any excess ammonia will react with the sulfur trioxide, which
forms as a result of the oxidation of sulfur dioxide in the presence of the
catalyst, yielding ammonia hydrosulfate. Ammonia hydrosulfate can corrode the
metal of the air preheater and trap flyash causing decreased heat transfer
performance and increased pressure loss. All of these potential problems
associated with SCR usage in U.S. electric plants will be investigated in the
SCS SCR project to determine design, cost, and operating parameters that may
mitigate these concerns. In order to address these uncertainties, a multiple
unit SCR prototype demonstration is planned between units 5 (75MW) and 6 (320
MW) of Gulf Power Company's Plant Crist near Pensacola, Florida. This
location allows access to flue gas from approximately three percent sulfur
coal under a variety of different N0X and particulate levels.
The prototype demonstration plant will include three SCR/air preheater trains
with 2% MW flue gas flow each. These reactors will be used to evaluate
current commercial catalyst offerings. Six smaller test units (0.05 MW each)
will be included to increase the flexibility of the commercial catalyst
testing and to allow testing of advanced catalyst formulations. These
prototype demonstration units are being proposed as the most cost-effective
approach to SCR evaluation. Larger test units are not believed to be
necessary because the major technical issues to be resolved with SCR in the
United States are related to the chemical effects of high-sulfur coal flue gas
and fly ash upon catalyst life and performance and not to the size of the SCR
installation.
CE-WSA/SNOx
The SCR process in combination with a S02 catalytic reduction process is for a
demonstration project proposed by Combustion Engineering, Inc., (CE) in a
joint venture with Snamprogetti USA, Inc., a part of ENI, the Italian National
Energy Company. The combined process is known as WSA-SNOX technology and is
planned for demonstration on the flue gas from Ohio Edison Niles Station
Boiler No. 2 located in Niles, Ohio. The quantity of flue gas that will be
processed is equivalent to 35 MW of electric generation. The WSA-SNOX Process
is designed to remove at least 90 percent of N0X emissions.
The WSA-SNOX process is capable of achieving higher NO reduction than a
stand-alone SCR process because much of the unreacted ammonia, known as "NH,
slip", is oxidized to nitrogen oxides and nitrogen in the subsequent S0?
catalytic reactor. There are no limitations as to the type of boiler to which
the WSA-SNOX process is applicable.
1-4
-------
B&W - SOx/NOx/ROx/ Box
Another combined SO-/NO emissions control process has been proposed by the
Babcock & Wilcox Company of Alliance, Ohio. Its S0X-N0X-R0X Box (SNRB)
process also accomplishes NO reduction through the use of the basic SCR
process -- reaction with ammonia in the presence of a catalyst -- to achieve
90 percent NO reduction. The simultaneous reduction of S02, NO , and
particulates takes place in a single unit, a hot temperature bagnouse. One of
the advantages of using the SNRB process is that it recovers the heats of
reaction of the pollutant reduction process, resulting in less derating than
other types of emission control devices.
A 5MW demonstration project of the SNRB process is planned at Ohio Edison
Company's R.E. Burger Station, Unit No. 5, in Dilles Bottom, Ohio. As with
the WSA-SNOX process, the SNRB process is expected to be applicable to all
boiler configurations.
B&W-Coal Reburnina for Cvclone Boilers
Babcock & Wilcox has also proposed NO control systems for cyclone boilers for
which B&W has been the exclusive supplier to the electric utility industry.
Although cyclone boilers have the advantages of burning lower grade coals,
allowing for the use of crushed coal and thus requiring less coal handling
equipment, creating less fly ash in the flue gas, and reducing total boiler
size compared to other PC boilers; they produce greatly increased turbulence
in the primary combustion zone which causes increased NO formation.
Coal-fired cyclone boilers make up 9 percent of the pre-i979 NSPS capacity but
contribute to 15 percent of the 1985 NO emissions from these units. This
increased turbulence prevents use of delayed or staged Low-NO combustion
systems with cyclone boilers due to the corrosive effect on furnace walls.
However, reburning with a supplementary fuel in a reducing atmosphere above
the primary combustion zone results in converting much of the NO in the
combustion gases to molecular nitrogen and carbon monoxide. To complete the
combustion process, a third combustion zone is created through the use of OFA
ports. Although fuels used for reburning have been oil or gas, the B&W
proposal plans to use ultra-fine pulverized coal as the reburning fuel. The
proposed reburning will use the same coal that is fed to the cyclone burners,
thus eliminating (1) the need to access an alternative premium fuel, which may
not be available, and (2) the associated expenses of purchasing, handling and
burning the alternative fuel.
B&W plans to demonstrate reburning technology at Wisconsin Power & Light
Company's Nelson Dewey Station Unit No. 2, located at Cassville, Wisconsin.
This unit is a 100 MW boiler equipped with three cyclone burners and is
representative of the majority of cyclone burners currently in operation.
Transalta Low NOx/SOx Combustor
Transalta Resources Investment Corporation of Calgary, Alberta, Canada has
proposed another technology for NO control in Cyclone Boilers, through
retrofit of the Low SO /NO (LNS) Burner. The LNS burner is designed to
release all of the NO and the precursors of NO into the gas stream where it
is converted to molecular nitrogen. The NO removal efficiency of the LNS
burner is expected to be about 90 percent removal, comparable to SCR removal
rates but at a fraction of the cost. The hot gas stream exiting the LNS
burner contains hydrogen, carbon monoxide and a small amount of unburned
carbon. Completion of the combustion process takes place with the use of OFA
downstream of the burner. Throughout the process, techniques are used to
minimize the formation of thermal N0X.
1-5
-------
Demonstration of the LNS burner is planned on the 33MW Unit 1 cyclone boiler
of Southern Illinois Power Cooperative near Marion, Illinois. The boiler will
be retrofitted with a coal pulverizer system and two LNS burners.
Repowerinq Technologies:
Repowering technologies comprise two major categories -- fluidized bed
combustion and gasification. Both types of repowering are often designed as
combined cycle units where the combustion gases can be used to run a gas
turbine. The remaining heat of combustion can be used to produce steam for
the existing steam cycle. In this combined cycle mode the original power
output of the original boiler is typically increased anywhere from 30 to 300
percent resulting in lower cost of electricity compared to single cycle
systems. An additional benefit of repowering is life extension of the current
plant.
In a fluidized bed a stream of air flows upward through a mass of granular
particles. Given sufficient air velocity, the particles are suspended in a
highly turbulent suspended state. When a small amount of coal is introduced
into the bed, it burns very efficiently and at a much lower temperature,
1600 F bed temperature, than in a pulverized coal boiler (above 2800 F). The
lower temperature is sufficient to sustain combustion but is below the ash
fusion temperature. The lower temperature also results in virtually
eliminating thermal NO formation. In addition it widely believed that the
majority of the fuel bound NO is converted to molecular nitrogen by the
catalytic role played by the calcium compounds suspended along with the coal
in the fluidized bed. (The calcium is derived from sorbents -- either
limestone or dolomite -- added to the bed to capture the sulfur in the coal.)
A fluidized boiler typically emits about 0.3 pounds or less of NO per million
Btu.
Fluidized bed combustion systems are designed to operate either at atmospheric
or elevated pressure. Atmospheric fluidized bed combustion (AFBC) is
configured either as a bubbling bed or as a circulating fluidized bed (CFB).
In a CFB coal, air and sorbent are injected into the fluidized bed to create a
primary combustion zone. Combustion is further enhanced by a secondary
combustion zone above the bed in which additional air is introduced. This
staged combustion contributes to NO reduction as described earlier in this
paper. Combustion gases and unburnld coal particles flow into an adjacent
cyclone separator resulting in unburned solids being recirculated to the
fluidized bed. This design contributes to more efficient combustion of the
coal, a less complex feed system, and lower amounts of sorbent usage than a
bubbling fluidized bed boiler.
SPS-CFB
Southwestern Public Service Company (SPS) of Amarillo, Texas, an electric
utility, is proposing to demonstrate a CFB repowering of the 250 MW unit NO. 3
at SPS's Nichols Station near Amarillo. The current PC boiler will be
completely replaced with a CFB boiler. This is twice as large as the largest
CFB currently under construction.
AEP-PFB
A pressurized fluidized bed (PFB), where combustion typically takes place at
10-15 atmospheres, can produce sufficient combustion gases to drive a gas
turbine as well as produce high quality steam for the unit's existing steam
cycle. Demonstration of repowering with such a combined cycle system has been
1-6
-------
proposed by American Electric Power Service Corporation, part of the American
Electric Power Company (AEP) -- one of the Nation's largest utilities. Plans
are for the project to take place in Units 3 and 4 of AEP's Philip Sporn Plant
in New Haven, West Virginia. The Philip Sporn Plant is over 35 years old and
the repowering is intended to extend its life at least another 30 years.
After the units are repowered with PFBC, net electrical generation capability
will be about 330 MW for the two units. Each unit is currently rated at
150 MW.
In the repowering of an existing boiler with integrated gasification combined
cycle (IGCC), the existing boiler is typically supplemented or replaced by
gasifiers which supply cleaned fuel gas to a combustor which drives a gas
turbine creating additional electric power capacity. Waste heat from the
combustion turbine exhaust gas stream is used to produce high quality steam,
using a heat recovery steam generator, to run the existing steam cycle. An
IGCC can be designed to achieve a more efficient heat cycle than a PC unit
resulting in a significant savings in the cost of electricity. One of the
advantages of IGCC systems is that they are modular in design and the capacity
of the existing unit can be increased as it is needed.
Very low NO levels (0.1 pound per million Btu or better) are achieved with an
IGCC because of high removal of the fuel-bound nitrogen in the coal and lower
combustion temperatures preventing the formation of thermal NO . Partial
combustion in the gasifier is taking place in a reducing atmosphere where
intermediate combustion products of the fuel-bound nitrogen -- hydrogen
cyanide and ammonia - - are produced. (No formation of thermal NO takes place
which requires an oxidizing atmosphere.) These intermediate products are
removed in the gas cleanup system downstream of the gasifier. Formation of
thermal NO is inherently lower than in PC boilers and is minimized in the gas
turbine combustor by combining the fuel gas with water to reduce combustion
temperature.
CE-IGCC
Combustion Engineering, Inc., is proposing to demonstrate a coal gasification
system to power a gas turbine coupled to a heat recovery steam generator to
provide steam to the existing steam cycle. The repowered system will add an
additional 40 MW to the system resulting in a doubling of existing capacity. A
novel hot gas cleanup system will be employed in parallel with cold gas
cleanup.
Conclusion:
The projects selected by DOE in its second Clean Coal solicitation include a
wide range of promising NO control technologies. The adoption by the
industry of those projects that are successfully demonstrated for commercial
application will depend on a number of unique individual factors the industry
may need to consider (e.g., regulation, site specifics). However, the
projects represent a suite of potentially cost-effective advanced techniques
for controlling NO in existing coal-fired facilities.
1-7
-------
(intentionally Blank)
1-8
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
POLICIES FOR NOx CONTROL IN EUROPE
Anna-Karin Hjalmarsson and Jan Vernon, IEA Coal Research
ABSTRACT
Over the last two years, there have been significant developments in N0X control
legislation and policies in Europe. The key factor driving these changes has
been concern about the impacts of N0X emissions on forests, in conjunction with
other acidic pollutants and with ozone. Combustion of fossil fuels is the main
source of man-made N0X, with vehicles producing the majority of emissions.
Nevertheless, most legislation adopted so far relates to stationary sources.
At international level, agreement has been reached on a N0X protocol to the UN
Convention on Long-Range Transboundary Air Pollution, whilst the European
Communities have now adopted emission standards for N0X from new combustion
plants over 50 MWt, and targets for reduction of N0X emissions from existing
large combustion plants.
Nearly all western European countries have now, or plan by 1990 to regulate
emissions of N0X from both new and existing combustion plants.
Policies for control of N0X emissions vary. Germany is completing a major
programme of Selective Catalytic Reduction (SCR) retrofits, and this is also the
favoured technology in Austria. Denmark, Finland, the Netherlands and Sweden
are investigating the use of SCR and Selective Non-Catalytic Reduction (SNCR)
for larger combustion plants. Other countries are focussing on the use of
combustion modification approaches. By 1990, there will be around 140 flue gas
treatment plants installed in Europe.
Preceding page blank i-'
-------
INTRODUCTION
The importance of the role of nitrogen oxides in acidification, forest damage
and other environmental problems has received increased attention in recent
years in Europe. In the 1970s, long-range airborne transport of sulphur
compounds seemed to be the main problem, with surface water acidification the
major environmental impact. More recently, as the range of recorded
environmental effects has widened to include damage to forests and other
vegetation, groundwater, building materials (including historic monuments) and
human health, concern has widened to encompass additionally nitrogen compounds,
hydrocarbons, ozone and heavy metals. Nitrogen oxide emissions from human
activities are key components in the acidic deposition and photochemical
oxidation issues. More specifically nitrogen compounds have been shown to take
part in long-range (>110 km) transport of acidic gases and aerosols. They are
one of the major causal factors in the formation and transport of ozone and
other oxidants on a regional scale across Europe.
THE BACKGROUND AND HISTORY OF EUROPEAN POLICY ON N0X CONTROL
The first scientific observations on which present-day knowledge of
acidification is based were made as early as 1947, when a Swedish researcher
began systematically to measure the chemical composition of precipitation at
ground level. This system was developed further and in 1956 the European Air
Chemistry Network was set up including about 100 stations. Continuing research
established a link between the increasing acidification of surface waters in
Norway and Sweden and long range transport of pollutants. The UN conference on
the Human Environment in Stockholm, 1972, was the first time long range
transboundary air pollutants received broad political attention. A principal
was adopted that States have a responsibility to ensure that energy production
and other activities within their territory are not carried out in a way which
1-10
-------
harms the environment or population of the other states. This principle was
enshrined in the 1979 UN Convention on Long Range Transboundary Air Pollution,
and a protocol to this convention signed in 1985 committed signatories to reduce
their S0£ emissions by 30% (2).
Work also started in the Organisation for Economic Cooperation and Development
(OECD) on a programme to measure the long range transport of air pollutants
(LRTAP). The work on LRTAP between 1972-1977 focussed on sulphur dioxide
emissions.
In the late 1970s, the first reports were received of damage to forests in
Germany associated with air pollution. Research began to indicate that N0X and
other pollutants, as well as SO2> might be implicated in forest damage. Faced
with considerable political pressure from the growing influence of the Green
Party, the German government introduced stringent emission standards for both
SO2 and NOx. In order to prevent competitive disadvantages for German industry,
proposals were made for similar legislation to be adopted by the European
Communities.
Concern about the impact of N0X grew and a number of different organisations
have started to work on the subject. In 1985 the Nordic Council of Ministers
initiated a project called "Critical Loads of Sulphur and Nitrogen", designed to
assess the maximum deposition of pollutants which could be tolerated without
harm to the environment. Under the UN Convention on Long Range Transboundary
Air Pollution a group for N0X was set up. In 1988 agreements were made in both
the EC and the United Nations on controlling NOx emissions.
The majority of European states now accept that air pollution and the damage it
causes are perhaps the most serious environmental problem facing Europe as a
whole. It has taken considerable research and political effort to reach this
point. For 20 years there have been complex interactions, nationally and
1-11
-------
internationally, between politicians and scientists. This has also involved a
growing number of ordinary people and environmental groups, as awareness of
environmental issues and concern about their impacts has grown. Green issues
have become increasingly important in European politics, with Green parties
gaining parliamentary seats in Germany, Sweden, Belgium, Italy and the
Netherlands. Pressures for stricter control of atmospheric pollutant including
N0X, remain strong.
PHILOSOPHIES FOR SETTING EMISSION LIMITS
Over the last two years, there have been significant developments in N0X control
legislation in Europe, both nationally and internationally. Nearly all western
European countries now have, or plan to introduce by 1990, regulations on
emissions from both new and existing plants (1). Most western European
countries have adopted targets for a reduction of total NOx emissions.
Combustion of fossil fuels is the main source of man-made N0X, with vehicles
producing the majority of emissions. Nevertheless most legislation adopted so
far relates to stationary sources, with controls over mobile emissions in Europe
introduced only in the FRG and Sweden.
Limits on N0X emissions may be of various types but most countries have chosen
to set standards for N0X emitted in flue gas. The actual levels of emission
standards are based on two different principles. One principle is to have a
target (mostly high) for the required reduction in total emissions and to set
the standards based on that target. Technology must then be found or developed
to meet the standards. Examples of this "technology forcing" approach include
the Federal Republic of Germany, Austria and Sweden. The other philosophy is to
choose a suitable available technology for N0X reduction and set emission limits
according to what can be achieved with that technology. In the UK for instance
there is a requirement that the best currently available system of combustion
1-12
-------
modification suitable for UK boilers and systems should be used, and a design
target rather than an emission standard is set.
An alternative to these two major philosophies is to set a target for reduction
of the total N0X emission from a group of sources (eg utilities) and to allow
the sources at least in theory to choose where to take action to reach this
target. This "bubble" approach has been adopted in Denmark, although
constraints on emissions from individual plants may also be applied.
DIFFERENCES IN METHODS OF SETTING STANDARDS
The way in which emission standards for N0X are set varies in many different
ways between the countries. The differences lie mainly in the categories of
plants subject to limits, in the units used for emission standards, the time
base over which emissions are measured and the levels of the standards.
The categorisation of a plant can be based on:
unit size
combustion technology
total amount of N0X emissions
fuel type
whether it is a new or existing plant.
The most common way of categorising plants is by unit size and fuel type along
with whether a plant is new or existing. The Federal Republic of Germany,
though, has added to these categories the actual combustion technology. For
example, for smaller existing plants there are different emission standards for
the same size of plants using the same fuel, depending on whether they use
fluidised beds, pulverised firing with dry bottom or with wet bottom. Instead
1-13
-------
of plant size, Sweden uses the total amount of emitted N0X as the main category.
Plants with larger N0X emissions are subject to stricter standards than those
emitting smaller amounts.
The units used for emission standards also vary between the countries which
makes comparison of standards between countries confusing.
The units used are:
mg/m3
g/GJ
PPm
The most commonly used unit is mg/m3 at a specified temperature, pressure and
oxygen content (generally 0°C, 1.013 bar on dry gas and with 02 content between
5 and 7%). The Scandinavian countries use g/GJ,. where the energy unit stands
for the efficient (lower) heat value of the input wet fuel (compared to the
American unit lb/MBtu where the heat value is the calorimetric). Levels in ppm
are used in the UK.
Standards are also set on different time bases, for example:
averaged over a month or a year.
hourly mean values (eg 95% of 48 hour mean values must be within 110% of
the standard).
Nine European countries, all in western Europe, now have national limits on N0X
emissions from coal combustion currently in force or agreed for future
implementation. The countries are Austria, Belgium, Denmark, Federal Republic
of Germany (FRG), Italy, Netherlands, Sweden, Switzerland and the United Kingdom
(UK) (see Table 1). Other countries such as Finland, have drafts or proposals
for legislation.
1-14
-------
EUROPEAN INTERNATIONAL AGREEMENTS AND STANDARDS
THE EC DIRECTIVE
In September 1988 the European Community (EC) formally agreed a directive on
"the limitation of emissions of pollutants into the air from large combustion
plants". It has taken nearly five years of negotiations to complete.
The directive specifies target reductions in aggregate emissions from existing
combustion plants compared with 1980 emission levels. It covers combustion
plants having a thermal input of more than 50 MWt, and existing combustion
plants are defined as those in place before July 1987.
The reduction targets for existing plants will give a 10% reduction in EC total
emissions by 1993 and a 30% reduction by 1998 compared to 1980 baseline figures
(see Table 2). Some countries are allowed an increase in emissions such as
Greece, Ireland, and Portugal, and the highest reduction targets are set at 40%
for Belgium, the FRG, France, Luxembourg and the Netherlands.
The EC Directive also includes emission limits for new plants which must be
implemented in each member country by 1990. National authorities may choose
whether to treat plants entering service between July 1987 and 1990 as new or
existing plants. The limits are 650 mg/m3 (6% 02, dry) for plants over 50 MWt
or 1300 mg/m3 for plants firing coal with volatiles over 10%.
UNECE CONVENTION ON LONG-RANGE TRANSBOUNDARY AIR POLLUTION
The United Nations Economic Commission for Europe (UNECE) Convention on Long-
Range Transboundary Air Pollution was signed in Geneva in November 1979 by 33
countries.
1-15
-------
In November 1988 a legally binding protocol was signed freezing N0X emissions at
1987 levels (or levels of any earlier year) by 1994. The signatory countries
are also required to use the best available technology to control emissions from
both stationary and mobile new sources. Twenty five member countries signed the
protocol, which must be approved at the annual meeting of the UNECE in June
1989.
NATIONAL EMISSION STANDARDS
The position of N0X emission standards in Europe is now very dynamic. Further
countries are planning to introduce emission standards on N0X and countries
which already have limits are working on extending the groups of plants to which
the limits apply or tightening the standards. In addition to the national
limits local decisions on special regulations may be applied in many countries.
Emission standards are set for different fuels such as coal and other solid
fuels, oil, natural gas and municipal waste.
COAL
Table 3 shows the size of plants covered by N0X emission standards and Table 4
shows the level of standards applied.
Standards for existing coal fired plants have been introduced in the FRG, Italy,
the Netherlands and Switzerland and are planned in Sweden and Austria. The FRG,
Netherlands, Sweden and Switzerland have set standards for plants of sizes below
50 MWt; Austria has legislation for plants over 50 MWt.
1-16
-------
The most stringent standards for existing plants currently in force are in the
FRG, where they vary between 200-1300 mg/m3. However from 1995 Swedish plants
will have to comply with even tighter standards at 50-200 mg/MJ (equivalent to
around 140-560 mg/m3 at 6% O2 dry). Standards in Switzerland and planned for
Austria are similar to the FRG levels, whilst those in Italy and the Netherlands
are less stringent.
N0X emission standards for new plants are applied to small boilers (5-10 MWt) in
the FRG, Italy, the Netherlands and Sweden. In most other countries the size
limit for new boiler emission standards is 50 MWt but in the UK limits have so
far only been set for plants over 700 MWt. Legislation covering smaller new
boilers is being drafted. As with existing plants, Sweden now has the tightest
N0X emission standards for new coal fired plants at 50 mg/MJ (around 140 mg/nr).
A limit of 200 mg/m3 applies to boilers over 300 MWt in the FRG and Switzerland.
Other countries with legislation on N0X emissions from new plants have limits
ranging from 400-800 mg/m3.
OIL AND NATURAL GAS
Almost all countries with national N0X emission standards for coal combustion
also have standards for oil and natural gas fired plants (see Table 5). The
emission levels for oil and natural gas fired plants are shown in Table 6. In
most cases the limits for oil and natural gas are set lower than or equal to
those for coal. However in the UK limits for coal are lower than limits for oil
firing, and in Sweden the standards for coal are tighter than those for either
oil or gas.
The lowest emission levels for new oil-fired plants are in Sweden (-140-
560 mg/m3) and in the FRG (200-450 mg/m3). On existing plants the lowest levels
will be in Austria (150-450 mg/m3) and in Sweden (-140-560 mg/m3) starting from
1-17
-------
the mid 1990s. Only the FRG and Italy (1200 mg/m3) and the Netherlands
(700 mg/m33) have currently applicable limits on existing oil-fired boilers.
Fewer countries have limits on natural gas burning and the levels for gas are
lower compared with those set for oil. The countries with N0X emission
standards for gas firing are Austria, Belgium, the FRG, the Netherlands, Sweden
and Switzerland. For new plants the limits are set between 100-560 mg/m3 and
for existing plants between 140-560 mg/m3.
WASTE INCINERATORS
The only countries with national N0X emission limits for waste incinerators are
the FRG and Austria. The limit in the FRG is 500 mg/m3 for all plants (at 11%
02 for feed rates >750 kg waste/h, 17% 02 for feed rates <750 kg/waste/h) and in
Austria 100 mg/m3 for plants larger than 750 kg waste/h.
IMPLEMENTATION OF N0X CONTROL LEGISLATION
FLUE GAS TREATMENT
In the FRG stringent N0X emission standards were set in 1983 and had to be
implemented on many plants by the end of 1989. The standards were set
independently of the status of N0X reduction technology and to meet them flue
gas treatment had to be used. A large amount of treatment plant has been
installed simultaneously (3). The standards for new boilers had to be
implemented very quickly, which meant that for some plants almost at
construction phase the design had to be changed and a deNOx plant added at the
last minute. The first deNOx system began operation in the end of 1985. By
1990 plants with a total capacity of 32,000 MWe will be equipped with flue gas
1-18
-------
treatment for N0X reduction in the FR6. The capacity will be divided between
different fuels as follows:
70% of the total coal-fired boiler capacity will then be equipped with flue gas
treatment, dominated by hard coal fired plants.
In Austria, almost all hard coal-fired utility plants are equipped with flue gas
treatment, giving a total capacity of 1400 MWe. Gas and oil fired plants with a
capacity of 800 MWe are also equipped with deN0x (4).
Two municipal waste incinerator plants are expected to be operating with deNOx
this year, one in the FRG and the other in Austria. Other countries with
commercial applications of flue gas treatment are Denmark and the Netherlands.
The distribution of systems for deN0x flue gas treatment anticipated to be in
commercial operation in Europe in 1990 can be seen in Table 7 and Figure 1.
Most of the total capacity of 35,000 MWe equipped with flue gas deNOx is
situated in the FRG. Over 90% of coal fired plants with deNOx in the FRG are
equipped with SCR. In Europe, the most common location of the catalyst is
before the air preheater and electrostatic precipitator, (high dust) comprising
62% of the capacity. A further 36% has the catalyst situated after the flue gas
desulphurisation unit (tail end) and in 2% the catalyst is situated between the
electrostatic precipitator and the air preheater (low dust). The high dust
location dominated the earliest plants but the tail end location has become more
popular recently for various reasons.
coal fired, dry bottom
coal fired, wet bottom
lignite, oil and gas fired
51%
39%
10%
1-19
-------
Work on SCR arid SNCR techniques has been proceeding in other countries. Pilot
plants and tests have been completed and are ongoing on SCR and SNCR in
countries like Denmark, the Netherlands, Italy and Sweden. Techniques are also
being developed for simultaneous S0X and N0X abatement in Denmark and the FRG.
In Denmark, Finland, the Netherlands and Sweden flue gas cleaning in combination
with advanced combustion control will be necessary to meet the planned and
proposed emission standards.
COMBUSTION CONTROL
Most other countries have so far been working mainly with combustion control
techniques. In the UK the main utility is assessing different techniques that
are suitable for different boiler configurations as part of a programme to
retrofit all existing major coal fired power stations with low N0X burners over
a period of 10 years. The programme will involve all generating units with a
capacity of 500 MWe or over and would affect about 25,000 MWe.
Work has also been undertaken in the FRG on combustion control technologies.
For most of the lignite fired boilers this is expected to be sufficient to meet
the emission standards. Techniques used for combustion control include staged
combustion, over fire air, and low N0X burners among others. In Austria,
lignite fired plants will have combustion modification. Combustion control is
also the main measure for gas and oil fired plants.
Other ways to reduce the emissions from stationary sources that may also be used
in the future could be to increase the total efficiency of the plant and thus
reduce emissions per unit of useful energy. As different combustion
technologies emit different amounts of N0X, it will become more important to
have emission levels as one criterion when choosing a combustion technology.
1-20
-------
That will favour technologies such as fluidised bed combustors and gasification
which are reported to produce low N0X emissions.
REFERENCES
(1) J. L. Vernon. Emission standards for coal-fired plants: air pollutant
control policies. London, UK, IEA Coal Research, IEACR/11, 72 pp, August
1988.
(2) The Nordic Council of Ministers. Europe's air, Europe's Environment.
Report of the Nordic Council of Ministers to the Nordic Council's
International Conference on Transboundary Air Pollution, Stockholm,
Sweden, 1986.
(3) VGB Kraftwerkstechnik. Stand der Stickoxidminderung durch
Sekundarmaznahmen. VGB Kraftwerkstechnik, No. 1, January 1988, pp 60-68.
(4) I. Hammer-Kossina, A. E. Hackl. Status of emission control on power
plants and large industrial boilers in Austria. In: NATO/CCMS meeting in
Duesseldorf, FRG, October 10-14, 1988. Vienna, Austria, Umweltbundesamt,
UBA-IB-126, October 1988.
1-21
-------
Figure 1
Estimated market share of deNOx
systems on coal-fired plants in
Europe in 1990
1-22
-------
Table 1 European countries with controls over N0X emissions
from coal combustion (1)
New plants Existing plants
current planned current planned
Austria
+
Belgium
+
Denmark
+
FRG
+
+
Italy
+
+
Netherlands
+
+
Sweden
+
Switzerland
+
+
UK
+
EC
+
* from 1994
** from 1995
1-23
-------
Table 2 N0X emissions reduction targets for existing plants
over 50 MWt in the EC (1)
Country
1980 baseline
(103 t N0X)
1993 target
% reduction
1998 target
% reduction
Belgium
110
-20
-40
Denmark
124
-3
-35
FRG
870
-20
-40
Greece
36
+94
+94
Spain
366
+1
-24
France
400
-20
-40
Ireland
28
+79
+79
Italy
580
-2
-26
Luxembourg
3
-20
-40
Netherlands
122
-20
-40
Portugal
23
+157
+178
UK
1016
-15
-30
Total
3678
-10
-30
1-24
-------
Table 3 Sizes of plants covered by N0X control regulations (1)
New plants
current
Existing plants
current planned
Austria
>50 MWt
>50 MWt
Belgium
>50 MWt
-
Denmark
>50 MWt
-
FRG
>1 MWt
>50 MWt
>1 MWt**
Italy
>100 MWt
>400 MWt
Netherlands
*
*
Sweden
*
*
Switzerland
>1 MWt
>1 MWt
UK
>700 MWt
-
EC
>50 MWt
-
* no minimum plant size stated in regulations
** CFB
1-25
-------
Table 4 Range of N0X emission standards for coal combustion plants
in European countries (1)
New plants
current
(mg/m3)
planned
(mg/m3)
Existing plants
current planned
(mg/m3) (mg/m3)
Austria
800
200-600*
Belgium
550-800
200-400**
Denmark
1150 (1)
FRG
200-500
200-1300
Italy
650
1200
Netherlands
400-800
1100
Sweden
140 (2)
140-280 (3)***
Swi tzerland
200-300
200-500
UK
670 (4)
EC
650-1300
(1) 400 mg/MJ
(2) 50 mg/MJ
(3) 50-200 mg/MJ
(4) 330 ppm
* from 1994
** from 1995
*** from 1995
Note: The standards in mg/m3 are based on standard temperature and
pressure at 6% O2 on dry flue gas, except for some types of
combustion technology in the FRG where the O2 level is between
5 and 7%.
1-25
-------
Table 5 European countries with N0X regulations on oil and natural
gas combustion plants
Oil
new
fired plants
existing
Natural
new
gas fired plants
existing
Austria
+
+ *
+
+ *
Belgium
+
+
FRG
+
+
+
+
Italy
+
+
Netherlands
+
+
+
+
Sweden
+
.J. **
+
UK
+
EC
+
+
* from 1994
** from 1995
1-27
-------
Table 6 Range of N0X emission standards for oil and natural gas
combustion plants in European countries
Oil fired plants Natural gas fired plants
new existing new existing
(mg/m3) (mg/m3) (mg/m3) (mg/m3)
Austria
450
150-450 *
350 150-300 *
Belgium
450 (150 **)
350 (100 **)
FRG
250-450
250-700
200-350 200-500
Italy
650
1200
Netherlands
300
700
200 500
Sweden
140-560 (1)
140-560 **(1)
140-560 (1) 140-560 **(1)
Switzerland
150-450
100-200
UK
564
EC
450
350
* from 1994 Note: The standards in mg/m3 are based on standard
from 1995 temperature and pressure at 3% O2.
(1) 50-200 mg/MJ
Table 7 Estimated number and capacity DeNOx plants in Europe in 1990
FRG Austria Rest of Europe Total
number capacity number capacity number capacity number capacity
(MWe) (MWe) (MWe) (MWe)
Coal
SCR 105 29 000 4 1 340 1 130 (1) 110 30 470
SNCR 5 1 000 2 440 1 130 (2) 8 1 570
Activated 1 200 1 200
carbon
Oil and gas
SCR 10 1 700 3 800 13 2 500
Waste
SCR 1 1 2
Total 122 31 900 10 2 580 2 260 134 34 740
(1) The Netherlands
(2) Denmark
1-28
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
ENVIRONMENTAL EFFECTS OF NITROGEN OXIDES
Ralph M. Perhac
Electric Power Research Institute
Over the past few years, the utility industry's main concern over air
quality and the environment has been with sulfur dioxide emissions.
The importance of nitrogen oxides, however, should not be overlooked
because they can bear on such environmental concerns as damage to
ecosystems through acid rain, human health, and visibility
degradation. But present utility concerns over NOx focus primarily
on its role in ozone formation. The fact that dozens of metropolitan
areas are in a state of nonattainment is a major concern to the
utility industry. As recently as 1987, EPA designated 65 urban areas
as nonattainment for ozone, and those 65 areas have an estimated
population of nearly 130 million people.
Ozone is a health concern, and humans appear to be particularly
sensitive to ozone exposure during exercise. The U.S. Office of
Technology Assessment estimates that about 80,000 people are exposed
to levels of ozone exceeding 9.12 ppm (the U.S. ambient standard)
while doing heavy exercise, for an average of about 4 hours per
person per year. The population exposed to concentrations exceeding
the standard increases to 13 million when performing moderate
exercise, for an average of about 6 hours per person per year, and 21
million are estimated to be exposed to levels exceeding the standard
while performing low exercise, for an average of 8 hours per person
per year. Ozone reduction in nonattainment areas is a matter of high
concern.
1-29
-------
The utility is concerned over health consequences. But is is also
concerned over proposals for more stringent NO2 legislation which
would be costly to the industry and to consumers but which might do
little to alleviate the NOx/ozone issue and, in fact, might
exacerbate it. The chemistry of ozone formation is such that a
simple reduction of the NOx precursor does not necessarily insure a
reduction in atmospheric ozone.
Figure 1 is an overly simplified picture of the atmospheric chemistry
of ozone formation. Despite the simplification, it does point out
the importance of volatile organic compounds (VOC) in allowing
accumulation of ozone. Without VOC, a steady N02-N0—O3 state is
reached, depending on sunlight intensity. Obviously, the ratio of
VOC to NO is critical in defining the ultimate atmospheric
concentration of ozone. High VOC:NO ratios will favor ozone
formation. Conversely, high NO:VOC ratios favor attainment of steady
state. But in the steady state situation, reduction of NOx emissions
could increase the role of whatever VOC is present thereby actually
leading to an increase in ozone. The situation is not simple.
Reduction is VOC will almost certainly result in ozone reduction;
reduction in N0X may or may not (Fig. 2).
Recently, EPRI decided to evaluate the efficiency of N0X control in
reducing ozone in an urban area. To do so, it applied the EPA Urban
Airshed 4 model to Philadelphia, Pennsylvania, a nonattainment urban
area. Data on air quality and meteorology came from a July 1979
study conducted by EPA. Detailed emission inventories for
hydrocarbons and N0X were included in the exercise. The model was
run against observed (1-hour average) ozone concentrations at ground
1-30
-------
level for a test-case day (July 19, 1979) - the base case. The model
was run for four emission reduction "scenarios" with results being
compared to the base case. The four "scenarios" were:
51 25% reduction of only VOC emissions
52 25% reduction of VOC plus 25% reduction of power plant NOx
53 25% reduction of VOC plus 25% reduction of all NOx (mobile
and stationary sources)
54 25% reduction of only power plant N0X emissions
The modeling results (in summary) are as follows:
51 ozone decreased by 0.01 - 0.02 ppm
52 ozone increased by 0.01 - 0,02 ppm
53 ozone increased in Some areas, decreased in others
54 ozone increased by 0.01 - 0.03 ppm
Obviously reduction in NOx yields uncertain improvement in ambient
ozone concentrations. VOC reduction alone leads to decreased
ozone. NOx reduction alone leads to increased ozone. Reducing both
NOx and VOC leads to more complicated results.
The above-cited example of a model exercise is not cited as a reason
for avoiding NOx control. It is given only to caution that any NOx
control should consider the complexities of atmospheric chemistry as
an important part in judging what results will be achieved.
Concern over ozone is not the only reason for the utility interest in
NOx. Nitrogen oxides play a role in acid rain. Acid rain is
essentially a dilute solution of sulfuric and nitric acids plus minor
contributions form carbonic and organic acids, the two strong acids
1-31
-------
commonly accounting for over 90 percent of the acidity. In the
eastern U.S. and western Europe, the sulfuric/nitric ratio, in terms
of acidity contribution is typically about 2.1. Sulfuric acid
contributions tend to be higher in eastern Europe and in Some
metropolitan areas where vehicular NOx emission are high (e.g., in
Los Angeles), the nitric acid component may exceed that of sulfuric
acid.
Nitric acid is obviously an important component of acid rain. It is
derived, essentially, by gas phase oxidation of N02 according to the
reaction:
no2 + OH + m = HN03
where m is any third body which absorbs energy, viz., N2 or 02. At
night, N02 would be oxidized by any ozone present to nitrate radical
which, in reacting with N2 would yield N20^ and ultimately nitric
acid. This reaction points out that NOx control does, indeed, bear
on the acidity of precipitation. Most U.S. attention to acid rain,
however, has focused on S02 rather than on NOx control for two
reasons: (1) the technical aspect of S02 control is less demanding
that for NOx, and (2) sulfuric acid accounts for nearly 2/3 of the
precipitation acidity, hence attention to control the major component
of acid rain was considered warranted.
Will a reduction in NOx emissions play a significant role in
alleviating any damage that might be attributed to acid rain. We now
have extensive data showing that rain acidity, at ambient
concentration, has little harmful effect on crops in terms of growth
or yield. In fact, the overall net impact of the nitrogen component
of the rain has a positive fertilizing effect. An NOx reduction,
1-32
-------
therefore, would be of little value in protecting crops because they
don't need protection.
The forest situation is still not entirely clear, but at least in
America, acid rain may not be a major contributor to observed forest
decline. The nitrogen component of acid rain has been cited as a
cause of forest decline as a result of excess nitrogen
fertilization. That concept, however, has not be substantiated.
Perhaps the major acid rain concern in the U.S. has been that over
aquatic systems, particularly concern over lakes which are not acid
but which are considered "sensitive" - the concern being that such
"low sensitivity" lakes will soon go acid under present ambient acid
rain conditions. We have applied an EPRI-developed model (ILWAS) to
many such lakes to assess the effect, not just of continued
deposition of rain at current acid levels, but of deposition of
increased acidity as a result of increased sulfuric acid loading. We
find that many lakes classed as "sensitive" have considerable
neutralizing reserve to counter even increased sulfuric acid input.
Because of its lesser contribution to precipitation acidity, nitric
acid is less of a concern than sulfuric acid. We find that lakes do,
in fact, show less response to nitric than to sulfuric acid, not only
because of the lower amounts of nitric acid in acid rain, but because
watersheds have natural sinks for nitrogen thereby limiting the
amount of nitric acid which actually enters the lakes.
A third environmental concern over NLOx emission is that related to
the global climate/greenhouse gas issue. Nitrous oxide (^0) is an
infrared absorbing gas and can play a role in any future climate
1-33
-------
problems we may face. Recent studies suggest that utility emissions
of N2 are not nearly as large as previously suspected. The subject,
as well as that dealing with global emissions inventories of N20, is
currently under study by EPRI.
Another environmental effect warranting serious concern is that
dealing with human health, specifically with pulmonary function, and
NC>2 inhalation. As a result of human clinical studies, we know that
NO2 produces a decrement in pulmonary function less than that from
exposure to ozone but more than that for S02. Admittedly, the N02
exposure levels used (a few tenths of a part/million) far exceed
typical ambient levels but studies, so far, have focused on acute
effects. The question of effects from chronic exposure to lower
levels of N2 is unanswered. Likewise the exacerbating role of
particulate matter in conjunction with N2 is uncertain. We have Some
evidence from animal toxicology studies that N02 in combination with
particles may result in synergism. Again, this issue is being
investigated by EPRI.
The chemistry associated with nitrogen oxides in the atmosphere is
not simple and is perhaps even less understood than that of sulfur
dioxide which has received much attention over the past decade. Some
oxidation reactions are not linear, hence predicting the results of
precursor reduction on concentrations of reactants is uncertain in
many instances. Furthermore, alleviating one environmental concern
may well exacerate another. Any consideration of N0X control,
therefore, might best be approached in a holistic manner simply
because atmospheric constitutents are not isolated from each other
and changes in one can affects many reactions - even those involving
other pollutants.
1-34
-------
CO
o
$>!
~
CM
o + o
o °
^ + o
o
o
CM
©
CM
o
CM
o
-------
Relative N0X emissions
Relative VOC emissions
8580.07 B
FIGURE 2
-------
n2o emissions from fossil fuel combustion
W.P. Linak, J.A. McSorley, R.E. Hall
Combustion Research Branch, MD-65
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
J.V. Ryan, R.K. Srivastava
Acurex Corporation
4915 Prospectus Dr.
Durham, NC 27713
J.O.L. Wendt, J.B. Mereb
Department of Chemical Engineering
University of Arizona
Tucson, AZ 85721
ABSTRACT
On-line nitrous oxide (N2O) measurements from six full-scale, coal-fired utility boilers
indicate direct N2O emissions of less than 5 ppm. Laboratory and pilot-scale experiments
conducted to further characterize direct N2O emissions are consistent with the field data
indicating on-line N2O concentrations below 5 ppm. Further sub-scale coal experiments
using air staging and natural gas reburning for NOx control show only slight increases in
N2O emissions. These results question the reliability of the existing N2O data base, much
of which is likely affected by a sampling artifact by which N2O can be produced in sample
containers, awaiting analysis, through an as-yet-unknown mechanism possibly involving
NO, SO2, and water. Time resolved measurements of N2O, NO, and SO2 from combustion
emissions, collected in stainless steel sample containers, show that N2O concentrations in
wet samples can increase to levels greater than 100 ppm in less than 24 hours. These
increases are reduced (but not eliminated) in dried samples. All samples show that NO
reacts (possibly forming NO2) within 4 hours. SO2 concentrations decrease with time at
rates related to the sample dryness. These results suggest the need for further research to
identify other direct and indirect sources of N2O.
INTRODUCTION
The atmospheric concentration of nitrous oxide (N2O) is reported to be increasing at a rate
of between 0.18 and 0.26 percent per year(l,2) from its present level of 303 ppb.(3) This is a
matter of concern because N2O has been implicated as both a "greenhouse" gas and a
participant in stratospheric ozone depletion mechanisms. Fossil fuel combustion has been
proposed as a possible major contributor to the measured increases in ambient N2O
concentrations, as these increases seem to track measured increases in ambient CO2
concentrations. Additionally, limited data of direct N2O emissions from fossil fuel
1-37
-------
combustion had indicated stack emissions exceeding 100 ppm, and an approximate average
N20-N:N0x molar ratio of 0.58:1.(3) These concerns, implications, and the limited data
available prompted the U.S. Environmental Protection Agency (EPA) to stimulate
collaborative national and international research programs with industry, academia, and
other government agencies to better characterize the anthropogenic sources of N2O.
While partial summaries of these programs are presented elsewhere in the form of
workshop proceedings(4,5), the purpose of this paper is to present the results of research
and testing designed to characterize the direct N2O emissions from stationary fossil fuel
combustion. This research and testing was sponsored by the U.S. EPA through its Office of
Policy, Planning, and Evaluation (OPPE) and Office of Air and Radiation (OAR), and was
performed by EPA's Air and Energy Engineering Research Laboratory (AEERL). A portion
of this work was sponsored by the U.S. Department of Energy (DOE).
Historical Data Base
A summary of the N2O data base from stationary fossil fuel combustion as it existed
following the second N2O workshop, which was held in September 1987, is presented in
Figure 1.(5) The open and half-filled symbols represent data from laboratory, pilot, and
full-scale combustion systems taken by several research groups who first attempted N2O
measurement. Also included in Figure 1 is a line representing a proposed N20-N:N0X
molar ratio of 0.58:1.(3) The fully filled symbols represent data taken by our group using a
laboratory-scale downfired coal combustor (2.3 kg/hr, 5 lb/hr) burning two western coals.
These data (fully filled symbols), indicating direct N2O emissions typically less tham 10
ppm, were presented but were not included in the workshop proceedings. At the time,
these on-line gas chromatograph/electron capture detector (GC/ECD) data were considered
suspect due to the limited number of data points, the untested sampling procedure, the
poor condition of the analytical system, and the large discrepancies compared to the
existing data base.
In an effort to improve the data quality, AEERL aquired a new dedicated GC/ECD system,
set it up in a laboratory, and changed the sampling procedure to that used by several other
research groups. This sampling procedure involved using stainless steel sampling
containers and then scheduling the N2O analyses. The results of these changes are
presented in Figure 2. Again, fully filled symbols represent the data taken by this group
burning the same two western coals, and are compared to the existing historical data (open
and half-filled symbols). Variations in stoichiometry and burner swirl were used to affect
NOx (and possibly N2O) concentrations. The new data were highly scattered, but agreed
better with the existing data. These data were presented at the third N2O workshop that
was held in June 1988. At that meeting Muzio et al., conducting EPRI/EPA sponsored
research, presented data suggesting the presence of an N2O sampling artifact. (6) They
presented evidence that indicated that N2O might be produced in sampling containers
awaiting analysis. They further hypothesized a mechanism of this formation involving
NO, SO2, and water. This evidence questioned the validity of all existing data which
involved container sampling.
Based on this evidence, AEERL renewed efforts to characterize, by on-line means, the
direct N2O emissions from several laboratory and pilot-scale combustion systems burning
a variety of coals, fuel oils, and natural gas. Additionally, we sought to compare the results
from these sub-scale systems to direct N2O measurements (by on-line techniques) from
1-38
-------
full-scale pulverized-coal-fired utility boilers. Concurrent with all on-line analyses, we
sought to characterize the time dependent evolution of N2O, NO, and SO2 in container
samples.
ON-LINE MEASUREMENTS
Laboratory and Pilot-Scale Experiments
Three AEERL combustion systems that were used for the sub-scale experiments included a
29 kW (100,000 Btu/hr) refractory-lined, downfired, coal-fired tunnel combustor
(Downfired Tunnel Furnace), a 733 kW (2,500,000 Btu/hr) gas/oil-fired, fire-tube package
boiler (North American Boiler), and an 879 kW (3,000,000 Btu/hr) gas/oil-fired, water-tube
package boiler simulator equipped with a low NOx burner/precombustor (Low NOx
Burner/Package Boiler Simulator). Details regarding the specific designs of these systems
have been presented elsewhere(7,8,9) and will not be reiterated here.
Although several analytical techniques in addition to GC/ECD are applicable to on-line
N2O analysis from combustion sources(4,6,10), most involve extractive sampling. The
sampling system used at AEERL included extractive sampling through a stainless steel
sampling probe, heated, 0.64 cm (0.25 in.) outside diameter (O.D.) Teflon tubing and
particulate filter, water removal by use of a refrigeration drying system, and then pumping
by means of a Teflon diaphragm sample pump through silica gel dessicant to continuous
emission monitors (CEMs). CEMs included measurement of O2, CO, CO2, and NO.
Immediately after the sample pump, a portion of the sample was directed to a 1 ml/six port
GC sampling valve for on-line N2O analysis by GC/ECD (Shimadzu Model GC-9A). SO2
was measured from a heated sample taken downstream of the filter and upstream of the
refrigeration dryer. The GC used to quantify N2O utilized 5 percent methane in argon
carrier gas at 20 cm3/min., a 3.8 cm (1.5 in.) long by 0.64 cm (0.25 in.) O.D. P2O5 precolumn,
a 3.66 m (12 ft) long by 0.32 cm (0.125 in.) O.D. stainless steel Poropak Super Q column at 35
°C, and a Ni63 ECD maintained at 330 °C.
Table 1 presents the combustion conditions and average on-line N2O and CEM data for 10
AEERL tests involving the three combustion units and seven fuels including four types of
coal. Evident from Table 1 is that the direct N2O emissions from all tests never exceeded 5
ppm. In fact several natural gas and No. 2 fuel oil data indicate N2O concentrations less
than the instrument detection limit of 0.24 ppm. The coal tests showed the highest levels
(2 to 4 ppm), and little difference could be discerned between the two fuel oil and natural
gas tests. Data from the low NOx burner/package boiler simulator, with and without air
staging strategies for NOx control, showed that, while significant NO reduction was seen
with air staging, no measureable effect was detected in corresponding N2O concentrations.
These data, however, are too close to the instrument detection limit for accurate analysis.
The other CEM measurements are comparable to those typical of full-scale units.
Experimental work on N2O emissions from laboratory-scale coal combustors has also been
completed at the University of Arizona (UA), under the sponsorship of the U.S. DOE. The
downfired laboratory combustor used in that work was very similar in size and design to
that described above for the AEERL tests, except that the pulverized coal flame was
"premixed," and the overall configuration resembled a plug flow reactor. The flue gas
1-39
-------
sample to be analyzed was withdrawn through a water cooled, water quenched probe. The
water was removed in a refrigerated knockout pot, and the sample was passed through a 1
ml gas sampling valve into a Shimadzu Model GC-8A GC equipped with a 3.66 m (12 ft)
long, 0.32 cm (0.125 in.) O.D. stainless steel column packed with Porapak Q, conditioned at
220 °C. An ECD, operating at 250 °C, required the use of argon with 5.22 percent methane
as the carrier gas. Complete details of the sampling and analysis procedure can be found
elsewhere.(ll) Table 1 also presents four data sets from the University of Arizona work,
representing exhaust conditions under baseline, air staging, and natural gas reburning
conditions for a Utah bituminous coal. These data are consistent with the on-line AEERL
data.
Figure 3 shows values of exhaust NO and N2O emissions as a function of stoichiometric
ratio for a Utah bituminous coal. N2O levels vary from 0.5 to 1.5 ppm. Time resolved
profiles during air staging are shown on Figures 3b (primary SR=0.86) and 3c (primary
SR=0.65), and demonstrate the interesting fact that staging causes an increase in N2O near
the staging point. However, N2O levels are still exceedingly low and so the effect is not of
practical significance. Figure 3d shows time resolved NO and N2O profiles along the
combustor axis during employment of reburning as a NOx control technique. We
observed an increase of N2O at the point of air injection, similar to what was observed
under air staged conditions. For all configurations, however, these on-line measurements
of N2O yielded extremely low values, and indicated that direct N2O emissions did not
appear large from pulverized coal combustion, either with or without combustion
modifications for NOx control.
Field Tests
Following the laboratory and pilot-scale experiments, AEERL sponsored a field study to
characterize the direct emissions of N2O from full-scale, coal-fired utility boilers. These
measurements were conducted by Acurex Corporation at six boilers ranging in size from
165 to 700 MW, including Circular, Triple Cell, and Tangential designs manufactured by
Babcock & Wilcox, Riley Stoker, and Combustion Engineering. These units burned
primarily a medium sulfur Alabama bituminous coal. The sulfur content of this coal is
reported to be in the range of 1.5 to 2.0 percent. The sampling and analytical methods used
were similar to those described above for the AEERL sub-scale experiments. Table 2
presents the average on-line N2O and CEM data for each of the six utility units. All on-
line N2O measurements were below 5 ppm. In fact, almost all of the on-line N2O
measurements were below the detection limits of the two GC/ECD systems used (1.2 and
3.6 ppm). These samples were taken at existing sampling locations, downstream of all
economizer and particulate removal equipment, either before or after the systems'
induced draft fan. All six units were operated at or near full load during the tests. Limited
SO2 data were collected due to problems with the SO2 analyzer.
A summary of the on-line data from both the sub-scale and full-scale studies is presented
in Figure 4. Again, these data (indicated by the fully filled symbols) are compared to the
previous data base (open and half-filled symbols). These results suggest that the direct
emission of N2O from coal-fired utility boilers is likely low (less than 5 ppm), that the
existing data are suspect due to an as-yet-undefined sampling artifact, and that no simple
N2O/NOX correlation exists. The data do not indicate, however, that other types of
1-40
-------
combustion sources are not major N2O emitters, or that indirect mechanisms of
atmospheric N2O formation from combustion emissions do not exist.
TIME RESOLVED MEASUREMENTS FROM SAMPLE CONTAINERS
Concurrent with all on-line activities, extracted samples were collected in stainless steel
sampling containers for time dependent analyses of N2O by GC/ECD, NO by GC/ thermal
conductivity detector (TCD), and SO2 by GC/flame photometric detector (FPD) techniques.
These samples were collected at three degrees of dryness including wet samples (before
refrigeration dryer), partially dried samples (immediately after refrigeration system), and
dessicated samples (after additional drying by use of a P2O5 filled canister). Samples from
the sub-scale experiments were analyzed at 1, 4, 24, 48, 168, and 336 hours. Samples from
the full-scale tests were analyzed at 1, 4, 48, 168, and 336 hours. Time-resolved NO
measurements were not taken for the full-scale tests due to limited availability of field
equipment. Additionally, limited time resolved SO2 data were produced from the full-
scale tests due to instrument problems in the field.
Samples for time resolved analyses were collected in 0.5 L stainless steel containers. These
containers were fitted with septums for sample extraction by gas syringe. A number of
samples were collected in glass containers of similar size for comparison. While the
available data for this comparison are very limited, they indicate similar sample evolution
trends in each type of container. However, the rate of N2O increase may be somewhat
lower in glass containers. Combustion gas and N2O calibration gas (10 ppm) samples of
varying moisture content were passed through beds of different desiccants and analyzed to
determine any effect of selective water removal on N2O. Three desiccants were tested
including phosphorus pentoxide (P2O5), magnesium perchlorate [Mg(C104)2l, and silica gel
(Si02). None of these desiccants was seen to affect the N2O concentration. P2O5 was
chosen as the desiccant of choice for these tests due to its almost quantitative ability to
remove water. Tests were also conducted to identify sorbents for the selective removal of
SO2 from the extracted gas samples. Sodium hydroxide (NaOH) and calcium hydroxide
(CaOH) have been examined as possible candidates. The data presented here, however, do
not include any sample processing for SO2 removal.
Figures 5 through 8 present the results of these time resolved analyses. In an effort to
conserve space, only selected data (indicative of general trends) will be presented. Figures
5a and 5b present the evolution of N2O and SO2 from samples taken during the Utah
bituminous laboratory-scale coal experiments. Figure 5c presents similar N2O evolution
data for the full-scale coal-fired utility boiler tests from unit D. It is interesting to note that
both the sub-scale and full-scale N2O data from wet or partially dry samples show rapid
increases in the first minutes and hours after sampling. Smaller (but non-negligible)
increases are seen in the dry samples. Maximum N2O values from wet samples are seen
after 24 hours. These values (200 and 75 ppm) are comparable to those presented by the
historical data and the proposed N20-N:N0X correlation (Figure 1). The maximum N2O
concentration is likely related to the initial NO and SO2 concentrations. This speculation
is supported by comparing the N2O and initial NO data from these two tests. Figure 5
shows the N2O concentrations in the wet and partially dry samples to slowly decrease after
24 to 48 hours. This reduction, however, may be due to slight air leaks into the sample
containers due to numerous septum punctures and the volume of sample removed. Steps
were taken in later data collection activities to limit the number of samples taken from
1-41
-------
each container and increase the number of containers used. Figure 5b shows that the
evolution of SO2 seems to be very much dependent on the sample dryness. These data are
presented out to 168 hours to increase the resolution at the lower times. NO data (not
presented) all show rapid removal/reaction (possibly forming NO2) within 1 to 4 hours
regardless of the moisture content of the samples.
Figures 6 and 7 present similar data for No. 5 and No. 2 fuel oils, respectively, from the two
smaller-scale combustion systems. Similar behavior, as compared with the coal data, is
evident with respect to both N2O and SO2. Again, NO was removed/reacted to below
detectable levels within 1 to 4 hours. It is interesting to note that, while the trends are
similar to those seen for coal, the magnitude of the artifact is much less for the two fuel
oils. The data presented are also consistent with the supposition that the resulting N2O
formation is dependent on the initial NO and SO2 levels (in addition to the moisture).
Note that N2O increases seen in the Package Boiler Simulator samples (where NOx control
was used) are significantly smaller than those seen in the North American Boiler samples
(without NOx control) burning the same fuels.
The evolution of N2O from samples from these same two units burning natural gas is
presented in Figure 8. SO2 emissions were for the most part below detectable levels. And
again, NO was removed from all samples within 1 hour. The N2O data are inconclusive.
Some of the data suggest that the artifact still exists at minimal SO2 levels, while other data
show this effect to be negligible. Further samples from natural gas flames are planned to
investigate this further.
CONCLUSIONS
Results from on-line (GC/ECD) N2O measurements from six full-scale utility boilers
burning a medium sulfur Alabama bituminous coal indicate direct N2O emissions of less
than 5 ppm. Additional results from laboratory and pilot-scale combustion experiments
using similar procedures and burning a variety of coals, fuel oils, and natural gas are
consistent with these full-scale N2O data. Further experiments using a laboratory-scale
coal combustor indicate that combustion modifications used for NOx control may cause
slight increases in N2O emissions. These increases, however, are minor, causing overall
increases of less than 5 ppm. These results question the reliability of the existing historical
data base, much of which is likely affected by a sampling artifact. These historical data had
indicated direct N2O emissions exceeding 100 ppm. Additionally, these past data had
suggested a linear N20-N:N0X molar ratio of approximately 0.58:1.
Time resolved analyses of samples collected in stainless steel containers suggest that N2O
can be produced in the sample awaiting analysis within minutes. While the mechanism
of this formation is unknown, it seems directly dependent on the initial concentrations of
water, SO2, and NO in the sample. This artifact can be reduced (but not eliminated) by
removing the moisture present in the sample by desiccation. Characterization of full-scale
combustion emissions in the field by container sampling is much more cost effective
compared to on-line analyses. Hence, efforts are continuing to examine techniques for the
selective removal of SO2 in order to minimize the effects of the artifact in container
samples.
1-42
-------
The data indicate that the direct emission of N2O from pulverized-coal-fired utility boilers
is likely to be low (< 5 ppm). The data, however, do not support the conclusion that other
types of combustion sources are not major N2O emitters, nor do the data indicate that
indirect mechanisms of atmospheric N2O formation involving the constituents of
combustion emissions do not exist. Identification of the mechanisms involved with the
sampling artifact may shed some light as to whether similar mechanisms could exist in the
atmosphere.
ACKNOWLEDGMENTS
Portions of this work were conducted under EPA contracts 68-02-4701 and 68-02-4285 with
Acurex Corporation, DOE contract DE-AC22-87PC78850 with the University of Arizona,
and Purchase Order 8D1713NATA with J.O.L. Wendt. The authors would like to thank
EPA AEERL's R.A Grote for his assistance developing the N2O GC/ECD system. The
authors would also like to thank the Acurex field team (R.K. Clayton, K.A. Krebs, R.M.
Machilek, and A. Sykes) for their efforts and quick response to the field test schedule.
Grateful acknowledgment is made to S.M. Wilson of Southern Company Services and
D.M. Burdett of Alabama Power Company for arranging the testing of the full-scale boilers.
The authors would like to thank EPA AEERL's J.C. Ford and RTI's J. Evans and C. Wall for
their helpful suggestions, comments, and active support of data quality. Finally, the
authors would like to acknowledge EPA's OPPE and OAR for partial project funding.
DISCLAIMER
The research described in this article has been reviewed by the Air and Energy Engineering
Research Laboratory, U.S. Environmental Protection Agency, and approved for
publication. The contents of this article should not be construed to represent Agency
policy nor does mention of trade names or commercial products constitute endorsement
or recommendation for use.
REFERENCES
1. R.F. Weiss. "The Temporal and Spatial Distribution of Tropospheric Nitrous Oxide." J.
Geophys. Res., 86,1981, pp. 7185-7195.
2. M.A. Khalil and R.A. Rasmussen. "Increase in Seasonal Cycles of Nitrous Oxide in the
Earth's Atmosphere." Tellus, 35B, 1983, pp.161-169.
3. W.M. Hao et al. "Sources of Atmospheric Nitrous Oxide from Combustion." J. Geophys.
Res., 92,1987, pp.3098-3104.
4. W.S. Lanier and S.B. Robinson. "EPA Workshop on N2O Emission from Combustion
(Durham, NC, February 13-14, 1986)." Chapel Hill, NC: Energy and Environmental
Research Corporation, EPA-600/8-86-035 (NTIS PB87-113742), September 1986.
5. J.C. Kramlich, et al. "EPA/NOAA/NASA/USDA N20 Workshop - Volume I:
Measurement Studies and Combustion Sources (September 15-16, 1987, Boulder,
1-43
-------
Colorado)." Irvine, CA: Energy and Environmental Research Corporation, EPA-600/8-
88-079 (NTIS PB88-214911), May 1988.
6. L.J. Muzio et al. "Potential Errors in Grab Sample Measurements of N2O from
Combustion Sources." 1988 Fall Meeting Western States Section/The Combustion
Institute, Dana Point, CA, October 17-18,1988, No. 88-70.
7. G.C. Snow and J.M. Lorrain, "Evaluation of Innovative Combustion Technology for
Simultaneous Control of SOx and NOx." Research Triangle Park, NC: Acurex
Corporation, EPA- 600/2-87-032 (NTIS PB87-188926), April 1987.
8. D.W. Pershing et al. "Effectiveness of Selective Fuel Additives in Controlling Pollution
Emissions from Residual-Oil-Fired Boilers." Research Triangle Park, NC: U.S. EPA,
EPA-650/2-73-031 (NTIS PB225037), October 1973.
9. J.A. Mulholland and R.V. Srivastava, "Low NOX/ High Efficiency Multistaged Burner:
Fuel Oil Results." J. Air Poll. Cont Assoc., 38(9), September 1988, pp 1162-1167.
10. Montgomery et al. "Continuous NDIR Analysis of N2O in Combustion Environments
with a Field Prototype Instrument." 1988 Fall Meeting Western States Section/The
Combustion Institute, Dana Point, CA, October 17-88,1988, No. 88-71.
11. J.O.L. Wendt and J.B. Mereb, "Nitrogen Oxide Abatement by Distributed Fuel
Addition." Quarterly Reports 1 through 4, Contract DE-AC22-87PC78850, DOE-PETC,
Environmental Control Technology Division, 1988.
1-44
-------
250
200
E
a.
a.
o
CM
150
100
O Acurex/EPA
~ Harvard
A MIT
O EPRI
a RWE
• Utah bituminous
¦ Montana sub—bit.
1250
Figure 1. Existing N20/N0X data from laboratory, pilot, and full-scale combustion systems
presented at the EPA/NOAA/NASA/USDA N2O Workshop.(5) Open and half-filled
symbols represent historical data. The N20-N:N0X correlation of Hao et al. (3) is
presented. Fully filled symbols represent data taken by this group (and presented at the
Workshop) burning two coals in a laboratory-scale combustor using an on-line GC/ECD
sampling/analysis technique.
1-45
-------
250
200
E
a.
a
o
-------
1200
g
1100
CL
a.
1000
CN
900
O
800
K
O
700
£
600
O
500
O
X
400
O
CN
z
300
k-
0
200
0
z
100
(a) Utah bit. §2
UA Furnace,
O
NO
*
N20x100
1000
a 900
a
f 800
CM
o
K
O
&
¦o
700
600
500
0
° 400 (-
X
1 300
0.9 1 1.1
Stoichiometric Ratio
1000
w
o
K
O
£
¦o
o
o
900
800
700 |-
600
500
400 -
300 -
200 -
100
0
(c) Utah bit. §2
UA Furnace
0.65/1.08
o 200
o
z 100
0
1000
(b) Utah bit. §2
UA Furnace
0.86/1.08
_L
0
NO
*
N20x100
-J—
_1_
_l_
0.4 0.8 1.2 1.6 2 2.4 2.8
Residence Time, sec
(d) Utah bit. §2
UA Furnace
1.10/0.90/1.06
O
NO
*
N20x100
0.8 1.2 1.6 2
Residence Time, sec
2.4 2.8
0.4 0.8 1.2 1.6 2 2.4 2.8
Residence Time, sec
Figure 3. NO and N2O concentrations from a laboratory-scale coal combustor as a function
of (a) stoichiometric ratio (at combustor exhaust), and residence time using air staging (b)
SR=0.86/1.08, (c) SR=0.65/1.08, and natural gas reburning (d) SR=1.10/0.90/1.06 strategies
for NOx control.
1-47
-------
250
200
E
a.
a.
o
-------
250
200 -
g 150
a.
a.
S" 100
(a) Utah bituminous
Tunnel Furnace
1000
(b) Utah bituminous
Tunnel Furnace
800
a
,o-»—-....
50 ,
—Q
E 600'
a.
a,
° 400
200 ~
•A...
...A-
_i_
50 100 150 200 250 300 350
Time, hours
100 150 200 250 300 350
Time, hours
0 (ffi ¦ '13'
25 50 75 100 125 150
Time, hours
175
-©- Wet #1
Wet #2
-E>- Refrigerated #1
Refrigerated #2
A- Dry (P205) #1
A Dry (P205) §2
Figure 5. Evolution of N2O and SO2 with time from coal combustion samples collected
and stored in stainless steel sampling containers at three degrees of sample dryness. Stack
emission samples were collected using the (a, b) Downfired Tunnel Furnace burning Utah
bituminous coal (SR = 1.42), NO(initial) = 757 ppm, and (c) Unit D (full-scale utility boiler)
burning medium sulfur Alabama bituminous coal, NO(initial) = 354 ppm.
1-49
-------
25
20
E 15
a.
a
o"
(N , „
z 10
5 -
a-4-
(a) No. 5 fuel oil
North American Boiler
50 100
150 200
Time, hours
250 300 350
-©- Wet #1
Wet §2
-Q-- Refrigerated #1
Refrigerated #2
•A- Dry (P205) #1
± Dry (P205) §2
25
20
E 15
a
a.
o"
SJ 10
: (b) No. 5 fuel oil
Package Boiler Simulator
500
400
300
CN
° 200
100
: (c) No. 5 fuel oil
Package Boiler Simulator
100 150 200
Time, hours
350
0 ©¦
%
25 50
75 100
Time, hours
j I
125 150 175
Figure 6. Evolution of N2O and SO2 with time from No. 5 fuel oil combustion samples
collected and stored in stainless steel sampling containers at three degrees of sample
dryness. Stack emission samples were collected using the (a) North American Boiler
burning No. 5 fuel oil (SR = 1.25), NO(initial) = 189 ppm, and (b, c) Low NOx
Burner/Package Boiler Simulator burning No. 5 fuel oil (SR = 0.66/1.29), NO(initial) = 60
ppm.
1-50
-------
(a) No. 2 fuel oil
North American Boiler
150 200
Time, hours
350
100
80
| 60
cs
o
<" 40
20
(b) No. 2 fuel oil
North American Boiler
l? A.
0 (ffil1 'E1 I J
0 25
I I I i I -L-lj-l li
I—L
50
75 100
Time, hours
125 150 1 75
(c) No. 2 fuel oil
Package Boiler Simulator
10 -
100 150 200 250 300 350
Time, hours
-©- Wet #1
Wet §2
-Q-- Refrigerated #1
Refrigerated §2
••A- Dry (P205) 01
• A- Dry (P205) #2
Figure 7. Evolution of N2O and SO2 with time from No. 2 fuel oil combustion samples
collected and stored in stainless steel sampling containers at three degrees of sample
dryness. Stack emission samples were collected using the (a, b) North American Boiler
burning No. 2 fuel oil (SR = 1.25), NO(initial) = 105 ppm, and (c) Low NOx Burner/Package
Boiler Simulator burning No. 2 fuel oil (SR = 0.65/1.26), NO(initial) = 64 ppm.
1-51
-------
(a) Natural gas
North American Boiler
50 100 150 200 250 300 350
Time, hours
25
20 -
E 15
a
a.
o
z 10 1-
:(b) Natural gas
Package Boiler Simulator
5 -
-©- Wet #1
Wet §2
Refrigerated #1
Refrigerated #2
•A' Dry (P205) 01
A' Dry (P205) #2
50 100 150 200 250 300 350
Time, hours
Figure 8. Evolution of N2O and SO2 with time from natural gas combustion samples
collected and stored in stainless steel sampling containers at three degrees of sample
dryness. Stack emission samples were collected using the (a) North American Boiler
burning natural gas (SR = 1.24), NO(initial) = 62 ppm, and (b) Low NOx Burner/Package
Boiler Simulator burning natural gas (SR = 0.73/1.25), NO(initial) = 50 ppm.
1-52
-------
Table 1
AVERAGE ON-LINE CONCENTRATIONS TAKEN FROM LABORATORY
AND PILOT-SCALE COMBUSTION SYSTEMS*
Unit
Fuel
SR**
N2O
EPS
NO
EPS
SO2
EPS
CO
pgm
02
%
CO2
%
Tunnel
Furnace
Utah
bit. coal
1.42
4.2
757
753
17
5.5
13.9
Tunnel
Furnace
Montana 1.43
sub-bit. coal
2.2
613
380
30
5.3
14.6
Tunnel
Furnace
Western
Kentucky
bit. coal
1.46
3.7
553
1650
12
6.8
12.6
Tunnel
Furnace
Pittsburgh
bit. coal
1.33
2.2
570
1450
5.7
12.9
N. Amer.
Boiler
Natural
gas
1.24
<0.24
62
0
0
4.4
8.9
N. Amer.
Boiler
No. 2 fuel
oil
1.25
0.30
105
58
3
4.4
12.3
N. Amer.
Boiler
No. 5 fuel
oil
1.25
1.3
189
236
16
4.4
14.1
LowNOx
Burner/PBS****
Natural
gas
0.73/1.25
1.10/1.25
<0.24
0.72
50
638
6
4
11
14
4.6
4.6
9.2
8.7
LowNOx
Burner/PBS
No. 2 fuel
oil
0.65/1.26
1.13/1.26
<0.24
0.27
64
536
130
130
2
2
6.1
6.0
12.0
12.0
LowNOx
Burner/PBS
No. 5 fuel
oil
0.66/1.29
1.10/1.18
0.26
0.73
60
682
270
270
25
25
5.4
4.1
12.2
11.0
UA Coal*****
Furnace
Utah
bit. coal
1.25
1.28
1121
-
300
4.25
14.0
UA Coal
Furnace
Utah
bit. coal
0.65/1.08
1.99
605
-
400
3.60
14.3
UA Coal
Furnace
Utah
bit coal
0.86/1.08
3.80
216
-
500
3.50
14.8
UA Coal
Furnace
Utah
bit. coal
1.10/0.90
/1.06
4.45
382
-
400
3.85
13.8
•All concentrations presented as measured (dry), no correction to constant percent 02-
"Stoichiometric ratio.
""'Missing data.
* """Package Boiler Simulator.
*****UA data corrected to 0 percent O2 (dry).
1-53
-------
Table 2
AVERAGE ON-LINE CONCENTRATIONS TAKEN FROM
FULL-SCALE UTILITY BOILERS*
Size
N2O
NO
SO2
CO
O2
CO2
Unit
MW
Class /Tvpe**
E£m
PPm
PPm
%
%
A
250
Pre-NSPS
Circular
Babcock & Wilcox
1.3
386
_#**
~
4.6
14.8
B
250
Pre-NSPS
Triple Cell
Babcock & Wilcox
<3.6
513
-
13.3
7.1
13.5
C
250
Pre-NSPS
Circular
Riley Stoker
<3.6
559
—
8.6
6.1
14.3
D
165
Pre-NSPS
<1.2
354
.
2.2
8.3
11.7
Tangential
Combustion Engineering
E
700
Pre-NSPS
0.7
374
.
30.7****6.0
13.1
Tangential
Combustion Engineering
F
165
Pre-NSPS
<1.2
319
930
3.1
8.1
11.9
Tangential
Combustion Engineering
*A11 concentrations presented as measured (dry), no correction to constant percent O2.
**A11 units burned a medium-sulfur Alabama bituminous coal.
***Missing data.
****CO trace showed numerous spikes.
1-54
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
MEASUREMENT OF N2O FROM COMBUSTION SOURCES*
L.J. Muzio", M.E. Teague
Fossil Energy Research Corporation
23342 C South Pointe
Laguna Hills, CA 92653
T.A. Montgomery, G.S. Samuelsen
UCI Combustion Laboratory
Department of Mechanical Engineering
University of California, Irvine 92717
J.C. Kramlich, R.K. Lyon
Energy and Environmental Research Corporation
18 Mason
Irvine, CA 92718
ABSTRACT
N2O emissions from fossil combustion have been reported to be 25-40% of the NOx levels.
At these levels, fossil fuels have been suggested to be a major anthropogenic source of
N2O. Recent tests have shown these measurements to be in error, the N2O being formed
by reaction between NOx. SO2. and H2O in the sample containers. Time resolved
measurements of gas samples stored in Tedlar bags, supported by chemical kinetic
calculations, indicate that the majority of N2O forms over a time period of 6 hours and
depends on the amount of SO2 present, initial NO level, and amount of water present.
A continuous infrared analyzer, suitable for characterizing N2O from combustion sources,
has been developed. The analyzer has been used to perform continuous on line N2O
measurements at nine utility boilers (pulverized coal and oil fired). The measurements
indicate that the N2O levels are generally less than 5 ppm and are not related to the NOx
levels in the flue gas.
* Work sponsored by EPRI (RP2154-16, RP2533-9, RP8005-4) and U.S. DOE AR&TD (DE-
AC22-84PC70771).
** Corresponding Author
1-55
-------
BACKGROUND
While global climate change has been associated with increased CO2 levels in the
atmosphere, concern is growing about the role of other trace gas species (CH4,
chlorofluorocarbons, carbon tetrachloride, N2O, etc.). This paper addresses the
measurement of nitrous oxide, N2O, from combustion sources and the implications of these
measurements on the potential role of combustion as a major anthropogenic source.
The mean global concentration of N2O is approximately 300 ppb and has been
increasing at a rate of 0.2-0.4 percent per year (1,2). In the troposphere, N2O, is a relatively
strong absorber of infrared radiation, and can contribute to the Greenhouse Effect. Being
stable in the troposphere, N2O is transported to the stratosphere. In the stratosphere N2O is
the largest source of stratospheric NO and the major natural chemical sink establishing the
stratospheric O3 concentration (3).
The increase of N2O in the atmosphere has been attributed to anthropogenic
sources, although the dominant man-made source of N2O is still uncertain. Weiss (2), Hao
et al. (4), and Tirpak (1) suggest that combustion of fossil fuels results in emissions of N2O
that can account for the observed increase in N2O. Previous N2O measurements from
combustion sources indicate that systems fired with natural gas do not produce significant
concentrations of N2O. However, substantial levels of N2O were reported from systems
fired with residual oil or coals (e.g., fuels containing nitrogen and sulfur). In fact, the
measurements of Hao et al. (4) and Castaldini et al. (5) suggest that the N2O emissions are
related to the emissions of NOx (NO + NO2); the N2O level being 25-40 percent of the NOx
concentration. Global N2O production from combustion sources has been calculated using
the relationship between N2O and NOx coupled with known NOx emissions from
combustion. Thus it is important to accurately establish this N2O/NOX ratio.
The N2O measurements from combustion sources have primarily been made by gas
chromatographic analysis of gas samples collected in glass or stainless steel containers.
Recent measurements have shown that these grab samples can undergo chemical reaction
in the containers creating N2O concentrations substantially higher than those originally
formed in the combustion process (6, 7). A sampling "artifact" exists, as a result, the N2O
measured can exceed the actual emission. This paper will: 1) discuss this artifact including
experimental observations and a proposed chemical mechanism, 2) propose measures to
minimize this artifact, 4) discuss the development of a continuous N2O analyzer, and 5)
discuss representative emission levels of N2O from practical coal-combustion measured
with the continuous N2O analyzer.
GRAB SAMPLE N2O MEASUREMENT ERROR
The artifact in measuring N2O with grab sampling techniques was initially identified in
the course of combustion experiments in a down-fired gas, bench-scale tunnel furnace. A
description of the furnace is available elsewhere (8). Gas samples were withdrawn with a
1-56
-------
water-cooled, stainless steel probe and introduced directly into 1000 ml Pyrex sample
flasks. The flasks were subsequently analyzed for N2O by gas chromatography with
electron capture detection (9).
During a set of experiments with the gas fired combustion tunnel, NH3 was doped
into the fuel to increase NOx levels, and SO2 was doped into the primary burner air. Grab
samples from this experiment showed approximately 300 ppm N2O. This finding was
surprising since N2O concentrations above 10 ppm have not been reported in natural gas
flames. Further experiments determined that the high N2O concentrations were associated
with the SO2; if SO2 was absent from the reactants, the exhaust N2O concentrations were 2-
3 ppm, consistent with previous results.
Further combustor tests showed that the N2O was being formed in the sample flasks
through a reaction mechanism involving SO2, NO, and H2O (6, 7). To verify this reaction
mechanism, Pyrex containers (250 ml) were filled with simulated flue gas mixtures: N2,
86.2%; O2, 4.3%; CO2, 9.4%; and NO, 600 ppm. One milliliter of water was placed in the
container to simulate the condensed water from the combustion products. For selected
samples, the pH of the added liquid was adjusted to various values by the addition of
sulfuric acid (this simulated the collection of combustion generated SO3 into the
condensate). In each case, the filled sample containers were allowed to age for
approximately 2 hours before analysis. The samples were then analyzed for N2O in the
same manner as the sample flasks from the combustion experiment. Figure 1 illustrates
results from four of these tests. In the first, no SO2 was introduced into the gas-phase, and
only distilled water was added for the liquid phase. No N2O was observed in the container.
Likewise, the N2O yields were insignificant when the 1.0 ml liquid phase was acidified with
H2SO4 to a pH of 1.6. However when 1500 ppm SO2 was added to the gas-phase,
significant N2O was measured in the container, with the liquid phase being either distilled
water or H2SO4 (pH 0.6).
The presence of SO2 is a necessary condition for the formation of N2O within the
container. The influence of SO2 concentration was tested by repeating the combustion
experiment described above with varying amounts of SO2 introduced into the primary air
and collecting the sample in either 1 liter glass flasks or 270 liter Tedlar bags. The N2O
analysis results shown in Figure 2 (7) indicate that values of SO2 below 500-600 ppm lead
to only a minor amount of N2O formation. Above 600 ppm, however, significant amounts of
N2O are generated. Since SO2 concentrations vary considerably between various fossil
fuel sources, the severity of container-formed N2O may also vary between the sources. The
reader should note that the quantitative aspects of this conclusion may change depending
on initial condition in the container (e.g., O2, NO, amount of condensate).
Additional tests were performed to better quantify the progress of the reactions
occurring in the sample flasks. Specifically, tests were conducted to determine the time
history of N2O, NO, NO2, and SO2 in the sample containers. Gas samples from the natural
1-57
-------
gas fired combustor were collected in 270 liter Tedlar bags and subsequently analyzed for
NO, NO2, N2O, and SO2. The 270 liter Tedlar bags provided sufficient sample to allow
samples to be withdrawn and analyzed as a function of time using continuous gas
analyzers. (N2O was analyzed with the continuous analyzer described in a subsequent
section of this paper.)
The primary variable for the test series was SO2 concentration, which was varied
from 710 ppm to 2500 ppm. The time histories of N20, NOx, and SO2 are shown in
Figure 3. As can be seen in Figure 3a, the majority of the N2O formation occurs in the first
six hours irrespective of the initial SO2 level. The accompanying time histories of NOx and
SO2 are shown in Figures 3b and 3c. As seen in Figures 3a-c, at the high SO2 condition,
virtually all of the NOx initially present in the Tedlar bag shows up as N2O in the final
product. At the lowest SO2 test condition (716 ppm), N2O continues to form until virtually all
of the SO2 is removed from the gas phase, after which the N2O stabilizes while the NOx
continues to decrease. At the intermediate SO2 level (1210 ppm), the N2O formation rate is
initially rapid becoming more gradual after 2 hours, as the NOx and SO2 are consumed.
For this intermediate case, there appears to be sufficient SO2 to allow the reaction to
proceed longer than at the low SO2 condition, but not sufficient SO2 to convert all of the
initial NOx to N2O, as at the higher SO2 levels.
CHEMICAL KINETICS
The gas phase and aqueous reactions between SO2, NOx, and H2O leading to the
formation of N2O have been documented in the literature. Martin et al. (10) discuss the
reaction mechanism and the general chemistry has also been discussed in the wet FGD
scrubbing literature (11, 12).
Lyon and Cole (13) proposed the detailed mechanism, shown in Table 1. An overall
path based on this mechanism is shown in Figure 4. In the sample flasks, the NO is oxidized
to NO2. NO2 and SO2 in the gas phase are taken to be in equilibrium with the aqueous
phase. In the aqueous phase, the NO2 and SO2 ultimately form N2O and H2SO4 through a
reaction sequence involving HNO.
The detailed mechanism was integrated for the same conditions as the experiments
conducted in the Tedlar bags (Figure 3). The results of these calculations are compared to
the experimental results in Figure 5a-c. In general, the agreement between the model and
the experiments is good. The characteristic time for N2O formation is basically the same
between experiment and model, with the experiment exhibiting somewhat faster N2O
formation than predicted by the model. However, there are three noteworthy differences
between the experiments and the model calculations. First, for the medium and high SO2
cases, the kinetics under predict the amount of N2O formed. At the high SO2 condition the
experiments exhibited virtually 100% conversion of the NOx to N2O, whereas the model
predicts 65% conversion of the NOx to N2O. Second, in the Tedlar bags the rate of
1-58
-------
disappearance of NOx decreased as the SO2 level decreased. Whereas the kinetic
calculations indicate little effect of SO2 level upon the NOx time histories. Thirdly, the
experiments show an SO2 threshold for N2O formation (Figure 2) whereas the calculations
indicate a fairly linear dependence of N2O formation with SO2 concentration. The threshold
with SO2 appears in the calculations if the amount of condensed water is increased.
In spite of the differences, the initial calculations suggest that the mechanisms shown
in Table 1 provide a reasonably good representation of the N2O formation in the sample
flasks.
SAMPLING TECHNIQUES TO MINIMIZE N2O FORMATION
In light of the potential formation of N2O in grab samples from combustion systems, a
question arises as to which sampling protocols can lead to valid N2O measurements. The
results presented in Figure 3 indicate that N2O is formed in the sample flasks over a time
period of 1-6 hours. Thus, one approach to obtain valid samples would be to perform the
N2O analysis on-line, avoiding storage of the samples.
On-line N2O analysis may not always be available, or warranted, for an initial
characterization of a combustion system. In these cases, it is appropriate to establish, and
utilize, a grab sampling technique capable of obtaining valid N2O samples from combustion
systems. The results discussed previously indicate that the conditions leading to N2O
formation in the sample flasks require the presence of 1) condensed H2O, 2) SO2, 3) NO,
and 4) a low pH resulting from the aqueous reactions between SO2 and NOX-
Three sampling methodologies have been investigated as possible approaches for
obtaining valid grab samples (7). Each method focuses on eliminating one or more of the
requirements necessary for N2O formation in the sample containers:
1. drying the sample: eliminates the aqueous phase
2. removinn SQg ahead of the sample flask: eliminates one of the key species
needed to form N2O
3. altering the pH of the aqueous phase: raises the pH to interfere with the
chemical mechanism leading to N2O formation in the flasks
In the tests reported by Muzio et al. (7), sampling approaches 2 and 3 (outlined
above) have been shown to produce valid grab samples for N2O. Drying the gas sample
through an ice bath to a 0 - 1*C dew point does not appear to adequately condition the
sample. With 750 ppm of initial NOx and 2500 ppm of SO2, the dried sample produced 26-
40 ppm of N20, whereas methods removing the SO2 or altering the pH yielded N2O levels
less than 3 ppm.
1-59
-------
CONTINUOUS N20 ANALYZER DEVELOPMENT
Under EPRI funding, the UCI Combustion Laboratory has been actively engaged in a
research program to develop a continuous N2O analyzer suitable for on-line measurements
from combustion sources. Development of the continuous N2O analyzer was initiated prior
to discovery of the artifact in the grab sampling measurement technique. With the high
levels of N2O reported from coal fired utility boilers, it was felt that a continuous N2O
analyzer was necessary to better characterize N2O emissions. The discovery of the artifact
in the grab sampling method increased the necessity for an on-line technique for valid N2O
measurements. The goal has been to establish a measurement technique which 1) has
minimal interferences, 2) is relatively simple, compact, reliable, and 3) suitable for both field
and laboratory use. A previous paper outlined the requirements and criteria for a
continuous analyzer utilizing infrared absorption for measuring N2O in combustion effluent
(14). The key points from this study include:
1. The 7.8 micron region in the infrared is most suitable for measuring N2O
in combustion products.
2. At 7.8 microns, the primary interfering species are SO2 and NO2.
3. SO2 and NO2 can be readily removed from the sample stream ahead of the
analyzer without affecting N2O levels.
Based on the criteria listed above, a field prototype instrument has been designed
and built in cooperation with HORIBA, Ltd. The analyzer is a nondispersive infrared
analyzer utilizing a 500 mm sample cell. A schematic of the analyzer is shown in Figure 6.
Replaceable optical filters allow the instrument to measure N2O either in the optical region
around 4.5 microns, or 7.8-8.5 microns. To date, the 7.8-8.5 micron region has been used.
To minimize interferences, the analyzer uses two Luft-type detectors in series. The first, or
primary detector, senses N2O and any interfering gases that absorb in the 7.8-8.5 micron
region. Since the first detector absorbs all of the radiation from the N2O bands, the radiation
reaching the second detector is only that due to interfering species. The second detector
then senses the interfering species and electronically compensates for their effect.
The initial specifications for the analyzer included two ranges: 0 to 250 ppm and 0 to
500 ppm. This specification was made at the time N2O concentrations up to 400 ppm were
anticipated to be emitted from coal fired systems. However, the analyzer is capable of
operating on a 0 to 25 ppm full scale range with sufficient signal to noise ratio to detect 0.5
ppm changes in N2O level.
Simulated combustion products were used for the initial evaluation of the prototype
continuous N2O analyzer. High purity gases were blended to 1) verify the linearity of the
instrument, 2) quantify the extent of interferences from CO2, CO, NO, NO2, and SO2, and 3)
verify the capability of removing NO2 and SO2 from the sample stream without affecting
N2O levels.
1-60
-------
The analyzer responds linearly to N2O concentrations ranging between 0 and 250
ppm, as well as between 0 and 25 ppm. As previously indicated, the analyzer has sufficient
signal to noise ratio to detect a 0.5 ppm change while operating with a range of 0 - 25 ppm,
full scale.
Subsequently, the extent of the interferences produced by typical coal combustion
products was evaluated. Interference curves were generated for CO levels between 0 and
500 ppm; CO2, 0 and 20 %; NO, 0 and 1250 ppm; S02, 0 and 4000 ppm. The interference
curves generated are shown in Figure 7. In general, the results agree with the previous
study, SO2 is the primary interfering species (14). The analyzer utilized some construction
materials that adsorb N02, which precluded generation of an interference curve for this
species.
Because of the NO2 adsorption within the analyzer and the extent of SO2
interferences, provision for NO2 and SO2 removal from the sample stream ahead of the
analyzer is included in the measurement system. Previous work demonstrated that NO2
and SO2 could be scrubbed from the gas sample without affecting the N2O (14). The SO2
and NO2 are removed using sodium carbonate and sodium sulfite solutions, respectively.
These results were verified with the prototype analyzer.
FIELD APPLICATION
After completing the initial laboratory evaluation and validation, the analyzer was
used to make continuous N2O measurements in utility boiler combustion flue gases. To
date, N2O measurements have been made at nine sites. Table 2 briefly describes each unit
tested (manufacturer, firing configuration, fuel, and rated load) and summarizes the NOx
and N2O emissions from each unit under its normal operating conditions. As seen in
Table 2, direct N2O emissions from utility boilers firing either oil or pulverized coal are low
and not a substantial fraction of the NOx emissions. In fact, the N2O emissions are
generally less than 5 ppm. As seen in Table 2, at two of the units, the N2O analyzer showed
levels of 6 and 11 ppm. These levels are likely due to the sampling system and not
necessarily a consequence of the combustion process. At each site, the N2O analyzer
system was integrated into an existing continuous gas analysis system, consequently the
sampling system varied from site to site. As explained below, depending on the
configuration of the system, N2O can potentially form in the sample lines.
Evidence of sample line generated N2O has been found with the continuous
analyzer. At one of the sites, several unheated sample lines were used in the sampling
system. Typically all of the sample lines were not in use at the same time; several of the
lines were dormant while a specific sample (or set of sample lines) was being monitored.
The dormant lines contained condensed water and residual SO2 and NO from the flue gas.
Figure 8 is a duplication of a strip chart trace recorded during a particular test at this site.
The arrows indicate when a change was made from one group of samples lines being
1-61
-------
monitored to another, a sharp peak in N2O level can be seen. These peaks can be
attributed to N2O formation in the dormant sample lines. Once the line has been flushed
completely with "new" sample, the N2O levels decrease and remain constant. Since
substantial N2O formation occurred in the dormant sample lines, the measured levels under
steady-state conditions may have also included a small amount of N2O generated in the
sample lines to the analyzer.
At units A and B, grab samples were also obtained and analyzed for N2O. Grab
samples were obtained with NaOH in the flask to inhibit N2O formation. As seen in Table 3,
the N2O levels of the grab samples were in good agreement with the values obtained with
the continuous analyzer (a gas chromatograph with helium ionization detector was used for
the grab sample analyses, the lower detectability limit of this system is 2 ppm).
CONCLUSIONS
The following conclusions can be drawn from the results discussed above:
• N2O measurements using grab sampling techniques are subject to errors due
to reactions between NO, SO2, and H2O forming N2O in the sample containers.
• Formation of N2O in grab samples can be inhibited by either scrubbing the SO2
from the flue gas ahead of the sample flask or altering the aqueous pH of the
sample by adding NaOH to the container prior to use.
• The continuous analyzer system designed and built for the detection of N2O is
capable of N2O analysis in combustion effluents at levels down to a few ppm.
• Use of the continuous analyzer at nine utility boilers indicates that direct N2O
emissions are low (generally, less than 5 ppm), and that the N2O levels are
not a substantial fraction of the NOx levels as previously suggested.
ACKNOWLEDGEMENTS
We are pleased to acknowledge the support of the Electric Power Research Institute
at Fossil Energy Research (Contract No. RP2533-9), the Electric Power Research Institute at
the University of California, Irvine (Contract Nos. RP2154-16 and RP8005-4), and the U.S.
Department of Energy's Direct Utilization Advanced Research and Technology
Development Program at Energy and Environmental Research (Contract No. DE-^AC22-
84PC70771). We would like to acknowledge and thank HORIBA, Ltd. for their cooperation
and support in the development of the field prototype N2O analyzer. The continued
encouragement and discussions with G. Offen (EPRI) and A. Kokkinos (EPRI) are also
acknowledged. K. Anderson, J.A. Cole, T. Grogan, J.M. McCarthy, T. McGrath, and J.
Mclnerney helped in the performance of the experimental tasks and modeling.
1-62
-------
REFERENCES
1. Tirpak, D.A. (1987). "The Role of Nitrous Oxide (N2O) in Global Climate and
Stratospheric Ozone Depletion," Joint Symposium on Stationary Combustion NOx
Control, New Orleans, LA, March 23-25.
2. Weiss, R.F.(1981). "The Temporal and Spatial Distribution of Tropospheric Nitrous
Oxide." Journal of Geophysical Research Letters. 86 (C8), pp. 7285-7195.
3. Kramlich, J.C. et al. (1987). "Mechanism of Nitrous Oxide Formation in Coal Flames,"
Joint Fall Meeting of the Western States and Japanese Sections of the Combustion
Institute, Honolulu, November; also submitted for publication in Combustion and
Flame.
4. • Hao, W.M., S.C. Wofsy, M.B. McElroy, J.M. Beer, and M.A. Toquan (1987). "Sources
of Atmospheric Nitrous Oxide from Combustion," Journal of Geophysical Research,
March.
5. Castaldini, C. et al.(1983). "Environmental Assessment of Industrial Process
Combustion Equipment Modified for Low NOx Operation," Proceedings of the 1982
Joint Symposium on Stationary Combustion NOx Control, Vol. 11,46.1-46.24, EPA-
600/9-85-0266, U.S. Environmental Protection Agency.
6. Muzio, L.J. and J.C. Kramlich (1988). "An Artifact in the Measurement of N2O from
Combustion Sources," Geophysical Research Letters, in press.
7. Muzio, L.J., M.E. Teague, J.C. Kramlich, J.A. Cole, J.M. McCarthy, R.K. Lyon (1988).
"Potential Errors in Grab Sample Measurements of N2O from Combustion Sources,"
presented at the Fall Meeting of the Western Sates Section of the Combustion
Institute, Dana Point, CA.
8. Muzio, L.J., G.R. Offen, A.A. Boni, and R. Beittel (1986). "The Effectiveness of
Additives for Enhancing SO2 Removal with Calcium Based Sorbents," Proc. 1986
Joint Symposium on Dry SO? and Simultaneous SOg/NOv Control Technologies. V.
I, EPRI CS-4966, Electric Power Research Institute, Palo Alto, CA, pp. 13.1-13.23.
9. Kramlich, J.C., R.K. Lyon, and W.S. Lanier (1988). "EPA/NOAA/NASA/USDA N20
Workshop, Volume I: Measurement Studies and Combustion Sources," Boulder,
Colorado, September 15-16, 1987, EPA-600/8-88-079.
10. Martin, L.R. et al. (1981). "The Reactions of Nitrogen Oxides with SO2 in Aqueous
Aerosols," Atmospheric Environment. 15. pp. 191-195.
11. Chang, A.G. et al. (1982). in Flue Gas Desulfurization. (edited by J.L. Hudson and
G.T. Rochelle), ACS Symposium Series 188, American Chemical Society,
Washington, D.C., pp. 127-152.
12. Narita, E. et al.(1984). "Formation of Hydroxylamidobis (sulfate) Ion by the Absorption
of NO in Aqueous Solutions of Na2S03 Containing Fe"-edta Complex," Ind. Eng.
Chem. Prod. Res. Dev.. 23. pp. 262-265.
13. Lyon, R.K. and J.A. Cole (1988). "Kinetic Modeling of Artifacts in the Measurement of
N2O from Combustion Source," Combustion and Flame, in press.
14. Montgomery, T.A., L.J. Muzio, G.S. Samuelsen, and G.R. Offen (1987). "Continuous
Infrared Analysis of N2O in Combustion Environments," presented at the Joint Fall
Meeting of the Western States and Japanese Sections of the Combustion Institute,
Honolulu, Hawaii. (Submitted to the Journal of the Air Pollution Control Association).
1-63
-------
200
>%
1—
¦o
Q. 100
Q.
o
CM
Z
0
Liquid
SC>2, ppm
Figure 1. Concentration of N2O within the 250 ml sampling containers after 2
hours with artificial combustion products (N2, 02, CO2), 600 ppm NO, and 1 ml
of liquid (6) (Copyrighted by, and used by courtesy of, the American
Geophysical Union. )
H20 H2SO4 H2SO4 H 2 O
(pH-1.6) (pH-0.6)
0 0 1500 1500
Figure 2. Influence of furnace SO2 Concentration on N2O observed within the
sampling containers.
1-64
-------
400
300
6
CL
& 200
O
100
1 (
S02 - 2^70 ppm
1
•
- r
S02- 1210 ppm
- f
Lm.mmm
S02 - 716 ppm
-
' i
f f
f
0 5
10 15
TIME (HOURS)
20 25
a. N20
800
SOS- 716ppm
SOS - 1210 ppm
. S02 - 2470 ppm
5 10 15
TIME (HOURS)
20
b. NOX
25
TIME (HOURS)
Figure 3. Time Histories of N2O, NOx. and SO2 during Reaction in 270 liter
Tedlar Bags: Experimental Results
(a. N20, b. NOx, c. SO2)
1-65
-------
n2o
2N02(g)
S02(g)
/ >
7
/ 7
.2N02(a) HS03+H+^-i-S02(a)+H20
• ,i / V / / / / / /
2N02+HS03+H+(+H2°)—*HNO+HN°3+H2SC>4
hno+hno->n2o+h2o
¦ V-tW / /
Figure 4. Possible Overall Chemical Path Leading to N2O Formation in the
Sample Flasks (7)
1-66
-------
400
300
200
100
a N20
10 15 20
TIME (hours)
800
b. NOX
600
8
400
200
>S02- 7!6ppm
,SOZ- 1210 ppm
, SCB-2470 ppm
15
TIME (hours)
TIME (hours)
Figure 5. Time Histories of N2O, NOx, and SO2 during Reaction in 270 liter
Tedlar Bags: Experimental versus Model Results (a. N20, b. NOX, c. S02)
1-67
-------
LIGHT
SOURCES
CHOPPER
FILTERS
REFERENCE
CELL
N20
+
INTERFERENCES
Figure 6. Field Prototype N2O Analyzer Schematic
8
z
2
OL
CL
5
HI
3!
>
3
O
UJ
25
20
15
10
0'
0.0
A C02 (0 • 20 %)
* CO (0 - 400 PPM)
0 S02 (0 - 4000 PPM)
0^
^ *—: >—
¦ *¦ - * ¦ —
0.2 0.4 0.6 0.8
NORMALIZED CONCENTRATION
1.0
Figure 7. N2O Analyzer Evaluation: Interferences (without SO2 and NO2
Scrubbing System)
1-68
-------
i I
75 —
CT>
PPM
n2o
50
25
0
2 min.
24 ppm N2O
(SPAN GAS)
-/A
| SWITCH TO X-80UTH PROBES
SWITCH TO Y-SOUTH PBQBE8
TIME
Figure 8. Strip Chart Trace demonstrating N2O formation in Dormant Sample Lines
-------
Table 1.
Chemical Mechanism, Rate Constants and Equilibrium Constants at 25°C (13)
(Rate constants are in units of L/mol/s or L2/mol2/s)
Gas Phase Reaction
1. no + no + o2 = no2 + no2
Liquid Phase Reactions
2. no2 + hso3-= N02'+ hso3
3. HS03 + HS03 (+H20) =
H2SO3 + HgSO^
4. 2N02 + H20 = HN02 + HNO3
5. HN02 + HSO3- = NOSO3- + H20
6. NOSO3-+ H+(+H20) =
HNO + H2S04
7. HNO + HNO = N20 + H20
8. N0S03" + HS03" + HN0(S03)2-
9. HN0(S03)2- + H+(+H20) =
H0NHS03- + H+ +HSO4-
10. HN0(S03)2" + H20 =
HONHSO3- + H+ + HSO4'
Equilibrium Processes
11. N02(gas) = N02(aq)
12. S02(gas) = S02(aq)
13. S02(aq) = H++ HSO3-
14. HN02 = H+ + N02-
15. HSO4-= H+S042-
Pate Constant
6.73 E+3
3.00 E+5
5.00 E+5
7.00 E+7
2.40 E+0
5.00 E+1
3.00 E+4
8.50 E+1
1.90 E-2
1.50 E-6
Henry's Law Constants
H = 0.01 M/atm
H = 1.30 M/atm
Equilibrium Constants
K = 1.54 E-2 M
K = 5.10 E-4 M
K = 1.20 E-2 M
1-70
-------
Table 2.
N2O Emissions from Utility Boilers
UNIT
DESIGN TYPE
Fuel
LOAD
RATED TEST
(MW)
02
%
Continuous Emission Data
NOx N2O
PPM PPM
@ 3% O?
A
Foster Wheeler
• horizontally-opposed
.coal
620
600
3.5
982
11"
B
Combustion Engineering
• tangentially-fired
coal
790
750
5.2
325
<1
C
Combustion Engineering
• tanqentially-fired
coal
790
650
5.7
390
<1
D
Babcock and Wilcox
• opposed-fired
coal
280
270
7.0
692
6
E
Riley-Stoker
• horizontally-opposed
coal
280
270
6.0
662
2
F
Combustion Engineering
• tanqentially-fired
coal
180
90
8.3
498
1
G
Combustion Engineering
• tangentially-fired
coal
790
542
6.0
393
1
H
Babcock and Wilcox
• face-fired
oil
215
215
5.4
268
<1
1
Babcock and Wilcox
• face-fired
oil
115
90
3.3
268
1
" THE SAMPLE CONFIGURATION MAY HAVE LED TO N2O FORMATION IN THE SAMPLE LINES
-------
Table 3.
N2O Measurement Method Comparison: Grab Sample versus Continuous Analyzer
UNIT
DESIGN TYPE
Fuel
LOAD
CONTINUOUS
GRAB SAMPLES
RATED TEST
02
N2O
N2O
(MW)
%
PPM
PPM
@ 3% O?
@ 3% O2
A
Foster Wheeler
coal
620 500
3.5
8
5
• horizontally-opposed
380
4.4
9
10,12
B
Combustion Engineering
coal
790 750
5.2
<1
<2
• tangentially-fired
* 2 ppm is the detectibility limit; grab samples obtained with flasks containing NaOH
-------
Session 2
COMBUSTION N0X DEVELOPMENTS I
Chairman: D. Eskinazi, EPRI
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
The application of combustion modifications for NOx-reduction to
low rank coal fired boilers
by
K.R.G. Hein
Rheinisch-Westf31isches ElektrizitStswerk AG
Kruppstrasse 5
4300 Essen 1
West Germany
ABSTRACT
Stringent environmental control regulations in the Federal Repub-
lic of Germany aim for the NO^-emission from large combustion
equipment not to exeed 200 mg/rrr of dry flue gas (about 100
ppm) as daily average values, in general. This goal is expected
to be reached by the end of 1989 for all utility boilers with a
capacity of 300 MWth and larger.
Due to the fact that the properties of the coals in Germany
differ distinctly from those fired in other countries a direct
application of the limited experience with large scale N0X-
reduction techniques gained elsewhere was not possible. There-
fore, a large number of research and development projects were
initiated in the last 5 years throughout the Republic which
concentrated mainly on the two areas, namely selective catalytic
reduction (SCR) in the flue gases and modifications to the com-
bustion system. By these activities it has been found that the
application of combustion modifications is an economically in-
teresting alternative.
Up to now brown coal fired boilers with an electrical capacity of
more than 7000 MW in total were successfully retrofitted with
various combinations of combustion modifications.
The paper deals with the development of the N0x-reduction tech-
niques for large coal fired utility boilers, describes the solu-
tions applied, and presents the results reached so far in boiler
operati on.
2-1
-------
1. INTRODUCTION
The last two decades are characterized by a growing concern of
the public about the quality of our environment. This obvious
tendency is in particular emphazised by the increasing indica-
tions of negative effects of manmade emissions on human health,
fauna, flora and even on our nonliving surrounding like e.g.
stone and concrete buildings.
This general understanding lead in the past to legal actions in
various highly industrialized countries in order to limit the
discharge of i. a. noxious gases from stationary and mobile
sources. The consequence was that industry either had to choose
for alternatives with reduced emission potential or - as occured
in the majority of the cases - was forced to install removal
and/or conversion techniques to reduce the discharge of harmful
species.
The combustion of fossil fuels and their derivatives, as being
one of the important processes for the generation of secon-
dary/tertiary energy, is also a producer of various noxious gas
compounds depending on the fuel used. Among these gases, the oxides
of nitrogen (NOx) are believed to play a major role in the acid rain
formation. Therefore, the NOx-emission is restricted by law in
many countries already.
In the Federal Republic of Germany a general limitation of the
emission of noxious gases from stationary sources with a capacity
of larger than 50 MW+f, took effect in 1983 by the introduction of
the large furnace order. Additional restrictions were set by a
resolution of the state ministers for Environmental Affairs in
April 1984, which were further tightened by local agreements e.g.
in the State of Northrhine Westfalia. Table 1 shows the permis-
sible values. Today only the values given under 3 (right) are
valid and have to be applied to both new and existing installa-
tions. These values must not be exeeded in daily average. In
addition, 97 % of the half hour average values have to be below
the 1.2fold and all of these must not exeed the 2fold of the
values quoted. With these conditions the German NCL-emission
standards are at present the most stringent regulation with
regard to either the actual permitted emission values, the short
integration times for reporting, or the capacity and age of the
installations concerned.
2. REDUCTION OF NOx-EMISSIONS
For the reduction of the NOx-emission from combustion systems
there are, in principle, four groups of options available:
N0„ can be removed from flue gases by the selective cata-
lytic reduction (SCR) technique with the addition of ammo-
nia (NHo) resulting in the harmless gases molecular nitro-
gen (N2T and water vapour^O). This technique has been
introduced on large scale into the power industry in Japan
and Germany.
2-2
-------
Also, wet chemical cleaning can reduce the N0x-concen-
trations in the flue gases. Several principles are offered
the majority of which remove N0X and S0X in combination.
Most of these options are still under development and
large scale experience is hardly availabe at present.
There are various selective non catalytic reduction (SNCR)
processes available, in which the NO-reduction is caused
by the reaction with N-H or C-H based chemicals. These
materials are injected into the combustion chamber at
certain temperature levels. Various large scale demonstra-
tions are known from the USA from Japan and - more recent-
ly - also from Germany.
Finally the NOx-emission can be diminished by the applica-
tion of several modifications to the combustion system.
The effect of these technical alternations can be twofold
namely either lowering the production of nitrogen oxides
in the early stages of combustion or destroying once
formed NO during the burn out phase of the fuel. In gene-
ral, both phenomena cannot be hardly distinguished. Com-
bustion modifications were proven to be the first option
of NOx-removal to be considered and numerous applications
are known worldwide. However, so far only limited removal
efficiencies have been reported and a combination with the
earlier mentioned options e.g. with SCR is necessary in
order to comply with very low limiting values like those
valid in the Federal Republic of Germany.
Only recently a substantial increase of NOx-removal efficiency
has been reached by combining different combustion modifications,
in particular, for low rank coals. For such a fuel, the German
brown coal, the principles applied and the results obtained will
be described in the following chapters.
3. PRINCIPLES OF NOx-REDUCTION DURING COMBUSTION
The base for the application of all combustion modifications is
the knowledge of the most important influencing parameters.
With regard to the origin of the nitrogen it is known that the
mechanisms of NOx-production are temperature dependent and that at
least for the common dry bottom coal firing systems the source for
more than 95 % of the emitted N0X is the fuel-bound nitrogen.
Apart from temperature the accessibility of the fuel nitrogen to
oxygen is an important parameter which, in turn, is governed by
the local availability of the oxygen and of the N-containing
species.
Although at present not all of the fairly complicated mechanisms
involved are fully clarified a simplyfied scheme of the fate of
fuel nitrogen during combustion can help to understand the va-
rious options for diminishing N0X, figure 1. On heat treatment by
either combustion or external sources the fuel-N can be partly
liberated with the released volatiles and can be found e.g. in
the stable species NH3 and HCN with the rest remaining in the
2-3
-------
char. Both types of nitrogen species - gaseous and solid - will
be eventually converted into NO under oxygen-rich conditions. An
oxygen-lean surrounding, however, offers the possibility for con-
version into molecular nitrogen. Also, once formed NO can be
converted into N2 by the introduction of various reducing part-
ners. As already mentioned in chapter 2 and as indicated in
figure 1 different species can be utilized.
From this scheme it can be deduced that the following alterna-
tions to the combustion process should result in lowering the
NOx-emission:
- reduce overall oxygen level,
- diminish local oxygen accessibility,
- add oxygen stagewise,
- reduce temperature, in general ,
- avoid extreme temperature peaks,
- favour formation of NO-reducing species, or
- add NO-reducing species at adequate flue gas
conditions.
Consequently, a low NO^-combustion system could be described by
the following steps, figure 2: After ignition, a pyrolysis zone
should support the liberation of volatile nitrogenous species. As
next step an oxygen-lean combustion zone will produce flue gas
with a low NOx-concentration. This region has to be followed by
the burnout zone in order to complete the combustion reaction. If
necessary, NOx-reducing species could be also introduced at
appropriate temperature levels.
4. RESULTS FROM UTILTIY BOILERS
After the introduction of the large furnace order the Rheinisch-
Westfalisches Elektrizitatswerk AG (RWE), with more than 26000
MWe the largest utility in the Federal Republic, was considering
various options for the N0x-removal, in particular, for the
boilers based on Rhinisch Drown coal. The total capacity of these
units is 9300 MWe with single boiler sizes ranging from 75 to 600
MWe.
The Rhinisch brown coal differs distinctly from other fossil
fuels, e.g. bituminous coals or lignites. A high moisture level
of 55 - 62 % on raw coal basis, a variable ash content of 2 - 20
%, and a low nitrogen concentration of 0.3 - 0,4 % are typical
for this fuel.
Consequently brown coal combustion requires specifically designed
fuel preparation circuits and boilers which in conjunction with
the fuel properties result in specific flue gas conditions. The
flue gas contains about 20 % water vapour resulting in maximum
flame temperatures below 1200 °C and fairly constant N0x-concen-
trations well below 800 mg/mJ (approx. 400 ppm).
After detailed investigations of the SCR-technique and first
successful trials with alternations to the combustion chamber in
2-4
-------
research rigs and also in prototype boilers already in 1984, RWE
decided to retrofit all brown coal fired units with combustion
modifications only.
Based on the principles given in the previous chapter various
technical solutions were tested and lead to the following com-
bustion modifications:
The most simple method, namely a reduction of the overall excess
air level showed in short tests for both coal fired boilers and
research rigs the expected results (fifl. 3), namely a lowering of
the NOx-emission by about 60 - 80 mg/m3 (approx 30 - 40 ppmv) for
the 02-range investigated regardless of the boiler capacity.
However, the general application of an excess air level reduc-
tion is boiler and fuel dependent and may be limited due to the
dangers of reduced burn out, slagging, fouling, and corrosion.
Apart from reducing the overall excess air level a decrease of
the local stoichiometry at the burner will also have a positive
effect on the N0x-emission. By this principle, known as air
staging, the air/fuel-ratio at the burner face can be reduced to
values of 0,8 - 0,9. The remaining combustion air is introduced
as burn out air into the downstream area of the combustion cham-
ber via one or more levels of additional ports. For this reason
air staging is often called over fire air (ofa) technique. The
principle of air staging is widely applied to single burners by
splitting the secondary air flow and also to the combustion
chamber.
As an example for the application of this technique fig. 4 shows
the retrofit arrangement for a 300 MWe boiler with 2 levels of
air staging. Fig. 5 gives some results gained with this arrange-
ment. It can be clearly seen that air staging through the upper
level of air ports is more effective for N0x-reduction than
through the lower ports. This clearly shows that an early air
injection results in again enhancing the deliberately delayed
combustion by the lowered stoichiometry at the burner. In this way
the retarded N0x-formation in the flame root is shadowed by an addi-
tional NOx-production in areas where still unreacted fuel is
found.
As mentioned earlier the major goal of air staging is the reduc-
tion of the stoichiometry at the burner in order to minimize the
change for N0X-production provided that sufficient oxygen is
still available to ensure stable ignition of the fuel. Secondly
the intended production of gaseous N-containing species in an
oxygen lean envrionment is promoted which favours the conversion
to harmless molecular nitrogen. In fig. 6 air staging tests in
different boilers burning a variety of coals are expressed as a
function of the important primary burner zone stoichiometry. The
data show the expected tendency; the scatter is due to the para-
meters of the test.
The results given in fig. 5 have also indirectly shown that
sufficient residence time is required for the combustion before
the final air through the staging ports should be added. In
particular, for retrofit situations this may be a vital condition
and can be met by concentrating the fuel inlet in the lower part
2-5
-------
of the combustion chamber. The effect on the N0x-emission can be
clearly seen from fig. 7, when for a 600 MWe boiler (fig. 4) the
original 6 fuel injection levels are reduced down to 4 and 3
levels by closing the upper ports and still keeping full load
operation. This shortening of the burner hight also
leads indirectly to an additional reduction of the 02-level at
the burner by increasing the local fuel concentration. This again
is positive because it further lowers the potential for N0X-
formation in the flame root and also enhances the already men-
tioned production of gaseous nitrogenous species in a oxygen-
deficient surrounding, thus improving the chance for NO-conver-
sion to N2•
In addition, fig. 7 shows that also with a high concentration of
fuel (only 3 injection levels in operation) air staging can
further improve the NOx-reduction.
As an important result of the above described combination of
excess air level reduction, shortening of the burner belt area
and air staging fig. 7 also clearly reveals that the envisaged
goal for the final NOx-emission of 200 mg/m3 (approx. 100 ppmv)
can be reached for this fuel by a certain combination of com-
bustion modifications only.
The N0x-reducing effect can be further emphasized by recircula-
tion of flue gases from downstream of the fly ash precipitator
into the combustion chamber. This can be realized by either
mixing with the combustion air or injecting the flue gas as
separate jets directly into the boiler. This principle reduces
the local oxygen accessibility by dilution and may also lower the
local temperature. In addition - in the case of separate injec-
tion -, the jet momentum is favourable for the mixing of the
retarded low N0X combustion process in order to counteract the
residence time increase and, hence, reduce the danger of higher
burn out losses. Also it is assumed that the low 02-content of
the flue gas jets may be favourable in oxygen lean regions prior
to the injection of the final staged air in order to reduce ex-
cessive CO, a consequence of the delayed combustion and a poten-
tial problem for operation and environment.
In fig. 8 the effect of flue gas recirculation is shown for 4
already partly converted boilers of different type and capacity.
Although the N0X start values differ in some cases the enhanced
N0x-reduction with increasing the amount of recirculated flue
gases is obvious. Also noticible is that the recirculation effect
helps to reach emission values below the envisaged 200 mg/m3.
Another possibility for creating an oxygen-lean primary zone in
the vicinity of the burner mouth is the separation of the
fuel/gas mixture leaving the mill into a fuel-rich portion which
enters through the burner and a remaining fuel-lean portion which
can be injected through additional ports into the oxygen-poor
combustion zone. Apart from supporting the fuel pyrolysis close
to the burner the downstream injection of the fuel-lean portion
will result in similar effects to those expected from the flue
gas recirculation. If this principle, often called bruden separa-
tion, is combined with some of the previously described com-
2-6
-------
bustion modifications also NOx-emissions below 200 mg/m3 can be
reached.
The following figures 9-12 show typical examples of boilers
retrofitted to meet a final N0X emission of 200 mg/m^ or less. To
date - end of 1988 - 25 brown coal fired boilers with a total
capacity of 7200 MWe are already converted, the remaining 2100
MWe will follow before the end of 1989. Fig. 13 summarized the
results achieved so far. From these data it is obvious that a low
N0x-emission can be reached by various combinations of modifica-
tions. The choice for a particular combination has to be made
individually in each retrofit situation dependent on local condi-
tions and economic considerations.
5. EFFECTS ON PLANT PERFORMANCE
Assessing the combustion modifications it has to be taken into
account that due to the retardation of the combustion process
ignition problems and, in particular, unburnt carbon losses may
occur, which could also cause malfunction of the electrostatic
preci pitator.
Often a steep rise of the CO-concentration of the flue gas can be
observed when lowering the oxygen level, figure 14. This, in
turn, may also result in major operational problems like deposit
formation and corrosion. For bituminous coal fired boilers it is
known that high CO- (and H2S-) concentrations in an 02-lean
surrounding at or near the heat transfer surfaces may give rise
to material wastage.
Finally it should be noted, that when operating a flue gas desul-
furisation scrubber behind the boiler the change of Op in the
flue gas should be considered. Also any increase in flue gas
exit temperature could cause problems with the connecting ducts.
Therefore, when utilizing combustion modifications various nega-
tive side effects on plant performance could occur. This may
limit the application of the otherwise economically attractive
way for reducing the emission of nitrogen oxides with the conse-
quence of reconsidering also other more expensive options in
order to fulfil the requirements of existing emission standards.
2-7
-------
1
after July 1, 88
2
as early
as possible
3
after Dec 31,89
brown coal
1 000
200
200
bitum.
coal
dry
bottom
1 3 00
200
2 00
wet
bottom
2 000
200
200
oil
700
150
150
gas
500
100
100
1 Large furnace order (1 July 1983 )
2 Resolution by the state ministers of environmental affairs (5 April 1984 )
3 Agreement between gouvernment and utilities in the State of
Northrhine- Westfalia (November 1984)
Table 1: Permissible NCL-emission for existing units > 300 MWtb in the
Federal Republtk of Germany (values in mg/m3 in dry fTue gas
at 6 % 02;for ppm on volume basis devide by 2,05)
2-8
-------
(02~lean) (02-rich)
(02~rich) (02-lean)
Fig. 1: Conversion of fuel-N in coal combustion (simplified scheme)
2-9
-------
Fig. 2: Low NOx-comt>ustion (scheme)
2-10
-------
research rigs
5 % 02 6
3: NOx-emission as function of excess air level,
(expressed as oxygen content upstream of the
preheater)
2-11
-------
Fig. 4: The principle of air staging
(300 MWe, example)
2-12
-------
Fig. 5: Effect of air staging
(300 MWe, approx. 20 % of total air through staging ports)
2-13
-------
1400
1000 —
600
§ 200
0
-o-
o
A
0 0-6
0,8
1,0
1,2
1,4 sp
Fig. 6: Effect of primary zone stoichiometry Sp
2-14
-------
Fig. 7: Effect of fuel concentration by shortening of the
burner belt area
2-15
-------
flue gas
Fig. 8: The influence of flue gas recirculation
2-16
-------
ferfttrl .1 1; I :rr=rH j
62 m
52 m
20 m —
flue gas recirculation
burn out air
flue gas recirculation
reduction of burner hight
.. _ ' / \ ; | , ~y
Fig. 9: NOx-reduction by combined combustion
modifications
(600 MWe, example)
modifications: - reduction of excess air
- air staging
- reduction of burner hight
- flue gas recirculation
2-17
-------
27 m
22 m
L:
ii
19 m
~ ~-
"mm
burn out air
(air staging)
flue gas recirculation
reduction of burner hight
/tii 11 IfiX! 1
i
V ry-
"-tvy-fii
Fig. 10: NOx-reduction by combined
combustion modifications
(150 MWe, example)
modifications: - reduction of excess air
- air staging
- reduction of burner hight
- flue gas recirculation
2-18
-------
Reproduced from
best available copy.
burn out air
(air staging)
bruden turner
(fuel-lean)
reduction of burner
night
"2fn burner
(fuel-rich)
Fig. 11; NOx-reduction by combined
combustion modifications
(150 MWe, example)
modifications: - reduction of excess air
- air staging
- reduction of burner hight
- bruden separation
2-19
-------
burn out air (2)
(air staging)
flue gas recirculation
(2)
burn out air (1)
(air staging)
flue gas recurcu-
lation (1)
reduction of burner
hight
Fig. 12: N0X reduction by combined
combustion modifications
(75 MWe, example)
modifications: - reduction of excess air
- air staging
- reduction of burner hight
- flue gas recirculation
2-20
-------
technical
realIsation
number of
boilers
NOx~emission (mg/m5)
200 300 400 500 600 700
-
37
1
I
A
22
1
i ' '
A+B
22
! i i
A+B+C
13
1 i
A+B+C+D
17
A+B+E
3
cb
i
A excess air level reduction
B air staging
C burner hight reduction
D flue gas recirculation
E bruden separation
Fig. 13: NOx-reduction by combustion modifications;
results from selected boilers
(150 - 600 MWe, brown coal)
2-21
-------
Fig. 14: N0X~ and CO-emission
2-22
-------
PREDICING BOILER AND EMISSIONS
PERFORMANCE BY COMPARATIVE
TURBULENT/LOW NOx BURNER TESTING
ON A LARGE TEST FACILITY
Joel Vatsky
Director, Combustion and Environmental Systems
Foster Wheeler Energy Corporation
Clinton, NJ
Charles Allen
Senior Mechanical Consulting Engineer
Arizona Pulbic Service
Phoenix, Arizona
The work described In this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
2-23
-------
ABSTRACT
Based on results from pilot- and full-scale testing, an 800 MW opposed-fired "cell-burner" boiler
is being retrofitted with Foster Wheeler Controlled-Flow/Split-Flame (CF/SF) low NOx circular
burners. The design conversion involves change-out of the burners, firing-walls, burner piping,
and burner/pulverizer control system. Pilot-scale testing in an 80 million Btu/hr., two-burner facility
showed no significant differences in combustion or unit thermal performance when firing with
either turbulent or CF/SF low-NOx burners. Subsequent field testing indicated good agreement
between measured and predicted boiler emission and thermal performance and provided a
reliable basis to establish boiler performance guarantees. This paper discusses the design, cost,
and schedule for the burner conversion and results from the pilot- and full-scale tests.
INTRODUCTION
In March, 1972 the State of New Mexico Environmental Improvement Board (EIB) promulgated
Air Quality Control Regulation 603. B. This regulation set an emissions limit for nitrogen oxides of
0.70 pounds per million Btu from existing, pre-1971 coal-fired steam-electric generating units. All
five of the coal-fired units at the Four Corners Station were determined to exceed this limit under
normal full-load operating conditions.
The highest NOx emissions were produced by two Babcock & Wilcox (B&W), opposed-fired,
supercritical, once-through boilers designated as Units 4 and 5. Each boiler is maximum
continuous rated (MCR) for a steam flow of 5,446,000 Ibs./hr. at 1000/1000 F and 3590 psig to
produce a turbine-generator output of 820 MWG (780 MWnet). These boilers were built in the
late-1960's and went into commercial operation in 1969 and 1970, respectively; i.e., 2-3 years
before the State NOx regulation was established. B&W boilers of this period of manufacture were
equipped with two- or three-nozzle "cell" burners specially designed to maximize combustion
intensity and produce extremely high heat inputs in a compact burner zone. Figure 1 shows the
three-nozzle cell burners on Units 4 & 5 illustrating the close vertical spacing of individual burner
nozzles. As a result of this close spacing and rapid fuel-air mixing, the cell burners on Units 4 &
5 operate with flame temperatures approaching adiabatic conditions and produce an intense
Burner Zone Liberation Rate (Q/BZS) of around 400,000 Btu/hr-sq. ft.(FWEC basis). Full-load
NOx emissions average 1.25 pounds per million Btu at normal operating excess oxygen levels
in the furnace.
2-24
-------
Figure 1 - Three Nozzle Cell Burner
Between 1973 to 1980, Arizona Public Service Company (APS), a joint owner and the operator
of Units 4 and 5, conducted several NOx-control related test programs and attempted to achieve
NOx compliance by operational methods. Initial NOx reduction efforts centered around staged
combustion, off-stoichiometric firing. This technique involved shutting off the coal flow to selected
burners or pulverizers while leaving the secondary air flow undisturbed. Although this approach
reduced NOx emissions by 20 - 30 percent, it also aggravated already serious operational
problems such as furnace slagging and bottom-ash clinkering, coal pulverizer overloading, coal
pipe pluggage, and tube metal overheating in furnace outlet superheater and reheater sections.
While some of these combustion problems were subsequently corrected or improved, further
site testing in 1978 and 1980 showed that the 0.7 pounds per million Btu emission limit simply
could not be met on Units 4 & 5 by burner out of service off-stoichiometric firing techniques.
Based on these testing results and conclusions from technical studies conducted by KVB and
the boiler manufacturer (B & W) advising against the application of overfire air ports or B&W
low-NOx burners on Units 4 and 5, APS petitioned the EIB for operating variances from the 603.B
regulation. The variances were requested on the basis that control technology was not available
for reducing NOx emissions without jeopardizing unit operation. The EIB agreed and granted
variances in 1980 and 1983. Pursuant to conditions of the variances, APS continued to investigate
significant changes and developments in the area of NOx control technology. Ongoing develop-
mental work by the EPA and EPRI during this time indicated that low-NOx burners (LNB) were
2-25
-------
the most attractive coal-fired retrofit option. Studies conducted in-house by APS in 1985
supported these conclusions and further suggested that shorter-flame design LNB may be
compatible with the restricted furnace burner zone of Units 4 and 5 and capable of lowering NOx
emissions to the regulation limit. The most promising of such burners was the Foster Wheeler
Energy Corporation (FWEC) Controlled Flow/Split-Flame burner. However, several major con-
cerns still remained regarding potential adverse operating effects and the design feasibility of
replacing the cell burners with circular LNB. Specific process areas of concern were:
• lengthened flames impinging on the furnace side- or division- walls;
• increased slagging and/or clinkering in the lower furnace due to potentially
closer proximity of the flame envelope to the walls;
• elevated furnace exit gas temperatures which could result if the LNB caused
delayed combustion; and
• reduced combustion efficiency from potentially less intense, slower mixing
flames and reduced oxygen availability.
To evaluate these issues APS elected to conduct a pilot-scale burner test program which was
conducted at Foster Wheeler's Combustion & Environment Test Facility comparing the flame
shape/length, emissions, and combustion performance of a high-turbulent burner (as presently
installed on Units 4 & 5) to the FWEC low-NOx burner. Additionally, fuel-supply system engineer-
ing studies were completed to evaluate the performance, design and costs of alternate firing
arrangements which would accommodate the new LNB design. This paper discusses the
comparative burner tests, follow-up field boiler performance testing, and the preliminary design
engineering effort ultimately leading to a decision to retrofit the Foster Wheeler LNB on Four
Corners Units 3, 4 and 5.
PRELIMINARY SYSTEM DESIGN STUDIES
As noted in the Introduction, the three-nozzle cell burners on Units 4 and 5 posed a special design
problem in that there was no commercial low-NOx equivalent. Two basic approaches were
apparent: 1) develop and design a low-NOx, three-nozzle cell burner based on FWEC concepts
which would fit into existing burner ports; or 2) widen the burner spacing to allow installation of
a standard FWEC LNB design. Although FWEC expressed confidence in its ability to design a
low-NOx "cell" burner, the high costs for performing exploratory developmental testing and the
uncertain NOx and combustion performance of a full scale retrofit led APS and the other
participant owners to pursue further firing arrangement studies. It was recognized that spreading
the burners would substantially complicate the design and installation, and significantly increase
2-26
-------
costs; however, this conventional-design approach (versus an R&D path) offered the highest
assurance of NOx compliance without compromising the performance, efficiency or reliability of
the boilers.
OBJECTIVE AND REQUIREMENTS
OBJECTIVE
The objective of the preliminary system studies was to evaluate the performance, design, and
costs of alternate firing arrangements for the Four Corners Unit 4 and 5 boilers which would
provide a widened spacing between all burner openings. The existing boiler-plant design utilized
nine pulverizers and eighteen, three-nozzle cell burners (i.e., 54 burner nozzles). Five pulverizers
supply the front wall and four mills feed the rear wall. Each pulverizer supplies a pair of cell burners.
The front and rear wall arrangement of existing burners is shown in Figure 2. A conventional
arrangement for FWEC low-NOx circular burners of the capacity needed on Units 4 and 5 required
vertical and horizontal spacings between burner openings of at least 8 feet.
^ «
amewAU.
« i*xt v run.
DIVISION MALL
13'4 /a-
'l
<>••0 /?•
9A4
^TYPICAL CELL
"1
i
1
M"
111
NEAR UAU.-exXSTXNS
Figure 2
REQUIREMENTS
The following requirements were established for any acceptable firing arrangement design:
2-27
-------
(1) The firing pattern be of a conventional design and "successfully demonstrated" (with re-
spect to NOx and combustion performance) on a large coal-fired utility boiler.
(2) Localized heat fluxes in the burning zone should not exceed those produced by the cell
burner array.
(3) Furnace exit gas temperature must not increase more than 50 F above that of the cell
burner.
(4) Boiler slagging and bottom-ash clinkering problems must not be aggravated.
(5) The pulverizer system and Unit equivalent operating availability cannot be significantly re-
duced.
FIRING ARRANGEMENT ALTERNATIVES
Initially five (5) different firing arrangements were identified and assessed for performance and
general design practicability. This "screening" revealed that two of the options either had
numerous major interferences with boiler structural steel and piping or resulted in unequal firing
rates among burners and imbalanced heat input into the two combustion chambers formed by
the full-height division wall. The following three options showed promise and were selected for
further detailed evaluation:
1) Nine pulverizers and 54 burners with a row of six burners relocated on the front firing wall
to an elevation approximately nine feet above the existing top row.
2) Eight pulverizers and 48 burners, arranged in four rows of six burners on each firing wall,
utilizing B&W MPS-89N pulverizers.
3) Eight pulverizers and 48 burners, configured as under Option (2) but utilizing capacity-up-
graded B&W MPS-89G series pulverizers.
A readily-apparent means of widening the burner spacing on most of the cells was to take center
burner nozzles and relocate them horizontally within the existing firing zone. Although this worked
well for the rear wall, as illustrated in Figure 3, the greater number of burners (30 versus 24) in
the front wall necessitated that a group of six burners be moved upward. This was the concept
of Option (1). Burner openings at elevation 53'-3" would be relocated to elevation 82'-11" as
illustrated in Figure 4. The widened burner spacing and lowered burner zone heat release rate
2-28
-------
- ( SIDE*ALL
t UNIT & FULL
DIVISION WALL
t SIDEWALL-
9'*0 1/2"
VTYPICAL CELL
REAP WALL-EXISTING
FRONT WALL-EXISTING
13'-4 /8"
7^i
flC^
2F
XSC
-"¦Sf
IIA_-t
©
I I
OB.
t UNIT l> FULL
OIVISION WALL
9'0 !/2"
A6m DIA.
7C
10
IA
74.'-3'
65 '-T
57'-7"
4«'-lI*
ELEV.
RELOCATED
BURNERS,
l3'-4 1/8"
, t UNIT fc FULL
yS DIVISION WALL
¦ 9-0 1/2"
50.
5Cx. 6Ca, 69^.
3A<)
3a,
6E ISO 160
4F
7F
REAR WALL-RELOCATED BURNER PORTS
ARRANGEMENT AFTER RELOCATION
OF BURNERS AT 53'-3* TO 82'-ll"
82'-I I*
74'-3-
65'-7"
57'-7"
53'-3"
l6'-\I"
ELEV.
Figure 3
Figure 4
for this firing pattern offered the optimum NOx reduction while retaining all nine mills. However,
major concerns with this concept were: 1) increased furnace exit gas temperature and 2)
substantial steel and piping interferences in the upper windbox with the relocated row of burners.
Options (2) and (3) eliminated the need to raise burners vertically on the furnace front wall by
simply eliminating one pulverizer and six burners from the design. The reason this concept
appeared feasible, in general, was that Units 4 and 5 were in the middle of being retrofit with new,
increased capacity B&W MPS-89N pulverizers. B&W quoted the new pulverizers with a nominal
capacity 20 percent above the original CR-77 mills; (i.e., 60 tons per hour versus 50 tons per
hour). Thus it appeared that eight MPS-89N mills would provide an acceptable boiler plant design
on the basis that full-load MCR steam output from Units 4 & 5 could be carried with one pulverizer
out of service. Furthermore, industry experience with the MPS-89N series pulverizers had shown
them to be very reliable. Going to Option (3) effectively restored the lost capacity of converting
from nine to eight pulverizers; however these mills required substantially larger top housings. To
evaluate the three options further, APS contracted with :
2-29
-------
CONTRACTOR & WORK PERFORMED
B&W
To assess changes in FEGT with the elevated row of front wall burners; and to evaluate the capacity of the MPS-89N
mills (and new P.A. fans) with different fuel characteristics.
BLACK & VEATCH
Performed detailed availability analyses on the entire fuel-supply system with various firing arrangements.
UNITED ENGINEERS & CONSTRUCTORS
Evaluated interference and clearance problems with the enlarged G-series pulverizers and reviewed the scope of
required burner piping modifications.
FWEC
Evaluated NOx reduction effectiveness and made a detailed assessment of physical interference problems for the
elevated row of front wall burners under Option 1; evaluated Option 2 configururation and performed detailed test
program.
The following discusses the results of those evaluations.
COMPARATIVE BURNER TESTS
TEST FACILITY DESCRIPTION
Comparative burner testing was performed on Foster Wheeler's Combustion and Environmental
Test Facility. Shown in Figure 5 the CETF is configured to simulate a commercial steam generator.
The conditions produced include:
• Furnace residence time is limited to about 2 sec. maximum between the
burner pattern centerline and the furnace exit.
• Furnace Exit Gas Temperature (FEGT) is about 2200F max. and can be varied
by adjusting total furnace absorption.
• Burner/furnace aerodynamics are similar to those of commercial equipment
as are furnace mixing patterns, due to both overall geometry and the two-
burner-high arrangement.
• The arch configuration, which is required to provide proper conditions for low
volatile fuel combustion is also important to produce the desired velocities/res-
idence times and FEGT. The reduced upper furnace cross-section is respon-
sible for the increased velocity and resultant decreased residence time.
Decreased upper furnace heat transfer surface, as compared to the same unit
2-30
-------
without the arch, reduces absorption above the burner zone, thereby increas-
ing FEGT to levels normally achieved in commercial equipment. Commercial
practices, used throughout the CETF system design have been mated with
research-oriented considerations wherever practicable to maximize the use-
fulness and flexibility of the system.
A direct firing system is utilized; consisting of a hot-primary air-swept ball mill supplying two 40
million Btu/hr burners. Commercial (Forney) gas ignitors, ignitor and main flame scanners, as
well as burner management and overall control system are employed.
Combustion gases flow upwards from the burners, over the nose and down through an
economizer, followed by a heat-pipe air heater and then to the stack. The furnace is a
refractory-covered water jacket operating with natural circulation and feeding a steam drum.
Independent water circuits supply the furnace, economizer and tube-wall test panel located on
the rear wall. Calorimetric measurements are made of each circuit in order to obtain absorption
duties.
2-31
-------
TEST PROGRAM OVERVIEW
The primary purpose of the comparative burner testing was to quantify any potential changes in
FEGT or heat transfer duties when substituting Foster Wheeler's Controlled Flow/Split-Flame low
NOx burner for the turbulent intervane burner. The latter was installed as original equipment on
Four Corner's No., 3 a 225 MW FW steam generator, and has flame characteristics nearly
identical to the high turbulent closely-spaced cell burners on Units 4 and 5.
Of secondary interest, although still significant, interest were changes in emissions and other
combustion conditions. This was the case because the CF/SF low NOx burner has been first in
service since 1979; having been retrofitted to Public Service New Mexico's 360 MW San Juan
#1. On that unit NOx was reduced about 55%, to below 0.45 lb/million Btu, with no changes in
boiler performance. Similarly, other retrofits and new unit applications yielded low NOx emissions
with normal boiler performance.
Nevertheless, APS elected to fund an intensive test program to obtain data specific to the
conditions at Units 3, 4 and 5 at the Four Corners Station; utilizing the same coal fired on those
boilers. These results, coupled with boiler performance data obtained on Units 3 and 5, would
enable Foster Wheeler to provide definitive guarantees for both emissions and boiler perfor-
mance.
A two segment test program was prepared to first, develop baseline data for the turbulent
intervane burner; and second to obtain equivalent data for the low NOx burner. The following
comparisons were made:
• Changes in flame characteristics (shape, length, flame temperature axial
gradient). These were observed both visually with a Land heat flux probe.
Figure 6 - Intervane Burner
2-32
-------
• Ability to match the CF/SF low NOx burner's flame characteristics (FEGT,
horizontal and vertical heat flux profiles) to those of the intervane burner while
achieving NOx reductions of at least 50% without decreasing carbon burn-out
efficiency.
• Changes in FEGT, or distributions of heat flux on the furnace walls caused by
possible differences in flame characteristics or slagging conditions.
• NOx reduction effectiveness of the CF/SF burner.
• Changes in combustion efficiency (CO and carbon burn-out efficiency).
A statistically-based test matrix was developed for each burner's test series. A total of 25 intervane
and 27 CF/SF burner tests were performed at full load to permit statistically significant results to
be obtained. Formal test data, for the statistical analysis comparisons were taken after each
burner's combustion and emissions performance was optimized and its flame shape adjusted
for compatibility with the furnace confinement.
BURNER DESIGNS TESTED
The FW Intervane Burner, shown in Figure 6, has been in use for many years on coal-fired boilers.
It produces a short, bushy flame with good combustion efficiency (low CO and unburned carbon),
but generates high NOx levels. A single air register is used to produce high secondary air swirl
which results in a strong recirculation zone in the flame. Coal and primary air are injected into
the furnace through an annular coal nozzle. The resultant rapid mixing between the fuel and
combustion air produces the combustion characteristics noted above.
Flame characteristics have a significant role in the design of a steam generator since they define
the furnace heat absorption profile. A short, intense flame will produce a high absorption rate in
the burner zone, yielding an FEGT commensurate with that absorption profile.
A burner which produces a long flame will generate a flatter absorption profile with, generally, a
higher FEGT. Since most boilers are designed for short, intense flame burners, such as the
intervane burner used on Four Corners #3 or the cell burner used on Four Corners #4/5,
retrofitting such a unit with a long flame burner could have severe operational and performance
consequences. Among these are:
• High FEGT, due to reduced furnace absorption, resulting in higher stack
temperatures; thereby reducing tube life and overall boiler efficiency.
2-33
-------
• Increased fouling of superheater and/or reheater surfaces; causing de-
creased heat transfer surface duties; again resulting in reduced efficiency.
• Flame impingement on the rear wall of front-wall fired units; causing slagging,
tube overheating and, potentially, tube failures.
• Overheating of superheater and reheater tubes causing reduced service life.
• Increased carbon carryover.
Consequently, in order to be compatible with the pre-NSPS furnace designs of Units 3,4 and 5,
an LNB must be capable of producing a flame envelope and absorption pattern similar to that of
original equipment turbulent burners.
The FW Controlled Flow/Split-Flame low NOx burner, shown in Figure 7, was developed in the
mid-1970's. During the CF/SF burner's developmental phase the following design philosophy
was adhered to:
• Flame length and envelope to be equivalent to that of the turbulent burner
designs that have been in historical use.
2-34
MANUAL ELECTRIC OUTER
REGISTER 8LEEVE DAMPER REGISTER
DAMPER
Figure 7 - Controlled Flow/Split Flame Burner
-------
• Thermal characteristics to be such as not to compromise boiler design criteria;
the burner could then be utilized in new boilers without modifying the boiler
design from that used with turbulent burners and in retrofit cases without
adversely impacting boiler performance.
• Combustion air flow and swirl to each burner to be independently controllable.
• Adjustable primary air/coal velocity to ensure optimum relation between
primary and secondary air steams.
• No increase in primary or secondary air pressure drop so that existing PA and
FD fans can be used.
• Burner capacity of cover the complete range of industrial and utility use:
approximately 30 to 300 million Btu/hr.
• Plug-in retrofitability, i.e. no pressure part changes, no burner piping rear-
rangement and no major windbox modifications when installed on existing FW
boilers.
The successful implementation of this philosophy has yielded good combustion efficiency, with
NOx reductions of 50-60%, excellent stability and no detrimental affects on boiler efficiency or
operation.
The burner's components are described below:
• Perforated Plate with Sleeve Damper: used to control secondary air flow on
a per burner basis. By measuring the pressure drop across the perforated
plate an index of air flow is obtained. The air distribution vertically and
horizontally within the windbox is thus optimized by adjusting the sleeve
dampers to obtain equal burner stoichiometries. This is a one-time optimiza-
tion after which the "open" position is fixed. The sleeve damper has "closed",
"ignite" and "open" positions and is used, instead of the main radial vane
register, to shut off the air flow when the burner is out of service. It is controlled
by an electrically operated linear drive, but is not modulated with load.
• Dual Series Registers: provide improved flame shape control and NOx reduc-
tion by two-staging the secondary air. A key mechanical reliability feature of
this register configuration is that the blades and drive mechanisms are set
back from the furnace wall and are well "shaded" from direct flame radiation.
Consequently, the registers operate at windbox temperature and do not
overheat, warp or bind. Additionally, once the flame is optimized for proper
2-35
-------
flame shape and low NOx, the registers are fixed. They remain in their optimum
position and are not modulated with load or closed when the burner is taken
out of service since the sleeve damper performs that function. The burner
essentially becomes a fixed register type with the option for adjusting the
register if a major fuel change occurs; the drive mechanisms being manual.
• Adjustable Coal Nozzle: allows primary air/coal velocity to be optimized
without changing primary air flow. The proper relationship between primary
and secondary air is important for both NOx control and good combustion.
Once optimized no further adjustment is required.
• Split Flame Nozzle: segregates the coal into four concentrated streams. The
result is that the volatiles in the coal are driven out and are burned under more
reducing (oxygen-starved) conditions than otherwise would occur without the
split flame nozzle. The volatiles in the coal contain a high percentage of fuel
bound nitrogen. Combustion under reducing conditions converts these vola-
tile nitrogen species into N2, substantially reducing NOx formation.
Succinctly, only the sleeve damper, used to shut off the secondary air flow, is moved when the
burners are taken in or out of service. Thus after optimization the burners become fixed register
types.
Coal for the APS test program was provided by Arizona Public Service from the Navajo Mine
located adjacent to the Four Corners Station; approximately 800 tons of coal was utilized during
the testing. The following section presents the results of the test program.
SUMMARY OF COMPARATIVE BURNER TEST RESULTS
The intent of data collection and analysis was to make comparisons at specific load and excess
O2 conditions and to repeat tests several times to determine the amount of variability in the
dependent variables due to independent variables changing in the system. A difference of means
t-test was the preferred test; however, during actual testing it was not possible to control the
independent varaibles (e.g. load, O2) as closely as desired. Furthermore, changes in furnace
cleanliness with time over the first two weeks of testing (due heat transfer surface seasoning from
the 22 percent ash coal) introduced further unanticipated variability in the data. Hence, the
intervane and low NOx burner data sets turned out to be widely scattered over the specified load
and O2 test range, particularly in FEGT and furnace and convection pass absorption rates. As a
result, the "sample" variances were large. This problem was addressed by applying techniques
of multiple regression analysis to correlate the response variable to statistically significant
2-36
-------
independent variables. This procedure subdivides the total variability in the dependent variable
data into variations explained by a postulated model and unexplained variability. The net result
was a reduction in unexplained variability, and consequently, a reduction in the uncertainty
interval for the response.
The evaluation and statistical analyses of the test data were jointly performed by APS and FWEC
personnel. The following summarizes the results:
Combustion and Thermal Characteristics:
Of particular interest is the potentially adverse effects of lengthened flames causing rear wall
slagging and increased convection pass temperatures on Units 3, 4 and 5. The differences in
flame characteristics between the two burner types were evaluated on the CETF by both analytical
and visual methods. The following summarizes the results both analytical and visual methods.
Rear Wall Flame Impingement:
The high-turbulent IV burner produces luminous flames which are short and bushy; flames were
well off the rear wall of the CETF furnace which has a depth of 16 feet, during the APS testing --
approximately 3-4' for the upper burner and 5-6' for the lower burner. The low NOx CF/SF burner
is a more flexible device due to the dual registers and adjustable coal injector which permit a wide
range of flame shapes to be developed. However, under optimum conditions selected for the
APS comparative testing, low NOx flames had the same shape/length as the IV burner flames
but with improved stability.
Rear Wall Slagging:
A specially designed tube wall panel was installed on the back wall of the furnace to determine
if any adverse changes in rear wall slagging were caused by the low NOx burner. The slag panel
was designed to produce a tube metal surface temperature equivalent to that of Unit 3. Numerous
observations revealed no change in rear wall slagging and in fact the slag panel's heat absorption
was lower with the CF/SF burner as compared to the IV burner. In general, gas temperatures
near the rear wall were 200F lower with the CF/SF burners as compared to the IV burners. Based
on these two observations no increase in rear wall slagging on Four Corners Unit #3 is expected.
Furthermore, qualitative observations of furnace refractory wall slag build-up indicated reduced
deposition during the CF/SF burner operation.
Convection Pass Temperature:
The temperature of most importance in boiler design/operation is the Furnace Exit Gas Temper-
ature (FEGT), which is the thermal fulcrum of the steam generator or boiler. It defines the furnace
conditions and performance capabilities of the convection pass by setting the heat head available
2-37
-------
for the superheater, reheater, economizer and air heater.Increased FEGT will result in reduced
steam generator efficiency by increasing spray de-superheating requirements and raising stack
gas temperature. If the FEGT increases beyond safe design limits, the superheater tube metal
temperatures can rise high enough to cause structural damage, and excessive ash build-up may
occur within the closely-spaced superheater tubing. Consequently, it is critically important that
any combustion modifications for NOx emission control purposes not raise FEGT beyond its
design range.
Two methods of evaluating any changes in FEGT were utilized during the APS testing program:
• Direct measurement using a multiple-shielded, high-velocity thermocouple
(MHVT). This method was marginally successful due to problems of ash
pluggage in the ceramic shields.
• Back calculation of FEGT from the air heater inlet gas temperature (convection
pass outlet) through the convection pass.
The back-calculation method, which is a standard technique used for commercial steam
generating equipment, utilizes the flue gas temperature as measured with thermocouples at the
boiler's gas exit (usually the economizer exit). Heat balance calculations are then used across
each of the intervening heat transfer banks until the FEGT is determined.
In the case of the CETF boiler, the only bank upstream of the exit is the gas cooler (essentially
a low pressure economizer). Since, by measurement, the unit load and excess air (and therefore
gas flow), gas cooler outlet temperature and gas cooler duty were known, we can then calculate
the gas cooler inlet temperature (Tgci). This is the critical temperature from which the FEGT must
be determined via back-calculation through the convection pass upper cavity and screen tubes.
The method used to determine convection pass differences due to burner type was as follows:
Calculate Tgci for each test and determine, statistically, the significance of any differences
between turbulent and low NOx burners. From the mean values of the respective Tgci's, calculate
the predicted mean value of each FEGT.
The statistical decision criteria, defined by APS, required acceptance of a null hypothesis that
there was no significant difference between the furnace-outlet gas temperatures produced by
each burner. The required statistical test performed on the mean Tgci at 75 million Btu/hr heat
input and 3.5% O2 dry revealed:
2-38
-------
Tgci (IV)
Tgci (CF/SF)
Mean
1727°F
1731°F
No. of Data
25
27
Pooled Variance, Sp
30°F
Degrees of Freedom
45
Student's1 (from samples)
0.133
*0.05,v (critical value)
1.68
Therefore, since t sample < to.os, the null hypothesis that the means of the two data sets came
from the same population (there is no difference in Tgci due to burners), is accepted.
Further, since FEGT is directly dependent, data point by data point, on Tgci there is no difference
between FEGT's between the burners.
The calculated FEGT's for the respective mean Tgci's were:
Burner Type
Tgci
FEGT
IV
1727
2139
CF/SF
1731
2148
The resultant difference in FEGT's (0.42%) is too small to have any functional effect on the Four
Corners units, if in fact there is a real difference of this magnitude.
EMISSION DATA:
• NOx:
The mean NOx emission rate at full load (80 million Btu/hr) and 20% excess
air (3.5% 02 dry) with the IV burner was 1.00 lb/million Btu.
The mean NOx emission rate from the CF/SF low NOx burner under the same
conditions was 0.41 lb/million Btu. Thus, the average NOx reduction produced
by the CF/SF burner was 59%.
• CO:
Under conditions representative of normal coal-fired utility boiler operation
2-39
-------
(i.e., above 3.0% excess O2, dry), CO levels were statistically equivalent for
both the IV and CF/SF burners: 20-25ppm.
• Unburned Carbon:
This parameter was calculated as percent efficiency loss, rather than percent
unburned carbon in the ash, since efficiency loss reflects the actual effect on
steam generator performance. UBC was one of the decision parameters
defined by APS for evaluation.The CETF test data show:
UBC for IV burner
=
0.256%
UBC for CF/SF burner
=
0.322
Increase
=
0.066%
The increase amounts to 25% over the baseline IV tests. It should also be noted that both values
are in the range of normal coal-fired utility boiler operation and of current combustion efficiency
conditions at Four Corners #3. Also, during the testing no effort was made to reduce unburned
carbon levels from the LNB to match those of the turbulent burner.
In summary, it was clearly demonstrated the CF/SF low NOx burner can reduce NOx nearly 60%
while maintaining a short flame, commercial quality combustion efficiency and no significant
change in furnace thermal characteristics (particularly FEGT) or boiler efficiency.
BOILER PERFORMANCE EVALUATIONS
APS requested FWEC to provide both emissions and boiler performance guarantees for Units 4
and 5 if they are converted to CF/SF low NOx burners. A reliable set of operating data was needed
to provide the basis upon which predictions, and subsequent guarantees could be made. The
methodology Foster Wheeler uses to evaluate emissions and performance capability of a unit to
be retrofited with low NOx burners is to compare measured values (NOx, CO. FEGT, etc) with
respective predictions from FW's proprietary calculations techniques. For this case comparisons
were made between the cell-burner equipped unit and that unit as if it were equipped with
intervane and then low NOx burners. In particular:
• Compare cell-burner NOx emissions with predicted emissions from the boiler
if it were equipped with the intervane burner.
2-40
-------
• Predict NOx emissions from the unit if it were retrofitted with CF/SF low NOx
burners.
• Compare measured boiler performance with predictions for both IV and
CF/SF burners as evaluated by FW's boiler performance computer program.
Since the CF/SF burner produces a flame shape that is not significantly different from that of
turbulent high NOx burners, and there is no evidence from several industrial and utility boiler
retrofits that the FEGT is increased, FWEC's boiler performance program does not differentiate
between the intervane and low NOx burners. The pilot scale test program, outlined above,
confirms the previous experience demonstrating that there is no significant difference in FEGT
between the intervane and CF/SF burners, although the latter reduces NOx by nearly 60%.
FWEC was contracted to do a complete boiler performance test on Four Comers #5 in order to
develop the necessary thermal and emissions data to enable guarantees to be offered. Table 1
summarizes key data from the as-found unit and compares it to predicted performance.
Performance is predicted for the unit with the existing configuration and with the new 8-mill/48-
burner configuration.
Results indicate that boiler performance would not be significantly changed by retrofitting the
CF/SF low NOx burners in the new configuration.
TABLE 1: BOILER PERFORMANCE DATA MEASURED VS PREDICTED
Unit
Configuration
Existing
Exisitng
New
No Mills l/S
9
9
8
Excess Air
18
18
18
FEGT
2394
2404
2405
RH Spray (1000 Ib/hr)
0
5.4
5.7
RH Temp
1029
1000
1000
NOx (lb/10 6 Btu)
1.27
1.32
<0.6
2-41
-------
PERFORMANCE EVALUATIONS
NOx Reduction Effectiveness
Foster Wheeler predicted that if the cell burners were all replaced with the CF/SF low-NOx burner
in a conventional widened burner firing pattern NOx emissions of Units 4 & 5 could be reduced
by 50 to 55 percent; or to an emission level at full-load of between 0.55 - 0.60 pounds per million
Btu.
Independent estimates of the NOx performance achievable with a full retrofit of low-NOx design
burners in the cell array was in the range of 0.85 - 0.90 lbs/million Btu. The reduced effectiveness
predicted for this approach was based on concerns over flow and temperature interferences
between closely-spaced adjacent flame envelopes such that proper staging could not be
achieved. However, Foster Wheeler's opinion is that a low NOx cell design based on CF/SF
Technology would provide lower NOx levels than these predicted by the independent estimates.
Pulverizer and P.A. Fan Capacity:
B&W analyzed pulverizer and primary-air fan performance on the basis of coal properties shown
in Table 2. Results from the analysis confirmed that MCR boiler output could be achieved when
operating with seven MPS-89N mills, in their "worn" condition at 70% fineness, and firing the
Design (worse-case) coal. At this condition the pulverizers would be operating at their full rated
capacity. However, Units 4 and 5 are normally operated at a derated, full-load output of 743
MWnet. This corresponds to approximately 95% of boiler MCR capacity. Additionally, less than
five percent of the time is a coal received with a heating value below 8500 Btu/lb. B&W's analysis
showed that during normal full-coal operation with eight pulverizers in service the mills would be
loaded to approximately 80 percent of rated capacity.
TABLE 2
COMPOSITION, WT. %
PARAMETER
DESIGN COAL
AVERAGE COAL
Ash Content
23.0
22.0
Moisture
16.01
13.0
Fixed Carbon
31.0
34.5
Volatiles
31.0
30.5
Nitrogen
1.2
1.1
Heating Value, Btu/lb
8500
8750
Grindability, HGI
48
50
2-42
-------
B&W also evaluated the capability of new TLT-Babcock primary air fans to supply the necessary
quantity of hot, primary-air against the increased resistance of an eight mill system configuration.
Based on one mill out of service operation and firing the "Design" coal, B&W calculated that the
new fans had margins of 20 percent on flow and 29 percent on pressure.
Effect on Furnace Exit Gas Temperature
Both Foster Wheeler and Babcock & Wilcox were asked to evaluate expected increases in furnace
exit gas temperature (FEGT) for a retrofit of their LNB in the 48 burner widened firing pattern
(Options 2 or 3) and with a relocation of six burners in the lower front firing wall to an elevation
nine feet above the existing top row (Option 1). The following results were calculated from each
manufacturer's Furnace Performance program:
Predicted Increase FEGT (F)
48 Burner Array
54 Burners with an Elevated Row
on Front Wall
B&W
12.0
20.0
FWEC
0.0
10.0
B&W noted that the slower combustion and "lazy" flames of its LNB accounted for the predicted
12 degrees F increase in FEGT with the widened 48 burner firing pattern. The Foster Wheeler
prediction for Options 2 or 3, based on FW's other field data, was subsequently supported by
the CETF pitot-scale testing.
Design FEGT at MCR conditions on Units 4 & 5 is around 2250F MHVT. General design practice
on high ash content coals and those with high slagging potential is to limit the FEGT to the Initial
Deformation Temperature (oxidizing) of the ash. A minimum value for T /.d.o. for Four Corners
coal is 2300F. Hence the approximate 10F increase in FEGT calculated by both manufacturer's
from elevating a row of front wall burners was judged acceptable.
Changes in Localized Heat Fluxes
For either the 8 or 9 mill widened burner firing patterns, it was expected that localized heat inputs
into the surrounding burner-zone waterwalls would change from those produced by the cell
burners.The general feeling was that the widened pattern of LNB would provide a more even
distribution of heat release. Detailed analysis confirmed this belief. Figure 8 shows the distribution
2-43
-------
of incident radiative fluxes predicted on the division wall from "slices" of the flame envelopes (near
the point of peak thermal emission rate) for the cell and "spread out" burner patterns. As
BURNER - D.W. FLUXES
(KW/M2)
Figure 8-lncident Radiative Flux Cell vs. Low NOx Burners
anticipated, directly opposite the concentrated heat release of the lower cell burner closest to
the division wall, the localized heat fluxes were significantly higher than those produced by the
widened burner spacing.
It should be pointed out that this analysis was only intended to be indicative of the changes to
be expected from the revised burner pattern and was not formulated to accurately account for
the total incident radiative flux on the wall surface. Peak total incident fluxes on the order of 500
KW/meter sq. (or around 160,000 Btu/hr - ft.sq.) would be expected in the highly confined burner
zone of Units 4 & 5.
Availability Losses with Removal of One Pulverizer
Black and Veatch Consulting Engineers (B & V) developed a sophisticated computer model,
using Monte Carlo simulation techniques, to analyze comparative pulverizer-system and unit
availabilities with installation of 8 or 9 MPS-89 pulverizers. The model incorporated the physical
arrangement and operating logic of the entire Units 4 & 5 fuel supply system. The analysis
2-44
-------
required estimates of the mean-time-between-failure (MTBF) and the mean-time-to-repair (MTTR)
for all of the fuel system components; this data was developed by APS based on plant operating
and maintenance records.
The model accounted for variability of coal properties, unit heat rate, and the availability of
first-point heaters and the balance of plant by use of statistical probability distributions. Other
features of the model included realistic accounting for unit minor and major overhauls, preven-
tative maintenance on the pulverizers, and storage capabilities of the coal silos. The base case
modelled nine MPS-89 N pulverizers on each unit, and compared its availability performance to
modified systems using eight MPS-89 N or G series mills. The results in terms of average
equivalent operating availabilities (EOA-%) are summarized below:
Subsystem
Nine MPS-89 N
Pulverizers
Eight MPS-89N
Pulverizers
Eight MPS-89G
Pulverizers
Pulverizers
99.83
99.09
99.58
BOP
76.22
76.22
76.22
Overall
76.15
75.82
76.04
Thus the effect of removing one pulverizer was an average decrease in unit EOA of 0.33 percent,
or an overall mean difference in energy production of 20,000 MWH. At APS' current and projected
costs for replacement power, the 20,000 MWH loss of energy production corresponded to a
differential annual fuel cost penalty of approximately $150,000. Two-thirds of this loss could be
recaptured by upsizing to MPS-89G pulverizers.
Sensitivity analyses were performed for the 8 mill, MPS-89N system design to assess the effect
of improved component reliability. Base MTBF values were doubled (i.e., failure rates were cut
in half) for each component. Results of the analysis showed that unit EOA was fairly insensitive
to component reliability within the estimated range of component failure rates. For example,
reducing coal feeder failures from 48 to 24 per year/unit only increased unit EOA by 0.03%. The
most sensitive operational problem in the analysis was coal silo plugging which stops the coal
flow to two pulverizers for an average of 20 hours. Silo plugging accounted for over 50 percent
(11,000 MWH) of the total lost generation.
2-45
-------
DESIGN REVIEWS
Interference Problems with Elevated Row of Front Wall Burners - The lower furnace of Units 4 &
5 is a two-pass design where fluid enters alternating tubes around the periphery in the first pass
and flows upward to an intermediate elevation where it is collected and rerouted back down to
the bottom of the furnace. The second pass outlet flows are collected at the same elevation as
the first pass outlet and routed through "mix bottles" for pressure and temperature equalization
before being introduced in a third pass to the top of the furnace. This enclosure mix area on Units
4 & 5 is approximately 10-15 feet above the existing top elevation of cell burners, on both the
front and rear walls. APS Engineering suspected that numerous interferences would occur in
trying to relocate a row of burners in this area.
Detailed review by APS and FWEC revealed the following major burner interferences in the mix
enclosure area:
1. The first pass horizontal outlet headers interfered with the burner registers and sleeve
damper assembly.
2. Supply tubes feeding the second pass downcomer interfered with the burner barrels and
register drives.
3. Two boiler-support diagonal braces [14WF158] were interfering with the outside pair of
burners.
In addition to these major interferences, several other problems were discovered including
conflicts with: a row of sootblowers; supply piping to the third-pass inlet headers; and windbox
internal trusses and platforms. B&W also advised that with burners located higher up in the
furnace, gas temperatures would be hotter in this area which might require moving the entire mix
enclosure section of the furnace upward; B&W indicated that a full circulation study would be
required. All of these problems stopped any further evaluation of Option (1).
PULVERIZER ARRANGEMENTS
Figure 9 shows the existing site general arrangement of pulverizers on Units 4 & 5. Burners on
the furnace rear wall are fed by the outside mills # 1, 2, 8 & 9 while the center mills #3-7 supply
coal to the front firing wall; mills #3, 4, and 7 supply the lower front wall. The obvious candidates
for pulverizer removal, for converting to an eight (8) mill system design, were one of the three
mills supplying the lower front wall. However, a thorough review by UE&C showed that the best
2-46
-------
mill for removal on each unit was mill #9. Removal of mills # 9 involved the least amount of burner
piping changes and provided the following additional benefits:
• The common silo serving mills 4-9 and 5-9 could be removed from service.
Maintenance and repair of this silo was always a problem since Units 4 & 5
were never scheduled for overlapping overhauls.
• Access for removal of roll-wheels on mill 5-9 was very tight due to the close
proximity of cable trays.
• An alternate pathway would be available to the pulverizer alley for maintenance
personnel and equipment access.
U E&C also reviewed the design feasibility of installing enlarged top housings/classifier assemblies
on the MPS-89N pulverizers for capacity upgrading. The outline dimensions of the enlarged top
housings were two feet higher and three feet larger in diameter than those of the MPS-89N mill.
The focus of UE&C's review was to identify any major interferences with boiler structural steel,
large high pressure piping, cable trays, or physical conflicts between adjacent mills. Results from
this design review were as follows.
UNIT 5
W&C6CIX ENCLOSURE
UNIT 4
WIMDBOX ENCLOSURE
MILL No 3. 4 k 7 ON EACH
UNIT SERVE TVC LOWER
FRONT FIRING WALL
HILLS SELECTED
FOR REMOVAL
*04 7 MILL No 7 COAL
PIPING REROUTE!
O 4-6 TO SUPPLY
^ EXISTING No 9
_ MILL BUffCRS
0-8
o-*
0*-3
o-'
SITE GENERAL ARRANGEMENT OF PULVERIZERS
ON FOUR CORNERS UNITS 4 15
Figure 9
2-47
-------
Several significant interferences were identified between the larger top housings and existing
cable trays and structural steel. Specifically, pulverizers # 4-1 and 5-2 interfered with cable tray
support steel and trays, which would require redesign of the cable trays system. Pulverizers #
5 on each unit interfered with horizontal structural steel necessitating that it be moved up 7'-6"
to eliminate the problem. Other minor conflicts were noted. Nevertheless, none of these
interferences were so severe as to preclude implementation of this design.
BURNER PIPING
In the cell burner firing arrangement, each pulverizer supplies coal to two (2), three-nozzle cell
burners, one on each side of the division wall. As indicated by the number/letter designations in
Figure 9, the pair of cells per mill are all non-symmetrical with respects to the division wall except
for pulverizer #4. As a result, the heat release across the furnace is not uniform during startup
or with any mill out of service except #4. In evaluating burner piping design for the new low-NOx
burner system, two basic alternatives were available. Both were based on an 8 mill, 48 burner
configuration. The first alternative was based on the maximum reuse of existing piping and
routings. It involved replacing only the piping necessary to supply the relocated burners and to
accommodate the tangential inlet connection on the FWEC low-NOx burner versus the center
inlet on the B & W cell burners. Piping replacement for this alternative included:
(a) All burner piping above the feeder deck at the front and rear furnace faces. The quantity
was estimated at 1050 ft.
(b) New burner piping from the outlet of mill #7 to connect with existing piping (from deleted
mill #9) supplying burners on the rear firing wall as shown schematically in Figure 9. The
amount was estimated at 400 ft.
(c) Sections of burner piping from front wall mills #3-6 below the feeder deck required by the
new burner locations. An additional 250 ft. was estimated for this modification.
The second alternative was to completely redesign the burner piping to an arrangement where
each pulverizer fired a horizontal row of burners. This type of design is standard for newer,
wall-fired, circular burner boilers. However, UE&C emphasized the significantly increased com-
plexity of this piping configuration and estimated that 3500 ft. of new, rerouted piping would be
required.
The obvious operational advantage of horizontal firing is a uniform heat release across the furnace
regardless of the number of pulverizers in service or their loading. Uniform heat distribution across
2-48
-------
the furnace during startups and at reduced loads with mills out of service minimizes FEGT
imbalances and extremes in superheater-outlet tube metal temperatures, particularly during
startups - before a coolant steam flow is established in the superheater tubing. Nevertheless the
initial cost differential projections by UE&C were $1.0 million per unit for upgrading to the
horizontal firing arrangement.
APS Engineering further evaluated the differences between the two alternatives and discovered
that Plant Maintenance was planning on replacing all of the existing bends and some sections
of straight-run burner piping, due to severe erosion, during the next two annual overhauls. This
fact substantially reduced the scope and cost differences between the two alternatives. Reas-
sessment of alternative No. 1 showed that its total or combined burner piping replacement was
around 2500 ft. when the maintenance scope was added. Total piping replacement for Alternate
No. 2 remained unchanged at 3500 ft.
A more detailed cost evaluation also showed that UE&C's preliminary estimates of alternative
No. 1 had not accounted for removal, handling, and reinstallation of existing piping below the
feeder deck in order to install new front wall piping. This omission plus the added maintenance
scope reduced the cost differences for upgrading to horizontal firing to $250,000 on Unit 4 and
$400,000 on Unit 5. Additional design and construction factors favoring a complete burner piping
redesign to horizontal firing were:
• Proper upflow flame rotation provided on all burners adjacent to the division-
and side-walls.
• An existing lower elevation of horizontal burner conduits, with short radius
elbows near the mill outlet, were eliminated.
• Piping fit-up and interface problems minimized.
• Speed and ease of erection improved.
• Eliminated the uncertainty about the condition of piping targeted for reuse and
the potential for substantial, costly field piping fabrication.
PULVERIZER/BURNER CONTROLS
The existing Burner Management System for the cell burner array was designed with nine (9)
identical burner-pair control logic sub-loops. Each sub-loop provides logic control on one
pulverizer and its pair of cell burners for operation of:
• Ignitors
2-49
-------
• Burner Air Registers (Not in Service)
• Pulverizer Outlet Isolation Valves
• Burner Impeller Position (Not in Service)
• Pulverizer Drive Motors
• Coal Feeders
• Mill Lube Oil System
Two gas ignitors are mounted between the closely-spaced nozzle openings on each of the 18
cell burners.
Installation of the new low-NOx burner system with its 48 circular burners (and individual ignitors)
in a widened burner firing pattern more than doubled the I/O requirements of the Burner
Management System. Since the existing system of electromechanical relays and timers was
outmoded and had no expansion capability, it had to be replaced. APS Control-System engineers
evaluated various options and selected a new, Bailey Net 90 microprocessor-based distributed
Burner Management System.
ECONOMICS OF CAPACITY-UPGRADED PULVERIZERS
Both short and long-term economics were evaluated to determine the cost-effectiveness of
capacity upgrading the MPS-89N mills. Since the new N-series mills were already installed on
Unit 5 the modifications involved upgrades for Unit 4 and retrofitting on Unit 5. The estimated
M&L cost to upgrade to larger top housings/classifiers and larger grinding zone elements on the
Unit 4 mills was $400,000; however, the projected costs for retrofitting the Unit 5 pulverizers was
$2.0 million. An additional $1.5 million was estimated to eliminate interferences and clearance
constraints caused by the larger top housings and cover increased design engineering and
construction inspection costs. Thus the total added capital costs for increased capacity pulver-
izers was estimated at around $4.0 million for both units.
The "savings" or return on this investment was the difference in energy production, and
replacement power costs, provided by the upsized pulverizers. This difference was estimated at
26,600 MWH annually, or around $200,000 per year in the short-term. Long-term economics
based on a present worth analysis of differential fuel costs showed a similar unfavorable
"break-even" point at around 19 years.
2-50
-------
SELECTED SYSTEM DESIGN
Based on the studies outlined above, and the successful pilot-scale comparative burner testing,
in mid-1987 the owners of Units 4 and 5 approved the following boiler-plant design modifications
for NOx control.
1. Deletion of one pulverizer from each unit and conversion to an 8-mill/48 burner configur-
ation utilizing MPS-89N pulverizers and FWEC circular LNB.
2. Replacement of essentially all burner piping, except for straight runs along the sides of the
boiler, to accommodate the revised burner pattern and provide horizontal firing on all pul-
verizers.
3. Replacement of furnace firing-wall tubing panels in the burner zone.
4. Installation of a new Bailey Net 90, programmable-logic distributed Burner Management
System.
5. Addition of Forney SIE UV and IDD2 flame detectors for ignitor and main coal flame scan-
ning.
SCHEDULE AND COSTS
SCHEDULE
Detailed design engineering for the Four Comers Units 4 and 5 NOx Abatement Project began
in September, 1987, shortly after project approval. APS' Generation Engineering staff performed
all detailed engineering work. Initial work involved development of specifications and drawings
for major equipment and materials procurement including the new burners, waterwall panels,
and Burner Management control system. The lead times for delivery of this principal equipment
are shown in Figure 10 along with the spans and interfaces for various phases of design
engineering, procurement, and construction; schedules shown for Unit 4 are "actuals".
To complete the detailed engineering phase of work on schedule, a project team of eight
engineers (3 ME, 3 CSE, 2 EE and 1 C-S) and five designers were assigned full-time. An
engineering Project Manager was also appointed to plan, coordinate, and direct work activities
and monitor progress against the plan.
2-51
-------
The magnitude and complexity of the project required that the construction work be planned
around 60-day major Unit overhauls scheduled for the spring of 1989 (Unit 4) and 1991 (Unit 5).
TAMCTCMT
t.O ENBDCER1NS
I.I PWF1 TMTMAflY CM.
1.8 OCT AXLED OCBXSN
a.o eoutpmn t material
pnocunoonr
a.i LOMOy aunoa
a.a UATERtMLL PANELS
a.s ».m. control art.
2.4 Bumo PIPINO AW
VALVES
3.0 CONSTRUCTION
5.1 PRE-CUTAflE
CONTROLS MO
ELECTRICAL
3.2 KOMNXCAL INSTALL
3.3 OUTAGE C t C
4.0 STARTUP/TESTZNB
1967 . .
MClilAWDMC JF^MJUAjBSE JFMAfciVl*B5SE JWfciKH|P»IE ]
TO «
IWR
SCHEDULE PLAN. NO* ABATEMENT PROJECT.
FOUR COWERS UNITS 4 1.5
Figure 10
COST
The finalized, total cost estimate for retrofitting the new low-NOx burner system on Unit 4 is $17.5
million. Unit 5 projected costs are slightly higher due primarily to escalation. The finalized estimate
is based on design engineering being 100% complete, and all equipment/material PO's and
construction contracts awarded. Costs were distributed as follows for major elements of the
project.
Unit 4 Costs ($000)
Component
Material
Labor
1.
Burners & Scanners
4,640
950
2.
Burner Zone Tubing Panels
860
1,200
3.
Pulverizer/Burner Net 90 Conrol System
1,450
750
4.
Burner Piping, Hangers, Valves
1,600
1,500
5.
Structural Modifications and Platforms
225
500
6.
New Ignitors
155
906
7.
Furnace View Ports
140
120
8.
Removal of Abando
150
9.
APS Engineering & Construction Administration Labor
1,260
2-52
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
REDUCTION OF NOX EMISSIONS FROM A 500MW
FRONT WALL FIRED BOILER
P Beard,* W J D Brooks,** K Johnson,***
K J Matthews,**** P Wells,***** J Vatsky,******
~Central Electricity Generating Board
Eggborough Power Station, Eggborough
Nr Goole, North Humberside. DN14 OBS United Kingdom
**Central Electricity Generating Board
Generation Development and Construction Division
Barnett Way, Barnwood, Gloucester. GL4 7RS United Kingdom
***Foster Wheeler Power Products Ltd
Olympic Office Centre, No 8 Fulton Road
Wembley, Middlesex. HA9 OTH United Kingdom
****Central Electricity Generating Board
Research Division, Marchwood Engineering
Laboratories Marchwood, Southampton.
S04 4ZB United Kingdom
*****Central Electricity Generating Board
Operational Engineering Division (Northern)
Beckwith Knowle, Otley Road
Harrogate, Yorkshire. HG3 IPS United Kingdom
******Foster Wheeler Energy Corporation
Perryville Corporate Park
Clinton, New Jersey. 08809-4000 USA
ABSTRACT
The Central Electricity Generating Board is committed to a policy
of NOx emission reduction and to this end all coal fired
500-660MW(e) boilers in the UK will be converted to low NOx
combustion systems before 1996. This paper reports the pilot
conversion of a 'large burner1 500MWe front wall fired unit at
Eggborough, England. The FWEC controlled flow split flame burner
was used for this conversion and data are presented showing the
NOx reduction effected and identifying the potential long term
2-53
-------
problems of such a conversion when normal UK coal supplies are
burned. A consistent NOx reduction of 40-45% has been shown.
The burners were installed in 1986 by Foster Wheeler Power
Products and the boiler is constantly being monitored for
resultant effects in a 3 year demonstration and proving period.
INTRODUCTION
The Central Electricity Generating Board (CEGB) has adopted a Low
NOx Retrofit Programme in which all large coal fired boiler plant
(500MWe and above) has been divided into four groups or
'families', based on similarities between the combustion system
components involved. The boilers at Eggborough Power Station are
included in Family 2, which comprises front wall fired furnaces
incorporating coal burners of approximately 58MWt capacity.
Thus Boiler 2 at Eggborough is now the prototype unit for Family
2, and the experience gained during the three year post
modification test programme could be directly applied to a further
11 boilers.
The purpose of the conversion is to assess the performance
available from proven technology when applied to existing plant
and to identify any operating or maintenance costs which may arise
from the conversion. Of particular concern is the effect of NOx
technology on plant which is firing United Kingdom (UK) coal with
a significant chlorine content, and on plant which already suffers
to some extent from fireside corrosion.
The ultimate aim is to provide plant which will comply with the
European Economic Community (EEC) large plant directives, if and
when these are fully ratified and accepted. Unit 2 at Eggborough
has an estimated life expectancy of 25 years beyond 1988.
SYSTEM INTENT
The main intention of the retrofit was to establish the NOx
reduction available from current combustion technology and to
assess any additional plant operating costs incurred. To this end
it was necessary to determine the performance and NOx output of
unmodified plant, and Nol boiler at Eggborough was selected for
this purpose. Because of the deterioration normal to a boiler as
operating hours increase in the 38 months between overhauls, it is
not possible to test the trial boiler "before and after"
2-54
-------
conversion, as the post overhaul performance differs substantially
from that produced pre overhaul, such factors as casing duct work
leakage affecting boiler performance considerably.
Boiler Nol (unmodified) and No2 (modified) were to be tested in
the immediate post overhaul state and with the exception of course
of the burners, were judged to be as nearly similar as
practicable.
To summarise, the objectives of the tests were:
1) To minimise the NOx emissions from the modified boiler by
careful optimisation of the system using the variables available.
2) To characterise and compare the performance of the modified
and baseline units.
3) To assess the long term performance of the modified boiler by
monitoring emissions and efficiency during normal commercial
operation.
4) To accurately identify any operational and maintenance cost
increase as a result of the modification.
CHOICE OF LOW NOX BURNER
Following an investigation of the available technology and
detailed discussion with burner manufacturers, the Foster Wheeler
Energy Corporation (FWEC) controlled flow split flame (CF-SF)
burner was selected as a suitable burner for the prototype
installation.
The main reason for this decision was that this design of burner
had been installed and operating in large boiler plant for some
years and had demonstrated NOx reduction of 50%+ at input sizes
both smaller and larger than the 58MWTh/burner required.
Subsidiary reasons were that, on US coals at least, there were
apparently few material or other burner problems experienced on
front wall fired boilers.
PLANT DESCRIPTION
There are four 500MW(e) units at Eggborough Power Station, North
Humberside. The boilers are of Foster Wheeler John Brown design,
2-55
-------
operating at 159 bars steam pressure (TSV) and 565/565 deg. C
steam temperatures. The furnaces of each boiler were originally
divided into 4 cells by a central water wall and two steam
division walls. The steam division walls were truncated early in
the station life (last unit was commissioned in 1969), leaving a
furnace divided only by a central water wall. The boiler is fired
by 24 coal burners arranged in three horizontal rows of 8 burners
on the front wall. Air supply is via a common windbox. The
pulverised coal is supplied by 6 Foster Wheeler D9 tube ball
mills, each mill feeding four burners. Full load can be attained
on 5 mills but it is normal to operate 6 mills for loads over
450MW.
The oil igniter burners are not original equipment, but are of tip
shut off design by Spectus, and have been in service since 1972.
Class G residual fuel oil is used at Eggborough for boiler warm up
and coal ignition/stabilisation duties.
The modifications to the boiler and replacement of the burners on
No2 unit were carried out by Foster Wheeler Power Products (FWPP).
The burner registers were manufactured by FWPP to FWEC designs at
the FWPP Hartlepool works, but the nozzle tip and coal
distribution nozzle castings were made in the USA by FWEC and
shipped to FWPP who completed the assembly. A considerable
increase in quarl throat diameter was required by the CF-SF burner
(43" compared with the 38" of the intervane turbulent burner) and
this necessitated a large amount of boiler front wall tube work
modification. Panels of 24 tubes each containing the tube sets
for 3 burner quarls were shop fabricated and were installed
complete. Eight such panels were required.
Also installed were 6 air ports complete with dampers, adjacent to
the boiler sidewalks. These ports could not be installed in
accordance with FWEC practice due to mechanical restrictions and
were therefore located one on each burner horizontal centre line,
the intention being to provide blanketing air for the boiler
sidewall if required. The Eggborough boilers have a history of
sidewall tube metal corrosion in well defined areas; this is
countered by the extensive use of co-extruded boiler tube, the
outer tube sheath being 25/20 Cr Ni stainless steel.
The FWEC CF-SF burner (1) consists of a register with radial
inflow swirl vanes, which swirl all the air entering the burner,
and an inner swirl vane system which swirls only the secondary air
inside the secondary/tertiary flow divider.
2-56
-------
The coal nozzle consists of a sliding tip and an outer nozzle
(fabricated originally, but subseguently cast in later US
installations: cast nozzles were adopted for Eggborough) which
divides the fuel into four axial streams. The tip movement varies
the amount of fuel admitted to an inner annulus and also to the
four outer axial streams. In order to meet the 3 8 month component
life reguirement, FWEC modified these nozzles, substituting an
annular conical outer nozzle containing four vanes which imparted
swirl to the coal flow. With these nozzles installed target NOx
reduction was not achieved and after a short period of operation,
axial flow split flame nozzles were substituted with a significant
improvement in performance. The boiler design did not readily
lend itself to the installation of overfire air ports (lack of
space above the top burner row). Because of the doubts regarding
furnace corrosion and the heavy expense which would be incurred by
an OFA installation it was decided not to fit such ports.
To provide additional security after a major change to the
combustion system the burners were fitted with flame monitors of
the Land Flamescan type which have proved successful in service,
discriminating reliably between flames and background radiation
levels. The flamescan system also proved useful in the coal flame
optimisation process.
RESULTS
The evaluation of the low NOx system proceeded in stages as
follows:-
1) The establishment of the baseline NOx and performance levels
by tests on Nol soon after the outage of 1986. This unit was
similar to No2, but unmodified as mentioned previously.
2) The optimisation and testing of No2 to establish the 'as
converted' performance levels.
3) Further regular test work which would identify any
deterioration in performance from No 2 as operating hours built
up.
4) The establishment of operating costs both as efficiency losses
and also as increased maintenance/component replacement levels
resulting from the conversion.
2-57
-------
FUEL
The normal Eggborough fuel supplies were used for the boiler test
work, variations in fuel composition from test to test being
identified by frequent sampling and analysis. On average the
station is supplied by coal from up to 27 mines but an attempt was
made to use only coal from 2 or 3 sources during boiler tests.
A typical (as received) test coal analysis was:-
Optimisation Instrumentation (Low NOx Conversion)
As the unit was basically a prototype it was considered necessary
to install the FWEC required level of post economiser gas analysis
instrumentation. FWEC consider this essential to the combustion
system optimisation process and therefore a 32 point probe
sampling system was installed from which 0 , CO and NOx levels
could be measured at evenly distributed points across the
economiser exit ductwork.
Test Instrumentation (Reference and low NOx boilers)
The formal boiler trials were made using gas analysis
instrumentation in the post ID fan ductwork, oxygen
(paramagnetic), carbon monoxide (infra red) and NOx
(chemilumincescent) gas analysers were used at this point. Carbon
in dust determinations were made by the normal station processes.
Several hundred items of plant performance data were logged at
minute intervals during each test using the station mainframe
computer. The data were transferred to hard disc for subsequent
off-line analysis.
Total moisture
Volatiles
Ash
Fixed Carbon
C.V. (kJ/kg)
Nitrogen
10%
27%
16%
47%
25.5
1.3%
INSTRUMENTATION
2-58
-------
UNMODIFIED BOILER NOX EMISSIONS
As mentioned previously, No 1 boiler was selected as the datum
unit and was tested in early 1986. The test programme was aimed
at establishing the basic combustion and heat transfer
characteristics of the boiler whilst recording NOx levels. Most
of the tests were conducted at or near 500MWe.
The NOx recorded levels are shown in Figure 1 and are corrected to
3% oxygen dry as are all NOx levels quoted in this paper.
The average NOx level from the load range 477-503MW was 748vpm,
but some variation was found with time due to fuel variation and
boiler casing deterioration. The maximum boiler load is
controlled by the draught plant capability which is indicated by
the CO level in the gases at the ID fan.
In conjunction with the exhaust gas analysis, p.f. distribution
and furnace wall gas analyses were performed, as were carbon in
dust level determinations.
The p.f. distribution was found to be good, but strongly reducing
conditions were found close to the boiler sidewalls. CO levels of
up to 4.9% being measured with all mills in service. The rear
wall CO levels were found to be low, with a maximum of .42%
measured.
NO 2 BOILER OPTIMISATION 1986 (Mk I BURNER)
The optimisation after conversion in Autumn 1986 was carried out
by FWEC and FWPP commissioning engineers, with close CEGB
assistance.
FWEC attempted to follow the procedure now well documented in
their publications. However, certain aspects of the normal
process were not possible because it was discovered that the
burner register pressure loss was too high. The station has been
proved to be short of fan power over the years as the original
plant did not possess adequate margins. Efforts have been made to
remedy the situation by tipping impellers and various other
draught plant modifications. No 2 is now limited to 485MW in the
summer and possibly 510MW in the winter. At the moment the FD is
the limit, but as the unit progresses through the 3 years between
outages and the casing leaks increase, the ID fan limit may be
2-59
-------
reached, although this now appears unlikely. (The 500MW unit load
is now expected, following turbine steam inlet modifications which
have restored the output to design levels after some years of a
450MW turbine restricted rating).
In accordance with normal FWEC practice the sleeve dampers were
adjusted in an attempt to balance the 0 and CO across the
furnace, but with detrimental effect.
All the trimming was therefore effected by the use of the moveable
coal tip, together with secondary and tertiary air vane
adjustment.
However, it should be noted that the burners were very stable and
produced only small levels of side and rear wall CO in this state
of modification. The settings were Outer Air Register (OAR)
50-60%, and the Inner Air Register (IAR) 25%.
The initial measurements after the optimisation phase indicated a
NOx reduction of 25-30% when compared with the unmodified unit.
This performance was disappointing and it was decided not to test
the boiler fully before the further modification.
Following these results, FWEC decided to substitute the split
flame tip. The change, however, was not possible until
August-September 1987 and then it was made in two stages as
replacement components became available.
NO 2 BOILER OPTIMISATION AND TESTS NOVEMBER 1987 (Mk II burner)
Immediately after the full complement of split flame tips were
installed the unit was re-optimised, but as the air register was
not modified in the conversion the pressure drop limitation
remained. Well over 100 tests were made during the optimisation
process. The resultant settings for the outer air registers were:
55deg. sidewall burners and 45deg. otherwise. The majority of
inner air registers were closed i.e., maximum swirl. All sleeve
dampers were fully open.
The NOx production did not appear to be significantly affected by
burner adjustments although flame stability certainly was
affected. The optimised settings quoted here gave satisfactory
stability, but did increase the pressure drop, again causing a
load restriction.
2-60
-------
NOX EMISSIONS
NOx levels were found to be substantially lower than those
produced in the "interim" phase and 42 0vpm or below was recorded
at 500MW as shown in Figure 2. This represents a NOx reduction of
4 3%, and this has been maintained or bettered by the boiler in
subsequent operation.
Following optimisation, the boiler was subjected to an intensive
test series, largely at high loads. Draught plant restrictions
prevented exploration of the NOx/excess air relationship at 500MW
and it was necessary to establish the excess air effect at
440-460MW, but with all mills in service the technique was simply
to vary the total airflow to a fixed fuel input unit load at a
constant furnace suction to prevent changes in boiler air ingress.
The effect on NOx emission for both the Low NOx and unmodified
boilers is shown in Figure 3.
During these tests the carbon in dust values were also measured by
analysing samples obtained by isokinetic sampling at the
precipitator inlet. These results are shown in Figure 4. The
relationships shown in Figures 3 and 4 enabled the costs in terms
of enhanced carbon in dust levels, to be evaluated on the basis of
test data. Operation to a NOx level of 390vpm would be expected
to produce a three-fold increase in the carbon in dust level.
Comparison of Figures 2 and 3 suggests that an increase in unit
load from 462MW to 500MW would be expected to increase NOx
production by about 8% at the design excess air level.
OTHER EFFECTS OF THE LOW NOX MODIFICATION
As mentioned earlier, the Eggborough boilers have a history of
fireside corrosion which has been countered by the selective use
of coextruded tube. The areas subjected to high rates of
corrosion can usually be detected by high (>1%) CO concentrations
at the furnace walls. The effects of the low NOx modification in
these retrofit circumstances have been to greatly increase the
area of furnace wall subject to reducing conditions. In fact, the
boiler walls are now affected by reducing gases on both side and
rear walls. CO levels of up to 9.0% were measured at the rear
wall. No furnace tube failures have occurred, however, but
accurate measurement of the metal loss rates awaits the
examination planned for the scheduled unit shutdown in 1989.
2-61
-------
The low NOx modification would be expected to produce some change
in the heat transfer pattern within the boiler. Compared with the
unmodified units, the modified boiler shows a reduction in heat
absorption by the furnace and an increase in heat absorption
within the convective zone. Whilst generally beneficial, in that
a problem in maintaining final steam temperatures has been eased,
the use of attemporation is very dependent on loading pattern and
fuel quality.
A number of erosion failures of the burner core tube near the
sliding tip have recently been experienced, 22 of the 24 core
tubes having been holed as a result of p.f. erosion. These
failures have occurred after 16000 hours operation, 9,500 hours
since the nozzle change, well short of the 25,000 hours (38
months) specified. The outage period in the USA is normally 12
months and thus there is limited experience within FWEC of burners
running for longer periods without examination and weld repair if
necessary. A modification involving higher grade erosion
resistant materials is being developed; meanwhile temporary
repairs are being made on a piecemeal basis to carry the unit
through to the 1989 outage.
CONCLUSIONS
1) The FWEC CF-SF Low NOx coal burner has been in operation on a
500MW boiler at Eggborough for three years producing a consistent
reduction of 40-45% in NOx emissions compared with an unmodified
unit.
2) An increase in carbon in dust levels has been observed
dependent on the level of excess air employed and NOx reduction
sought.
3) Changes in the heat transfer pattern in the boiler have been
measured. Final steam temperatures can be achieved more readily
on the unit, than was the case prior to the modification.
4) Significant deterioration in furnace atmospheres near the
water walls have been found but no tube failures have occurred.
5) Burner core tube metal erosion failures have occurred at
16,000 hours, well short of the specified burner component life
(38 months = 25,000 hours). Solutions to the problem are being
engineered.
2-62
-------
6) During the period since conversion, Unit 2 has operated as
required by Grid Control except for periods during testing and
modification. No major changes in operating procedures were
necessary.
ACKNOWLEDGEMENTS
The authors wish to express their gratitude to the Station Manager
and staff of Eggborough Power Station for their assistance in all
stages of the execution of this project.
This paper is published with the permission of the Central
Electricity Generating Board.
The work described in this paper was not funded by the US
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
REFERENCE
1. Vatsky, J. (1982) High Capacity Low NOx Coal Burner for
Retrofit and New Units. Power Engineering. January 1982.
2-63
-------
£
Q. 800
A January tests
• May tests
+ July tests
¦ September tests
x December tests
Average value (477-500mw)
748vpm S.D. 31
7« 0, (wet) at I.D. fan
Figure 1 NOx emissions from Unit 1 Boiler (unmodified)
with 6 mills in service at Unit loads of 477
to 503MW.
500
E
a.
O 400
300
¦ = 'A' side
• = 'B' side
N0X = 23.06 (02) + 342.3-
2.8
3.0 3.2
7. 02 (wet) at economiser
3.4
Figure 2 NOx emissions from Unit 2 Boiler (Low NOx) at
different excess oxygen levels.
2-64
-------
Figure 3
800
700
E
CL
>
>N
k_
TD
600
500
¦Unit 1 at Wmw
400 -
300
-Unit 2 at 462mw
Design excess air level
7. 02 (wet) at economiser
Comparison of NOx emissions from Unit 1 and
Unit 2 boilers at various excess oxygen levels.
2-65
-------
% Carbon in dust
Comparison of carbon in dust levels from
Unit 1 and Unit 2 boilers with emitted NOx
level.
2-66
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
N0X EMISSIONS RESULTS FOR A LOW-NOx PM BURNER RETROFIT
R. E. Thompson and G. H. Shiomoto
Fossil Energy Research Corp.
23342 C South Pointe
Laguna Hills, CA 92653
D. E. Shore and M. D. McDannel
Energy Systems Associates
15991 Red Hill Ave., Suite 110
Tustin, CA 92680
D. Eskinazi
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94303
ABSTRACT
The retrofit application of combustion modifications to pre-NSPS boilers has
received widespread attention by many electric utilities as a result of environ-
mental pressures to further reduce national N0X emissions. Although pilot-scale
tests suggest that such retrofits are promising, full-scale demonstrations are an
essential step in determining actual commercial viability. Boiler emissions
results are reported for the first of several full-scale low-N0x demonstration
projects now being conducted by the Electric Power Research Institute (EPRI). In
this project, the PM low-N0x firing system was retrofit to a 400 MWe tangen-
tially-fired boiler at Kansas Power and Light's Lawrence Energy Center Unit 5.
The PM firing system included low-N0x burners and two levels of overfire air.
Test results with a low sulfur bituminous coal indicate that N0X reductions of
30 - 50? were achieved at high loads (200 - 335 MWe) under carefully controlled
test conditions. Overfire air was found to be the dominant parameter in achiev-
ing these N0X reductions. Post-retrofit N0X emissions levels varied from 180 ppm
to 400 ppm across the load range in comparison to tuned baseline (pre-retrofit)
emissions of 350 ppm to 400 ppm. At low loads (less than 200 MWe), N0X
reductions diminished. For the higher N0X reductions, 0o requirements increased
by up to one percent. No change in unburned carbon or CO emissions was observed.
Long-term tests are now underway to evaluate emissions performance under normal
day-to-day operating conditions.
INTRODUCTION
Nitrogen oxide (N0X) emissions from new coal-fired utility boilers have been
regulated since the passage of the Federal Clean Air Act and establishment of the
New Source Performance Standards (NSPS) in 1971. Utility boilers, which
represent about one-third of the total U.S. N0X emissions, may be required to
further reduce N0X emissions. Because almost two-thirds of these utility N0X
emissions result from the pre-NSPS coal-fired boilers, the Electric Power
Research Institute (EPRI) and its member utilities are sponsoring an extensive
program to evaluate and demonstrate low cost, retrofittable N0X controls for pre-
NSPS coal-fired utility boilers. The low-N0x burner retrofit project at Lawrence
Unit 5 is one element of this program.
2-67
-------
excellent burnout characteristics, the low heating value limits the maximum load
to approximately 335 MWe. A typical coal analysis is given below:
PROXIMATE ANALYSIS ULTIMATE ANALYSIS
As
Dry
As
Dry
Received
Basis
Received
Basis
% Moisture
12.79
xxxxx
t
Moisture
12.79
xxxxx
% Ash
8.84
10.14
%
Carbon
60.60
69.49
% Volatile
36.73
42.12
%
Hydrogen
4.38
5.02
% Fixed Carbon
41.64
47.74
%
Nitrogen
1.29
1.48
100.00
100.00
%
Sulfur
0.91
1.04
%
Ash
8.84
10.14
%
Oxygen
11.19
12.83
100.00
100.00
Lawrence Unit 5 is generally typical of other pre-NSPS CE designs which was an
important factor in its selection. Specifically, the furnace dimensions are 50'-
8" wide by 40'-2" deep and 110' high yielding a volumetric heat release and plan
area representative of other CE units of similar capacity. However, the net heat
input per plan area is near the low end of the range of most CE units of similar
vintage when firing at a derated capacity of 335 MWe with the low sulfur Wyoming
coal. This reduced heat release rate influenced the baseline N0X emission which
were 10 to 15% lower than many other large coal-fired CE units.
PM Fi ring System
The PM firing system included new windboxes with integral ("close coupled") OFA
ports, separate compartments for additional ("separated") OFA, a new secondary
air control system, and a partial replacement of the coal piping. Each of the
four new windboxes contain five elevations of PM burners with each burner consis-
ting of a fuel-rich and fuel-lean coal nozzle. The PM burner system achieves
low-NOx emissions by controlling the local stoichiometry in the near-burner
region where coal ignition and combustion of the volatile matter occurs. Fuel-
rich and fuel-lean combustion zones are created at the burner front by aerody-
namic segregation of the pulverized coal in the approaching coal pipe as shown in
Figure 2. The close coupled overfire air ports in combination with the two PM
coal nozzles per burner elevation results in a taller windbox than the original
windbox as shown in Figure 3. Separated overfire air ports were installed
approximately six feet above the new windboxes as shown in Figure 4. The burners
and both sets of OFA ports can be tilted _+ 30° with respect to horizontal. The
separated OFA ports also have yaw tilt capability of + 10° relative to the firing
circle. The PM firing system is described in more detail in References 1, 2 and
3.
6aseous Emissions Measurement System
Gaseous emissions measurements were made for NO , 0?, CO, CO05 and SO2 using the
continuous gas analysis system shown in Figure 5. A total of ten sample probes
were installed in each economizer exit duct to evaluate the combustion uniformity
at the boiler outlet. Flow meters in the test trailer were adjusted to achieve
balanced flow from all 20 sample probes. The gas analyzers used at Lawrence 5
are listed in Table 1.
2-68
-------
Retrofit low-NOx combustion systems are based on the same concepts currently
being applied to new units. However, new boilers have larger furnaces which are
designed to accommodate the operating characteristics of low-NOx combustion
systems. A major element of the EPRI demonstration program is to determine the
impact of low-NO^ operation on the performance and maintenance requirements of
boilers with typical pre-NSPS furnace designs.
PROJECT OBJECTIVES AND SCOPE
The EPRI demonstration project at Lawrence Unit 5 involved the retrofit of the PM
low-NOx firing system on a 400 MWe tangentially-fired boiler. The primary objec-
tives were to assess the overall performance and reliability of the PM system and
to determine the maximum N0X reduction potential without an adverse impact on
plant operations. Other issues besides emissions were addressed including boiler
thermal efficiency, steam temperature control, furnace exit temperature, slagging
and fouling, and operating flexibility.
The full-scale evaluation of the retrofit PM burner system consists of three
major testing activities: (1) Pre-Retrofit Baseline Testing; (2) Post-Retrofit
Low-N0x Testing; and (3) Long-Term Evaluation.
Pre-retrofit baseline testing was conducted to determine the emissions and boiler
performance of the existing Combustion Engineering (CE) burners. Post-retrofit
tesing focused on characterization of the PM burner emissions and performance
with and without overfire air (OFA). Finally, optimization tests were completed
in preparation for a six to eight month long-term evaluation which was initiated
in January 1989. The long-term tests will evaluate day-to-day variations in
emissions and boiler performance during normal load dispatch operations as
opposed to the carefully controlled conditions estabilshed during the short-term
N0X characterization tests. This paper discusses the results of the emissions
tests conducted by Fossil Energy Research and Energy Systems Associates up to the
beginning of the long-term tests (EPRI RP 2916-3). The PM firing system retrofit
and evaluation of boiler thermal performance was performed by CE under separate
contracts (Reference 1, 2, 3).
Long-term day-to-day evaluation of the PM system and extrapolation of results to
the pre-NSPS tangential-fired boiler population will be completed in 1989 and
reported subsequently. As such, while the authors have attempted to generalize
results obtained to date, some results may be specific to Lawrence Unit 5.
SYSTEM DESCRIPTION
The unit selected for retrofit with the PM low-N0x firing system is the Kansas
Power & Light Lawrence Energy Center Unit 5 located in Lawrence, Kansas. Unit 5
is a tangentially-fired boiler rated at 400 MWe (2,805,000 lb/hr steam flow at
1005°F and 2620 psig, with 1005°F reheat). The single furnace design is fired
from the four corners at five burner elevations for a total of twenty pulverized
coal burners; see Figure 1. The original burners were conventional CE design
with +30° tilt for reheat steam temperature control.
The boiler was designed to burn a Midwest (Kansas) high volatile A bituminous
coal with a heating value of 12,750 Btu/lb. However, the coal now being fired is
a high volatile C bituminous coal from Wyoming's Powder River Basin having a
heating value of 10,580 Btu/lb. Although this coal is very reactive and yields
2-69
-------
GASEOUS EMISSIONS INSTRUMENTATION
Species
Analyzer
Analyzer Method
N0X
Teco 10
Chemiluminescent w/ss converter
°2
Teledyne 326
Electrochemi cal
co/co2
Horiba PIR2000
Nondispersive infrared
S0?
Western Research 721 A/D
Nondispersive ultraviolet
BASELINE TEST RESULTS
The baseline pre-retrofit test series characterized the emissions and performance
of the original equipment burners for direct comparison to the PM firing system
that was subsequently installed. Baseline boiler performance was monitored by
examining key characteristics such as 1) combustion efficiency (CO emissions, fly
ash carbon content and excess air level), 2) the ability to maintain the proper
superheat and reheat steam temperatures, 3) fouling or slagging in the furnace or
convective section, 4) radiant heat flux and 5) other emissions.
NO^ emissions from a tangentially-fired boiler are affected by a variety of oper-
ating parameters including: excess air, load, burner tilt position and secondary
air damper settings. Baseline testing of Lawrence Unit 5 under load following
conditions revealed a substantial variability in N0X emissions over a one week
test period. An example of this variability for a typical weekday is illustrated
by the solid data points in Figure 6. Differences in operator preference for
excess air level and secondary air damper settings were the primary cause of the
wide N0X emissions range. The effect of excess air variability is particularly
evident at low loads, where high excess air levels are often utilized to maintain
minimum air flow, windbox pressure or adequate steam temperature.
The tuned performance data in Figure 6 represents the range of N0X emissions
resulting from operation with the O2 level adjusted for efficient operation and
uniform fuel/air flow to the boiler using the extensive set of gaseous instrumen-
tation installed for the test program. With tuned operation, emissions were
relatively insensitive to boiler load; a characteristic exhibited by many tangen-
tially-fired boilers. In addition, the Wyoming coal exhibited very good burnout
characteristics with fly ash carbon content remaining well below 1% and typically
of the order of 0.5%.
Secondary air damper settings had a significant influence on N0X emissions at
higher loads. Full load N0X emissions could be varied from 250 ppm to 425 ppm at
3% O2 by adjusting burner damper and tilt settings as shown in Figure 7. Lower-
ing the burner tilts and increasing the burner zone fuel/air mixing increased the
N0X emissions by 16% relative to the tuned performance baseline conditions (-14°
burner tilt and normal fuel/air mixing). Conversely, a 30% N0X reduction was
attained by biasing the secondary air flow between burner elevations plus adjust-
ing individual dampers to minimize fuel/air mixing. No adverse impact of this
operating mode could be discerned based upon furnance flame zone and inlet con-
vective section obervations, gaseous emission measurements, and boiler perfor-
mance (load, steam temperature, etc.). Although the substantial N0X reduction
2-70
-------
achieved with the original equipment burners illustrates the potential of opera-
tional combustion modifications with little or no hardware expense, the reduction
that can be achieved on other units will depend on the boiler design, coal type,
and operating practices.
POST-RETROFIT TEST RESULTS
Emissions Characterization/Optimization
The post-retrofit test program included PM firing system emissions characteriza-
tion tests and optimization tests. The characterization tests defined the rela-
tionship between the PM firing system emissions and a number of boiler operating
parameters including:
• load
• C>2 level
• burner tilt
• OFA flow
• primary and secondary air damper settings
t windbox/furnace differential pressure
• OFA tilt (vertical and horizontal)
• OFA injection (port) location
The characterization tests also established a satisfactory operating envelope for
the PM firing system that was used in defining subsequent optimization tests and
the assessment of potential application to other tangentially-fired units.
N0X emissions data from characterization tests comparing the original equipment
(baseline) and the PM burners operating without OFA are summarzied in Figure 8.
The objective of these full load (300 MWe) tests was to isolate the PM burner
effect from the OFA staged combustion effect by making a direct burner-to-burner
comparison. An average N0X reduction of approximately 40 ppm or about 10 to 15%
was measured. However, since the PM burner was designed to operate with the OFA
ports open, these tests without OFA may have resulted in non-ideal burner throat
velocities and local air/fuel ratios which could alter N0X emissions.
Subsequent tests were conducted to determine the N0X dependence on the overfire
air flow rate (expressed as a percentage of the total combustion air flow).
Overfire air flow was estimated by damper position and flow area calculations
since direct measurement of both "close-coupled" and "separated" OFA flow was not
possible. Test results for a typical 300 MWe firing configuration are shown in
Figure 9. A N0X reduction of 70 ppm occurred for approximately every 10%
increase in OFA flow rate. OFA flow rates greater than 20% resulted in N0X
levels less than 200 ppm with a maximum NO reduction of over 50% obtained with
maximum OFA (estimated to be above 25%). Figure 9 clearly shows the dominance of
the OFA system in reducing N0X emissions. Approximately 80% of the PM firing
system NO reduction is attributable to overfire air which emphasizes the need
for a well designed and flexible OFA system, possibly with booster fans and sep-
arate ducting in some applications. This system should include provisions for
the direct measurement of OFA flow rate and fine tuning of the OFA system to
accommodate site specific constraints.
It should be emphasized that the absolute emission levels shown in Figure 9 are
specific to the Lawrence 5 boiler firing configuration and coal characteristics
for steady state and carefully controlled operating conditions. An adverse
impact on boiler performance and efficiency can result when operating at a high
OFA flow rate to achieve minimum N0X emissions.
2-71
-------
The N0X emissions characteristic of the PM firing system is influenced by a
number of interdependent parameters which affect OFA mixing and heat release in
the upper furnace region. An important element of the emissions characterization
testing was a parametric investigation of other secondary operating parameters
such as burner tilt, OFA tilt, and burner damper settings. Although burner tilt
is primarily used to regulate reheat steam temperature and cannot be directly
controlled to minimize N0X emissions, it was important to define its influence.
The relative angle between the burner flow and the OFA flow affects N0X
emissions, the minimum O2 or smoke threshold, boiler efficiency and the uniform-
ity of On at the economizer outlet. As the angle of separation between the OFA
and the burner tilts increased, the N0X emissions decreased as shown in Figure 9
for a series of full load, high OFA flow tests. A more effective staged combus-
tion condition occurs with a large separation angle which delays the mixing of
the OFA flow and main burner flow. More importantly, the data indicates that
little additional NO.. reduction occurs when the OFA tilt is raised above +10° for
a burner tilt of -14 , typical of full load pre- or post-retrofit operation.
However, the data in Figure 10 is presented for a specific O2 of 3.5% and does
not reflect the impact on minimum O2 and boiler efficiency. Large OFA to burner
separation angles adversely impacted the smoke point and combustion uniformity at
the boiler exit which partially offset the gains in N0X. This trade-off between
N0X, minimum 02, and combustion uniformity can become an important issue depend-
ing upon whether a unit is equipped with motor driven OFA tilts. A unit with
motor driven OFA tilts would provide the opportunity to maintain a desired angle
relative to the burner flow as the burner tilts moved to control reheat steam
temperature across the load range. With mechanically operated OFA tilts, one
fixed position must be selected as a compromise between N0X emissions, minimum
O2, combustion uniformity and boiler efficiency across the load range of the
unit.
The nonuniformity in combustion noted previously was most evident at high OFA
flows and occurred in two different forms. A large difference or "split"
developed between the north and south furnace halves exceeding 2% in some
instances compared to nominal differences of less than 1%. The split was occa-
sionally accompanied by an oscillation in the 0? level that required above
average settings to avoid smoking. The O2 oscillations were as large as 1.5% Oj
about the mean and sometimes caused smoking on the furnace half with the lower
O2. As a result, an above average O2 was required to preclude smoking during
these oscillations. On several occasions, testing was aborted due to this unex-
pected behavior. Typical O2 fluctuations during normal operation were 0.2 to
0.5% O2 (peak-to-peak).
The 02 nonuniformity and oscillation appears to be associated with a large separ-
ation angle between the OFA flow and the burner flow which typically occurs as
load increases and burner tilts rotate down. However, the exact cause of the O2
split and oscillation is not fully understood. It is believed to be related to
changes in upper furnace mixing patterns brought about by the overfire air.
Reducing the overfire air flow and raising the O2 level usually reduced the
magnitude of the fluctuations and alleviated the smoking. An 02 split of up to
1.7% was occasionally noted with the original equipment burners as the burner
tilts varied. However, the O2 fluctuations appear to be only associated with the
use of overfire air. Long-term testing is expected to provide more insight
regarding the day-to-day impact of this behavior. Operational adjustments to
overfire and O2 level are occasionally required, but it should be noted that the
unit's ability to meet load demand has not been compromised.
The optimization phase of the post-retrofit test program examined trade-offs
between N0X emissions parameters (such as OFA flow) and boiler performance
2-72
-------
parameters (such as operating O2 level, boiler efficiency, carbon utilization, CO
emissions, steam temperature control, and dynamic boiler response). The objec-
tive was to define a long-term operating mode that minimized N0X emissions with
little or no degradation in unit performance. One of the important trade-offs
with the current low-NOx firing system is the dependence of the minimum O2 opera-
ting point (i.e. smoke threshold) on the overfire air flow rate. A series of
tests was conducted at 300 MWe to define this dependence which is shown in
Figure 11. The family of slanted lines define the NO vs. 0o characteristic for
various amounts of OFA flow (0, 7, 15 and 27%). The lowest R0X level that could
be achieved at each OFA flow was limited by the minimum O2 or smoke threshold
indicated by the cross hatching. A significant N0X reduction occurred with
increasing OFA offset by only a minor increase in minimum O2 (<0.5%) until the
OFA flow reached approximately 20%. At higher OFA levels, the N0X reduction
benefits diminish as the minimum O2 increases. In addition, a decrease in boiler
efficiency also occurs with the increased minimum O2. For example, the one
percent increase in minimum O2 as the OFA increased from 0 to 27% OFA represents
an efficiency loss of approximately 0.4%.
The typical minimum O2 characteristic shown in Figure 11 is dependent upon boiler
load, coal properties, firing configuration, and other boiler operational para-
meters (tilt, etc). All of these factors must be considered in establishing the
recommended minimum 02 and associated nominal 02 operating level. In some cases,
it may eventually become necessary to utilize very high OFA flows (depending upon
future regulations) with some sacrifice in minimum 02. Other factors that must
be considered involve flame detection and safety considerations, carbon carryover
in the ash, ash disposal/sale requirements, etc. OFA testing at Lawrence Unit 5
was not pushed to these limits because of the adverse impact on boiler efficiency
and fuel consumption.
Overall NO^ Reduction Potential
A summary of pre-and post retrofit N0X emissions over the load range under tuned
and carefully controlled conditions is shown in Figure 12. Typical full load
(300-320 MWe) post-retrofit N0X emissions were 180 to 225 ppm. This represents a
30 to 50% reduction relative to the pre-retrofit baseline. As shown in
Figure 11, maximum NO reductions were achieved at the expense of increased 02 of
up to 1.0%.
Similar comparisons at 200 MWe indicate a 25 to 40% reduction in N0X relative to
the pre-retrofit baseline. However, optimum conditions during normal operation
will require automatic OFA tilt.
While testing focused on high load conditions, low load test results suggest
little or no N0X reduction is achieved below 200 MWe. This is believed to be
caused by one or more of the following: 1) decreased OFA flow and penetration;
2) minimum air flow stops on the combustion controls; 3) increased 02 require-
ments; and 4) non-optimum OFA tilt position.
Data generated during the characterization and optimization testing was used to
define a long-term firing practice. Systems performance will be evaluated over
5-8 months beginning in January 1989 based on continuous emission monitoring
data, utility operating experience, and occasional repeat performance tests.
Boiler fireside inspections will be made. Finally, an economic feasibility study
which will extrapolate/generalize results for the pre-NSPS coal-fired population
will be conducted. Results of these activities will be presented separately.
2-73
-------
PARTICULATE EMISSIONS DATA
Particulate emissions measurements at Lawrence Unit 5 included total particulate,
ash carbon content (loss on ignition or LOI), particle size distribution, fly ash
composition, and fly ash resistivity. Comparisons were made between the particu-
late emissions with the original equipment burners and the PM firing system. The
primary objective of these measurements was to determine if the particulate
emission characteristics with the PM firing system had changed in any manner that
could affect the performance of a particulate control device. Although these
measurements were made upstream of the wet scrubber at Lawrence 5, data was
gathered to allow an assessment of the impact (if any) on the collection
efficiency of an electrostatic precipitator (ESP). The results of the LOI, total
particulate, and particle sizing tests are presented below.
For both the baseline and post-retrofit tests, carbon content of the fly ash
remained low. Average carbon content for the baseline full load particulate
tests was 0.4%, while the average for the post-retrofit, full load tests was
0.3%. These very low values indicate essentially complete combustion with both
types of burners reflecting the excellent burnout characteristics of the Wyoming
coal.
Total particulate tests obtained during baseline operation showed an average of
4.0 grains/dscf ash loading, while data from recently completed post-retrofit
tests averaged 2.7 grains/dscf. The results indicate that particulate emissions
were lower during the post-retrofit tests compared to the baseline tests.
Further analysis of these results is planned once final post-retrofit coal
analyses are completed.
The primary focus of the particle sizing tests was on the fraction of particles
smaller than 1 micron, since these are the particles that are most difficult to
collect with an ESP. Particle size tests were performed using both a cascade
impactor and an electrical aerosal analyzer (EAA). The impactor sizes particles
aerodynamically with eight cut sizes from approximately 0.3 to 10 microns The
EAA sizes particles according to their electrical mobility and has an effective
measurement range of 0.03 to 1.0 microns. The results of the cascade impactor
tests indicate that no significant change occurred in particulate in the range of
0.3 to 10 microns. Preliminary results from recent EAA measurements suggest that
fine particulate emissions below 1.0 microns may be as much as four to five times
higher for low N0X firing than with the original equipment burners. This result
is contrary to prior correlations relating increased fine particulate with higher
N0X emissions. However, little or no sub-micron particulate data for low-N0x
burners is available for comparison to the Lawrence 5 data. Further analysis of
the EAA ash samples and comparison to pilot scale fuel rich low-N0x particulate
data are planned.
CONCLUSIONS
To date, the evaluation of the emissions and performance characteristics of the
PM firing system under short-term, carefully controlled conditions has been
completed. While further work is planned, valuable insights on the potential
retrofitability of this system has been gained. Key results follow:
• N0X reductions of 30 to 50% were achieved with the PM firing system at
high loads (200 - 335 MWe) under controlled test conditions.
2-74
-------
• While only limited low load tests were conducted, little or no N0X
reduction was measured under these conditions.
• Overfire air is the dominant factor in the N0X reductions achieved.
Specifically, about 80% of the total NO reduction appears to result
from OFA. As a result, design of the OFA system is critical and provi-
sions for site specific fine tuning of the OFA flow and upper furnace
mixing are recommended.
• For Lawrence Unit 5, typical full load NO^ emissions with 20% overfire
air were 180-200 ppm at an average operating 02 level of 3.5% and 200-
225 ppm at 4.0% O2. Further N0X reductions were achievable with
greater OFA flow rates of up to a maximum of 25%. For these condi-
tions, minimum Op levels were increased by up to 1% resulting in a
reduction in boiler efficiency.
• The minimum O2 level for the PM firing system was approximately 0.5 to
1.0% O2 higher than the original equipment burners.
• Total particulate emissions and ash carbon content were relatively
unchanged while burning a very reactive low sulfur mid-western coal.
The long-term PM system evaluation, now being initiated, will define: 1) achiev-
able, day-to-day NO reduction levels; and 2) boiler performance impacts under
normal load dispatch conditions. These results, together with a planned economic
feasibility study, will be used to formally extrapolate the results obtained at
Lawrence 5 to the pre-NSPS pulverized coal, tangential-fired boiler population.
Acknowledgements
The authors acknowledge Mr. Preston Tempero, Plant Manager and the rest of the
Lawrence Energy Center personnel for their cooperation and assistance in obtain-
ing the resulted presented. The technical support and cooperation of Combustion
Engineering is also acknowledged.
This project was funded, in part, by the Kansas Electric Utility Research
Program.
References
1. R. Lewis. "Boiler Performance Evaluation of a 350 MW Low-N0x PM Firing
System Retrofit." Proceedings: 1989 EPRI/EPA Joint Symposium on Stationary
Combustion N0X Control, San Francisco, CA.
2. M. McCartney. "Manufacturer's Update on N0X Control Technology."
Proceedings: 1989 EPRI/EPA Joint Symposium on Stationary Combustion N0X
Control, San Francisco, CA.
3. P. Tempero. "Installation and Operation of Low NO PM Burners." Presented
at the Fifty-ninth Annual Engineering Conference of Missouri Valley Electric
Association, Kansas City, M0, March 1988.
2-75
-------
Figure 1. KPL Gas Service, Lawrence Unit 5 - 400 MWe
Figure 2. PM Burner System
2-76
-------
33
m
-31'
SS.
a
isr
Ain
COAL
GAS
AMI
GAS
COAL
OAS
Am
OAS
COAL
OAS
Am
GAS
COAL
OAS
AVI
GAS
COAL
Ain
M
~
S
B
a
jg.
fi
S
~
J=3
a
a.
B.
SEPARATED-
| OFA
CLOSE-
COUPLED OFA
GAS
FUEL-RICH
FUEL-LEAN
OAS
FUEL-LEAN
FUEL-RICH
GAS
Ain
GAS
FUEL-RICH
FUEL-LEAN
GAS
FUEL-LEAN
FUEL-RICH
GAS
All
GAS
FUEL-RICH
FUEL-LEAN
A*1
i
-35'
CONVENTIONAL
LOW-NO,
Figure 3. Comparison of Conventional and PM Burner Configurations
Separated
Overfire Air
Nozzles ^
Top of
Close Coupled
Overfi re Ai r
Nozzles
Figure 4. Furnace View of Separated OFA Nozzles (Offset)
2-77
-------
Figure 5. Flow Schematic of the Continuous Gas Analysis System
-------
700
600
500
CM
<=> 400
ro
(Sj
E
O.
Q.
300
200
100
1
1 1
•
•
•
1
1
• •
•
•
•
•
•
m
. •
•
Tuned
Performance
• • ••
v t • a • •
-
^j|| Tuned Performance
i
• Load Following Data —
(Typical Work Day)
l i
100
200
LOAD (MWe)
300
400
Figure 6. Baseline N0X Emissions
2-79
-------
500
450
High NO
400
16% Increase
Tuned Performance
Baseline
o
n.
Q_
350
300
250
200
30% Decrease
Low NO
150
100
High NO: Max. fuel/air mixing,
full down burner tilts
Baseline: Normal fuel/air mixing,
-14 deg. burner tilts
Low NO: Min. fuel/air mixing,
biased fuel/air flow,
-14 deg. burner tilts
50
2 3 4 5
ECONOMIZER EXIT 02, %
Figure 7.
Full Load Baseline N0X Emissions Variations
2-80
-------
500
1
1 I
1 1 1
450
-
HI
400
111
JBIH
IS
350
-
IKKKFJSk
300
qPjIIP'
5-9
n
300 MW
-------
Figure 9. N0X Dependence on Overfire Air Flow
2-82
-------
500
450
400
350
300
250
200
150
100
50
0
"i 1 1 i i r
Burner Tilts
-20°
1 1 r
310 MWe
-27% OFA
3.5% 02
i I i i I I I I I I I L
-30 -20 -10 0 10 20 30
OFA TILT, DEGREES
Comparison of NO Dependence on OFA Tilt and for Burr
2-83
-------
500
450 -
400
300 MWe
-14° Burner Tilt
+20° OFA Tilt
350 -
300 —
Estimated OFA
® 250 -
Q_
Q.
200 -
150
100
50 -
Minimum On
ECONOMIZER EXIT 0?, %
Figure 11. Typical Minimum O2 vs. OFA Flow Trade-off
2-84
-------
700
600
500 -
OJ
o
co
E
Q.
400 -
300 -
200 -
Post-Retrofi t
Tuned Low-N0x
loo -
I
100
200
LOAD, flWe
300
400
Figure 12. Comparison of Pre- and Post-Retrofit N0X Emissions Data
2-85
-------
i
(Intentionally Blank)
2-86
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
RETROFIT AND BOILER PERFORMANCE EVALUATION OF THE LOW
NOx PM FIRING SYSTEM AT KANSAS POWER & LIGHT
R. D. Lewis, A. F. Kwasnik, R. C. LaFlesh
Combustion Engineering, Inc.
Windsor, CT
D. Eskinazi;
Electric Power Research Institute
Palo Alto, CA
ABSTRACT
A tangentially-fired boiler, Kansas Power & Light Lawrence Unit #5, was retrofit
in July, 1987 by Combustion Engineering with the low NO PM Firing System. This
retrofit was done as part of a comprehensive EPRI low N$ combustion
demonstration project.
The objective of the test program was to establish the minimum achievable NO
levels with the low NO firing system consistent with no adverse impact on
boiler performance or operation. This paper summarizes the effects of the PM
firing system on boiler performance and operation.
Results indicate that NO emissions reductions of 35% to 50% were achieved with
no adverse impacts on boiler operating characteristics such as start-up, minimum
load, and flame stability. While changes in boiler thermal performance were
measured, these changes have had no significant impacts on power generation.
For example, waterwall heat absorbtion was increased which resulted in a
decrease in the furnace outlet gas temperature (FOT) of less than 50°F.
However, this change in absorbtion (which can affect outlet steam temperature)
was easily accommodated via the automated windbox tilt feature integral in the
design.
Additionally, at operating conditions promoting the higher achievable NO
reductions (low emissions), there was a greater tendency to produce higher CO
emissions and smoke which was alleviated with higher excess air levels. These
higher excess air levels increased dry gas losses, and, as a result, reduced
overall boiler efficiency by 0.2-0.3%.
Further testing is planned, beginning in 1989, by EPRI and KP&L to characterize
NO emissions over an extended time period with the unit operating under normal
load dispatch.
INTRODUCTION
EPRI sponsored a comprehensive series of full scale tests of the PM firing
system at Kansas Power and Light's (KP&L) Lawrence Energy Center. The test
program was comprised of boiler performance and emissions tests. This paper
presents a summary of the results from the boiler performance testing conducted
by Combustion Engineering as the boiler OEM because of their specialized
expertise in this area. Gaseous and particulate emissions tests were completed
by an independent contractor, Fossil Energy Research Corporation (FERCO) .
Combustion Engineering retrofitted the PM firing system at the KP&L Lawrence
Unit #5 in the spring of 1987 under a separate contract to KP&L.
2-87
-------
BOILER CONFIGURATION
Lawrence Unit #5 is a controlled circulation, radiant-reheat, single cell
furnace of balanced draft design. The unit is designed for a maximum continuous
rating (MCR) of 2,805,000 Lb/hr of steam flow at a predicted superheat outlet of
1005°F/2620 PSIG and a power output of 400 MWe. At this MCR rating, a reheat
inlet flow of 2,450,000 Lb/hr at 638°F/612 PSIG is heated to a predicted reheat
outlet of 1005°F. Reheat outlet temperature is controlled by fuel nozzle tilt.
Superheat outlet temperature is controlled by desuperheating spray. Pertinent
boiler specifications and typical coal analysis are shown in Table 1. A side
elevational view of the unit is depicted in Figure 1.
Lawrence Unit #5 was designed in 1967, prior to the implementation of any NO
regulations. As such, it did not include any NO control techniques such as low
NO burners or overfire air (0FA). There were five coal elevations as well as
associated normal auxiliary air and warm-up oil compartments as shown in Figure
2. The unit has a nominal name plate rating of 400 MWe, based on the original
design Midwestern bituminous coal which had a higher heating value (HHV) of
12,770 Btu/lb and a Hardgrove grindability of 63. Full load operation was
limited to 300-330MWe during these tests due to both the addition of a wet
scrubber and a change in coal type fired since the original unit design and
construction. The Wyoming Powder River Basin (HHV of 10,750 Btu/lb, Hardgrove
grindability of 50) coal currently used limits pulverizer capacity as compared
with the original design coal.
(2)
The PM firing system retrofit included four (4) completely new windboxes,
separated OFA compartments, a new secondary air control system, a partial
replacement of the coal piping and larger fan blades on the pulverizer
exhauster. The new windboxes, shown in Figure 2, include five elevations of the
PM firing nozzles. An elevation of PM firing nozzles includes a fuel rich
{Cone} and a fuel lean {Weak} nozzle. Successive elevations of CONC and WEAK
fuel nozzle sets have also been invterted (ganged) relative to each other to
further emphasize the fuel rich and fuel lean arrangement. The new windboxes
also include two elevations of integral or "close coupled" (CC) OFA with
vertical tilt capacity. The separated (SEP) OFA includes two individually
controlled compartments and has manual vertical and horizontal tilt adjustment
capability. Note also, in Figure 2, that the pre-retrofit natural gas firing
capability for Unit #5 has been integrated into the PM firing system windbox
design. The burner windbox has normal automatic vertical tilt capability used
for reheater steam temperature control. The main windbox for the PM firing
system is four feet taller to provide for space for the close coupled overfire
air. The extra windbox envelope for the PM firing system has been accommodated
by stretching the original windbox height while keeping the original windbox
centerline constant. The separated overfire air windbox at each corner (4
total) is approximately four feet tall with the bottom of the separated OFA
windbox located 6 feet above the top of the close coupled OFA compartment.
The PM firing system at Lawrence #5 was specifically designed for a retrofit
application. Changes to existing structural steel members were minimized and
the only pressure part changes required were in the furnace waterwalls. A new
unit application of the PM firing system would normally increase the vertical
centerline spacing between successive fuel elevations to augment the NO ...
reduction capabilities of the system. This would require a 25-50% increase
in windbox height and thus associated changes in structural members and either
an increase in furnace overall furnace height and/or a rearrangement of steam
side surfaces. For the Lawrence #5 retrofit the fuel elevation centerline
spacing did not change. The extra windbox height was used to accommodate the
2-88
-------
close coupled OFA which was designed for 10%, or 2/5 of the total overfire air.
Retrofits on other units, using this more compact design, may or may not require
a windbox extension depending on whether overfire air preexisted, amount of
overfire air in the retrofit and existing structural configuration.
New larger fan blades for the pulverizer exhausters were installed both to
upgrade the pulverizers (boiler was load limited due to coal quality degradation
from design) and to accommodate the increased pressure drop associated with the
PM firing system coal pipe elbow separators.
Installation of the PM firing system commenced five weeks prior to the actual
outage with the installation of temporary support steel and rigging of hoist
lines ' . During the outage, a 20 ft. x 20 ft. hole was cut in the right
(north) side of the boiler waterwall and a supported rail system was set up
through the hole. The existing windboxes were cut free, lowered to the rails
and removed from the boiler. The new shop fabricated windboxes were raised to
the rails, slid into the boiler and raised into place. The new windboxes, with
their associated tube corners, are approximately 40 ft. tall and weigh more than
30 tons apiece. The installation was essentially completed during the allocated
8 week outage. (The actual unit start-up was delayed several weeks to complete
a planned turbine overhaul.) Other retrofit activity, which included a new
secondary air control system, the separated OFA windboxes and ductwork, the
partial coal pipe replacement and the fan blade replacement for the pulverizers
was also completed in parallel.
TEST SCOPE
Both pre-retrofit baseline tests and post-retrofit PM characterization tests
were conducted. The baseline tests were completed during March and April 1987.
The post-retrofit PM characterization tests were conducted during May 1988.
During these test periods parametric studies of the boiler performance were
evaluated. Key gaseous emissions data, taken by FERCO are summarized along with
boiler performance results. Details of the emission test results are presented
elsewhere (1). The boiler performance parameters assessed included:
1. Boiler Thermal and Combustion Efficiency
2. Furnace Outlet Temperature and Distribution
3. Furnace and Convective Steam Sectional Heat Absorptions
4. Steam Temperature Control and Operating Flexibility
5. Air Heater Performance
6. Air and Fuel Flow Distribution
7. Firing System Operation (stability, turndown, etc.)
Tests were conducted to evaluate the effect of the following key operating
parameters on the above-mentioned performance parameters:
1. Amount of Overfire Air
2. Overfire Air Location: Separate, Close Coupled, and Combined
3. Fuel Nozzle and Overfire Air Tilt
4. Excess Air Variation
5. Furnace Waterwall Conditions
6. Load
Comprehensive boiler instrumentation was required in order to fully characterize
boiler performance. The instrumentation package was mainly comprised of the
following:
2-89
-------
Feedwater flow orifice
Water and steam temperature thermocouples located at the high
pressure heater, economizer, superheater sections, and
reheat sections (25 total)
Suction pyrometers (furnace gas temperature)
Water/steam pressure transducers
Waterwall chordal thermocouples (36 total)
Superheater/reheater element thermocouples (68 total)
Flue gas analyzers ^^ ) '^y FERCO)
Fuel and ash sampling systems u>y FERCO)
Primary air, secondary air and fuel distribution instrumentation
C-E Automated Data Acquisition and Reduction System (ADARS)
The computerized ADARS system allowed immediate evaluation of major boiler
sectional heat absorption circuits such as the economizer, waterwalls, low
temperature horizontal superheater, superheater pendants, and finishing
superheater plus the reheat radiant wall and finishing reheat section. The
steam cycle flow chart for Lawrence No. 5 is shown in Figure 3.
Unit efficiency was calculated using C-E's standard boiler efficiency software.
The heat absorbtion performance and analysis was completed with the aid of C-E's
proprietary Reheat Boiler Performance Program (RBPP). This program develops
heat balances about the total boiler envelope and about individual heat transfer
elements. For this test program the RBPP was run in reverse. That is, acquired
field data was input to the program. Using this data and running the program in
reverse allowed the furnace outlet temperature to be calculated and sectional
heat absorbtion and thermal performance to be evaluated.
The RBPP program is structured in a modular fashion performing the calculations
in a predetermined sequence. The calculations begin with the boiler efficiency
which is dependent on the fuel analysis, and the general data module which is
dependent on the turbine heat balance. The calculations continue as the flue
gas flows through the boiler: furnace performance is calculated followed by the
convection pass, followed by the air heater. The program's calculation routine
is divided into nine (9) modules which are performed in the sequence listed
below:
1. Efficiency
2. General Data
3. Mill
4. Net Heat Input
5. Furnace Outlet Temperature
6. Upper Furnace
7. Interface Between Upper Furnace and Steam Generator
8. Steam Generator
9. Air Heater
UNIT CONDITION
Analysis of data from the baseline series of tests showed that, except for a few
minor operational problems, the "as found" boiler performance at Lawrence #5 is
typical of tangentially fired utility boiler's of that vintage (1967). The
operational problems noted included pluggage in a superheater desuperheat spray
valve, an air heater gas side outlet temperature imbalance of approximately
100°F at low (200 MWe and below) unit loads, a left to right flue gas 0^
imbalance on the order of 1.5% O2, and a low fuel distribution at one
2-90
-------
corner/elevation. These operational issues were taken into account, as best as
possible, throughout testing and during data analysis.
An extensive series of baseline tests were conducted. However, since PM
performance is best evaluated on a relative basis by comparing post-retrofit
results with the baseline, baseline results are discussed below.
TEST CONDITIONS
The following sections review data obtained from parametric testing of the PM
firing system; data from the baseline parametric test series is included for
comparison.
Parametric tests were carried out at loads ranging between 120 MWe and 330 MWe.
The minimum test load was established by the inability to measure turbine steam
bypass flow at lower loads.
The maximum test boiler load was set by pulverizer and I.D. fan operating
limits.
Excess combustion air was parametrically evaluated over a range between 10% (2%
O2) and 27% (4.5% O2). Additionally, fuel nozzle tilt variations were tested
over their total cycling range (± 25°) and overfire air vertical nozzle tilts
therefore were tested between -5° and +20° (from horizontal). Since the PM
firing system was designed for operation with both close coupled and separated
OFA, OFA location (close coupled and/or separated) was evaluated. Total OFA
quantity was also varied from minimum (4% of total combustion air for nozzle
cooling only) to greater than 30%.
Additional tests were run to characterize how waterwall cleanliness impacts on
boiler performance with the PM system; tests were conducted immediately after
waterwall sootblower operation ("clean") and a subsequent test was conducted
after ten hours of steady state operation without the sootblowers in service
(i.e. representative of "dirty" wall conditions).
EFFECTS ON BOILER THERMAL AND COMBUSTION EFFICIENCY
For deepest NO reduction, boiler efficiency with the PM firing system was
observed to be 0.2 - 0.3% lower than boiler efficiency with the baseline firing
system. The reason for the efficiency change is that a higher than baseline
operating excess air level (approximately 0.5 - 1.0% 0^ higher) was required for
PM operation at Unit #5. The higher excess air level increases dry gas loss
and, as a result, decreases boiler efficiency.
The observation was made throughout the PM tests that in some instances, both CO
emissions and smoke emissions increased when the PM system was operated at
equivalent to baseline 0„ levels. These observations were, of course, made when
comparing both systems at low 0^ levels.
The reason for this increase in 0£ with the PM system is postulated to be a
slight increase in the probability and amplitude of 0^ imbalances at the
economizer outlet. Both plant 0_ instrumentation and multiple-point 0- sampling
during the parametric trials confirmed the existence and magnitude of the
imbalances. Under conditions where the imbalance was significant, local 0.
levels at the economizer outlet would fall below 2%, resulting in increased CO
emissions. 0^ imbalances observed on other tangentially-fired boilers are
2-91
-------
typically on the order of 1-1.5%. Further information on the imbalance is
contained in reference 1.
It should be noted that carbon conversion efficiency, as determined via
particulate sampling and subsequent chemical analysis, remained essentially
constant between the baseline trials and the PM trials. Typically, less than
1.0% carbon-in-fly ash was observed throughout the entire test program. This
result is attributed to the fact that the Lawrence #5 coal was a high reactivity
Western bituminous with an "as received" volatile matter content of 36-40%. A
highly reactive coal is less sensitive to changes in combustion environment
(time/temperature history, firing system configuration, etc.) than a less
reactive (lower volatile matter) coal. It would be desirable to test less
reactive coal types at Lawrence #5 to evaluate the generic applicability of this
result.
EFFECT ON FURNACE OUTLET TEMPERATURE AND DISTRIBUTION
Calculated furnace outlet gas temperature (FOT) for normal recommended PM system
operation (combined close-coupled/separated OFA) was lower on average (40-60°F)
as compared with the baseline firing system FOT. Greater FOT changes (from
baseline) were noted when the PM system was operated with separated OFA only
(55°-90',F lower FOT on average) or when the PM system was operated with close
coupled OFA only (100° - 140°F reduction in FOT).
It should be noted that good correlation was established between measured
(suction pyrometry) FOT's and FOT value's calculated using C-E's Reheat Boiler
(RHBP) computer code. Since the RHBP generates consistently more accurate FOT
information, calculated FOT values were used exclusively in making comparisons
between pre- and post- PM retrofit FOT's.
Calculated FOT's for a number of test cases are presented in Figure's 4, 5A, 5B,
and 5C. Figure 4 presents calculated furnace outlet gas temperatures in the
baseline case, vs. furnace NHI, or net heat input. NHI, in boiler practice
(which roughly correlates with load) describes the heat available to the furnace
from the higher heating value of the fuel corrected by subtracting radiation
losses, unburned combustible, latent heat of water in the fuel, and adding
sensible heat in the air for combustion. PM, under normal operation (with
separated and close coupled OFA) has an indicated FOT reduction of 40°-60°F on
average from baseline, as shown in Figure 5A. PM, with separated OFA only
(Figure 5B) shows a 60°-90°F reduction and finally, PM with close coupled OFA
only (Figure 5C) shows a 100°-140°F reduction.
The above result is consistent with firing system tilt observations made during
testing. With the tilts automatically providing steam temperature control, the
baseline firing system fuel nozzle tilts ranged between -10° to -15°F from
horizontal, while for the PM firing system the fuel nozzle tilts automatically
ranged between -5° to -10°, which, all other factors considered equal, would
equate to the PM tilt system compensating for lower average FOT's compared to
baseline.
Figure 6, which shows typical predicted waterwall heat absorbtion and furnace
outlet temperature versus windbox stoichiometry, provides a basis for
understanding a reduction in FOT with PM. Windbox stoichiometry in Figure 6, is
defined as total furnace stoichiometry with excess air minus overfire air. (For
example 25% OFA with 15% excess air yields 1.15 - (0.25) (1.15) - 0.86 windbox
stoichiometry.) As the windbox stoichiometry decreases towards 1.0, there is
less gas dilution which yields higher flame temperatures and increased waterwall
2-92
-------
heat absorbtion subsequently lowering FOT. As the windbox stoichiometry
decreases below 1.0, the combustion process is delayed which lowers flame
temperatures and combustion occurs higher in the furnace which yields decreased
waterwall absorbtion and an increase in FOT.
The baseline windbox stoichiometry was approximately 1.15; PM with combined
overfire air windbox stoichiometry was approximately 0.85; and PM windbox
stoichiometry with either close coupled or separated overfire air alone was
approximately 0.95.
Based on the above analysis, operations which yield a windbox stoichiometry
close to 1.0 should result in the lowest FOT and highest waterwall absorbtion.
This is borne out in practice as PM operation with either close-coupled or
separated OFA (stoichiometry -0.95) had relatively large FOT reductions
(60°-145°F) on average as compared with PM with combined OFA (stoichiometry
-0.85, 40°- 60°F reductions on average).
It should be noted that, since the FOT calculations are site specific to
Lawrence Unit #5, one should not assume that the absolute values of the above
cited FOT's will be generally applicable to the T-fired boiler population at
large. The above results, however, do reinforce the need to review retrofit PM
system impact on FOT on a case-by-case basis.
In addition to the above, test data was obtained to quantify point-to-point
furnace outlet gas temperature distributions under pre- and post PM retrofit
conditions. Figure's 7A and 7B present typical examples of data from both the
baseline and PM firing system parametric tests. The results show generally
uniform gas temperature profiles with a peak-to-peak differential typically less
than 200°F left-to-right and/or front-to-rear. The gas temperature profiles
are slightly more uniform in the baseline case than the PM case, however, the
magnitude of the change noted with PM operation was insignificant and will not
have any negative impact on normal furnace operation.
BOILER SECTIONAL HEAT ABSORBTIONS
Figure 8 presents typical waterwall and economizer sectional heat absorbtions
for the baseline tests and several PM firing system configurations including: 1)
PM without OFA; 2) PM with close coupled OFA only; 3) PM with separated OFA
only; and 4) PM normal operation. With the as-found baseline configuration the
furnace waterwalls absorbed 39% of the total unit heat absorbtion. The
waterwall absorbtion (radiative) increased with the PM firing system retrofit
with a maximum waterwall absorbtion of 44% observed with both the close coupled
OFA only and separated OFA only PM configurations. With the normal PM
configuration (combined OFA) the waterwall absorbtion was more in line with the
baseline level at 40% (a 1% increase over baseline). These actual results are
consistent with changes in FOT presented in Figure 5A, 5B, and 5C and
explanation given earlier (Fig. 4). The changes in economizer (convection)
absorbtions, as shown in Figure 8, are inversely proportional to the changes in
waterwall absorbtion. Thus, if more heat is absorbed in the waterwalls, less
heat is available for the economizer.
Other steam sections were also instrumented and changes were monitored and shown
in Figure 8. These other sections include the LT superheat, SH division panel,
SH platen, HT SH pendant, RH radiant, and HT RH pendant sections.
2-93
-------
It should be noted that, in general, for both baseline and PM, increased excess
air resulted in decreased waterwall absorbtion as the furnace gas temperatures
were decreased by dilution as would be expected. Also, increased furnace
waterwall slagging resulted in decreased furnace absorbtion, again as expected.
STEAM TEMPERATURE CONTROL
Full load steam temperature control has also not changed except for the nominal
fuel nozzle tilt position previously discussed. Waterwall cleanliness and rate
of slag accumulation remained essentially unchanged. The variation in FOT due
to slag accumulation was approximately 100°F (equivalent to 3% change in
waterwall heat absorbtion) for both firing systems. This 100°F change in FOT
due to slag variations is not unusual; 300-400°F changes have been observed on
units with thick slag accumulations.
EFFECTS ON LOAD, START-UP, ETC.
Flame stability, boiler start-up and shut down procedures, load ramp rate and
load following operation after PM retrofit were identical to pre-retrofit
baseline operation.
Minimum achievable boiler load was unaffected by the installation of the PM
system; the unit is capable of operating at 15%-20% of the design maximum
continuous rating (MCR) of 400 MWe.
AIR AND FUEL FLOW DISTRIBUTION
Figures 9A & B present the primary air velocity and fuel massflow distribution
for the PM firing system upstream of the elbow separator. Both the fuel and air
corner-to-corner distributions on all five fuel elevations are within normal
design tolerances (± 5% tolerance for each corner for both air and fuel splits).
The baseline tests provided similar results with only one elevation/one corner
providing a fuel distribution out of specification. This flow imbalance was
corrected during the PM installation outage. The primary air flow was
approximately 12% higher with the PM system due to the pulverizer exhauster fan
blade retrofit.
The PM elbow separator performance is shown on Figures 10A & B. These fuel and
air splits shown are as per design with an approximately even air velocity split
and a fuel mass flow split of 80/20 for the C0NC and WEAK nozzles, respectively.
Additionally, it should be noted that the coarser coal particles favored the
C0NC nozzle (53% minus 200 mesh and 1.5% plus 50 mesh) rather than the WEAK
nozzle (94% minus 200 mesh and 0% plus 50 mesh); this result is not unexpected
with the elbow separator design.
The pressure loss associated with the installation of the PM elbow separator was
determined to be less than 1.0 inch W.G.
Testing indicated that operating with balanced dampers on the windbox fuel and
air compartments and at a slightly higher than baseline windbox pressure
(approximately 1.0 inch W.G. higher) yielded an improvement in the occasional
backpass 0^ imbalance.
2-94
-------
TYPICAL EXCESS AIR LEVEL
With the PM firing system the typical 0^ values range between 3.0-3.5%,
depending on OFA quantity. This compares with pre-retrofit baseline typical 0^
values of 2.5 - 2.7%. This increase in excess oxygen is due to an increased
propensity for an 0^ imbalance to occur with PM. This is especially true for
operations yielding the highest NO reductions (i.e. with significant quantities
of OFA, particularly separated OFA, high tilt differential between the fuel and
OFA nozzles.
NO EMISSIONS
x
Nominal baseline results were approximately 350-400 ppm (all NO values
corrected to 3% 0^). Typical PM firing system NO values were approximately 180
- 225 ppm @ 3% 0„ and 20% total OFA. Typical PM $iring system NO values were
approximately 150-180 ppm @ 3% 0^ and 25% total OFA. Approximate¥y 70% of this
reduction can be attributed, based on comparative tests conducted, to overfire
air while approximately 30% was due to the PM windbox modification.
A detailed summary of NO emissions from this program is contained in
reference 1.
Combustion Engineering is currently working with EPRI, KP&L, and FERCO in order
to establish recommendations for long term boiler operation. These
recommendations will be integrated into the six month duration emission's
monitoring period planned for the unit in 1989.
REPORT CONCLUSIONS
The following summarizes key findings reviewed in this report:
• A low NO PM firing system appears to be a viable retrofit option for many
existing T-fired boilers. An economic/application study is planned by EPRI
to quantify this issue.
• The retrofit system achieved NO reductions on the order of 35 -
50% (from a baseline level of 3?0 - 400ppm) at full load under tuned
short-term test conditions.
• At maximum NO reduction operation operating excess air level's increased
as a result ol the PM retrofit (-0.5. to 1.0% 0„ increase); overall boiler
efficiency was reduced slightly (0.2% - 0.3%) after PM retrofit as a result
of the additional required excess air when operated at very low NO^ levels.
• Waterwall heat absorbtion increased slightly (less than 1%) after PM
retrofit when the PM system was operated as recommended (using combined
close-coupled and separated OFA); a change in fuel nozzle vertical tilt
(+5° towards horizontal) accommodated this change.
• Furnace outlet temperature (FOT) after PM retrofit under recommended
operating conditions was 40° - 60°F lower on average than baseline
pre-retrofit FOT.
• Waterwall cleanliness and rate of slag accumulation was unchanged as a
result of the PM installation.
2-95
-------
• Carbon in fly ash levels were less than 1%, both pre and post PM retrofit,
for the reactive coal tested.
• The PM system had no adverse affects on boiler start-up, minimum load, or
flame stability.
ACKNOWLEDGEMENTS
First and foremost we would like to acknowledge Mr. P. Tempero, Plant Manager,
and the operating and engineering staff at Kansas Power & Light's Lawrence
Energy Center. Their commitment to and assistance throughout the program was a
tremendous benefit.
We would also like to acknowledge the Fossil Energy Research Co. and Energy
Systems Associates staff who put in many long days during the testing. This
acknowledgement also applies to the Combustion Engineering Technical Services
staff involved with the program.
We would also like to acknowledge the many C-E and KP&L people that were
involved in the system design and retrofit.
This project was funded, in part, by the Kansas Electric Utility Research
Program.
REFERENCES
1. Thompson, R., et al, NO Emissions Reduction for a 350 MW Low NO PM
Burner Retrofit, Proceedings: 1989 EPRI/EPA Joint Symposium on Stationary
Combustion NO Control, San Francisco, CA.
x
2. McCartney, M.S., et al, 1987 Update on NO Emissions Control Technologies
at Combustion Engineering, Proceedings: 1$87 EPRI/EPA Joint Symposium on
Stationary Combustion N0x Control, New Orleans, LA.
3. Kokkinos, A., et al, Part I: Feasibility Study of a Low NO Retrofitable
Firing System with U.S. Coals, Proceedings: 1982 EPRI/EPA Joint Symposium
on Stationary Combustion N0^ Control, Dallas, TX.
4. Tempero, P., Installation and Operation of Low NO PM Burners, Presented at
the Fifty-ninth Annual Engineering Conference of Missouri Valley Electric
Association, Kansas City, MO, March 1988.
5. Kuether, R.C., et al, Cut NO with New Burner System, 1988 Electric Utility
Planbook.
2-96
-------
TABLE 1
KANSAS POWER & LIGHT COMPANY
LAWRENCE UNIT NO. 5
GENERATOR RATING
MAXIMUM STEAM FLOW
SUPERHEAT STEAM
TEMPERATURE/PRESSURE
FURNACE WIDTH
FURNACE DEPTH
FIRING SYSTEM
PULVERIZERS
400 MWb
2,805,000 Ib/hr
1005°F/2620 psig
50' - 8"
40' 2-1/8"
TILTING TANGENTIAL
FIVE 743RS (WITH EXHAUSTERS)
FUEL ANALYSIS:
(As Recieved!
MOISTURE
105
VOLATILE MATTER
38.0 %
FIXED CARBON
42,9 %
ASH
8.6 %
H
4.50%
C
61.91%
S
0,47%
N
1.27%
0
12.75%
ASH
8,60%
MOISTURE
10.50%
HHV
10,884 Btu/lb
2-97
-------
Pendant Convection RH —
Pendant Convection SH
It
— Platen-Type Superheater
Steam Drum
Downcomers
Panel-Type
Superheater
Figure 1: Side Elavation of Kansas Power & Light Gas Services
Lawrence Energy Center Unit No. 5 as Originally Built
2-98
-------
6'
Firing Zone
ibs^i
u 1 1 u
II—aaa—II
11—f ^—11
AIR
COAL
GAS
AIR/OIL
GAS
COAL
GAS
AIR/OIL
GAS
COAL
GAS
AIR/OIL
GAS
COAL
GAS
AIR/OIL
GAS
COAL
AIR
Windbox
SIDE ELEVATION
UNMODIFIED
WINDBOX
2'
a
B
a
a
SEPARATED
OFA 0FA
CLOSE COUPLED
OFA
GAS
CONC.
COAL
WEAK
COAL
GAS
WEAK
COAL
CONC.
COAL
GAS
AIR
CONC.
COAL
WEAK
COAL
GAS
WEAK
COAL
CONC.
COAL
GAS
AIR
GAS
CONC.
COAL
WEAK
COAL
AIR
SIDE ELEVATION
PM FIRING SYSTEM
MODIFIED
WINDBOX
Firing Zone
PM Elbow
Separator
(Cone)
(Weak)
Burner
Front
Coal
Pipe
Figure 2: Schematics of Unmodified & PM Modified Windboxes at Kansas
Power & Light Lawrence Unit No. 5
2-99
-------
Figure 3: Water & Steam Cycle at Kansas Power & Light Lawrence No. 5
2-100
-------
a
E
,
o
«j
c
3000 -
2900-
2800-
2700-
2600 -
2500-
2400-
2300-
2200 -
2100 -
2000 -
1900 -
1800 -
1700 -
1600 -
1500
1000
—"?
1400
_l
150
—I 1 1 1 1 1—
1800 2200 2600 3000 3400 3800
Net Heat Input (10® Btu/Hr)
l_ I I
200 250 300
Approximate Megawatts
-------
A. PM Full Load - Normal Component Overfire Air Operation
3000 n
2700 -
P 2400 -
Baseline Mean (Replotted from Fig. 4) ,
3?
PM Normal Operation
1 1 1 1 1 1 1
2900 2940 2980 3020 3060 3100 3140 3180
Net Heat Input {10® Btu/Hr)
Full Load - PM With Separated Overfire Air Only
3000 t
E
P 2400 •
O
S 2100 -
1800-
Baseline Mean (Replotted from Fig. 4)
z\
"1 60 °F
PM With Sep. OFA Only
1 1 1 1 1 1 1—
2900 2940 2980 3020 3060 3100 3140 3180
Net Heat Input (10® Btu/Hr)
Full Load - PM Firing With Close Coupled Overfire Air Only
3000
Baseline Mean (Replotted from Fig. 4)
2700
1500
PM With CC OFA Only
I I i I I I I
2900 2940 2980 3020 3060 3100 3140 3180
Net Heat Input (10s Btu/Hr)
Figure 5: Comparison of Baseline Calculated Furnace Outlet Temperature
to the FOT with PM with Different Overfire Air Configurations
2-102
-------
3000
Note: Overall Stoichiometry Held Constant at 1.2
2900 -
WW Absorbtion
2800 "
2700 -
2600 -
F0T
- 950
2500
T
.9
T
1.0
1.1
T
1.2
Windbox Stoichiometry
Figure 6: Predicted Furnace Outlet Temp. & Absorbtions vs. Windbox Stoichiometry
2-103
-------
A. Baseline Traverse
Furnace Width - Ft.
B. PM Firing System Traverse
Furnace Width - Ft.
Figure 7: Typical Baseline and PM System Furnace Outlet Suction Pyrometer Traverses
2-104
-------
Figure 8: Sectional Heat Absorbtion Summary
2-105
-------
120
100
¦ SE CORNER 1
¦ SW CORNER 2
¦ NW CORNER 3
H NE CORNER 4
ELEVATION
(A) PRIMARY AIR VELOCITY DISTRIBUTION
La
3
a
5 c
La 3
2 -*
S r"
V2
<
5
2 3 4
ELEVATION
(B) COAL MASS FLOW
¦ SE CORNER 1
~ SW CORNER 2
¦ NW CORNER 3
H NE CORNER 4
Figure 9: Primary Air Velocity and Coal Mass Flow Distribution
PM Firing System Upstream of PM Separator
2-106
-------
120
SE CORNER SW CORNER NW CORNER NE CORNER
CORNER
(A) PRIMARY AIR DISTRIBUTION
1 00
z
SE CORNER SW CORNER NW CORNER NE CORNER
CORNER
(B) COAL MASS FLOW
Figure 10: Primary Air Velocity and Coal Mass Flow Distribution
PM Firing System Downstream of PM Separator
2-107
-------
(intentionally Blank)
2-108
-------
Reproduced from
best avsileble copy.
The work described in this paper was not funded by the U.5. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
ENEL'S ONGOING AND PLANNED NO
CONTROL ACTIVITIES X
A.Benanti - D.Bonolis ^^2)~ R,Tarli ^ ^ ^ ^A. Baldacc i
A. Piantanida - A.Zennaro
ABSTRACT
EIJEL is implementing activities to improve the environmental impact
of thermal power stations. Particularly as regards NO emissions,
ENEL intends to adopt mainly low-NO^ burners and to install abate-
ment systems of the SCR type.
The paper describes the program under way, the design criteria
adooted for NO reduction in the furnace, and the first results
X
obtained with demonstrative plants.
The steam generators that will undergo retrofitting are of both the
tangential and front-firing types, and burn fuel oil, gns and coal.
The capacities of the units involved are 320 KV/ and 660 MY;.
1. FOREWORD
In 1388 over 67% of the overall electricity demand (1S1 TUh) in Italy
was met with ENEL's power stations fired with fossil fuels (oil, coal
and natural gas). Coal-fired stations supplied 27 T\/h, thus covering
14% of the demand.
In the field of environment protection, from some years EMEL has
followed an advanced policy. The environment has become one of the
main items that influence ENEL's basic choices and plf-r.ning.
One of the main targets is the drastic reduction of the emissions
from both new and operating plants. In this connection, ENEL has to
comply with three international protocols on reduction of emissions
signed by the Italian Government: the Helsinki protocol of 1S85, the
ECE-ONU protocol and declaration signed in Sophia on 31 October 1988,
and the European Community Directive for large combustion plants.
The first protocol covers the overall sulfur dioxide emissions, which
' Ente IJazionale Energia Elettrica - Italy
F. Tosi F.T.C. - Italy
( 3)
v Ansaldo Componenti - Italy
2-109
Preceding page blank
-------
in 1993 shall have to decrease by 30;S as compared to the 1S80
emissions. For the share of its competence, ENEL v/ill comply with the
Helsinki undertakings three years in advance, that is, in 19S0 as
requested by a Decree of tine Italian Ministry for the Environment.
In Sophia two protocols were signed for reduction of the overall
national emission of nitric oxide; the first protocol was signed by
almost all the western and eastern countries, and establishes that
the present value of No emissions shall not be exceeded in the
X
future. The second protocol was signed only by twelve countries,
among which Italy, and establishes that by 1993 the overall national
emissions of NO shall decrease by 30% versus the emissions of any
year chosen from 1980 to 1986.
The CEE directive sets forth severe limits for emissions from each
new large combustion plant, whereas for existing plants it sets forth
limits to the overall emissions.
To meet the requirements concerning NO emissions, as a priority EJCZL
intends to make all the actions necessary to reduce emissions—
through modifications to the combustion system—from all planned and
operating plants, and then to adopt abatement systems of the high-
dust SCR type particularly in plants fired with low sulphur content
oil.
ENEL's Board of Directors is just now deciding on a first tranche of
modifications to be made on a group of plants witnin 1993.
Subsequently, all the plants will be modified making every effort to
minimize plant unavailability. This requirement is essential owing to
the difficulty ENEL will have in meeting the increase in power demand
in the next five years as a consequence of abrogration of the nuclear
programs and of the delays in the construction of new conventional
plants.
The paper illustrates the efforts made by ENEL jointly with the
national steam generator manufacturers (Ansaldo and Tosi) to reduce
nitric oxide emissions through modifications to the combustion
system.
2. ACTIVITIES UNDER WAY FOR NO REDUCTION THROUGH COMBUSTION
MODIFICATIONS X
Many activities have been planned and are under way to reduce NO
x
emissions from ENEL's boilers in operation or under construction
through modifications to the combustion system. The boilers being
modified are manly of two types, that is, the front-firing type of
Ansaldo make under the licence of Babcock & Wilcox, USA, and the
tangential firing type of F.Tosi make under the licence of Combustion
2-110
-------
Engineering.
Table 1 shows the configuration of the burners of the operating
boilers that will be modified. The modifications will cover an
overall installed capacity of 23815 MW, of which 29% from coal.
The units can be divided into nine different groups, depending on the
combustion system arrangement (five for oil-fired units and four for
coal-fired units). Furthermore, twelve new multifuel (coal, oil and
gas) units are under construction for an overall capacity of over
6,500 MW. For these units the most recent available technologies have
been adopted to reduce NO production in the furnace.
X
The program therefore provides for a series of activities that vary
depending on the burner arrangement and on the type of fuel, and
taking into account the possibility of simultaneous utilization of
different fuels.
For a better understanding, the activities mentioned above are
described separately for the two different types of boiler, (front-
rear firing and tangential firing.)
3. ACTIVITIES UNDER WAY ON FRONT-REAR FIRING BOILERS
The activities belong to two categories, namely:
(a) Retrofitting of Existing Boilers (Cell Burners)
Most of the existing boilers adopt cell burners. The cells can be
of different types:
- cells of two or three burners installed in coal designed units,
which can be fired v/ith coal or oil
- cells of three burners installed in oil units
- cells of three burners installed in oil designed units, which
can be fired with oil or gas.
The solutions selected have not yet been qualified in full scale
plants; therefore, for each solution it was deemed advisable to
nrocede in phases:
- selection of a reference boiler to be modified as prototype
- characterization and operation tests
- analysis of the results and definition of the final solution;
- modifications to the existing boilers.
(b) Modifications to New Plants
The solutions chosen belong to the best proven technology of the
Ansaldo/Babcock & Wilcox, though they have not yet been adopted in
Italy. The solutions did not require a campaign of experiments on
prototype.
2-111
-------
3.1 Low-NO Combustion Tests with Cell Burners Units
x
The oroRram of low-NO combustion tests with cell burners is divided
X
into the following three campaigns.
POWER UNIT FUEL CAPACITY CELL BURNER SCHEDULE
STATION (MWe) NUMBER NUMBER
PER UNIT PER CELL
La Casella 3 oil 320 6 3 1987-88
Rossano C. 4 gas 320 6 3 1988
oil 1S89
Vado Ligure 4 coal 320 15 2 1990
oil 1990
3.1.1 Low-NO combustion tests with oil units
x
In the years 1987-88 two types of tests were performed at La Casella
Station:
- tests in normal operation conditions (18 burners in service)
- tests with 12 burners, of which the six upper burners were used as
NO ports.
Furtner long-duration tests are still under way to carry out a more
exhaustive analysis on the combustion atmosphere and on the furnace-
corrosion .
Fig. 1 illustrates the lay-out of the gas taps and of the corrosion
panels installed on the boiler walls.
As concerns emissions, we note that:
- Operation with twelve burners and air immission from the upper
burners reduces the NO concentration as comnared to normal
X x
operation (18 burners in service, Fig. 2).
- Gas recirculation lowers NOx in both burner arrangements (18 and 12
burners in service, Fig. 3). We note that with gas recirculation up
to 25% the reduction is considerable; beyond this value the
reduction is quite low.
- Opening of the upper air registers has moderate influence on
emissions, because the air leakage through the closed registers is
considerable, that is, about 15% of the total combustion air versus
26-28% with registers open at 50%. With closed registers the lower
burners operate in under-stoichiornetric conditions (stoichiometric
ratio about 0.88%).
As concerns NO^ production in the furnace, the atmosphere near the
waterwalls was' analyzed by measuring the 0 , CO, H S and NO
2 2 x
2-112
-------
concentrations. The flue gas is extracted and conditioned for the
analysis through adequate sampling lines. Gas extraction is performed
through a penetration in the membrane, which allows tapping of the
gas along the walls.
A first evaluation of the NO values shows that:
- The NO profile along the different walls varies greatly depending
on theX type of combustion (18 or 12 burners) and on the gas
recirculation. In particular, we observe similar NO values at
different elevations only with 18 burners, whereas witn two-stage
combustion the NO concentration increases with elevation. This
means that at the considered elevations combustion is not yet
completed. At any rate, the overall NO concentrations are lower
than those measured with 18 burners. It should be noted that gas
recirculation besides reducing the NO flattens the concentration
x
profile.
- The trend of the 0 concentration allows an exact evaluation of the
reducing zones in the furnace. With 18 burners too, these zones are
localized on the side walls, where the ultrasonic tests carried out
on the tubes indicate the lower residual thicknesses. It is
interesting to note that the 0^ concentration on the front and rear
walls is very high, probably owing to the combustion air coming
from the burners that is diverted along the walls.
- These typical trends are common to the various boiler loads;
obviously, at lower loads N0_ is lower and 0^ is higher.
During the tests with 18 and 12 burners the metal temperatures and
the thermal fluxes were measured, the latter by means of chordal
thermocouples located where the maximum thermal flux is expected.
Thus, it was possible to observe that with 12 burners in operation
the flux increase due to the higher heat release in the combustion
zone is limited and that the operation temperatures of the metal are
still below the design limits.
3.1.2 Low NO combustion tests with gas-fired cell burners
The tests were performed on unit 4 of the Rossano Gal a'oro station in
November-December 1988. Three series of tests were performed:
- Baseline tests (18 burners in operation)
- Tests with 12 burners, orientable spuds, increased nozzle diameters
and the same arrangement as above, and uooer burners used as NO
. x
ports.
- Tests with 12 burners, orientable spuds, increased nozzle
diameters, different nozzle arrangement, and upper burners used as
NO oorts.
x A
During all the tests the vibrations of the buckstays and the
2-113
-------
pressure oscillations in the furnace were measured; the values
obtained were low and therefore did not pose any problem.
A synthesis of the results is shown in the diagrams, where all the
NO and CO concentrations are referred to dry gas with a 3% 0^
concentration.
We note that:
- operation with 12 burners and air immission through the upper
burners entails a considerable reduction in the emission as
compared with the 18 burners arrangement (see Fig. 4). At rated
load, with 0^ = 0.8% and with the gas recirculation dampers closed,
the NO emissions decrease from 690 mg/Nm (18 burners) to about
490 mg/Slm (12 burners with both gas nozzle arrangements).
- Opening of the gas recirculation dampers allows a further
considerable reduction in the emissions (Fig. 5).
- Opening of the upper air registers has little influence on the
emissions (Fig. 6).
This behaviour can be explained with the fact that even with closed
upper registers the air flowing into the furnace through the latter
is considerable; a first evaluation (based on the measurement of the
differential pressure between windbox and furnace) shows that about
20% of the overall combustion air flows through the "closed"
registers. Therefore, the lower burners operate in severe
understoichiometric conditions (stoichiometric ratio 0.8-0.85% with
02 = 0.8%)
- The influence of the excess air on the NO and CO emissions is
_. „ x
shown m rig. 7.
- Orientation of the gas spuds (only standard nozzle arrangements
were tested) influence NO emission significantly.
- The thermal fluxes and i^he metal temperatures measured with the
12-burner arrangement confirm the results of the tests with the
oil-fired cell burners described above. The two-stage combustion
carried out with 12 burners does not jeopardize structural
stability of the boiler pressure parts.
3.1.3 Low NO combustion tests with coal
x
In Italy, forteen 320-MV boilers of this type are in operation; the
boilers are fired with coal and oil and are equipped with two or
three register cell burners. The particular configuration of this
system, which is characterized by a very reduced vertical pitch and
the small size of the furnace, do not allow replacement of the
existing burners with low-NO burners or opening of ports for the
post combustion air.
On the basis of the technology developed by Babcock & Wilcox in
2-114
-------
cooperation with EPRI and experimented on reduced scale, only the two
lower burners of the three register cells and the lower burner of the
two register cells are operated, whereas the upper burner is used for
immission of the post-combustion air (Ref. 1). For a full-scale
demonstration of this technology, use will be made of unit 4 of Varto
Ligure station (320 MW).
The once-through UP boiler was designed by Babcock & Wilcox and built
by Ansaldo, and is equipped with 30 burners arranged in 15 two-
register cells.
The design is already completed; therefore, the modifications will be
carried out during the scheduled outage in autumn '89, and :ae tests
will be performed in 1S90.
The test campaign aims mainly at evaluating the following aspects of
the new combustion system before modifying all the remaining boilers:
- effectiveness of the new system in significantly reducing t,'0
emissions without adverse effects on boiler operation (in botn coal
and oil firing).
- Study on short-term corrosion in order to assess the effect of the
reducing atmosphere in the combustion zone in terms of chemical
attack on the waterwall tubes. Therefore, gas samples will be drawn
systematically to analyze the content of potentially corrosive
components, and the initial and final thickness will be Measured of
calibrated and instrumented samples. The whole test campaign should
last 6-8 months.
As regards the three-register cells their performance will be
assessed with a 35-MY/th scaled down .model installed in a test rig.
Design of the modified solutions provides for installation of the
equipment for natural gas combustion in order to develop a multi-fuel
combustion system, following ENEL's policy of maximum fuel
diversification.
3.1.4 Preliminary conclusions
On the basis of the results of the oil and gas tests, we can make the
following observations and outline the actions aiming at attaining
the final plant arrangement for continuous low-NO operation.
- Staged combustion does not jeopardize mechanical stability of the
waterwall tubes; therefore, modifications are not required for the
pressure parts of existing boilers.
- The upper burners will be replaced with adequate air registers that
operate as after-air ports, to obtain improved air inlet for a more
efficient combustion and a better control of the stoichiometric
conditions in the burner zone. The latter aspect will become even
2-115
-------
more important should the corrosion tests indicate a high corrosion
rate. which can be reduced by controlling the stoichiometric
conditions in the furnace.
3.2 New Boilers Under Construction
The boilers discussed here have the following original design
characteristics:
(a) Four UP boilers, of Babcock & Wilcox design, for Erindisi Sud
S tation:
- capacity: 660 MW
- fuel: coal and oil
- 56 Babcock & Wilcox burners of the dual-register type.
(b) Two natural circulation boilers for the Tevazzano Station:
- c ap ac i ty: 320 MW
- fuels: coal and oil
- 30 Babcock & Wilcox burners of the dual-register type.
During construction of the boilers it was decided to snake the
necessary modifications to obtain maximum reduction of the tfO
X
emissions and at the same time to have tne possibi.l i l;y of tiring
as well.
The first 660-MW unit, which had already undergone tne hyriraul i c
test, was equipped with NO oorts and with the svstern for mixing
» * i. i. X*
recirculated flue gas and combustion air. The original burners were
modified by installing in the throat a multi-nozzle ring for natural
gas firing.
The remaining 660- and 320-KW units we re further modified oy
replacing the original dual register burners with Babcock £• V/t Icox
low-NO burners of the XCL type.
The X&j burner was modified to add trie components required for
natural gas firing; a full-scale prototype was experimented on a test
rig of Riley Stoker in the USA in early 1988.
The first startup of unit 1 is expected in April 1S90 for the 55C-MW
boilers, and in Autumn 1990 for the 320-MW boilers.
ACTIVITIES UNDER WAY ON TANGENTIAL-FIRED BC7LER5
Mere too, the activities can be grouped into the two following
categories:
(a) Retrofitting of Existing Boilers
To attain the low-NO emission goal , EITKL has prepared a program
that utilizes an advanced system for low NO combustion that is
being developed by Combustion Engineering. Tine program covers the
2-116
-------
Reproduced from
best available copy.
following phases:
- Laboratory experiments of an advanced combustion system.
- Analysis of the results and application of the advancer; system
to a full-scale boiler.
- Tests for characterization and operation of the boiler before
and after the modifications.
- Analysis of the results and selection of the final solutions
for application to the other boilers.
(b) Modifications to New Plants
Because of the advanced state of design of Flume Santo, it was
essential to proceed with tangential system v/i to. proven
experience. The burners were purchased from C.2. and installed in
the Fiume Santo unit.
The schedule for Gioia Tauro permits a review of the results from
the Fusina demonstration before a final design is chossn.
4.1 Combustion Tests
The test Program for the advanced
out as follows:
Boiler Fuel
C.E. Boiler Coal
Simulation Facility
Fusina Coal
4.1.1 Fusina project
4.1.1.1 Low-NO Combustion Svstem
x
combustion system will be carried
LTV; (th) Schedule
15 MV.'(e) Nov.-Dec. 19s8
150 I-"W 1 S8
The Fusina Station is e'quipped with four TOSI-CE tangential-firing
boilers of the multi-fuel type but firing mainly coal. Of the
boilers, two are rated 320-MW and, two 160 MW. One of the 160-MV
boilers will be equipped with a low-NO combustion svstem consisting
of: K
- Overfire Air Ports (OFA)
- Concentric Firing System (CFS)
- Vertical Grouping of Coal Nozzles
The system delays mixing of the fuel and of the combustion air thug
reducing NO emissions to values that should range around 40C mg/r,'
-------
that had to be studied and solved through laboratory research before
the final design. In particular, the following items were
investigated:
- Reduction obtained in NO emissions with different nozzle
x
groupings.
- Optimization of the air flow through the intermediate auxiliary nir
compartments to attain the best operation condition.
- Influence on the IJO of the amount of air introduced through the
conventional OFA por^s and through the separate OFA ports (HSHP).
- Effects on the un'ournt of the coal nozzle grouping and of the riSPi?
OFA installation.
- Variation of the required pressure drop for the HSHP OFA as a
function of the distance from the furnace outlet.
- Optimization of the burner nozzles.
The laboratory tests are being performed on the 5C-r;st1 lion Etu/h
boiler simulation facility (3SF) at the KDL laboratory of Combustion
Engineering in Windsor, USA. Several Different configurations of cosl
nozzle groupings and of the HSHP OFA are being compared. The first
coal used was American coal having similar characteristics to the
coal now being burned at the Fusina station in terms of fixed
carbon-to-nitrogen ratio, nitrogen in the fuel, anc -=s h fusion
temperature.
4.1.1.2 Future Programs for the Fusina Station
The obtained results will be applied to the Fusina 160-.'!!/ boiler that
will operate as prototvoe for the new low-NG combustion svstem. A
X "
test campaign for coal firing was carried out to have 3 finger-
printing of tne boiler operation.
In the corning months these tests will be performed with oil ana gas
firing and with different coals. Then, the modifications will be
carried out, of which the most importat are:
(a) Installation of separate OFA ports on the four corners of the
boiler, designed for 25-30% of the total combustion air. To
optimize combustion it will be possible to orient the air nozzle';
either vertically or horizontally.
(0) Re-positioning of the coal nozzles by adequate grouping,
reproportioning of the auxiliary air compartments, installation
of conventional OFA ports in the upper part of the winribox.
Through the OFA ports 10-20% of the total combustion air will be
introduced. The intermediate air nozzles can be oriented on the
horizontal plane in order to divert the combustion air towards
the waterwalls (CFS system).
2-118
-------
(c) Installation of a booster fan to give sufficient pressure to the
air flowing through the HSHP OFA ports.
4.1.2 Fiurne Santo Station (2 x 320 H\v~)
The two boilers for the new 320-MV units of the Fiurne Snnfo station
were in advanced construction when it was decided to .-nod if/ t.ie
combustion svstem in order to limit NC production. The choice f^]1
X
on the installation of PM coal burners of Japanese technology
supplied by C.E. (Ref. 2) Since the new windbox housing the
tangential burners was higher, the furnace height was increased by
2.5 rn. Other modifications, such as replacement: of the mi 1 1
exhausters with the pressurized primary air system and a fe.v minor
changes to the reheater surfaces were also included.
The boilers are being assembled and it is expected that +-he first
unit will begin operation in 1990.
4.1.3 Gioia Tauro station (4 x 660 MW)
The boilers of this station will be equipped with the system being
installed in Fusina demonstration unit.
4.1.4 Oil - and gas-fired boilers
ENEL intends to modify the combustion system of this type of boilers
already in operation, mainly by adopting the OFA and the flue-gas
recirculation techniques.
5. CONCLUSIONS
ENEL's efforts in the field of NO reduction entailed an extended
x
activity program carried out; jointly with tne ri^.t ior;r. L ste? .;n
generator manufacturers.
The different types of combustion systems and the fact that the
boilers are generally designee! for two or more *"ypes of fuel
sometimes prevent the utilization of proven technologies, mainly as
regards retrofitting where the solutions are conditioned co tne
Intrinsic characteristics of the boiler, to the unit unavailability
and to the costs required to perform the modifications. Extensive
programs were therefore implemented to demonstrate the validity of
the solutions hypothesized for the various types of steam generators.
The first studies and results have proven that the goals set by ENEL
can be attained with the programs under way.
2-119
-------
REFERENCES
1. Clark W.J., La Rue A.D., Liang A.D. and Eskinazy D. Large Scale
Testing and Development of the B&V7 Low N0_ Cell Burner.
Presented to Joint Symposium on Stationary Combustion HO
Control, New Orleans, 1987.
2. Y. Takahashi, T. Kumirnoto, K. TokudaT. Kawarnura, S. Xaneko.
Develooment of Pulverized Coal Fired Low MO PI I Burner Technical
¦—1 ....I. ii ¦ in in— .I i... ii i—..I ¦ i-i i. 'X ' "" — ¦- ¦
Review, October 1931.
2-120
-------
LA CASELLA - UNIT NO. 3
GAS TAPS AND CORROSION PANELS LOCATION
3790.7 12514.61 3790.7
3771
¦ I j 2514.6 j 3771.8
+1 -|l._
3790.7 I 2514.6 I 3790.7
4> +>
1.4 |2133.6j 3981.4
41 +
-¥ -f4 —
| 2000 | 6334
-|!> -f9 4-10
2799 | 2B14 j 2214 j 2231
3981
.4 j 2133.6j 3981.4
Rear
Wall
Left Side
Wall
Front
Wall
- FIG. 1 -
3771.8 I 2514.61 3771.8
+1 np
6334
* -4s
-45,+5 —
f, f , —
-45 _JJ0 _jjl
2231 | 2214 I 2814 I 2799
/ I
/ !
' I
n n / 1 1
^
Right Side
Wall
Gas taps
| Corrosion Panels
-------
LOW NOX OIL COMBUSTION TESTS
LA CASELLA POWER STATION-UNIT NO. 3 (320 MW)
1100
1000
900
800
700
m
| 600
oo
E
w 500
K
O
z
400
300
200
lOO
o
100 140 180 220 260 300 340
LOAD (MW)
- FIG. 2 -
LOW NOX OIL COMBUSTION TEST
LA CASELLA POWER STATION-UNIT NO. 3 (320 MW)
GAS R£CXRC. DAMPER OPENING (%)
- FIG. 3 -
2-122
~ 18 BRNS - NO GR
+ 18 BRNS - MAX GR
O 12 BRNS - NO GR - UPP.REG.CLOSED
A 12 BRNS - NO GR - UPP.REG.50% OPEN
X 12 BRNS - MAX GR
-------
800-
700 -
LOW NOX GAS COMBUSTION TESTS
ROSSANO POWER STATION-UNIT NO. H (320 MW)
~ 18 BURNERS
+12 BURNERS - MODIFIED SPUDS
C 12 BURNERS - STANDARD SPUDS
LOAD (MW)
- FIG. 4 -
LOW NOX GAS COMBUSTION TESTS
ROSSANO POWER STATION-UNIT NO. 4 (320 MW)
500 -
~ 18 BURNERS
+ 12 BURNERS - MODIFIED SPUDS
O 12 BURNERS - STANDARD SPUDS
T
10
30
T
50
GAS RECIRC. DAMPER OPENING (*)
- FIG. 5 -
2-123
-------
LOW NOX GAS COMBUSTION TESTS
ROSSANO POWER STATION-UNIT NO. H (320 MM)
800-
700-
600-
500-
9 GR = 0%
-GR = 20%
300-
-o- en = 50%
200-
100-
n J r
10 30 50
UPPER REGISTER OPENING <*)
- FIG. 6 -
LOW NOX GAS COMBUSTION TESTS
ROSSANO POWER STATION-UNIT NO.4 (320 MW)
800-
600-
cn 500-
s
Z
400-
200-
100-
~ NOx 18 BURNERS
+ CO 18 BURNERS
O NOx 12 BURNERS
A CO 12 BURNERS
0.4
0.6
0.8
02 (%)
- F16. 7 -
1.2
2-124
-------
BURNER CONFIGURATION OF ENEL's EXISTING BOILERS
(CAPACITY > 200 MWe)
OIL FIRED BOILERS
CH1VASS0
Unit
5 (*)
PIACENZA
Units
3,4
LA CASELLA
Units
1,2,3,4
0STIGL1A
Units
1,2,3,4
TURBIGO LEV.
Unit
1 (*)
Unit
2
Units
3,4
SERM1DE
Units
1,2,3,4
TAVAZZANO
Units
5,6
MONFALCONE
Units
3,4
PORTO TOLLE
Units
1,2,3,4
TORREV. SUD
Units
2,3,4
TORREV. NORD
Units
1,2,3,4
ROSSANO
Units
1,2,3,4
TERMINI 1.
Units
4,5
PR10L0 G.
Units
1,2
TOTAL CAPACITY (MU)
COAL FIRED BOILERS
LA SPEZ1A
Units
1,2,4
Unit
3
VADO L1GURE
Units
1,2,3
Unit
4
PIOMBINO
Units
1,2,3,4
FUSINA
Units
3,4
BR1ND1S1 NORD
Units
1,2
S. F1L1PP0
SULC1S
Unit
Unit
Units
Units 3,4
5
6
1,2,3 (•>
TOTAL CAPACITY (MU)
FRONT-REAR BURNERS
Two
register
cells
330
1,280
640
320
2,570
Three
register
cells
1,280
320
1,280
640
3,520
600
990
640
320
2,550
Parallel
Flow
1,280
640
640
2,640
5,200
(*) Units equipped with circular burners on the front wall (total
capacity 1 ,220 MU).
- Tab. 1 -
2-125
-------
-------
Session 3
COMBUSTION NOx DEVELOPMENT II
Chairman: R. Hall, EPA
3-i
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
PILOT EVALUATION OF REBURNING FOR CYCLONE BOILER NOx CONTROL
H. FARZAN and L RODGERS
The Babcock & Wilcox Company
Alliance, Ohio
G. MARINGO
The Babcock & Wilcox Company
Barberton, Ohio
A. KOKKINOS
Electric Power Research Institute
Palo Alto, California
J. PRATAPAS
Gas Research Institute
Chicago, Illinois
ABSTRACT
There are currently no commercially-demonstrated combustion modification techniques for cyclone
boilers which reduce NOx emissions. The emerging reburning technology offers cyclone boiler owners
a promising alternative to expensive flue gas cleanup techniques for NOx emission reduction.
Reburning involves the injection of a supplemental fuel (natural gas, oil, or coal) into the main furnace in
order to produce locally reducing stoichiometric conditions that convert NOx produced in the main
combustion zone to molecular nitrogen, thereby reducing overall NOx emissions. This paper presents
the encouraging pilot-scale results obtained using natural gas, fuel oil, and pulverized coal (PC) as
reburning fuels. At a reburning zone stoichiometry of 0.9,67% NOx reduction for gas and oil reburning
and 57% for coal reburning were achieved. Flue gas recirculation (FGR) was introduced at the
reburning burners, and improved mixing and NOx reduction. FGR was more effective with PC
reburning. Carbon monoxide (CO) emissions levels were low (less than 30 ppm) throughout the various
optimum test conditions. Although the carbon content of the fly ash increased, the overall combustion
efficiencies were insignificantly lower for all the reburning fuels. Furnace exit gas temperatures (FEGTs)
increased by less than 50°F during reburning operation. Capital costs for rebum technology on a
200-MW cyclone-fired unit can range from about $22/kW (gas) to $41/kW (coal). Ten-year levelized
incremental busbar costs were sensitive to assumed rebum fuel prices. Lower projections were 1.7
mills/kWh (coal), 2.3 mills/kWh (gas), and 3.3 mills/kWh (oil). Babcock & Wilcox (B&W) is presently
starting a full-scale 100-MW utility cyclone reburn demonstration project with Wisconsin Power & Light
per the U.S. Department of Energy's (DOE's) Clean Coal II program.
3-1
-------
INTRODUCTION
The Electric Power Research Institute (EPRI) and the Gas Research Institute (GRI) contracted with the
Babcock & Wilcox Company (B&W) to perform a pilot-scale evaluation of the reburning technology for
cyclone boiler NOx emissions control. These pilot tests were justified via a previous EPRI-sponsored
(Project RP-1402-30) engineering feasibility study of reburning for cyclone boilers performed by B&W.
The feasibility study revealed that the majority of cyclone-equipped boilers could successfully apply this
technology in order to reduce their NOx emission levels by approximately 50 - 70% (1). The major
criteria that substantiated this potential was that sufficient furnace residence time does exist within these
boilers in order to apply the technology. Thus, based upon this conclusion, the next level of
confirmation - pilot-scale evaluation - was justified. These pilot tests involve evaluating the potential of
natural gas, oil, and coal as the reburning fuel in reducing NOx emissions. This paper focuses on the
results obtained during the evaluation of these three reburning fuels.
There are presently 105 operating, cyclone-equipped utility boilers representing approximately 14% of
pre-NSPS coal-fired generating capacity (over 26,000 MW). However, these units contribute
approximately 21% of the NOx emitted since their inherent turbulent, high-temperature combustion
process is conducive to NOx formation. Although the majority of cyclone units are 20 - 30 years old,
utilities plan to operate many of these units for at least an additional 10 - 20 years. The potential of
future acid rain control and the midwest location of the majority of these units may target these boilers
for NOx emission control.
Cyclone-equipped boilers have a unique configuration that prevents application of standard low-NOx
burner technology - that is, that the combustion occurs within a water-cooled horizontal cylinder
attached to the outside of the furnace. Furthermore, other conventional NOx reduction techniques such
as two-stage combustion cannot be applied to the full extent due to associated cyclone operational
concerns (cyclone corrosion). The use of selected catalytic reduction (SCR) technology offers promise
of controlling NOx from these units, but at high capital and operating costs. Reburning is therefore a
promising alternative NOx reduction approach for cyclone-equipped units at more reasonable capital
and operating costs.
Reburning technology involves injection of a second fuel into the main furnace (above the cyclone
region) to produce a secondary combustion zone where a reducing atmosphere exists. These local
chemical reducing conditions convert NOx to molecular nitrogen, thus destroying a portion of the NOx
produced in the primary cyclone combustion zone. The reburning technology had never been applied
(bench-/pilot-/commercial-scale) to cyclone-equipped boilers. Although numerous reburning studies for
natural gas-/oil-/PC-fired boilers have been performed, it is difficult to extrapolate these results directly to
cyclone units due to inherent differences in cyclone boiler design and operation. The characteristics of
cyclone units that may affect NOx reduction, solids deposition, and efficiency of back-end equipment
when applying this technology include:
• Higher initial primary NOx levels (600 - 1400 ppm at 3% O2)
• Lower char/ash carryover to the main furnace due to the cyclone slagging capabilities
• Different convective pass and downstream particulate equipment design due to the
lower fly ash loadings
3-2
-------
The purpose of the pilot-scale tests was to substantiate/expand on the results obtained from the
engineering feasibility study and to assess the potential NOx emission capabilities of the process. This
information will aid in identifying the potential commercial application of the technology. The following
summarizes the specific objectives of the test program:
• Evaluate the major process parameters such as reburning/main combustion zone
stoichiometries, reburning fuel type and quantity, and reburning/burnout zone
residence times.
• Determine effects of reburning:
- The degree of slagging/fouling in the upper furnace and convection pass
- Combustion efficiency (based upon unburned combustibles and CO emissions)
- Corrosion potential
- Changes in furnace exit gas temperature
• Maximize NOx reduction while maintaining combustion conditions compatible with
design and operation of cyclone-equipped boilers.
BACKGROUND
The cyclone furnace consists of a cyclone burner connected to a horizontal water-cooled cylinder called
the cyclone barrel (Figure 1). This auxiliary furnace was an attractive alternative to pulverized coal (PC)
by providing the ability to burn low-grade coals, significantly reducing fly ash in the flue gas, requiring
less fuel preparation equipment, and allowing for a reduction in total boiler size. Crushed coal and air
are introduced through the cyclone burner into the cyclone barrel. The larger coal particles are thrust
out to the barrel walls where they are captured and burned in the molten slag layer which has formed,
while the finer particles burn in suspension. The mineral matter melts, exits the cyclone furnace from a
tap at the cyclone throat, and is dropped into a water-filled slag tank. The flue gases and remaining ash
leave the cyclone and enter the main furnace.
No commercially-demonstrated combustion modifications have significantly reduced NOx emissions
without adversely affecting cyclone operation. Past tests with combustion air staging achieved 15 - 30%
reductions (2)(3). Cyclone tube corrosion concerns due to the resulting reducing conditions were not
fully addressed because of the short duration of these tests. Further investigation of staging for cyclone
NOx control was halted due to utility corrosion concern. Additionally, since no mandatory Federal/State
NOx emission regulation was enforced, no alternative technologies were pursued.
The recent emergence of the reburning technology offers a promising alternative to conventional
combustion controls and SCR systems. During the early 1980s, Tokyo Electric Power Company
(Japan) and Babcock-Hitachi K.K. (BHK) jointly developed a technology known as in-furnace NOx
reduction (IFNR) to reduce NOx emissions from natural gas-, oil-, and coal-fired furnaces. B&W is the
exclusive licensee of BHK's IFNR process. The technology, a version of the process widely known in
the industry as reburning, is based upon extensive laboratory research and has been proven in
subsequent pilot- and full-scale operation in Japan. BHK has applied IFNR as a retrofit technology to
ten wall-fired utility boilers in Japan. The retrofit units are natural gas, oil, and coal fired units ranging
from 175 to 700 MW in size. BHK's pilot- and full-scale experience has demonstrated typical NOx
3-3
-------
emission levels shown in Table 1. Adapting this technology to the inherent characteristics of cyclone
boilers was reviewed/determined during the aforementioned B&W engineering feasibility study. Based
on the results of the study, it is estimated that a nominal 60% NOx emissions reduction could be
achieved from cyclone boilers with the use of reburning. Table 1 identifies these anticipated levels.
REBURNING PROCESS DESIGN CRITERIA
The reburning process employs multiple combustion zones in the furnace, as shown in Figure 2. The
main combustion zone is operated at a reduced stoichiometry and has the majority of the fuel input (70 -
85% heat input). The majority of investigations on natural gas-/oil-/coal-fired units have shown that the
main combustion zone of the furnace should be operated at a stoichiometry of less than 1.0. This
operating criteria is impractical for cyclone units due to the potential for highly corrosive conditions, since
most cyclones burn high-sulfur, high-iron content bituminous coals. To avoid this situation and its
potentially catastrophic consequences, the cyclone main combustion zone was determined to be
operated at a stoichiometry of no less than 1.1 (2% excess O2).
The balance of fuel is introduced above the main combustion zone (cyclones) in the reburning zone
through reburning burners. To protect the tubes around the reburning burners in the reburning zone
from fireside corrosion, some air is introduced through these burners. The burners are operated in a
similar fashion to a standard wall-fired burner. The furnace reburning zone is operated at
stoichiometrics in the range of 0.85 - 0.95 in order to achieve maximum NOx reduction based on
laboratory/actual boiler application results (4)(5).
The balance of the required combustion air - totaling 15 - 20% excess air at the economizer outlet - is
introduced through overfire air (OFA) ports. These ports are designed with adjustable air velocity
controls to enable optimization of mixing for complete fuel burnout prior to exiting the furnace.
Pilot- and field-scale studies by BHK and other researchers (4 - 7) have defined acceptable limits of
residence times in the reburning zone. A 50 - 60% reduction can be achieved at residence times
greater than 0.45 second. In order to complete combustion (based on B&W/BHK experience with
staged combustion), about 0.65 second residence time is required. Thus, a total of about 1.1 seconds
is required between the reburning ports and furnace exit.
PILOT REBURNING TEST PROGRAM
Technical Objectives
The technical objectives of this project are to demonstrate NOx reductions of nominally 50 - 60% while
maintaining acceptable cyclone/boiler operating conditions. Three reburning fuels were evaluated while
operating under various anticipated full-scale simulated reburning conditions. Table 2 summarizes the
various ranges of reburning criteria that were evaluated for NOx reduction capability. Main
cyclone/reburning burners fuel splits, reburning fuel type, furnace stoichiometries, and furnace
residence times were varied. Additional variables that were evaluated include mixing, corrosion
potential, fireside deposition, and combustion efficiency. The results will be utilized to confirm/expand
the Phase I engineering feasibility study.
3-4
-------
The major areas of technical uncertainty that were identified in the feasibility case studies and were
evaluated during the pilot tests for all reburning fuel types include:
• NOx reduction potentials of the reburning fuels when operating in a cyclone boiler
environment of high initial primary NOx levels and low char carryover to the main
furnace (high char carryover increases available, unconsumed oxygen in the
reburning zone)
• Optimization of process parameters for cyclone application
• Effects on increased solids deposition with coal reburning in the upper furnace and
convective section
• Corrosion throughout the furnace
• Unburned combustibles and FEGT changes
Research Facility
B&W's 6-million Btu/hr Small Boiler Simulator (SBS) was utilized to perform the pilot-scale cyclone
reburning tests (Figure 3). The SBS is fired by a single, scaled-down version of B&W's cyclone furnace.
Coarse pulverized coal (44% through 200 mesh), carried by primary air, enters tangentially into the
burner. (Pulverized coal had to be utilized in the SBS instead of crushed coal to obtain complete
combustion in this small cyclone.) Preheated combustion air at 700°F enters tangentially into the
cyclone furnace. The larger coal particles are captured and burn in the molten slag layer formed within
the cyclone furnace, while the finer particles burn in suspension. The mineral matter melts, exits the
cyclone furnace from the tap at the cyclone throat, and is dropped into a water-filled slag tank. Only 15 -
20% of the ash leaves the cyclone with the flue gases and enters the furnace.
The furnace is water-cooled and simulates the geometry of B&W's single-cyclone, front-wall fired
cyclone boilers. It consists of four separate water-cooled sections. The lower part of the furnace can be
converted to a PC-firing scheme. The SBS facility has been operating for a total of six years, with the
last three years being operated in the cyclone configuration. This cyclone facility has been proven to
simulate typical full-scale cyclone units via furnace/convective pass gas temperature profiles and
residence times, NOx levels, cyclone slagging potential, ash retention within the resulting slag,
unburned carbon, and fly ash particle size. A summary of these comparisons is shown in Table 3.
The inside surface of the furnace is insulated to yield a FEGT of 2265°F at the design heat input rate of
6-million Btu/hr. A water-cooled tube bank simulates the flue gas time/temperature history inherent in
full-scale cyclone convective passes. The tube bank consists of four separate sections for simulating a
secondary superheater, reheater, primary superheater, and economizer. Each section consists of a
water-cooled jacket and tubes to quench the flue gases. All four sections are connected to a common
atmospheric drum. This use of convective tubes to cool the gas, in conjunction with the cyclone
furnace, makes this a unique facility among pilot-scale combustors.
Two reburning burners were installed on the SBS furnace rear wall at an elevation above the cyclone
burner/barrel. The facility is capable of firing natural gas, oil, or coal at the reburning burner region. The
3-5
-------
multi-fuel rebuming burners are designed to accommodate the required velocities for furnace
penetration and also to allow for enough flexibility for varying mixing characteristics. Each burner
consists of essentially two zones with the outer zone housing a set of spin vanes while the inner zone
contains the rebuming fuel injector. Air and flue gas recirculation (FGR) can be introduced through the
outer zone. The natural gas/oil burners also have the capability to bias the air/FGR flow between the
inner/outer zones.
OFA ports are available on both the front and rear walls of the SBS at three elevations, with each
elevation containing two ports. Locating the OFA ports at different elevations assists in assessing the
effects of residence time on fuel burnout and NOx reduction.
The SBS furnace and convection pass sections are equipped with numerous observation ports at
different elevations to allow for complete evaluation of the process under investigation. In-furnace
probing is performed at the port locations in order to determine temperatures and gas/solids
composition.
Two air-cooled deposition probes are available in the convective section (simulating secondary
superheater and reheater tubes) in order to allow for fouling (deposition) studies to be performed.
These probes are equipped with thermocouples for measuring metal wall and inlet/outlet air
temperatures. The probe metal temperatures are maintained at typical boiler tube temperature in order
to assure meaningful ash deposition results. The effect of ash deposition on heat transfer is determined
by energy balance calculations for each probe. In addition, a simulated commercial sootblower is
available to determine the required sootblower pressure necessary to remove the deposits and restore
maximum heat flux potential.
PILOT-SCALE RESULTS
The pilot-scale cyclone project reported herein includes the baseline and rebuming tests. The baseline
tests are performed under normal cyclone operating conditions and identify the benchmark data to
which the subsequent rebuming test results are compared. The rebuming tests involved the
installation/operation of two rebuming burners and two OFA ports (at two different elevations). The
rebuming burners have the flexibility of varying the degree of mixing via changing reburn fuel injector
type, spin vane location/angle, and addition of FGR. Critical data collected for both the baseline and
rebuming phases include NOx, CO, O2, and unburned combustible levels. In addition, gas temperature
profiles throughout the furnace were measured. All the tests were performed while utilizing a
Pennsylvania Kittanning seam coal as the main cyclone fuel. The same Kittanning coal ground to 85%
<200 mesh was utilized during the coal rebuming tests. The coal, oil, and natural gas analyses are
shown in Table 4. The following paragraphs describe the results of these tests.
Baseline Tests
Figure 4 illustrates the NOx emission levels obtained during the baseline tests. Operating the cyclone at
6-million Btu/hr results in a baseline NOx level of 920 ppm at 3% excess 02. NOx emissions increased
by approximately 40 ppm per each percentage point increase in excess 02. Reducing furnace load to
4.3-million Btu/hr decreased the NOx emission level to 850 ppm at 3% excess 02. As excess 02
3-6
-------
changes, the slope of the resulting NOx curve is the same as that observed at full load. Firing natural
gas in the cyclone at 6-million Btu/hr resulted in NOx emissions of 455 ppm at 2% excess O2 (typical
operating excess O2). Reducing the oxygen to 1% resulted in the same NOx level as that observed at
2% 02 (455 ppm), but increasing the oxygen to 3% excess 02 reduced the NOx level to 392 ppm.
These resulting NOx levels can be explained via the cyclone exit gas temperatures and the various
mechanisms of NOx formation.
Cyclone exit temperatures were measured utilizing an optical pyrometer. At 6-million Btu/hr (coal firing),
the temperatures changed from 2950° to 2850°F at 2 to 4% 02, respectively. At 4.3-million Btu/hr, the
same trend was observed at 2 to 4% 02 (2800° to 2700°F). Natural gas firing at 6-million Btu/hr
showed temperatures of 2640° to 2570°F from 1 to 3% 02. The various trends of NOx emission levels
versus excess oxygen can be explained by the different mechanisms of NOx formation. During natural
gas firing, thermal NOx is the major mechanism of NOx formation. Thus, NOx levels decreased as the
excess oxygen increased since the cyclone exit temperature was also observed to decrease. During
coal firing, fuel NOx is another contributor along with thermal NOx. Since fuel NOx increases with
increasing excess oxygen, the overall NOx levels were observed to increase with higher 02.
Reburnlnq Tests
The two rebuming burners were located at the rear furnace wall of the SBS. Kittanning coal was fired in
the cyclone during all test phases and the cyclone was operated at 65 to 85% of total load under excess
air conditions. Reburning fuel portions provided the remaining 15 - 35% heat input. In order to obtain
various in-furnace reburning zone stoichiometries (0.85 - 0.95), the reburning burners were operated at
substoichiometric conditions. The balance of air was then introduced through OFA ports located in the
upper furnace. Under optimized test conditions, reburning burner stability was observed during each of
the reburn fuel test phases. No indication of excessive CO levels (at the stack) or burner instability was
observed during any of the optimum test conditions.
The reburning burners were first adjusted for optimum NOx emission levels via burner hardware.
Changing the swirl component exiting the burner (via spin vanes in the outer zone) had an effect on
resulting NOx levels. Reducing the amount of swirl within this system provided more reburning fuel
penetration and improved NOx reduction capability. In addition, FGR could be introduced to the burner
and an improvement in NOx reduction was also observed under this condition. Over 50% NOx
reduction was achieved with natural gas, oil, and coal reburning at optimum conditions of natural gas,
oil, and coal reburning burners. The optimum burner settings for each reburning fuel were determined
based upon NOx reduction capability, flame stability, and CO emission levels.
NOx Emissions
A 40 - 75% NOx reduction (from the baseline NOx level) was achieved during reburning under various
test conditions. These results are reported as overall reductions and consist of basically three
components:
• NOx reduction via lower heat input at the cyclone burner
• NOx reduction via substitution of main combustion zone coal input with oil or natural
-------
gas, thus reducing the total fuel nitrogen content to the furnace (oil and gas reburning
tests only)
• NOx destruction via the reburning process
The following results are based upon the overall NOx reductions obtained.
Reburning Zone Stoichiometrics. Figure 5 shows that NOx emissions decreased with decreasing
reburning zone stoichiometry for the three tested reburning fuels. The reburning burner throat
stoichiometries were set at 0.4 for coal reburning and 0.2 for oil and natural gas reburning. Therefore, a
greater portion of reburning fuel is required for coal reburning to achieve a given reburning zone
stoichiometry. Varying the amount of natural gas and oil reburning fuels from 16 to 28% of total heat
input changed the reburning zone stoichiometry from 0.95 to 0.85, respectively. To achieve the same
reburning zone stoichiometry during coal reburning tests, 22 to 36% reburning coal had to be introduced
to the furnace. Nitrogen-free natural gas provided the best NOx reduction. NOx concentrations ranged
from 420 to 235 ppm while varying the reburning zone stoichiometry from 0.95 to 0.85 during gas
reburning operation. From the baseline NOx emission level of 925 ppm, these NOx emission levels
correspond to a 55 to 75% reduction. During No. 6 fuel oil reburning tests, NOx reductions of 42 to 73%
were achieved at reburning zone stoichiometries of 0.95 to 0.85. Pulverized coal reburning reduced the
NOx levels 40 to 68% for the same range of reburning zone stoichiometry. For 50% NOx reduction from
baseline conditions, 15% natural gas or 25% coal is required.
Flue Gas Recirculation. Figure 6 shows that NOx emissions decreased with FGR rate to the reburning
burners. In these tests, cyclone and reburning burner stoichiometries and fuel portions were constant.
Reburning fuel portions were 22% for natural gas or oil reburning and 28% for coal reburning. As
explained before, the reburning burner throat stoichiometry was set at 0.2 for gas and oil reburning and
0.4 for coal reburning. These reburning fuel portions provided the reburning zone stoichiometry of 0.9.
Addition of FGR helps to improve the mixing between furnace combustion gases and the reburning fuel.
With coal reburning, NOx emissions were more sensitive to FGR than natural gas and oil reburning.
This could be due to the presence of coal nitrogen in the reburning coal portions. Without FGR, some
NOx is being formed through the volatile flame attached to the reburning burner. When FGR is added,
in addition to improved mixing, NOx formation by the volatile reburning flame may be reduced.
Therefore, coal reburning is more sensitive to FGR flow rate. This hypothesis will be confirmed through
future investigations.
Cyclone Burner Stoichiometry. The effects of varying the cyclone burner stoichiometry and percent
reburning fuel were investigated; the results are plotted in Figure 7. Although B&W recommends that
minimal cyclone operation changes be employed, various cyclone stoichiometries were tested during
this project in order to complete the technology database. Figure 7 is based upon maintaining a
constant reburning zone stoichiometry of 0.9. As the cyclone stoichiometries were varied between 1.0
to 1.2, the percentage of reburning fuel to the reburning burners (versus coal to the cyclone to keep a
constant 6-million Btu/hr load) was changed accordingly to achieve the reburning zone stoichiometry of
0.9. The natural gas input varied between 13 to 31%. The figure shows that NOx levels decreased
from 420 to 260 ppm as the cyclone stoichiometry was increased from 1.0 to 1.2, respectively.
During coal reburning tests as the cyclone stoichiometry increased from 1 to 1.2,17 to 37% coal had to
3-8
-------
be introduced to achieve the reburning zone stoichiometry of 0.9. The NOx levels were almost
insensitive to the cyclone stoichiometry. During pilot-scale coal reburning tests, the same coal was
utilized at the cyclone and reburning burners, but with different grind size. Since the total heat input was
constant at 6-million Btu/hr, the total fuel nitrogen input to the furnace was not changed at different
cyclone stoichiometries. These results indicate that the reburning zone stoichiometry is the controlling
parameter in NOx reduction in the reburning zone.
In-furnace NOx measurements were taken throughout the SBS (nine sampling ports located on the side
furnace wall) during both the baseline and reburning test phases. Baseline NOx levels were uniform
throughout the test facility, thus substantiating that all of the NOx generation occurs within and/or
immediately upon exiting the cyclone. Operating in the natural gas reburning mode (cyclone burner at
77% of load and 2% excess O2; reburning burners input at 23% of load), NOx levels at an elevation
between the cyclone exit and the burners were 900, 743, and 450 ppm at the right side, left side, and
center of the furnace, respectively. While the right-side/left-side NOx levels agree with the baseline
results, the 450 ppm at the center port indicates that some of the reburning fuel is being recirculated
below the reburning burners. During coal reburning (cyclone burner at 72% load and 2% O2; reburning
burners at 28% load), NOx levels of 900, 860, and 830 ppm were measured and recirculation was not
observed. Measuring the NOx levels directly above the reburning burners showed that the majority of
NOx reduction had occurred. These results substantiate that good mixing between the reburning fuel
and combustion gases does exist.
Pilot Furnace Temperature Profile
SBS furnace temperatures were measured during both baseline and reburning phases to determine the
technology's potential effect on temperature variations. Figure 8 illustrates the resulting FEGTs under
various operating conditions. The data indicate that while utilizing reburning, rear-wall OFA ports, a
cyclone stoichiometry of 1.1, zero percent FGR, and maintaining a constant 6-million Btu/hr furnace heat
input, approximately a 50°F FEGT increase (from baseline) was observed. However, when 10% FGR
was added to the reburning system, a temperature quenching phenomena occurred and a 50°F FEGT
decrease (from baseline) resulted. A + 50°F variation in FEGT is considered to have a minimal (if any)
impact to boiler performance.
The in-furnace probing showed no significant temperature variations between the baseline/reburning
conditions, except that again a quenching effect occurred in the reburning zone when FGR was added.
Combustible Loss
Unburned carbon and CO emissions were measured at both the stack and throughout the furnace
during the baseline and reburning phases. An inherent cyclone characteristic is that the majority of the
combustion occurs within the cyclone itself. Since the cyclone will continue to be operated in an excess
air mode, this combustion characteristic will not be altered. However, the amount of unburned char that
does not burn within the cyclone will now enter a reducing environment in the reburning zone, with the
remaining combustion air not to be introduced until the OFA ports. When coal and fuel oil are used for
reburning, additional unburned carbon may exist since the reburning fuels are introduced into the
reducing environment of the reburning zone. Although they volatilize and partially burn, final burnout will
3-9
-------
be delayed until the burnout zone. If FGR is introduced, unburned combustible levels increase since the
burnout zone residence time decreases due to increased mass loading through the furnace and the
associated lower gas temperature profile whithin the reburn zone region. Efficient mixing of the air
introduced through the OFA ports will help alleviate this concern and any potential CO emission
problems.
Numerous measurements were taken to establish a database and to validate the trends of variation of
unburned combustibles with different reburning zone parameters such as fuel split, FGR, and reburning
fuel type. Table 5 illustrates the comparison of baseline and reburning tests at optimum conditions with
and without FGR. Isokinetic samples of the fly ash were withdrawn from the stack of the SBS and
analyzed for combustibles. In addition, total mass loadings of the fly ash were measured. Table 5
shows the carbon content of the fly ash and percentage of ash at the convection pass to the total ash
input to the boiler at baseline conditions. During natural gas and oil reburning tests, the ash went down
since these reburning fuels did not contain ash. On the other hand during coal reburning tests, ash
loading almost doubled since ash from the reburning coal portion, unlike the cyclone, was not removed
as slag. Total combustion efficiencies were calculated from ash percent in the convection pass, carbon
content of the fly ash, and coal analysis. The overall change of combustion efficiencies from the
baseline condition is less than 0.1% for natural gas and oil reburning and 0.13% for coal reburning. This
is a minimal impact and provides a strong justification that the unburned combustible potential
associated with the reburning technology could be controlled to acceptable levels.
Further analyses were performed to calculate the individual combustion efficiencies of cyclone and
reburning fuels. It was assumed that natural gas burns completely. Therefore, the cyclone fuel burnout
was calculated from the total combustion efficiency and fuel split during natural gas reburning tests.
Knowing the cyclone fuel burnout, then reburning fuel burnout could be calculated during oil reburning
and coal reburning tests. The results indicate that up to 99.79% of the coal reburning fuel was burned.
CO levels were low (less than 30 ppm) at the stack during the baseline tests and there was no apparent
increase when the reburning technology was applied. In-furnace probing at the reburning zone revealed
areas of high CO (>1000 ppm) due to the substoichiometric condition of this region. Upon introduction
of OFA, the CO emissions were dramatically reduced - as stated above, less than 30 ppm CO was
measured at the furnace exit. Thus, it is apparent that good mixing between the OFA and combustion
gases did indeed exist.
Corrosion Potential
Since the reburning zone must be operated under substoichiometric conditions, corrosion potential
within this region was investigated. By operating the cyclone in an excess air mode, the majority (if not
all) of the sulfur from the coal in the main combustion zone is converted to SO2. Due to the reducing
atmosphere in the reburning zone, H2S measurements were performed. High concentrations of H2S
can be conducive to increased rate of tube corrosion. H2S concentrations at baseline and reburning
conditions are illustrated in Table 6. Multiple measurements were performed in the furnace, and results
are presented in a range of H2S concentrations. Up to 60 ppm of H2S were measured at the SBS
during baseline conditions. H2S levels did increase up to 90 ppm during gas reburning where no
additional sulfur was added with the reburning fuel. Fuel oil utilized for reburning contained 0.78%
3-10
-------
sulfur, and H2S levels were compatible to those observed during gas reburning. When coal was
utilized, however, up to 265 ppm of H2S was measured. The impact of these levels of H2S on tube
wastage has yet to be determined. It is encouraging that only a small percentage of S02 from cyclone
flue gases is converted to H2S. In addition, when sulfur-bearing fuels were utilized for reburning, only a
small fraction of the reburning fuel sulfur converted to H2S. Up to 200 ppm of H2S for oil reburning and
900 ppm of H2S for coal reburning would be detected if all of the reburning fuel sulfur were converted to
H2S. Further evaluations will predict corrosion rates within the various furnace region during reburn
operation.
FULL-SCALE UTILITY APPLICATION ECONOMICS
An economic analysis was performed in order to estimate the total capital and levelized revenue
requirements for retrofitting and operating a reburning system to reduce NOx emissions from a base
case 200-MW unit. Costs associated with this process included: acquisition and handling of the
reburning fuels, installation and operation of the reburning system, and boiler impacts and
countermeasures. Prime concern within this task is to evaluate the potential of this technology on a
commercial scale based upon dollar practicality. High priority exists concerning cost comparisons
between utilizing various reburning fuels (gas, oil, or coal) in this process. The basis for the costs
utilized in this evaluation are B&W cyclone reburning proposal cost estimates that have been prepared
for numerous cyclone reburn proposed demonstration projects. These proposals have included utilizing
each of the three reburn fuels.
The EPRI economic premises for electric power generating plants were used to estimate the above-
stated objectives. The information reported herein is an expansion to a previous B&W/EPRI-sponsored
engineering feasibility study that included an economic analysis (EPRI Project RP-1402-30). Table 7
summarizes the economic evaluation per each reburning fuel type.
The following listing summarizes the major equipment components utilized for each of the reburn fuels
evaluated:
Maior Reburning Control System Components
All Reburn Fuels
• Reburn Burners
• Overfire Air (OFA) Ports
• Tube Wall Openings/Replacement Wall Panels
• Piping/Ductwork to Reburn Burners/OFA Ports
• Burner/Combustion Control System
• Cyclone Gravimetric Feeders
• Cyclone Secondary Air Monitors
Coal Reburning
• Pulverizer/Gravimetric Feeder
• Coal Handling System
3-11
-------
• Coal Silo
• Structural Steel/Pulverizer Enclosure
Oil Reburninq
• Positive Displacement Pumps
• Oil Storage Tanks
Gas Reburninq (where gas is currently not on-site)
• Gas Substation
• 10-Mile Tie-In to Nearest Pipeline at $300,000/Mile
Capital costs and 10-year levelized busbar power costs were sensitive to reburn fuel type, fraction, and
price. Approximately 70 - 90% of the associated 10-year levelized cost is attributable to the fuel cost.
Variations in prices for gas, oil, and coal in different demand regions will influence the economics of
reburning with these alternative fuels. Prices (1987 dollars), ranging from $2.50 - $3.50/10® Btu for gas,
$3.00 - $4.00/10® Btu for oil, and $1.70/10® Btu for coal, were evaluated. Two gas availability scenarios
were also considered - gas on-site and 10-mile tie-n to nearest pipeline at $300,000/mile.
The results presented in Table 7 give some indication of the variability in costs as key cost parameters
are altered. For gas reburning, the installed capital costs range from $22.4/kW - if gas is available on-
site - to about $44/kW - if the assumed $3 million gas-line cost is borne solely by the power plant.
[Note: In many cases the gas supplier willextend gas service at no direct cost to the user, but will factor
this cost into the contracted transportation charges (rate base). In this case, the capital cost would be
the same as the gas on-site situation.]
The 10-year levelized costs for 15% gas reburning are shown to increase from 2.3 mills/kWh at
$2.50/106 Btu gas to 4.1 mills/kWh at $3.50/10® Btu gas. These prices translate into gas-oil price
differentials of $0.80 and $1.80/10® Btu, respectively. The gas reburning busbar costs do not include
any credits for reduced coal handling/inventory, ash disposal, or maintenance as a result of 10% gas
substitution.
Oil reburning is projected to cost about $28/kW on the 200 MW plant with 10-year levelized costs
ranging from 3.3 to 4.9 mills/kWh at assumed oil prices of $3.00/10® Btu and $4.00/10® Btu,
respectively, and 19% oil firing.
Finally, capital costs for pulverized-coal reburning are estimated at $41.4/kW. Assuming the 25%
reburn coal fraction is the same fuel as that currently fired in the cyclone burners, the 10-year
incremental busbar cost is estimated at 1.7 mills/kWh.
CYCLONE REBURNING DEMONSTRATION
B&W is presently starting the final stage of its three-phase program to evaluate and commercially
demonstrate reburning technology in cyclone-fired boilers. As stated earlier, Phase I involved an
EPRI/B&W-sponsored engineering study, while Phase II was the pilot-scale testing performed at B&W's
Alliance Research Center sponsored by EPRI/GRI/B&W and reported herein. Consequently, based
3-12
-------
upon the positive results of the first two phases, Phase ill of the program is in progress and involves a
full-scale demonstration of coal reburning at Wisconsin Power & Light's 100-MWe Nelson Dewey
Station Unit 2 (see Figure 9). The performance goals are to reduce existing uncontrolled NOx
emissions by greater than 50% and exhibit no serious impact on cyclone combustor operation, boiler
efficiency, fireside corrosion or deposition, and ash removal system performance.
To accomplish these objectives, the project is divided into three phases:
Phase I - Design and Permitting
Phase II - Construction and Start-up
Phase III - Operation and Disposition
Each phase is broken down into tasks and subtasks as shown in the Work Breakdown Structure (WBS)
shown in Figure 10. Project initiation is anticipated for Summer 1989 with system start-up in Fall 1991.
An actual one year demonstration test duration is scheduled. The contract for this overall 43-month
project is presently being negotiated under the DOE Clean Coal II solicitation. The program is being
supported by the DOE, Wisconsin Power & Light, B&W, EPRI, the State of Illinois, and a consortium of
utilities.
Wisconsin Power & Light's Nelson Dewey Unit 2 is a B&W single-wall-fired RB-type cyclone boiler
equipped with three cyclone furnaces. The major components that will allow for the implementation of a
coal reburning retrofit demonstration are:
1) Reburning burners capable of firing pulverized coal
2) Flue Gas Recirculation (FGR) to reburn burners
3) Dual-Air Zone Overfire Air (OFA) ports
4) Furnace wall tube panels for reburn burners/OFA ports
5) Air monitors, dampers, and damper drives within air/gas recirculation ducts and flues
6) Burners/OFA/FGR Control Management System
7) B&W MPS-67 Pulverizer with Gravimetric Feeder
8) Reburn fuel handling system
9) Gravimetric Feeders to replace Cyclone Volumetric Coal Feeders
10) B&W Continuous Monitoring Diagnostic System-140
11) Continuous Gaseous Emission Monitoring System
12) Miscellaneous equipment, wiring, piping and ducting, etc.
3-13
-------
In summary, the demonstration project includes collecting baseline emissions and performance data,
along with performing a general boiler system assessment, at Nelson Dewey Unit 2 prior to the coal
reburning retrofit. The coal reburn system will be designed, fabricated, and installed based upon B&W's
pilot-scale combustion tests, physical and numerical flow modeling, and experience/knowledge of full-
scale burner/OFA port/control system retrofits. Parametric/optimization and performance tests will
assess the potential of the technology from both the resulting emission reductions and boiler
performance capability aspects. Finally, readiness for commercialization will be determined from both a
technical and economic viewpoint.
CONCLUSIONS AND RECOMMENDATIONS
• A 40 - 75% overall NOx emission reduction is achievable in cyclone-equipped units
via the reburning technology. This overall NOx reduction is attributed to three
different mechanisms: 1) NOx destruction in the reducing environment of the
reburning zone via reburning process, 2) during gas and oil reburning, secondary fuel
input to the reburning zone contributes a small percentage of NOx formation (little or
no fuel-bound nitrogen in fuel), and 3) reduced load and oxygen level at the cyclone.
Typical uncontrolled NOx emission levels from cyclone units are 600 - 1400 ppm at
3% 02.
• For a 50% NOx reduction, 15% natural gas or oil and 25% coal are required.
• The lower in-furnace reburning zone stoichiometry (0.85 - 0.95 range) provided the
best overall NOx reduction.
• Flue gas recirculation (FGR) to the reburning burners improved the mixing
(turbulence) characteristics between the combustion gases/reburning fuel and
consequently improved the NOx reduction capabilities. FGR was more effective
during coal reburning than during natural gas or oil reburning. Where applicable, this
tool could be beneficial in future applications.
• CO emission levels were low (less than 30 ppm) throughout the various optimal test
conditions and, thus, were of no concern during the reburning operation.
• Total combustion efficiency insignificantly decreased less than 0.1% for natural gas
and oil reburning, and 0.13% for coal reburning. This is a minimal impact.
• Furnace exit gas temperatures (FEGTs) increased by less than 50°F during reburning
operation.
• The cyclone itself must be operated under excess air conditions in order to minimize
corrosion potential within the cyclone barrel. Accurate air/fuel control is also essential
to alleviate this potential concern.
• H2S concentrations in the reburning zone were 90 and 265 ppm for natural gas and
coal reburning, respectively. Only a small portion of sulfur in the coal was converted
to H2S.
• The nominal costs to apply reburning to a baseloaded 200-MW cyclone unit to
achieve a 50% NOx reduction with different reburning fuels are estimated as follows
(total capital costs, 10-year busbar power cost):
3-14
-------
- Gas (On-site or pipeline extension factored in rate base) - $22.4/kW, 2.3 mills/kWh
- Oil - $28.4 kW, 3.3 mills/kWh
- Coal - $41,4/kW, 1.7 mills/kWh
The corresponding gas and oil fuel price differentials (compared to coal) used to
determine these values are $0.80 and $1.30/10® Btu, respectively. The coal reburn
fuel is assumed to be the same as the main cyclone coal.
• If the capital cost of a 10-mile tie-in to an existing pipeline is passed on directly to this
plant, the capital and 10-year levelized power costs increase to about $44/kW and
3.1 mills/kWh, respectively.
• The reburning fuel choice has a major impact on the economics of this process. Site-
specific consideration of the availability and price of alternative fuels, the availability
of capital, and NOx reduction target will influence the attractiveness of any one
option.
ACKNOWLEDGMENTS
The authors extend their appreciation to the following B&W personnel for their help in the management
and performance of the SBS testing: John Doyle, John Rackley, Cliff Eckhart, Mike Holmes, Suzanne
Barnes, Vic Burgess, Gene Staffer, John McCoury, and Terry Wilson.
REFERENCES
1. Maringo, et al., "Feasibility of Reburning for Cyclone Boiler NOx Control," 1987 EPA/EPRI Joint
Symposium on Stationary Combustion NOx Control, New Orleans, Louisiana, March 23-27,1987.
2. EERC, "Evaluation of In-Furnace NOx Reduction and Sorbent Injection on NOx/SOx Emissions of
U.S. Designed Pulverized Coal-Fired Boilers," 1st and 2nd Technical Review Panel on In-Furnace
NOx Reduction, Salt Lake City, Utah, 1984 and 1985.
3. Okigami, et al., "Three-Stage Pulverized Coal Combustion System for In-Furnace NOx
Reduction," 1985 EPA/EPRI Joint Symposium on Stationary NOx Control, Boston,
Massachusetts.
4. Wendt, et al., "Reduction of Sulfur Trioxide and Nitrogen Oxides by Secondary Fuel Injection,"
14th International Symposium on Combustion.
5. Takahashi, et al., "Development of Mitsubishi (MACT) In-Furnace NOx Removal Process,"
U.S.-Japan NOx Information Exchange.
6. Chen, et al., "Controlling Pollutant Emissions From Coal and Oil Through the Supplemental Use
of Natural Gas," 1986 EPA/EPRI Symposium on Dry S02 and Simultaneous S02/N0x
Technology, Raleigh, North Carolina, June 2-6,1986.
7. McCarthy, et al., "Pilot-Scale Process Evaluation of Reburning for In-Furnace NOx Reduction,"
Final Report on EPA Contract 68-02-3925.
3-15
-------
LEGAL NOTICE
This paper was prepared by The Babcock & Wilcox Company (B&W) as an account of work sponsored
jointly by The Electric Power Research Institute, Inc. (EPRI), and The Gas Research Institute (GRI).
Neither EPRI, or GRI, members of EPRI or GRI, B&W, nor any person acting on behalf of these:
(a) makes any warranty, express or implied, with respect to the use of any information,
apparatus, method, or process disclosed in this paper or that such use may not
infringe privately-owned rights; or
(b) assumes any liabilities with respect to the use of, or for damages resulting from the
use of, any information, apparatus, method, or process disclosed in this paper.
SECONDARY
COAL CHUTE
CRUSHED COAL
1/4" SCREEN MESH
TERTIARY
AIR INLET
SCROLL
BURNER
CYCLONE BARREL
SLAG SPOUT
OPENING
Figure 1 Cyclone Furnace
3-16
-------
BALANCE OF AIR
1.15 -1.20 OVERALL
STOICHIOMETRY
15 -30% HEAT INPUT
0.2 - 0J STOICHIOMETRY
FLUE GAS RECIRCULATION
(OPTIONAL)
OVERFIRE "
AIR PORTS.
REBURNING _
BURNERS --
70 - 85% HEAT INPUT
(CRUSHED COAL)
1.1 STOICHIOMETRY
CYCLONES
BURNOUT
ZONE
REBURN
ZONE
MAIN
COMBUSTION
ZONE
3 - 4% EXCESS 02
0.85-0.95
"STOICHIOMETRY
ZONE
>
Figure 2 Reburning Technology
Figure 3 Small Boiler Simulator (SBS) Schematic
3-17
-------
1000
900
800
O
g 700
CO
o
H
a eoo
IU
H
O
UJ 500
oc
a
o
O 400
\ 300
o
z
200
100
NSPS EMISSION LEVEL
~ KITTANNING COAL AT 6-MILLION BTU/HR
A KITTANNING COAL AT 4.3-MILLION BTU/HR
• NATURAL GAS AT 6-MILLION BTU/HR
Figure 4
2 4
EXCESS OXYGEN.PERCENT
Baseline N0„ Emission Levels in the SBS
600
500
« 400
e
o
UJ
o
UJ
oc
CC
o
o
E
Q.
0.
0.95
REBURNING ZONE STOICHIOMETRY
0.90
0.85
300
200
100
o«
XX
V\
GAS
CYCLONE AT 10% EXCESS AIR
10% FGR
6-MILLION BTU/HR LOAD
40
60
80
UJ
O
CC
UJ
a
z"
o
H
a
z>
a
UJ
CC
28
16 18 20 22 24 26
OIL & NATURAL GAS, PERCENT
I I
100
30
20
25 30
COAL, PERCENT
Figure 5 N0X Emission Levels With Reburning
35
3-18
-------
600
200
100
CYCLONE AT 10% EXCESS AIR
REBURNING ZONE STOICHIOMETRV 0.9
~ GAS REBURNING
O
o
OIL REBURNING
COAL REBURNING
l
4 8 12 16 20 24
FGR, PERCENT OF TOTAL FLUE GAS
Figure 6 Effect of Rue Gas Recirculation (FGR) on N0X Emission Levels
soo
O
e
3
H
O
Ut
cc
cc
o
o
E
o.
400
300
200
100
REBURNING ZONE
STOICHIOMETRV OF 0.9
~ GAS REBURNING
O OIL REBURNING
O COAL REBURNING
0
0.98
I
1.02
1.06 1.10 1.14
CYCLONE STOICHIOMETRY
1.18
1.22
Figure 7 Effect of Cyclone Stoichiometry and Percent Reburning Fuel on N0X Emission Levels
3-19
-------
2400
2300
2200
BASELINE
REBURNING
2100
CYCLONE AT 10% EXCESS AIR
FULL LOAD CONDITIONS
~ GAS REBURNING
O OIL REBURNING
O COAL REBURNING
2000
I
I
Figure 8
6 8 10 12 14 16
FGR, PERCENT OF TOTAL FLUE GAS
Operational Effects on Furnace Exit Gas Temperature (FEGT)
18
20
3-20
-------
Reproduced from
best available copy.
Figure 9 Wisconsin Power & Light Company's Nelson Dewey Station Unit 2
3-21
-------
Figure 10 Project Work Breakdown Structure
3-22
-------
Table 1
N0X EMISSIONS WITH IN-FURNACE NOx REDUCTION
Main Fuel/Reburninq Fuel NO., Level (ppm @ 3% Oo)
Natural Gas/Natural Gas
15 -
40
Fuel Oil/Fuel Oil
40 -
60
Pulverized Coal/Pulverized Coal
50 -
100
Future Cyclone Application
240 -
560
* Based on a 60% NOx reduction from
existing cyclone
emission levels.
Table 2
SUMMARY OF REBURNING CONDITIONS EVALUATED DURING PILOT TESTS
Fuel
Fuel Split
Stoichiometry
Residence Time
(Assume Plug Flow)
Main Combustion
(Primary) Zone
Kittanning Coal
70 - 85%
1.0 - 1.2
0.1 second
Reburninq Zone
Natural Gas,
No. 6 Fuel Oil,
Kittanning Coal
15 - 30%
0.85 - 0.95
0.5 - 0.8 second
Burnout Zone
1.05 - 1.2
0.6 - 0.9 second
3-23
-------
Table 3
COMPARISON OF BASELINE CONDITIONS FOR SBS FACILITY AND COMMERCIAL UNITS
Cyclone Temperature
Residence Time
Furnace Exit Gas Temperature
NOx Level
Ash Retention
Unburned Carbon
Ash Particle Size (MMD; Bahco)
SBS
>3000°F
1.4 sec at full load
2265°F
900 - 1200 ppm
80 - 85%
<1% in ash
6-8 microns
Typical Cyclone-
Fired Boilers
>3000°F
0.7 - 2 sec
2200° - 2350°F
600 - 1400 ppm
60 - 80%
1 - 20%
6-11 microns
Table 4
FUEL ANALYSIS
Kittanninq Seam 2
As-Received
Dry
Fuel Oil
Proximate Analysis, %
Moisture
Volatile Matter
Fixed Carbon
Ash
Gross Heating Value,
Btu per lb
Btu per lb (M&A Free)
Ultimate Analysis, %
Moisture
Carbon
Hydrogen
Nitrogen
Sulfur
Ash
Oxygen (Difference)
Total
5.62
35.35
49.31
9.72
12,580
5.62
69.93
4 .77
1.42
2.55
9.72
5.99
100.00
37.45
52.25
10 .30
13,330
14,860
74.09
5.05
1.50
2.70
10.30
6.36
100.00
18,509
88 .25
10.70
0.22
0.78
0.05
0.00
100.00
Natural Gas:
88.27% methane, 5.27% ethane, 1.72% propane, 3.6% nitrogen,
1.14% others
3-24
-------
BASELINE
GAS REBURN
NO FGR
10% FGR
OIL REBURN
NO FGR
10% FGR
COAL REBURN
NO FGR
10% FGR
Tables
COMPARISON OF COMBUSTION EFFICIENCIES
CARBON,
%
0.3
2.3
4.5
3.0
5.4
1.6
3.4
ASH, % IN
CONVECTION
PASS
18.2
14.2
14.2
14.2
14.2
32.7
32.7
TOTAL
COMBUSTION
EFFICIENCY
99.99
99.96
99.92
99.95
99.91
99.94
99.87
CYCLONE
FUEL
BURNOUT, %
99.99
99.95
99.90
99.95
99.90
99.95
99.90
REBURNING
FUEL
BURNOUT, %
N/A
100
100
99.95
99.93
99.91
99.79
Table 6
REBURNING NOx CONTROL FOR CYCLONE BOILERS FIRESIDE CORROSION • H2S CONCENTRATION
(ppm)
MEASURED
CYCLONE OUTLET
BELOW REBURN
REBURN ZONE
BASELINE
0
40-55
0-60
GAS REBURN
50
25-90
OIL REBURN
0
14-93
COAL REBURN
98
0-200
0-265
Table 7
ECONOMIC EVALUATION FOR APPLYING REBURNING TO CYCLONE BOILERS*
TOTAL ESTIMATED 10-YEAR LEVELIZED
REBURNING
FUEL
FUEL COST
(S/106 Btu)
MAIN/REBURN
FUEL SPLIT
CAPITAL COST
REQUIRED ($/kW)
BUSBAR POWER
COST (mills/kWh)
GAS
2.5
85/15
22.4
2.3
3.5
85/15
22.5
4.1
2.5
85/15
43.9"
3.1"
3.5
85/15
44.1"
4.9"
OIL
3.0
81/19
28.4
3.3
4.0
81/19
28.6
4.9
COAL (SAME
1.7
75/25
41.4
1.7
AS MAIN FUEL
* BASED ON 200-MW UNIT OPERATING AT 65% CAPACITY WITH 50% REDUCTION.
CYCLONE BURNER OPERATES AIR-RICH (1.1 STOICHIOMETRY) AND REBURN ZONE
FUEL-RICH (0.93 - 0.97 STOICHIOMETRY).
** ASSUMES J3-MILLION GAS PIPELINE COST.
3-25
-------
(intentionally Blank)
3-26
-------
Application of Reburning to a
Cyclone Fired Boiler
R. W. Borio, A. F. Kwasnik, D. K. Anderson
Combustion Engineering, Inc.
Windsor, CT 06095
D. A. Kirchgessner
Air and Energy Engineering Research Laboratory
United States Environmental Protection Agency
Research Triangle Park, NC 27711
R. A. Lott
Gas Research Institute
Chicago, IL 60631
A. Kokkinos
Electric Power Research Institute
Palo Alto, CA 94303
S. Durrani
Ohio Edison
Akron, OH 44308
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
3-27
-------
ABSTRACT
Cyclone fired units typically produce the highest nitrogen oxide (NO ) emissions
of the commonly employed boiler designs. Although cyclone units represent about
14% of the pre-new source performance standards (NSPS) coal fired generating
capacity, they contribute about 21% of the NO emitted by pre-NSPS units.
Reburning is currently viewed as the most promising in-furnace NO reduction combustion
technology for cyclone fired boilers. Based on the above the U.S* EPA, Electric
Power Research Institute, Gas Research Institute, and Ohio Coal Development Office
have cosponsored Combustion Engineering's demonstration of reburning on a large
cyclone fired boiler. Key elements of the program include reburn system design
and retrofit, followed by short and long term test segments, and an overall
economic evaluation of reburning applicability. Coal is the primary boiler fuel
and natural gas will be used as the reburn fuel.
The paper reports the progress principally of the design related phases of the
program. Reburn system design criteria are presented as well as the methodology
and results of the cold-flow modeling evaluation which was used to identify
existing boiler aerodynamics and help to determine the optimum configuration for
reburn fuel and additional air injectors. Mathematical modeling was conducted to
determine potential effects of a reburn system on boiler thermal performance.
Finally, experience has been cited regarding effects of low air/fuel
stoichiometric conditions on boiler tube wastage.
INTRODUCTION
Statement of the Problem
It is widely acknowledged that NO plays an important role in the formation of
acid rain (1,2,3). One important source of NO is utility boilers, particularly
those which burn coal and to a lesser extent tftose which fire oil or natural gas.
Of particular interest are coal fired cyclone boilers since they contribute 21% of
the NO emitted by pre-NSPS coal fired units even though they represent only 14%
of the pre-NSPS coal fired generating capacity. In-furnace NO reduction
technologies represent one commercially attractive means for reducing NO .
Reburning represents a particular in-furnace NO reduction technique for which
cyclone fired boilers are uniquely well suited.
Given this background the U.S. EPA, Gas Research Institute (GRI), Electric Power
Research Institute (EPRI), and the Ohio Coal Development Office (OCDO) have
sponsored a program for Combustion Engineering (C-E) to demonstrate reburning on a
cyclone fired boiler. The Consolidated Natural Gas Company (CNG) is a significant
contributor to the program as natural gas will be utilized as the reburn fuel.
Ohio Edison is a key participant in the program and will be providing Unit No. 1
at their Niles Station, a 108 MW cyclone fired boiler, as the demonstration site.
Working with C-E as subcontractors in this program are Energy Systems Associates
(ESA), Mitsubishi Heavy Industries (MHI), and Physical Sciences, Incorporated
(PSI).
Program Objective
The overall objective of this program is to successfully demonstrate the operation
of a reburn system on a cyclone fired boiler. Figure 1 shows the key tasks in the
program and a schedule for their planned implementation. Progress to date
includes completion of the technology review task and most of the reburn system
design task. Current activities are mainly centered on the reburn system
fabrication and installation task. This paper will serve as a foundation for
3-28
-------
future technical papers while specifically reporting on the progress made within
Tasks 1 and 3 (See Figure 1). In this regard it will present important criteria
to be considered in the design of a reburn system, it will report on the use of
cold-flow modeling as a site specific engineering design tool, and it will address
the question of potential impacts of reburn system operation on boiler
performance.
Reburn Technology
The concept of reburning and the postulated chemical -reactions which account for
the reduction of NO in the reburn zone have been addressed elsewhere in the
literature (4,5,6,7^ and will not be reiterated in detail here. Briefly,
reburning is an in-furnace technique for reducing NO by creating a slightly
reducing zone downstream of the primary combustor as shown schematically in Figure
2. The reducing zone is created by introducing fuel into a zone with insufficient
oxygen available to fully combust the fuel. The presence of a reducing zone
creates intermediate nitrogen-containing species which subsequently react with
previously formed NO to form the desired product, molecular nitrogen. Any
unburned fuel leaving the reburn zone is subsequently burned to completion in the
burnout zone.
Reburning can be used on all types of fossil fuel fired boilers (i.e., wall,
tangential, and cyclone) and, in fact, has been successfully employed on a number
of oil fired boilers in Japan where oil has been used as the reburn fuel (8).
Wall fired and tangentially fired boilers employ more conventional burners (as
contrasted to a cyclone boiler) making possible the opportunity for other
in-furnace NO reduction technologies involving burner modifications in
conjunction with staged air combustion. Since a cyclone combustor may not be able
to tolerate significant changes to its operation, such as lower excess air,
without the possibility for creating other problems, it is limited to an
in-furnace N0_ reduction technology that does not depend on significant changes to
its present mode of operation. Reburning does not require that any significant
operational changes be made to the primary combustor. The key requirement is that
the fuel feed rate be reduced in the primary combustor with an equivalent amount
(on a Btu basis) of fuel being injected in the reburn zone, usually not more than
20% of the total fuel input. The excess air, and hence the air/fuel stoichiometry
within the cyclone combustors, can be held constant thereby not creating the
potential for cyclone operational problems.
REBURN SYSTEM DESIGN
Technical Review
A successful reburn system should meet the theoretical criteria for effective NO
reduction, it should not adversely affect normal boiler operation, and it should
be competitively economical to install and operate.
The reburn system being designed for Ohio Edison's Niles Station, Unit 1, will
fire natural gas as the primary reburn fuel, but the reburn fuel injectors have
been designed to accommodate oil as a future reburn fuel. The reburn fuel feed
system is designed for a nominal fuel flow rate of 20% of total fuel fired (on a
heat input basis). Recirculated flue gas will be used as a carrier to inject fuel
into the reburn zone; the recirculated flue gas injection system will be designed
for a nominal flow rate of 10% of flue gas flow leaving the stack. The system for
injection of additional air downstream of the reburn zone will be nominally
designed for an air flow rate that will permit original overall boiler
stoichiometric conditions to be maintained. Note that the stoichiometry within
3-29
-------
the cyclone combustor will riot be altered from normal operating conditions.
Parametric testing will be carried out under Task 5 to determine optimum flow
rates of reburn fuel and recirculated flue gas as well as optimizing other reburn
system operational parameters for effective NO reduction.
x
Overall design objectives for the reburn system were:
(1) To meet theoretical criteria for effective NO reduction while
.... x
minimizing any impact on normal boiler operation.
(2) To incorporate operational flexibility within the design to permit
optimized field performance
The goal of Task 1 in the program was to develop a list of key design criteria
commensurate with the overall design objectives for the reburn system. The
criteria would consist of theoretical criteria for effective NO reduction and
they would identify key practical, commercial considerations for reburn system
design, installation, and operation.
Key design criteria for the reburn zone were determined to be:
o Inject reburn fuel into as high a temperature zone as possible
o Maintain average stoichiometry between 0.90 and 0.95
o Permit a small amount of 0„ to promote formation of OH and H
radicals
o Maintain a residence time between 0.5 and 0.7 seconds
o Maximize entrainment, mixing, and dispersion of reburn fuel
o Avoid direct fuel impingement on boiler walls
o Minimize the number of required boiler penetrations
o Locate fuel injection nozzles to minimize boiler/structural steel
modifications
o Provide for maximum flexibility of reburn fuel jet direction and
flow rates
o Provide a fuel flow rate control system with automatic load
following capability
o Provide safeguards for fail-safe operation
The first five criteria, above, deal with theoretical considerations while the
last six deal with practical, commercial considerations.
Injection of reburn fuel into a high temperature zone enhances NO reduction by
favoring higher chemical reaction rates; however, reburn fuel should not be
injected before the bulk of the primary fuel has burned to completion. If
injected too early in a coal fired boiler, natural gas, as the reburn fuel, would
preferentially burn before the coal char particles have burned to completion.
This could increase the possibility for unburned carbon while, additionally, not
permitting all the char bound nitrogen to be released prior to the reburn zone. A
stoichiometry in the range of 0.90 to 0.95 has been found to represent a
reasonable balance between achieving a desirable stoichiometry from the standpoint
of NO reduction chemistry and a stoichiometry that will not exacerbate ash
deposition and/or boiler tube wastage (8). Though the reaction kinetics for NO
reduction in the reburn zone are quite fast, requiring on the order of 0.1 second,
the bulk of the residence time in the reburn zone is required to achieve good
mixing of the reburn fuel with the bulk flue gas. The naturally occurring small
amounts of oxygen in the bulk flue gas entering the reburn zone, along with that
in the recirculated flue gas which is used to carry and mix the natural gas, are
sufficient to promote the desired formation of OH and H radicals. Effective and
rapid mixing of reburn fuel ensures that all NO entering the reburn zone will
3-30
-------
contact the intermediate nitrogen-containing species so that maximum NO reduction
is possible. Effective mixing must be achieved in such a way that there is no
direct fuel impingement on boiler walls which could exacerbate tube wastage or
iron-related ash deposition by creating low local stoichiometries or worse yet a
condition where extremes occur between highly oxidizing and highly reducing
atmospheres. Other practical considerations involve minimizing the number of
boiler penetrations and the avoidance of unnecessarily costly boiler modifications
relative to the number and placement of reburn fuel injectors. The number and
placement of reburn fuel injectors must not create thermal or structural boiler
problems. The reburn fuel injection system should have sufficient flexibility to
permit on line adjustment to maintain optimum mixing as a function of boiler
operational variables, such as load changes, that could alter gas flow patterns
within the reburn zone. Since the amount of reburn fuel required will likely
change as a function of boiler load, a control system should be provided which
will provide automatic load following capability. The reburn fuel control system
should also have permissives which must be satisfied to ensure its safe operation.
Key design criteria for the burnout zone were determined to be:
o Inject burnout air in as low a temperature zone as possible
commensurate with obtaining fuel burnout before entering the
first convective surface,
o Provide for rapid mixing of air to minimize pockets of unburned
fuel.
o Avoid direct air impingement on furnace walls.
o Minimize final excess oxygen commensurate with obtaining good fuel
burnout.
o Provide for a residence time in the range of 0.6 to 0.8 second,
o Minimize the number of required boiler penetrations commensurate
with obtaining good mixing,
o Locate burnout air injectors to minimize boiler structural
modifications while providing good mixing,
o Provide for maximum flexibility of air jet direction and flow,
o Provide an air flow rate control system with automatic load
following capability,
o Provide safeguards for fail-safe operation.
For the burnout zone, unlike the reburn zone, air should be injected in as low a
temperature gas as possible to prevent the reformation of NO . However, lower
temperatures could prevent complete burnout of the uncombusted fuel which leaves
the reburn zone, and a balance must be struck between the dual objectives of
minimizing NO reformation and complete combustible burnout. Rapid and thorough
mixing in the'Surnout zone is necessary. Although the reaction between fuel and
oxygen is quite rapid, the bulk of the recommended 0.6 - 0.8 second residence time
is needed to achieve effective mixing rather than for combustion reaction time per
se. Direct impingement of air on furnace walls should be avoided, more for
reasons of preventing local temperature increases than for any concern about the
presence of an oxidizing atmosphere. The amount of air should be just sufficient
to achieve desired fuel burnout; an overabundance of excess air will contribute to
dry gas losses and will increase the potential for NO reformation in the burnout
zone. The rationale for employment of the practically oriented burnout zone
design criteria is basically the same as that presented earlier for the reburn
zone.
Cold-Flow Modeling,
As previously discussed, the aerodynamically controlled mixing processes of the
injected reburn fuel (hereafter designated as UFI for Upper Fuel Injector) in the
3-31
-------
reburning zone and the air (hereafter designated as AA for Additional Air) in the
burnout zone are major factors in determining the effectiveness of NO reduction
in a reburn system. Since the penetration and dispersion of the UFI and AA jets
are a function of the gas flow patterns in the furnace, the injection process must
be characterized in terms of both injector and furnace operating parameters for
the boiler in question.
Based on established methods for physical flow modeling (9), the purpose of the
isothermal flow modeling program was to develop and screen UFI and AA injection
concepts which would provide effective penetration and mixing within Ohio Edison's
Niles Station Unit 1. Key components of the cold-flow modeling experimental
program included:
o Fabrication of a one-ninth scale isothermal flow model of Niles
Unit 1.
o Characterization of the baseline furnace aerodynamics entering the
reburn zone, within the reburn zone, and within the burnout
zone.
o Characterization of the UFI and AA injection process as a function
of furnace load; UFI and AA injector size, number, and
location; and injection tilt, angle,and velocity.
o Identification of UFI and AA injection configurations for
subsequent design, fabrication, and installation.
Experimental Test Program
The isothermal flow modeling was performed at C-E's Kreisinger Development
Laboratory. The one-ninth scale model of the Niles Unit, Figure 3, was built
primarily of clear plastic, encompassing the entire furnace from the cyclone
combustors to the vertical furnace outlet plane. The cyclone combustors were
designed to produce the correct swirl number and momentum (axial and tangential)
entering the primary furnace. Upper furnace radiant and convective heat transfer
surfaces were also modeled. A header system fed by a high pressure blower
controlled the introduction of smoke or tracer gas to any one or a combination of
UFI or AA injection nozzles. Flue gas flow through the model was simulated by
drawing air through it with a large induced draft fan.
An initial series of isothermal flow modeling tests characterized the as-found or
baseline gas flow characteristics of the boiler. Following the establishment of
the baseline reference data, flow modeling of the reburn system consisted of two
screening level tests for both the UFI and AA jets:
o Screening Level 1 - Flow visualization (with smoke) of a large
number of reburn fuel/burnout air injector configurations.
o Screening Level 2 - Mixing study tests on best configuration
candidates from Level 1.
o Level 3 - Detailed three-dimensional velocity flow modeling on
best configuration candidates from Level 2.
In Screening Level 1, smoke flow visualization tests were performed for each
candidate injection system at simulated full and 70% load furnace operating
conditions. Each injection configuration was evaluated at three injection
velocities, three tilts, and a number of yaws. After initial selection of the
best injection configurations, Screening Level 2 testing was performed consisting
of methane tracer gas injection with concentration measured with a laser
absorption spectrophotometer. Final injection configurations were determined from
results of detailed velocity profile measurements using three dimensional
(five-hole pitot tube) analysis techniques. Details of the instrumentation used
3-32
-------
to carry out these measurements are given in Reference 9.
Test Results
Baseline furnace velocity fields were measured and the flow patterns determined at
the five test planes shown in Figure 4. These data were obtained under flow
conditions simulating boiler operation at 100% and 70% MCR (Maximum Continuous
Rating). The data were then used to formulate potential injection configurations
for the UFI and AA systems. For example, Figure 5 shows the results for a zone of
particular interest, the entrance to the reburn zone.' As shown, this plane is
characterized by two high velocity areas along the rear wall of the boiler,
corresponding to the flow originating from the two lower (of four) cyclones. The
outlet from the lower cyclones is partly below the dividing wall; consequently, a
large portion of the gases exiting these cyclones passes unimpeded under the
division wall into the secondary furnace.
Fourteen UFI configurations were evaluated at three different yaws in the first
reburn fuel injector screening test series. Based on the results from these tests
it was found that, at all velocities tested, three injectors along the rear wall
were insufficient to cover the flow plane. Five injectors along the rear wall
were found to be the best overall. However, with this configuration, the two
outermost injectors suffered from jet wall attachment when injecting straight into
the furnace. Yawing these injectors toward the center of the unit eliminated this
problem. Injection from the side wall also provided generally good distributions,
but no better than with the more economical five rear wall injector
configurations.
The addition of pant-legs to the ends of the injection nozzles was found to be an
effective means of enhancing, the dispersion of the reburn fuel jets. Pant-legs
were found to significantly improve the dispersion of the jet near the rear wall.
The most effective use of yawing was when each of the three inboard upper nozzles
was split and yawed, using pant-legs, while the lower nozzles were unencumbered
and allowed to penetrate to near the division wall. Since it would not make sense
to put pant-legs on the outermost sets of injection nozzles (being located next to
the side walls), these nozzles, both upper and lower, were yawed toward the center
of the furnace.
Based on the results from Screening Level 1, the Screening Level 2 test matrix for
the reburn fuel injectors was developed. The mixing studies carried out in the
Screening Level 2 tests supported most of the conclusions from the initial smoke
flow visualization studies and permitted the selection of an optimum UFI jet
configuration. At this point it was determined that the penetration and
dispersion performance of the injectors was a function of the flow field into
which they were injected. Injectors that were firing into the lower velocity
segments along the rear wall of the furnace at a simulated velocity of 100 ft/sec
(30 m/sec) were capable of penetrating all the way to the division wall, while
those that injected into the higher velocity zones, associated with the two lower
cyclones, could not. It was found that an injection velocity of 300 ft/sec
(91 m/sec) was too high, resulting in jet impaction on the division wall almost
directly across from the point of injection. Reducing the recirculated flue gas
flow rate below 10% generally resulted in reduced levels of dispersion. It was
found that tilting the nozzles down improved the overall dispersion of the jet at
the outlet of the reburn zone, while tilting upward reduced the dispersion. The
configuration depicted in Figure 6 was chosen as the recommended UFI
configuration.
In determining the location of the air injector jets in the burnout zone, the
potential for injection sites was limited to the side walls and one centrally
3-33
-------
located on the front wall because of interferences on the front wall of the unit.
The choice of candidate AA injection configurations/locations was also guided by
the need to inject air into a zone which contained numerous internal boiler
obstructions in the form of cyclone burner hanger tubes. Figure 7 shows the
location of the AA injectors evaluated.
Each AA configuration was evaluated at 150 and 300 ft/sec (45.7 and 91.4 m/sec),
three tilts (-20°, 0", +20°), and configuration specific yaws ranging between plus
and minus 20°. During all AA injection tests the recommended UFI injection
configuration was installed on the model and was in service.
From the smoke flow visualization tests it was clear that the injection trends
were consistent with those of the UFI injection system tests; i.e., the jet
penetration and dispersion increased as the AA jets were tilted into the flow and
yawed for maximum dispersion.
Although air injection from Locations I, A, and B (Figure 7) provided reasonable
dispersion to the central regions of the burnout zone, the overall dispersion was
no better than the combination of A and B alone. Location I was eliminated
because the difficulty of installation would not be justified by its performance.
The recommended AA configurations (Locations A and B on each side of the unit) are
depicted in Figure 8. Each injector is divided into three compartments: the
upper two are designed to yaw from straight-ahead to an extreme of a 40° included
angle, and the bottom compartment can yaw plus or minus 10°. All compartments can
tilt in unison plus or minus 25° from horizontal.
Injection from Locations C and F resulted in jet attachment along the rear wall
with poor dispersion. Down tilting (into the bulk gas flow) from any given
location usually provided better mixing, but extreme downward tilting could cause
undesirable recirculation of air into the reburn zone. Injection from the
remainder of the locations (D, G, H, E) did not result in better dispersion and
had the further disadvantage of a lower residence time within which to attain
complete mixing.
Following selection of the recommended UFI configuration and the recommended AA
configuration, the combination of these integrated configurations was then
subjected to steady state gas mixing studies at 70% and 100% loads and to complete
three-dimensional flow velocity characterization tests. The results from these
tests were then used to assess the impact of the reburn system on original
baseline furnace aerodynamics.
IMPACTS OF REBURN SYSTEM ON BOILER PERFORMANCE
Boiler Thermal Performance
Since operation of a reburn system requires that the main fuel be reduced by up to
20% of its original value and that an equivalent amount of reburn fuel be injected
downstream with recirculated flue gas as a carrier, it is reasonable to expect
that some changes will occur in boiler gas and steam-side thermal performance. A
series of mathematical models were utilized in conjunction with operating data
supplied by Ohio Edison and some baseline information taken at the plant to
investigate:
o Furnace heat absorption profile
o Convection pass performance
o Boiler efficiency
3-34
-------
o Boiler circulation; departure from nucleate boiling.
The following steps were taken to evaluate boiler thermal performance:
o Calculate or obtain physical data for the boiler components (e.g.,
heating surfaces, tube diameters, tube arrangement, tube
material, free gas areas),
o Set up computer programs to calculate boiler efficiency,
cyclone/furnace performance, convection pass performance, and
air heater performance,
o Calibrate programs with baseline data. Determine required
calibration factors to match baseline data,
o Calculate baseline boiler performance,
o Calculate boiler performance with reburn.
o Compare boiler performance with reburn to baseline performance.
Baseline Boiler Performance
In order to determine the impact of the reburn system on boiler performance it was
first necessary to establish the current or baseline performance for reference
purposes. The basis for making this calculation was operating data supplied by
Ohio Edison together with some actual gas temperature data taken at the plant.
The first step in calculating baseline performance was to establish boiler
efficiency; this was done using the heat loss method. The calculated losses and
resultant efficiency are shown in Table I. Knowing the boiler efficiency and the
output of the unit allowed the coal input to be calculated. Based on the coal
analysis shown in Table II, combustion calculations were performed to establish
the gas and air weights. Sufficient information was now available from the above
gas and air weights which, in conjunction with physical data taken from boiler
drawings, allowed the convection pass program to be run. The convection pass
program was run backwards to determine: (1) furnace exit gas temperature, (2)
effective surfaces, and (3) intermediate steam and gas temperatures.
Furnace/cyclone performance calculations were performed next using C-E's lower
furnace program. Program inputs were varied until certain conditions were met
relative to a particular cyclone combustion efficiency, a particular gas
temperature actually measured in the unit, and the furnace outlet temperature that
was back-calculated by the convection pass program.
A heat absorption baseline profile was then generated using C-E's lower furnace
program and is shown by the solid line in Figure 9; conditions for this
calculation were 108 MW and 12% excess air. The heat absorption rates shown are
perimeter average rates. Where heat transfer surfaces are more or less uniformly
covered with refractory or ash deposits, the local rates should be reasonably
close to the average rates. Where tube sections are not covered with refractory
or ash deposits, local rates could be much higher than the average rates.2 The
calculated average rate for the cyclone is approximately 83,000 Btu/hr-ft , and
for the primary furnace and screen tubes the rates are approximately 54,000 and
44,000 Btu/hr-ft , respectively.
The total lower furnace heat absorption can be calculated by multiplying the heat
absorption rates from the profile by the EPRS (Effective Projected Radiant
Surface) and by correcting for casing heat loss. If the heat absorbed by the
evaporative surface in the convection pass is added in, the sum should equal the
heat absorbed by the fluid from the boiler inlet to the steam drum outlet; this
was checked and was found to be in agreement within 2%.
3-35
-------
Boiler Performance with Gas Reburn
Boiler performance with gas reburn was calculated in the normal forward design
mode; i.e., the programs were run in the following order: (1) cyclone/lower
furnace, (2) convection pass, (3) air heater, and (4) boiler efficiency. Process
flow quantities were determined from assumptions regarding reburn natural gas flow
rates, recirculated flue gas flow rates, additional air flow rates, and the exact
locations of the above injectors.
For predicting performance with reburn, the lower furnace program was run with the
firing rate in the cyclones reduced by 20%. Excess air was maintained at the same
level as that of the base case, 12%. Boundary surface conditions (waterwall
deposits) were varied in the secondary furnace: (1) in one case they were kept at
the same condition as the back-calculated value for the base case, and (2) in the
other case it was assumed that there would be about 30% less thermal resistance
because of the decreased amount of coal being fired and the expected lower gas
temperatures and changes in ash deposit characteristics.
The calculated heat absorption profile for the reburn case is shown with a dotted
line in Figure 9. The profile indicates a 10% reduction in overall waterwall heat
absorption with reburn where it is assumed that the thermal resistance of ash
deposits will remain the same as for the base case. If it is assumed that the
thermal resistance drops by about 30% in the secondary furnace, then the overall
waterwall heat absorption would be about 5% less with reburn than for the base
case .
Utilizing the output from the lower furnace program the convection pass program
was then run to show superheater and reheater performance. The effective heating
surfaces calculated from the base line data were input into the program. A
weighted fuel analysis (80% coal + 20% natural gas) was used to calculate changes
in gas properties. Primarily because of higher gas mass flows, due to the
recirculated flue gas, more heat is picked up in the convection pass with reburn
than for the base case. Slightly more heat is also picked up by the air heater
with reburn.
One consequence of picking up more heat in the convection pass is that superheater
spray water increases. However, the calculated increase in superheat spray is
within the capability of the unit even under a worst case scenario. Parametric
testing might well show that the maximum amount of recirculated flue gas, for
example, is not needed to achieve effective mixing; decreases in recirculated flue
gas flow rate will have a direct effect on heat transfer in the lower furnace and
convection pass, specifically more heat would then be picked up in the furnace and
less in the convection pass.
As shown earlier in Table I the boiler efficiency with natural gas reburn will be
about 0.5% less than the base case primarily due to greater moisture from fuel
losses; i.e., the higher hydrogen content of the natural gas results in more water
vapor being formed than when firing coal.
Two other boiler thermal performance related questions were addressed, namely the
effect of reburn on boiler circulation and the effect of reburn on departure from
nucleate boiling (DNB). DNB is defined as the occurrence of film boiling under
which the tube inside (water side) heat transfer coefficient drastically
deteriorates and tube overheating/failure usually results.
A computer program was used to perform boiler circulation calculations. In
general the program balances the pressure drops of the multiple parallel circuits
based on available thermal heads between the downcomers and risers. Both baseline
3-36
-------
and reburn cases were investigated. The results of this study have shown that
tube offsetting, for purposes of making openings for fuel and air injectors, would
have an insignificant effect on circuit flow and exit circulation ratio.
Relative to the question of DNB, evaluation of the division wall between the
primary and secondary furnaces was addressed since it sees the heaviest duty. The
criterion for evaluation of DNB is specification of the maximum allowable steam
quality which depends on pressure, heat flux, and mass flow of water/steam. To
avoid DNB the actual circuit steam quality must be kept less than the maximum
allowable steam quality and with adequate safety margin. Based on calculations it
has been determined that 56% steam quality (or less) ensures a DNB free condition.
The actual steam quality in the highest duty location, the division wall, is
calculated to be well under 10% with reburn. The occurrence of DNB is not seen as
a problem.
Impacts of Reburn on Boiler Tube Life
The presence of a slightly reducing condition in the reburn zone has raised
questions about the potential for tube corrosion when operating a boiler with a
reburn system. In response to this concern two things have been/will be done:
(1) previous relevant experience has been examined, and (2) precautionary measures
will be taken during all field testing to monitor tube wastage.
Mitsubishi Heavy Industries (MHI) has operated units in Japan which have their
version of reburn systems (MACT) installed on them (8). Experience gained from
these tangentially fired MACT-equipped units (all of which employ oil as the
reburn fuel) has shown no more wastage than normally expected after as many as 3
years of operation (Table III). One of the units co-fired oil and coal, another
fired oil and gas, and the third fired LNG as the main fuel. Though none of the
units was a cyclone fired boiler it seems reasonable to believe that the chemical
conditions in the reburn zone should be quite similar regardless of boiler type if
held to the same air/fuel stoichiometrics of 0.90 - 0.95. Additionally it seems
reasonable to assume that the use of natural gas as a reburn fuel rather than
sulfur-containing oil should produce results which are equal to (in the worst
case) or better than those achieved with oil.
Table IV shows data collected from long term corrosion testing on three coal fired
boilers where staged air, low NO systems have been in operation for about 2 years
(10). A combination of integral test panels and corrosion probes have produced
data which show wastage rates averaging between 1.5 and 3.9 mils (38 and 99 pm)
per year. These data indicate, as do those from MHI, that wastage rates produced
from low stoichiometric firing/reburn zones are not significantly different from
those typical of conventionally operated units.
A combination of corrosion probes and ultrasonic testing (U.T.) will be used
during testing of Ohio Edison's Niles Unit 1 boiler. Corrosion probes have the
advantage of being removable at any time without shutting down the boiler.
Ultrasonic testing has the advantage of being able to monitor a large number of
tubes without the need for penetrations in the boiler. C-E has extensive
experience with the use of corrosion probes and U.T. measurements as reliable
means for monitoring tube wastage (11, 12).
SUMMARY
Reburning is an in-furnace NO reduction technology that is uniquely well suited
for cyclone fired boilers. A reburn system has been designed for Ohio Edison's
Niles Unit 1 which meets the theoretical criteria for effective NO reduction.
x
3-37
-------
The number and location of reburn fuel and additional air injectors to achieve
effective mixing have been determined from extensive cold-flow modeling tests and
a further consideration to minimize the impact on boiler/plant modifications.
Mathematical modeling has determined that there will be a slight shift in heat
absorption patterns in the boiler when the reburn system is operated but that the
changes are well within the framework of boiler operational tolerances for
satisfactory and reliable boiler performance. Based on relevant experience it has
been determined that boiler tube life was not significantly changed by operating
either with reburn systems employing oil as the reburn fuel or with staged-air low
NO systems. A combination of corrosion probes and U.T. measurements will be
employed to carefully monitor tube wastage. Progress to date continues to
indicate that reburning should be a commercially viable in-furnace NO^ reduction
technology for cyclone fired boilers.
3-38
-------
References
1. R. I. Bruck, (1987), "Decline of Boreal Montane Forest Ecosystems in Central
Europe and Eastern North America - Links to Air Pollution and the Deposition
of Nitrogen Compounds," Proceedings: 1987 Joint Symposium on Stationary
Combustion NC>x Control, Volume 1, EPA-600/9-88-026a (NTIS PB89-139695).
2. C. Hakkarinen, (1987), "An Overview of Environmental Issues Related to
Nitrogen Oxides in the Atmosphere," Proceedings: 1987 Joint Symposium on
Stationary Combustion NO Control, Volume 1, EPA-600/9-88-026a
(NTIS PB89-139695). x
3. A. H. Johnson, T. G. Siccama, (1983), "Acid Deposition and Forest Decline,"
Environmental Science Technology, 17:294a-305a.
4. J. Kramlich, T. Lester, J. Wendt, (1987), "Mechanisms of Fixed Nitrogen
Reduction in Pulverized Coal Flames," Proceedings: 1987 Joint Symposium on
Stationary Combustion NO Control, Volume 2, EPA-600/9-88-026b
(NTIS PB89-139703). X
5. C. Kruger, G. Haussmann, S. Krewson, (1987), "The Interplay Between Chemistry
and Fluid Mechanics in the Oxidation of Fuel Nitrogen from Pulverized Coal, "
Proceedings: 1987 Joint Symposium on Stationary Combustion NO Control,
Volume 2, EPA-600/9-88-026b (NTIS PB89-139703). X
6. M. Toqan, J. Teare, J. Beer, L. Radak, A. Weir, (1987), "Reduction of NO^ by
Fuel Staging," Proceedings: 1987 Joint Symposium on Stationary Combustion
NO^ Control, Volume 2, EPA-600/9-88-026b (NTIS PB89-139703).
7. J. Freihaut, W. Proscia, D. Seery, (1987), "Fuel Bound Nitrogen Evolution
During the Devolatilization and Pyrolysis of Coals of Varying Rank,"
Proceedings: 1987 Joint Symposium on Stationary Combustion NO Control,
Volume 2, EPA-600/9-88-026b (NTIS PB89-139703). X
8. Y. Takahashi, et al. (1982), "Development of MACT' In-Furnace NO Removal
Process for Steam Generators," Proceedings of the 1982 Joint Symposium on
Stationary Combustion NO Control, Volume I, EPA-600/9-85-022a (NTIS
PB85- 235604) . X
9. D. D. Anderson, J. D. Bianca, J. G. McGowan, (1986), "Recent Developments in
Physical Flow Modeling of Utility Scale Furnaces," Industrial Combustion
Technologies, American Society for Metals.
10. J. M. Ferraro, P. S. Natanson, R. M. Vaccaro, (1982), "Long Term Optimum
Performance/Corrosion Tests of Combustion Modifications for Utility Boilers,"
Proceedings of the 1982 Joint Symposium on Stationary Combustion NO Control,
Volume I, EPA-600/9-85-022a (NTIS PB85-235604). X
11. A. L. Plumley, J. Jonakin, R. E. Vua, (1966), "Review Study of Fireside
Corrosion in Utility and Industrial Boilers," McMaster University, Hamilton,
Ontario, May.
12. C. F. Holtz, A. L. Plumley, (1979), "Improved Marine Boiler Reliability -
II," Soc. of Naval Architects & Marine Engineers Annual Meeting, New York,
NY', September.
3-3S
-------
Acknowledgements
In addition to those listed as coauthors of the paper, many others are responsible
for the progress reported in this paper. R. D. Stern of EPA/AEERL and S. J.
Wiersma of GRI have contributed to the paper directly and to the technical
progress represented to date in this paper. J. H. Pohl of ESA, S. L. Johnson of
PSI, and Y. Nakajima of MHI have directly contributed to the technical progress
from which this paper was derived. R. D. Lewis, R. C. LaFlesh, J. L. Marion, A.
L. Plumley, A. L. Gazsi, and C. Y. Sun, all of C-E, have made significant
contributions to the reburn program in areas which are represented in this paper.
J. McGowan, Prof, of Mechanical Engineering at the University of Massachusetts and
a consultant to C-E, has contributed significantly in the area of flow modeling.
The authors gratefully acknowledge the contributions of each of these people.
3-40
-------
TASK NO TASK TITLE I 1987 I 1988 I 1989 i 1990 I 1"1
IAS* IMU. I AoN IIILt I I I I I I I I I I I M I I I I I I I I I I I M I l I l I l I I l l l l l I l I I l l l l l l
1 TECHNOLOGY REVIEW
2 BASELINE CHARACTERIZATION TESTS
3 REBURN SYSTEM DESIGN
4 FABRICATE/INSTALL REBURN SYSTEM
5 PARAMETRIC GAS REBURN TEST
6 LONG TERM PERFORMANCE TEST
CO
^ 7 APPLICATION STUDIES
8 SITE RESTORATION (OPTIONAL)
9 PROJECT MANAGEMENT/REPORTING
Figure 1
PRELIMINARY OVERALL PROJECT SCOPE AND SCHEDULE
* THE OPEN AREA IN THE BAR REPRESENTS THE ACTUAL 30
DAY OUTAGE WHEN THE UNIT IS OFF LINE
-------
Additional — ¦
Air Injectors
Upper Fuel
Injectors ~~
Cyclones
Burnout
Zone
Reburn
Zone
Main
Combustion'
Zone
^ Zone
>
Figure 2
SCHEMATIC OF REBURNING PROCESS
3-42
-------
Figure 3
PHOTOGRAPH OF ONE-NINTH SCALE FLOW MODEL
3-43
-------
Figure 4
TEST PLANE LOCATIONS
3-44
-------
Figure 5
NORMALIZED VELOCITY PROFILE AT ENTRANCE
TO THE REBURN ZONE
3-45
-------
±10° Yaw
Fixed
Yaw
±10° Yaw
Pant-Leg
Fixed Yaw
Fixed
Yaw
±10" Yaw
Pant-Leg
Fixed Yaw
Fixed
Yaw
±10° Yaw
Pant-Leg
Fixed Yaw
Fixed
Yaw
±10° Yaw
±10° Yaw
Fixed
Yaw
±10° Yaw
Directed
20« off Walls
Directed
20° Off Walls
All Compartments Can Tilt in Unison ± 25° from Horizontal
Figure 6
RECOMMENDED UFI ARRANGEMENT
3-46
-------
CYCLONE
OUTLETS
Figure 7
AA INJECTOR LOCATIONS EVALUATED
3-47
-------
0° to -200
00 to +20°
0» to -200
00 to +20°
Yaw
Yaw
Yaw
Yaw
±100 Yaw
±10° Yaw
All Compartments Can Tilt In Unison ± 25° from Horizontal
Figure 8
RECOMMENDED AA ARRANGEMENT
3-48
-------
100,000 r
CO
I
ft*
co
90,000
80,000
70,000
« 60,000
u.
tr
z
jg 50,000
<
o
40,000
30,000
20,000
10,000
Cyclones
_L
BASELINE
REBURN
GC
d
00
Primary Furnace
X
L J
K/A X = 17
Scrns
I
X
Secondary Furnace
» '
X
J
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Slice No.
Figure 9
FURNACE HEAT ABSORPTION RATE
OHIO EDISON - 108 MW
(12% XA)
STRIP AVERAGE
-------
Table I
CALCULATED BOILER EFFICIENCY
108 MW
Dry Gas Loss
Moisture from Fuel Loss
Moisture from Air Loss
Radiation Loss
Ash Pit Loss
Miscellaneous
Total Losses
Boiler Efficiency
Stack Temp. °F
aseline Reburn
Coal 80% Coal/20% N.G.
2.84 2.63
4.47 5.32
0.07 0.06
0.24 0.25
0.74 0.62
0.50 0.50
8.86 9.38
91.14 90.62
267 269
Table II
COAL ANALYSIS - % BY WEIGHT
Ultimate
Proximate
Moisture
7.45
Moisture
7.45
Hydrogen
4.48
Volatile Matter
35.05
Carbon
63.00
Fixed Carbon
44.14
Sulfur
3.26
Ash
13.36
Nitrogen
1.12
Oxygen
7.33
Total
100.00
Ash
13.36
Total 100.00 HHV (Btu/lb) 11559
3-50
-------
Table III
EFFECT OF REBURN ON TUBE WASTAGE RATES (MHI)
Plant Name
Joban Joint EPCo.
Nakosa P/S Unit 8
Yuka Yokkaichi
Kyushu EPCo.
Shinkokara P/S Unit 5
Main Boiler Fuel
Reburn Fuel
Date of
Commercial Operation
Boiler Type
Max. Evap. (lb x 10~6hr)
Operating Hours
Wall Thickness Loss (in)
Overfire Air
Mid Burner (Front)
(Right)
(Right)
Below Burner
Oil/Coal (0.3% S)
Oil (1.8% S)
9-9-1983
Supercritical
4.29
25,912
0.004
<0.001
0.003
+ 0.001
Oil, Gas
Oil (2.6% S)
6-25-1982
Subcritical
0.62
27,320
LNG
7-1-1983
Supercritical
4.3
7,389
+ 0.003
+ 0.004
0.005 0.004 0 — —
_ _ — 0.008 0.008
0.008
0
0
0.004
0
+ 0.004
0
+ 0.016 +0.008 0.001 — — _ _ _ _
— No Measurement
0 No Change in Wall Thickness
Boiler
Table IV
LONG TERM CORROSION TESTING
Mill Creek No. 3 Conesville No. 5 Hunter No. 2
On-Line
Exposure Time
2 Years
(15,000 Hours)
2 Years
(12,000 Hours)
9 Ypatq
(~ 18,000 Hours)
Elevation/
Location
Wall Loss,mils*
Just Below Nose
~ 15ft Below Nose
4.0
3.4
0.8
2.0
3.1
-0.5**
I™?' Middle
Zone Bottom
4.4
3.7
3.5
2.1
1.8
2.9
2.5
1.7
2.1
Hopper
Elevation
4.8
1.5
0.8
Average
3.9
1.9
1.53
*One Mil = 0.001 Inch
(To Find Corrosion Rate (Mils/Yr), Divide by Exposure Time).
**Gain Was Due to Several Tubes Having High Thickness Values During
the Second Series of Measurements (No Reason Could Be Found to
Discount the Data).
3-51
-------
(intentionally Blank)
3-52
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
DESIGN METHODS FOR LOW-NOx RETROFITS
OF
PULVERIZED COAL FIRED UTILITY BOILERS
S. Morita
K. Kiyama
T. Jimbo
K. Hodozuka
Thermal Power Design Department
Babcock-Hitachi K.K. Kure, Hiroshima, 737, Japan
K. Mine
Thermal Power Engineering Department
Babcock-Hitachi K.K. Tokyo, 100, Japan
ABSTRACT
Babcock-Hitachi K.K. has partially reconstructed an existing pulverized coal fired
350 MWe boiler by changing conventional low-NOx Dual Reyister type burners with
Hitachi-NOx Reduction (HT-NR) type burners. The HT-NR burner was developed, as pre-
sented in the 1985 EPRI/EPA Joint Symposium on Stationary Combustion NOx Control,
to meet the concept of promoting "In-Flame" NOx Reduction. And two 200 MWe retro-
fitted boilers have been changed from conventional low-NOx Dual Register burners to
HT-NR burners, as presented in the 1987 EPRI/EPA Joint Symposium on Stationary com-
bustion NOx Control, which have achieved further NOx reduction without increased
carbon in fly ash and also without an increase of excess furnace air. This 350 MWe
retrofit has marked the second trial whereby a HT-NR technique has been carried out
for a retrofit on a full-scale utility coal-fired boiler, and a reasonable NOx
figure was confirmed. This paper introduces the design concept for several kinds
of low-NOx retrofits with our recent 90 x 106 Btu/h pilot test results and the
above field experiences.
Preceding page blank 3-53
-------
INTRODUCTION
At present the low-NOx combustion method employing fuel-rich reactions, such as
air-staging low-NOx burners or _Two S,tage Combustion (TSC), is the most practical
method in reducing fuel NOx from pulverized coal fired utility boilers. The
following items will be discussed to evaluate the NOx, combustion efficiency,
excess air and boiler performance etc.
Furnace effect with or without Two Stage Combustion
Burner effect promoting In-Flame NOx Reduction
Fuel property Volatile hydrocarbon effect to decompose
Nitrogen Oxides
DESIGN CONCEPT
NOx Performance
Figure 1 shows the behavior of NOx (mainly NO) formation in the boiler furnace.
Burner Zone NOx Formation. Primary NOx formation N0x,o may be described by the
equation 1.
NOx,o = krexp [(BHR)( |J- )al] SR > SRo
krexp [(BHR) ( )a2] SR < SRo (1)
0 (a^ > ag > 1, SRo = 1.0 ~ 1.2)
The Burner Zone Heat Release Rate (BHR), has direct effect on the average flame
temperature. Thermal NOx formation, by Zeldovich's mechanism, strongly depends on
the oxidizing flame temperature shown in Figure 2. Therefore, in general, the
burners should be optimally arranged to decrease BHR if the furnace does not have
Two Stage Combustion equipment. (SR Burners zone Stoichiometric Ratio) In
the equation 1, factor Iq depends on the content of fuel nitrogen in coal (N^af),
and depends on the reduction efficiency of the total fixed nitrogens (= NO + NH3 +
HCN) in the burner flame. That is,
kl = 5l(Ndaf)6l'[1_nN]
(0 < nN < 1), ("daf" dry, ash free basis)
It should be strongly remarked that the reduction efficiency of the total fixed
¦nitrogens (n^) is promoted by the hydrocarbon intermediates evaporated from coal
particles, and accelerated by increasing the temperature of the fuel-rich flames.
Rapid ignition near the burner throat and finer coal particle distribution can pro-
mote an increase in the temperature at the initial stages of combustion under
enhanced fuel-rich conditions.
3-54
-------
n / FR DP x ,oX
"N " " eXP Y1 75^ " T2 W0 * T3» <3>
(R> Y^» ^2' Y3 > ^ — Burner Performance Factor)
As described in equation 3, FR (= [Fixed Carbon]/[Volati 1 e Matter]) and Dp (average
coal particle diameter) are the important fuel factors for low-NOx retrofits.
Finer pulverizing, however, would require larger pulverizers or an increase in
their power consumption.
The HT-NR burner, shown in Figure 3, is one of the modern low-NOx burners proven in
field operation to maximize nNs keeping total combustion efficiency. A valuable
feature of the HT-NR burner is its ability to produce rapid ignition and con-
sequently high temperature reducing flames with Flame Stabilizing Ceramic Rings
fitted to the edge of the fuel nozzles, as a result of which reducing agents are
generated effectively. And the External Air Guide Sleeve is fitted with a collar
which contributes to maintaining the external (outer) air apart from the reducing
combustion zone.
NOx Decay in the Post Flame Zone. NOx can be reduced in the space between the
burner zone and after-air (overfire air) injection ports, if the furnace is adopted
with a Two Stage Combustion. This post-flame NOx decay phenomena will be promoted
by enough residence time of combustion gas from the burner zone to after-air
(overfire air) injection ports.
In the Figure 1, therefore, NOx behavior may be described by the equation 4.
NOx* = N0x,o[l - nj] (4)
depends on the Combustion Gas Residence Jime, RT^, from burner zone to after-air
(overfire air) injection ports. Therefore,
RT i
Hi = 1 - exp (- a3 ) - (5)
RT0
(a3 > 0)
Modern low-NOx furnaces can have enough RT^. But on many existing old style low-
NOx furnaces, proper attention to RT^ was not paid during the design/construction
stage.
NOx Reformation by After-Air (Overfire Air) Injection. If oxidizable nitrogen com-
pounds such as NH3 or HCN remain before after-air (overfire air) injection, a great
deal of these species may be converted to NOx, and a part of nitrogen in chars
(Char-N) will also be converted to NOx during heterogeneous combustion of the
residual coal chars by after-air (overfire air) injection.
Nitrogen in raw coals was effectively evaporated to gaseous substances by rapid
3-55
-------
ignition and increasing the temperature of the fuel-rich flame separated from oxi-
dizing flame. Consequently, the total fixed nitrogen (= NO + NH3 + HCN) was mini-
mized before after-air (overfire air) injection.
Given this aspect, the final NOx emission (NOx, final) from the boiler furnace is
described by the equation 6, 7 and 8.
NOx,final = NOx* + Reformation (6)
Reformation = l^exp [- ( —— )a^ • ( )a^] (7)
SR0 R'o
k2 = l2 (Ndaf)&2 CI - nN] - (8)
(a4» a5 > °)
So, k.j (i = 1,2) is an important fuel factor in the Fuel NOx reduction efficiency
for low-NOx techniques using fuel-rich reactions, regardless of whether the boiler
furnace has Two Stage Combustion equipment or not.
Combustion Efficiency
Delayed mixing of air-coals could decrease the combustion efficiency in many cases.
Figure 4 shows the multi-stage model of coal combustion in the furnace for two
cases. Curve A and B are calculated for single stage combustion (without after-air
ports) and curve A1 and B' are calculated for Two Stage Combustion (burner zone
stoichiometric ratio = 0.9 ~ 1.0). Coal A has a higher Volatile Matter and coal B
has a lower Volatile Matter. Higher volatile coal (generally FR < 1.2) has a rela-
tively large combustion rate constant of the heterogeneous char burning in the com-
plete combustion zone (from after-air injection ports to furnace exit). Therefore
Two Stage Combustion retrofitted boilers (from Single Stage Combustion to Two Stage
Combustion) can make less effect on combustion efficiency, if higher volatile coals
are fired. On the other hand, the Two Stage Combustion retrofitted boilers fired
lower volatile coal (generally FR > 1.5, especially FR > 2.2) will make a more or
lesser decrease of combustion efficiency.
Also conventional delayed air-fuel mixing type low-NOx burners could affect the
unburned carbon loss. Therefore the "trade-off" relationship of NOx and unburned
carbon loss could be unavoidable for lower volatile coal firing without finer coal
particle modification and/or without increased excess air. Only increasing the
temperature of fuel-rich flames, separated from the oxidizing zone in the burner
flame, may cast off the skin of the "trade-off" defect.
CASE STUDIES FOR LOW-NOx RETROFITS
Figure 5 shows several cases of modern low-NOx retrofits of the existing low-NOx
furnace.
3-56
-------
• Case I (A + B ) : only replace of conventional low-NOx burners
with modern low-NOx burners (HT-NRs)
o Case II ( A -~ C ) : only addition of after-air (overfire air)
ports
• Caselll ( B ->• D ) : B + after-air ports
• Case IV ( C > D ) : C + replace of conventional low-NOx burnrers
with modern low-NOx burners (HT-NRs)
o Case V ( A D ) : A + replacement of conventional low-NOx
burners and addition of after-air ports
Figure 7 shows the pilot evaluation of the above retrofit cases, all data was
obtained by Babcock-Hitachi K.K.'s new 90 x 10^ Btu/h test facility (see Figure 6).
Three kinds of coals were tested in each case. The following results were
obtained as shown in Figure 7.
® Not only Two Stage Combustion but also burner replacing achieved
remarkable NOx reduction.
• The HT-NR effect on NOx reduction, depending on volatile content of
coal, causes no harmful effects on combustion efficiency under the
same burner stoichiometric ratio, excess air and pulverized coal
fineness.
Therefore net NOx reduction efficiency may be described in Table 1.
FIELD EXPERIENCES
Background
As previously reported in the 1987 Symposium in New Orleans, two identical 200 MWe
P.C. fired boilers have been retrofitted by replacing conventional Dual Register
type burners with HT-NR type burners. This retrofitting corresponds to the above-
mentioned case IV ( C + D ), and 35 ~ 45 % of NOx reduction efficiency by the
retrofitting was confirmed for coals which had 31 ~ 52 % Volatile Matter (dry, ash
free basis) under the same condition of the burner zone stoichiometric ratio,
furnace excess O2 and pulverized coal fineness before and after retrofitting. And
noteworthy the carbon in fly ash was reduced slightly after this HT-NRs retrofit-
ting. Minimum NOx emissions,70 ~ 75 ppm (corrected to 6 % O2) leaving economizer,
were confirmed when firing a higher volatile coal (Volatile Matter (d.a.f.) = 52 %)
under very low carbon in fly ash (< 3 %).
The second trial of the retrofitting case IV ( C + D ) was executed at the 350
MWe P.C. fired boiler in mid 1987. This boiler furnace has a stricter physical
volumetric space both of RT^ and combustion completion space from the after-air
3-57
-------
injection ports to the furnace exit. Because the original design was centeralized
around the conventional Two Stage Combustion only for a Japanese domestic coal and
it had only been operated for higher volatile (Volatile Matter (d.a.f.) = 50 %)
Japanese coals. For recent economical reasons, however, coal must be blended with
imported coals (Volatile Matter (d.a.f.) = 30 ~ 40 %), and NOx emission from the
boiler furnace has to be less than 180 ppm (corrected to 6 % O2) in commercial
operation required under the conditions of the same excess air (or excess O2) and
allowable unburned carbon loasses.
Specification and Retrofit Items
Figure 8 shows the boiler furnace configuration.
The entire furnace enclosure is formed with membrane walls and these are 3
horizontal rows of 6 burners in the front wall and 2 horizontal rows of 6 burners
in the rear wall, total 30 burners. The burners, before-retrofit, were conven-
tional low-NOx Dual Register type, then partially reconstructed to Hitachi-NR
(HT-NR) type. The burner throat size was enlarged to meet the change of coal pro-
perties during this retrofit. Details of this retrofit are shown in Figure 9 and
Table 2.
About 2 months outage was allowed for the retrofit.
Operating Results
The project of retrofitting was conducted in three stages : before-retrofit com-
bustion testing, replacing, after-retrofit combustion testing using the same coal
property as before-retofit1s.
Table 3 shows the coal specifications during the combustion tests.
After the combustion test, this unit was put into service with the same low-NOx
performance as the after-retrofit combustion testing.
Combustion Performance. Figure 10 shows NOx emissions and carbon in ash from the
350 MWe unit before and after retrofitting. The coal property (a blend of Japanese
coal and Canadian coal) is the same as before and after retrofit. Average value of
FR (= [Fixed carbon]/[Volati1e Matter]) was 1.3. Some Mill Balls were maintained
at this retrofit. So the pulverized coal fineness, by percentage of weight through
a 200 mesh sieve, was 70 % before retrofit and changed to 65 ~ 70 % after retrofit
in this combustion test.
About 30 % of NOx reduction efficiency (nN0x) was confirmed under the condition of
the same carbon in fly ash before and after retrofit. And there was no need to
increase excess air to maintain the same carbon in fly ash as before the retrofit.
Figure 11 shows the NOx reduction efficiency (nNOx) in this retrofit as compared
with the 200 MWe retrofit. A reasonable NOx reduction efficiency rate was
confirmed as shown here.
3-58
-------
Figure 12 shows tuned NOx emission performance levels at each boiler load (turbine
output, MWe). The HT-NR burner promoted the flame stabilization and made a more
visible "clear-cut" flame than that of before retrofit. This confirmed complete
flame stability on the Flame Stabilizing Ceramic Rings at 79 MWe (22.5 % of boiler
load) - 350 MWe (100 % of boiler load) without oil supporting. Output voltage of
flame signal by flame detectors (Hitachi Dual-Signal type) at each burner was com-
pared before retrofit with after retrofit, as shown Figure 13.
Boiler Performance. Figure 14 shows the results of the steam temperatures before
and after retrofit. There was no significant difference before retrofit and after
retrofit. And this HT-NR retrofit produced no undersirable effects on the boiler
performance concerning the steam generation.
Slagging and Sootblowing. The sootblowing patterns and operating intervals were
not changed, but the ash deposition around the burner throat was remarkably reduced
by this HT-NR retrofit. This result might be of great use in measures against
fire-side wall corrosion of furnace, that have been through long term inspection
stages since the HT-NR retrofit.
CONCLUSION
The partial reconstruction of conventional low-NOx burners by using the Hitachi-
NR burner's design concept has been focused upon throughout this paper by present-
ing the 350 MWe retrofitted boiler in a series of 200 MWe retrofitted boiler. As a
result, the low-NOx capability of the HT-NR burner was brought to light. Babcock-
Hitachi K.K. confirms that these retrofitted applications of HT-NR burners will be
greatly adaptable to many existing P.C. fired boilers with or without Two Stage
Combustion and, will certainly be of greater use and more beneficial to modern
low-NOx boilers.
3-59
-------
ACKNOWLEDGMENT
The authors would like to thank their customers for their kind support in granting
BHK the opportunities to install new technological innovations in commercial util-
ity boilers. Additionally we would like to acknowledge the assistance by Messrs.
Nakashita, Kuramashi, Fukayama, Maeda and Inada of BHK in preparing the paper.
REFERENCES
1. Narita, T., Morita, S., et al. Development of the low-NOx burner for the
pulverized-coal-fired In-Furnace NOx reduction system. Paper presented at 1985
Joint Symposium on Stationary Combustion NOx Control, Boston Massachusetts, May
6-9, 1985.
2. Narita, T., Morita, S., et al. Operating experiences of coal fired utility
boilers using Hitachi NOx Reduction Burners. Paper presented at 1987 Joint
Symposium on Stationary Combustion NOx Control, New Orleans, March 23-26, 1987.
3. Azuhata, S., Morita, S., et al. A study of gas composition profiles for low NOx
pulverized coal combustion and burner scale-up. Twenty-first Symposium
(International) on Combustion, the Combustion Institute, 1986/pp. 1199-1206.
4. Glass, J.W. and Wendt, J.O.L. Mechanisms governing the destruction of
nitrogeneous species during the fuel rich combustion of pulverized coal. 19th
Symposium (International) on Combustion, pp. 1243-1251, 1982.
5. Muzio, L.J. and Arand, J.K. Gas phase decomposition of nitric oxide in combus-
tion products. 16th Symposium (International) on Combustion, pp. 199-208, 1976.
6. Wendt, J.O.L., Pershing, D.W., Lee, J.W., and Glass, J.W. Pulverized combus-
tion : NOx formation mechanisms under fuel rich and staged combustion condi-
tions. 17th Symposium (International) on Combustion, pp. 77-87, 1979.
3-60
-------
F'ce
AAPs
BNRs
E
N0X•f1nal
NOx
Re-formation
NOx ,0
RT,
Fig. 1 Behavior of NOx formation in furnace
NOx,0 ={krexp[(BHR)(J{j- ) al]
— SR > SR„
krexp[(BHR)(|£- ) a2]
0
— SR > SRQ
(SRo a 1.0 ~ 1.2)
(al > a2)
t
SRo
Stoichiometric Ratio, SR (-)
Fig. 2 BHR effect on NOx,o
3-61
-------
HIGH SWIRL REGISTER
GUIDE SLEEVE
FLAME
STABILIZING
RING
Fig. 3 HT-NR Burner Structure
Difference of Combustion
Efficiency between "Single"
and "Two" Stage Combustion
Curve
Coal
TSC
— A ~
FR = 1
Wi thout
— A' —
Wi th
1
1
eo
1
a
FR - 2
Wi thout
— B' —
With
Unburned Carbon
Fig. 4 Behavior of Unburned Carbon
3-62
-------
Single Stage
Combusti on
Two Stage
Combusti on
in
0)
c
C_
0) 3
fO C CO
P2
O 3 U
CO O)
4-)
C X i/»
OJ o •«-
> z cr>
C 1 O)
¦
¦
< >
< >
* >.
: CE>!
5 5
< >
< >
J
m
o o
ro
3
V
% ^
t- )
01 t-
c 0>
i- c
3 L.
C CO 3
L. CO
o> x
-o o cc
o z z
s: ¦ i
~
¦C >
[¦>,
¦ m
R
t %
r ;
c >
p
"
o zc
—1 '
V
c >
V
(Furnace has the same volume as each case.)
Fig. 5 Cases of Low-NOx Retrofit
Furnace Type
Water Jacket, Vertical
Heat Input
90 x 106 Btu/h
F'ce Dimension
Ui dth
Depth
Hei ght
4.5 m
3.3 m
20 m
Burner
Arrangement
2 Burners x 3 Rows, Opposed
Total 12 Burners
AAP Arrangement
2 Ports x 3 Rows, Opposed
Total 12 Ports
Pul veri zer
MPS Rol 1 er type Mi 11 x 1
Tube Mill x 1
Fig. 6 Test Facility
3-63
-------
— 6
.£
I/)
<
I 2
l_
rtJ
° 0
400
8 300
E
cl
Q.
x 200
100
o Dual Register Burner
• : HT-NR Burner
Coal-A
(FR = 1.21, N = 0.94 %)
Fineness: 200 mesh pass
77 - 81 %
Ex. 02 : s 3.5 %
\0
6'"
i
CASE IVi
i
W-
0.8 0.9
1.0 1.1
1.2
Stoicliiometric Ratio (-)
(Burner Zone)
o
JD
U
—
0.8 0.9 1.0 1.1 1.2
Stoichiometric Ratio (-)
(Burner Zone)
-------
^ uaai
aMCOCK
-------
Coal + Air -~
9 Details of Burner Retrofit
-------
Load : 350 MWe
Coal : Blended
/2
I /
5(3.n
3(3.6) 1 (M)
7(3*5.)
{Eco. 0?,%)
0.85
0.90
0.95
1.00
Stoichiometric Ratio {-)
(Burner Zone)
Fig. 10 NOx Emission and Carbon in Ash
3-67
-------
Fineness (%, 200 mesh pass)
Fig. 11 NOx Reduction Efficiency Relative to
Pulverized Coal Fineness
Fig. 12 NOx Emission and Carbon in Ash on
each Boiler Load
-------
(FRONT WALL) (REAR WALL) 1
No. 1 Mi 11 (Top row)
No. 4 Mi 11 (Top row)
1_.
r
r No. 2 Mill (Middle row)
_ No. 3 Mill (Bottom row)
M
^ 6
oj 4
2 2
° 8
S 6
Sr 4
3
o 2
r No. 5 Mill (Middle row)
#1 #2 #3
#5 #6 BNR
#1 #2 #3 #4 #5 #6 BNR
a) 350 MWe
5 Mi 1 1 s Oper.
No. 1 Mill (Top row)
r No. 2 Mill (Middle row)
J
£ 6
> 4
U 2
No. 4 Mill (Top row)
J
#1 #2
#4 #5 #6 BNR
#1 #2 #3 #4 #5 #6 BNR
b) 175 MWe
3 Mi 11s Oper.
"1
L.
Q : Before Retrofit
After Retrofit
ANN. Point
Output Voltage during Light-Oil Firing
Fig. 13 Output Voltage of Flame Detectors
3-69
-------
Superheater output
<9 m-°-
Reheater output
g Q—O
g—U <
O
Before Retrofit
After Retrofit
100 150 200 250 300 350
Turbine Output (MWe)
Fig. 14 Boiler Performance
3-70
-------
Table 1
NOx Reduction Efficiency nNQx (Coal-B)
^""^^Correction
basis
Case
Same excess air
Same UBC
I
35 ~ 40
35 - 40
II
35 ~ 40
< 10
III
30 ~ 40
< 10
IV
30 - 45
30 ~ 45
V
50 ~ 65
35 ~ 45
n.,„ - NOx (before) - NOx (after) „ lnn i*\
N0* NOx (before) x 100 i%>
Fineness : 200 mesh pass » 80 (i)
Table 2
Retrofit Items
— ______
"before"
"after"
Turbine Output
350 MW
The same
Combustion Method
Two Stage Combustion
The same
Burner
Type
Conventional Low-NQx
Burners
(Dual Register type)
Modern
Low-NOx Burners
(Hitachi-NR type)
See Fig. 9
Arrangement
front: 6 Bnrs x 3 Rows
Rear : 6 Bnrs x 2 Rows
The same
(no modification)
After-
Air Ports
Type
Circular type
Dual flow type
Arrangement
6 Ports x 1 Row,Opposed
The same
3-71
-------
Table 3
Coal Specification during the tests
Base
Unit
(Blended Coal)
GCV
A.D
cal/g
6,220
Moisture
W.B
%
6.6
VM
D.B
%
36.6
FC
D.B
%
47.7
Ash
D.B
%
15.7
FR (= FC/VM)
-
-
1.3
HGI
-
-
47
C
D.B
%
66.45
H
D.B
%
5.27
0
D.B
%
10.69
N
D.B
%
1.05
S (Total)
D.B
%
0.37
S (Comb.)
D.B
%
0.21
S (Uncomb.)
D.B
%
0.16
3-72
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
NEW APPROACH TO NOx CONTROL
OPTIMIZATION OF NOx AND UNBURNT CARBON LOSSES
M. Kinoshita
Mitsubishi Heavy Industries, Ltd.
Assistant Manager
Combustion Engineering Section
Nagasaki Shipyard & Machinery Works
1-1, Akunoura-machi, Nagasaki, Japan
T. Kawamura
Mitsubishi Heavy Industries, Ltd.
General Manager
Boiler Engineering Department
Power Systems Engineering Division
Power Systems Headquarters
Shin-Tamachi Bldg. 34-6, Shiba 5-Chome
Minato-ku, Tokyo, Japan
S. Kaneko
Mitsubishi Heavy Industries, Ltd.
Acting Manager
Power Plant Engineering Department
Nagasaki Shipyard & Machinery Works
1-1, Akunoura-machi, Nagasaki, Japan
M. Sakai
Mitsubishi Heavy Industries, Ltd.
Manager
Power Systems Engineering Research & Promotion Laboratory
Nagasaki Research & Development Center
5-717-1 Fukahori-machi, Nagasaki, Japan
ABSTRACT
Mitsubishi Heavy Industries, Ltd. has been engaged in the research and development
of new technology concerning the prevention of air pollution. In the field of low
NOx combustion technology, we have already developed and utilized many counter-
measures, for example, super-low NOx PM burners, MACT system and Advanced MACT
system which have already been introduced at the former symposia.
In order to offer the low cost NOx control option, MHI has developed and
commercialized the method of optimizing the emission of NOx and unburnt carbon
losses by controlling the fineness of pulverized coal.
This paper introduces the outline and operating experiences of the Mitsubishi low
NOx PM burner (including MACT) combined with "MRS Pulverizer" which can control the
fineness distribution of pulverized coal to achieve the most efficient low NOx
combustion at minimum operating cost.
3-73
-------
INTRODUCTION
The Japanese government regulation on NOx emission from thermal power plant has
been amended four times since it was first introduced in 1973 and has become
severer each time as shown in Figure 1. Especially in the last decade, the
regulation on pulverized coal firing has been strengthened, and a low NOx level
close to that of oil firing is now being required.
To control the NOx emission, MHI continuously developed and put into practical use
the new low NOx firing technologies one after another as shown in Figure 2. But to
achieve low NOx level with low unburnt carbon losses, relatively high initial and
operating costs were required.
Low NOx pulverized coal firing technology in the 19901s will be required to
accomplish the followings from the view-point of environment preservation and
resources and energy saving.
Operating economy by optimization of NOx and unburnt carbon losses.
Establishment of low Nox firing technology for a wide variety of solid
fuels such as low volatile coals and petroleum cokes.
Realization of the above technologies at low costs.
MHI succeeded in comnercializing the Advanced MACT system as the first step of new
system development meeting the above need for low costs and presented it in the
1987 Synposium. As the second step, we aimed at optimizing the fineness of
pulverized coal particles and succeeded in developing a new pulverizing and firing
system, which consists of Mitsubishi low NOx firing technology and hiqh-performance
"MRS Pulverizer".
The new system is the most efficient, economical and flexible system that meets the
future requirement for even lower NOx and unburnt carbon losses. It has already
been adopted to twenty-eight boilers and has been highly rated for its superior
performance.
EFFECT OF PULVERIZED COAL PARTICLE SIZE ON COMBUSTION
COMBUSTION MECHANISM OF PULVERIZED COAL
Figure 3 shows a flame model of pulverized coal burner. Pulverized coal introduced
into the furnace "from the coal compartment with primary air is heated by the flame
and the radiant heat from high-temperature slags on furnace walls, being ignited and
burnt to form the primary combustion zone. The primary combustion zone is a
combustion zone of volatile matter in coal where QU, H2 and CO volatilized from
coal particles mix with oxygen of the primary air and form flames around the
particles. The secondary combustion zone is mainly a combustion zone of char where
the unburnt gas and char flowing from the primary combustion zone mix and burn with
the secondary air blown in from the auxiliary air compartment.
Combustion of char is achieved by the diffusion of oxygen through the surface of
char or through pores in char, and its rate of combustion is very low as compared
with that of volatile matter. Therefore, the combustion time of char accounts for
80 ^ 90% of the total combustion time of coal as also can been seen in combustion
model of coal particle (Figure 4).
3-74
-------
The points to be noted for pulverized coal combustion in the flame model of Figure
3 are ignition, burn-out and NOx emission characteristics. These characteristics
are all affected by coal particle size because the speed of devolatilization and
diffusion of 02 into char are both affected by the surface area of the coal
particle.
COAL PARTICLE SIZE AND BURN OUT TIME
Combustion of pulverized coal particles in a furnace can be expressed by the
following equation.
_d_
de
i
Yr t
<4- DP3)
__-L + JL
K Kc Kf
-TrDp2-K-P
(1)
(2)
where,
K : Overall combustion rate coefficient
K.r : Combustion rate coefficient when oxygen diffusion rate in gas boundary
layer is dominant
K : Combustion rate coefficient when chemical reaction rate on particle
surface is dominant
P : Partial pressure of oxygen
Dp : Diameter of coal particle
Specific gravity of coal particle
Combustion time
Overall combustion rate coefficient K varies greatly with coal properties as well
as particle size and combustion zone gas temperature. Therefore, to keep unburnt
carbon losses low, it is necessary to optimize furnace design and pulverized coal
equipment to suit the characteristics of K of pulverized coal particles, coal
particle diameter Dp (fineness), residence time in furnace 0, gas temperature
distribution and oxygen concentration distribution in the furnace.
Figure 5 shows the unburnt carbon ratio calculated by the equations (1) and (2)
above and using the calculation results of locus of flame axis and gas temperature
and oxygen concentration distribution obtained by in-furnace fluid heat transfer
simulation. At a height about 5 meters above the burner toward furnace outlet,
combustion progresses rapidly, but becomes slow around 10 meters, and hardly
progresses over"20 meters because of decrease in gas temperature. This indicates
thatreduction in coal particle diameter by improving the fineness is more
effective to increase combustion efficiency of pulverized coal than lengthening
residence time in the furnace.
COAL PARTICLE SIZE VS. NOx AND UNBURNT CARBON
In order to confirm the effect of coal particle size on NOx and unburnt carbon
losses, combustion test has been conducted in MHI Laboratory. Pulverized coal were
classified into seven groups of different particle sizes and fired separately in 2
kg/h basic combustion test furnace shown in Figure 6.
3-75
-------
The test results are shown in Figures 7 to 9 and the properties of the coals used
for the tests are shown in Table 1. As seen from Figures 7 and 8, NOx emissions
increase with an increase in particle size and unburnt carbon considerably
increases when particle size becomes larger than 100 mesh. This indicates that
removing coarse particles larger than 100 mesh (150 u) is very effective in
reducing unburnt carbon losses.
PARTICLE DIAMETER AND IGNITION STABILITY
Figure 9 shows the relation between particle diameter and ignition distance. As
can be seen in this Figure, the smaller the particle diameter the smaller the
ignition distance. However, making particle diameter smaller than 200 mesh
produces little effects.
SUMMARY
Increasing fineness is effective for all of the followings.
Reduction in unburnt carbon losses
Reduction in NOx emission
Improvement in ignition stability
EFFECT OF FINENESS ON CONVENTIONAL PULVERIZER
Pulverized coal fineness of a conventional pulverizer with a fixed type separator
is normally 70 ^ 80% through 200 mesh and generally more than 5% residue on 100
mesh. Attempt to improve fineness to reduce unburnt carbon losses involves the
following problems.
High Vibration
High fineness operation increases the amount of recirculated load in
the mill and make the coal on the grinding table finer to make the
roll to slip and cause very high vibration.
Remarkable decrease in mill capacity
Capacity decreases by 25% to reduce 100 mesh residue to half.
Remarkable increase in power consumption
Power consumption increases by 30% to reduce 100 mesh residue to
half.
Remarkable increase in primary air to coal ratio
For the same mill throughput, a larger size mill with higher air
requirement must be selected. Therefore primary air to coal ratio
increases by 30% to reduce 100 mesh residue to half.
Therefore about 80% of through 200 mesh had been considered to be the practical
limit for conventional pulverizers and 100 mesh residue could not be reduced below
approx. 3%.
For these reasons, all the mills MHI delivered before 1985 were designed for
pulverized coal fineness of below 80% through 200 mesh as shown in Figure 10.
Development and commercialization of "MRS Pulverizer" has totally changed this
concept and has realized the economical operation of higher than 85% through 200
mesh.
3-76
-------
DEVELOPMENT OF HIGH PERFORMANCE "MRS PULVERIZER"
In order to realize the most effective and economical low NOx combustion by
optimizing the coal particle size distribution, MHI has developed "MRS Pulverizer".
CONSTRUCTION AND OPERATING PRINCIPLE OF "MRS PULVERIZER"
"MRS Pulverizer" is, as is shown in Figure 11, equipped with a "Mitsubishi Rotary
Separator" of forced rotation type instead of a conventional cyclone separator of
fixed type and its structure and operating principle is as follows.
Raw coal, which is fed through the coal feed pipe onto a rotating bowl, is
centrifugally transferred to the outer circumference of the bowl. The coal is then
introduced between a roll and a bullring segment, which lines the bowl, and ground.
Grinding load is applied to the roll by means of a hydraulic cylinder or a spring.
The pulverized coal is carried up to the upper part of the mill by hot air, which
flows from the outer circumference of the bowl.
In its upper part, the mill is equipped with a rotary separator, which performs
centrifugal and impinging classification by means of rotating vanes. The coarse
particles are separated from fine particles and fall onto the bowl to be reground.
After passing through the rotary separator, the fine powder of required size is
carried into a pulverized coal pipe through an outlet port.
CLASSIFYING PRINCIPLE OF "MRS PULVERIZER"
The classifying principle of "MRS Pulverizer" can be simply explained as follows
(Figure 12). Crushed in the grinding and drying chamber, coal particles are
pneumatically carried into the separator, where the particles impinge against the
rotating separator vanes, and according to the balance of forces within a
separating zone, only fine particles enter the inside, whereas the coarse are
ejected away towards the outside. The coal particles existing in this separating
zone are under the influence of centrifugal (F) and centripetal (R) forces caused
by the separator vanes and the air flow respectively.
It can be theoretically said that coal particles are ejected out of the separator
for F > R and they enter into the separator for F < R. F and R are expressed by
following equations when a particle is assumed as sphere.
(3)
(4)
where,
p.p
: Diameter of coal particle
P
,p : Density of particle and air
a
v
v
Rotational speed of vanes
Radial velocity of air
Y
Rotational radius of vanes
3-77
-------
CQ : Coefficient of resistance
Figure 13 shows the values of F and R when particle diameter "Dp" is varied.
Value of "Dp" at the intersection of two lines, or when F=R, is called theoretical
classification diameter.
In order to optimize the design of "MRS Pulverizer", transparent model shown in
Figure 14 was used to visually investigate the classifying phenomena at the vanes
as well as the series of actual grinding tests by 3 t/h test mill in MHI Research
and Development Center.
PERFORMANCE OF "MRS PULVERIZER"
Following three outstanding features of "MRS Pulverizer" have been confirmed as a
result of actual operation.
Extremely high fineness
Due to the optimum design of rotary separator, such a high fineness as more
than 95% through 200 mesh can be obtained without high vibration, an example
of operation results being shown in Figure 15. If separator RPM is adjusted,
operation can be made without any problem even at 100% through 200 mesh.
Effective classification of coarse particles
Figure 16 shows the relationship between the amounts of through 200 mesh and
residue on 100 mesh of pulverized coal. Residue on 100 mesh in a "MRS
Pulverizer" is reduced to less than one-fifth of that in a corresponding
conventional. As can be seen in Figure 16, residue on 100 mesh can even be
reduced to zero if through 200 mesh is adjusted at higher than 90%.
High capacity and lower power consumption
Due to the effective and sharp classification capability of "MRS Pulverizer",
recirculated load in the mill has been considerably reduced. Therefore for
the same fineness, capacity of "MRS Pulverizer" is higher and power
consumption is lower compared with conventional pulverizer as shown in Figures
17 and 18.
LABORATORY COMBUSTION TEST RESULT
In order to compare the combustion characteristics of pulverized coals prepared by
"MRS Pulverizer" with that of conventional pulverizer, following two kinds of
laboratory combustion tests have been performed.
COMBUSTION TEST IN 2 kg/h BASIC COMBUSTION TEST FURNACE
Test Conditions
Pulverized coals of different fineness level have been prepared by both type of
pulverizers, MRS and conventional, and tested in 2 kg/h basic combustion test
furnace shown in Figure 6. Property of the coal used for the test is shown in
3-78
-------
Table 2. Pulverized coal fineness distribution for each test is as shown in Figure
19.
Test Results
Test results are summarized in Figures 20 to 21. Following conclusions can be
derived from the Figures.
For the same excess 02 level, NOx and unburnt carbon losses of MRS
pulverized coal are both lower than that of conventional mill.
Unburnt carbon losses are approximately proportional to residue on 100
mesh for the same excess 02.
COMBUSTION TEST IN 1.5 t/h TEST FURNACE
Test Condition
In order to re-confirm the result of 2 kg/h test, additional larger scale
combustion test has been conducted in 1.5 t/h test furnace shown in Figure 22.
The properties and fineness of the coals used for the test are shown in Table 3.
Test Result
Test results are summarized in Figures 23 and 24. Following conclusions can be
derived from the Figures.
For the same excess 02, "MRS Pulverizer" can achieve approx. 40% lower
unburnt carbon losses.
For the same excess 02, "MRS Pulverizer" can achieve approx. 15 ppm
lower NOx emission.
For the same unburnt carbon losses, "MRS Pulverizer" can achieve approx.
35% lower excess 02 and 30 to 35% lower NOx emission.
APPLICATION OF "MRS PULVERIZER" TO LOW NOx BURNERS
DESCRIPTION OF BOILERS
Low NOx combustion system with "MRS Pulverizer" has been adopted in twenty-eight
boilers, of which twenty have already been in operation successfully.
Names of plants, burner type and fuel used are shown in Table 4.
OPERATION RESULTS
Figure 25 shows the comparison of unburnt carbon losses of "MRS Pulverizers" and
conventional pulverizers at various plants. It can well be said that unburnt
carbon losses were successfully reduced to one-third to one-fourth with the use of
"MRS Pulverizer".
Figure 26 shows the relationship of NOx emission and unburnt carbon losses for both
type pulverizers. For the same unburnt carbon losses, NOx emission has been reduced
by approx. 50% with "MRS Pulverizer".
3-79
-------
OPTIMIZATION OF NOx AND UNBURNT CARBON LOSSES
From the above operation result, NOx emission and unburnt carbon losses had been
both reduced by the application of low NOx combustion system with "MRS Pulverizer".
It is most remarkable that above improvement has been accomplished with reduction
of auxiliary power consumption as shown in Figure 27.
RETROFIT TO "MRS PULVERIZER"
One of the most remarkable advantage of the low NOx firing system with "MRS
Pulverizer" is that the existing mills can be easily converted to "MRS Pulverizer"
to achieve the optimization of NOx and unburnt carbon losses at minimum
modification cost.
RETROFIT PROCEDURES TO "MRS PULVERIZER"
Conventional mills with cyclone type separator can be easily converted to "MRS
Pulverizer" simply by replacing the existing separator with "Mitsubishi Rotary
Separator" as shown in Figure 28.
MERIT OF RETROFIT
Figure 29 shows the merit of modifying existing mills to "MRS Pulverizers" for 500
MW coal fired unit now under study. Merit of modification is 50% reduction of NOx
emission together with $ 736 THOUSAND annual saving by improved combustion
efficiency and reduced auxiliary power consumption including mill, IDF and FDF.
CONCLUSION
The low NOx burner system with "MRS Pulverizer" is one of the most innovative
technology which MHI has developed and commercialized on her own. The technology
has already been adopted in many plants including in Belgium, China and Hongkong as
well as in Japan and we are sure that it will be able to fully meet the
environmental requirement of all countries of the world in the future.
The work described in this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
REFERENCES
1. Araoka, M., et al., Application of Mitsubishi "Advanced Mact" In-furnace NOx
Removal Process at Taio Paper Co., LTD. Mishima Mill No.18 Boiler, March 1987.
3-80
-------
Govermental
Regulation
1000
Private
Voluntary
Regulation
300
200
100
1st 2nd 3rd
4th (Only for Small Facilities)
5th
lO'in hi >
480
150
100
II II
C COAL j
400
(OIL)
(GAS)
300
130
60
200
(6* Ex.0!)
(4*Ex.O!)
(5&EX.O2)
*70 '72 '74 '76 '78 '80 '82 '84 '86 '88
Year
Fig. 1 Trend of Governmental Regulation on NOx
Emission for New Large Plants in Japan
700
600
500
400
300
200
100
0
Fig. 2 History of Low NOx Technology for
Pulverized Coal Fired Boilers
SGR SGR Burner
PM PM Burner
MACT - Mitsubishi Advanced
Combustion Technology
A-MACT Advanced-MACT
MRS Mitsubishi Rotary Separator
PM +A-MACT
PM + A-MACT with
"MRS Pulverizer"
1
J I I I 1 I I I I I I i I I I L-
'70 '72 '74 '76 '78 '80 '82 '84 '86 '88 '90 '92
Year
3-81
-------
Vapor of 1ry Combustion
Volatile Matter Zone 2ry Combustion Zone
Aux. Air
Compartment
Coal
Compartment
Burning Out
Fig. 3 Flame Model of Pulverized Coal Burner
<=> •
Ash
+Unburnt
Carbon
e=^> o Ash
Volatile Matter
Combustion
Zone
Char
Combustion
Zone
100
80
60
40
20
0
-
100
200
300
Volatile
Matter
Char
Ash
Fig. 4
Combustion Time (ms)
Combustion Model of Coal Particle
Fuel Ratio
4.5
Air
Temparature
1ry 82°C/2ry 312°C
NHI/PA
1.75X 106Btu/ft2h
Fineness
Unburnt Carbon Ratio
Fig. 5 Relation between Fineness and Unburnt Carbon
3-82
-------
Pulverized Coal
Ash
Fig. 6 2kg/h Basic Test Furnace
1000
800
600
' 1
1 ! /
__
A
1 1 ! i
1
48# 65# 100# 150# 200# 270# 325#
Fineness of Pulverized Coal
50
10
0.1
'•^325#-
48#~65#V
\
\150#~200#
-• 200#—270#
600 700 800
NOx (ppm)
900
Fig. 7 Coal Particle Size vs. NOx, Unburnt Carbon Fig. 8 NOx vs. Unburnt Carbon in Ash
3-83
-------
Fineness of Pulverized Coal
Fig. 9 Ignition Distance and Particle Size of
Pulverized Coal
100
90
80
70
60
50
Comerciatization of MRS
1950 1985 1986
1987 1988 1989
Fig. 10 Trend of Design Fineness of Pulverizers
Supplied by MHI
Raw Coal
Pulverized Coal
to Burner
Mitsubishi
Rotary Separator
Primary1
Air
Fig. 11 Construction of MRS Pulverizer
3-84
-------
^ \ Rotating
V \ Direction
Fig. 12 Classifying Principle of MRS Pulverizer
Particle Diameter Dp
Fig. 13 Theorical Classification Diameter
Fig. 14 Transparent Model of MRS Pulverizer
3-85
-------
Fig. 15 MRS Speed vs. Fineness Fig. 16 Thru 200 Mesh and Residue on 100 Mesh
70 80 90 100
Fineness (thru 200 Mesh) (96)
Fig. 17 Fineness vs. Mill Capacity
1.5r
I 1.4-
'•& 1.2
B
1.0-
0.8L
Conventional Mill
70 80 90
Fineness (thru 200Mesh)
Fig. 18 Fineness and Mill Power Consumption
3-86
-------
\ Conventional Mill
\
\
\
75
thru 200 Mesh
19 Fineness Distribution of Coals Tested
15r
10-
§ 5-
f
\
Conventional Mill i
MRS Pulverizer"
T
\
700 725 750
NOx (02=6%Kppm)
Fig. 20 NOx and Unburnt Carbon
15r
10
o "MRS Pulverizer"
• Conventional Mill
5 10
Residue on 100 Mesh {%)
Fig. 21 Residue on 100 Mesh vs
Unburnt Carbon in Ash
15
3-87
-------
CO
n
o
Si 150j-
100
6.0
4.0-
2.0"
Q.ojr
Conventional Mill J
'MRS Pulverizer"
"MRS Pulverizer"
Fig. 22 1.5t/h Coal Firing Test Furnance
Fig.
2 4 6
Excess O2 (.%)
23 NOx and Unburnt Carbon in
Ash vs Excess O2
\ Conventional Mill
"MRS Pulverizer"
100 150 200
NOx [02=6% Kppm)
Fig. 24 NOx and Unburnt Carbon in Ash
3-88
-------
10
* 8
Ui
«
£ 6
c
o
S 4
0
1 2
• "MRS Pulverizer"
O Conventional Mill
UCI:Unburnt Carbon Index
UCI= (Furnace Size Factor) X (Ash) (?g) X [pue| |[at;0]
Fuel Ratio=Fixed Carbon(%)/Volatile Matter(^)
Ex.0! = 2.5%
V
10 15 20 25 30
UCI
Fig. 25 Mill Type and Unburnt Carbon in Ash
_ 5
I 4
V)
C
= 3
S
o
S 2
o
*-»
§ 1
02 = 3.5%
2 = 4.0%
Conventional
02 = 4.2 Mi||
02 = 2.7% 02 = 5.0%
~°—o— "MRS Pulverizer"
02 = 3.4%
150 200
NOx (ppm)
250
Fig. 26 NOx, Unburnt Carbon in
Ash and Mill Type
Conventional
Mill
MRS
Pulverizer
Fig. 27 Decrease of Power Consumption
by Adoption of "MRS Pulverizer"
3-89
-------
"IE
L
i—)
j
m
MRS'
Vr*
I
¥0-
Conventional Mill
MRS Pulverizer
: I Retrofit Parts
Fig. 28 Retrofit Procedure to MRS
*
<3 3.0-
2.0
800
600
400
^Retrofit to "MRS Pulverizer"
—i 1 1 1 i__
u
aa
200
$736Thousand/year
$416Thousand/year
^ Merit by
||| Decreasing :¦
'M. Ex. Air „'jS192 Thousand
' '/year
- ,J2"
ja-'*-* " ' Merit by Decreasing
Unburnt Carbon in Ash
Fig.
3.0 3.5 4.0 4.5 5.0
Eco. Outlet 02 (%)
29 Merit of Retrofit to "MRS Pulverizer"
3-90
-------
Table 1 Coals Used for Particle Combustion Test
Item —-
Coal Kind
A
B
C
Heating Value(kcal/kg)
(HHV-S-M Free)
5780
6620
6650
V)
'35
Surface Moisture {%)
1.9
3.4
7.0
-------
Table 4 Supply List of Low NOx Burner with "MRS Pulverizer"
Fuel
Plant Name
Boiler
Mill
Mill
Capacity
(t/h)
Numbers
Start of
Commercial
Operation
Actual Result of Coal
Fineness {%)
Low NOx
Capacity
Type
Furnished
Residue on
100 Mesh
Through
200 Mesh
Technology
Kanebo Co., Ltd.
Hofu Factory
130t/h
523RP
-MRS
7.1
2
Aug. 1985
1.2
83.6
A-MACT
Asahi Chemical Co., Ltd.
Nobeoka Factory #3 Unit
180t/h
643XRP
-MRS
15.0
2
Apr. 1986
0
96
A-MACT
Taio Pulp Co., Ltd.
Mislima Factory #18 Unit
350t/h
703RP
-MRS
17.6
4
Apr. 1986
0
89
A-MACT
Teijin Ca, Ltd
Tokuyama Factory
80t/h
483RP
-MRS
6.1
2
Apr. 1986
0
88
A-MACT
Ehime Paper MFG.
lyomishima Factory
95t/h
523RP
-MRS
7.2
2
May 1986
0.8
86.2
A-MACT
Showa Sangyo, Ltd.
70t/h
463XRP
-MRS
7.2
2
Aug. 1986
0.2
86.4
A-MACT
Teijin Co., Ltd.
Mihara Factory
130t/h
563RP
-MRS
9.0
2
Sep. 1986
0
88
A-MACT
Idemitsu Engineering
Co., Ltd.
130t/h
563XRP
-MRS
9.7
2
Oct. 1986
0.4
87
A-MACT
Kanegafuchi Chemical
Co., Ltd Takasago Factory
270t/h
623RP
-MRS
10.4
3
Apr. 1987
0.2
90.9
A-MACT
Coal
Matsuda Ca, Ltd.
11Ot/h
X2
523RP
-MRS
8.1
4
Sep. 1987
0.4
83
A-MACT
Chubu Electric Power
Co.. Ltd. Test Plant
15t/h
303RP
-MRS
1.7
1
Sep. 1987
0.1
87
A-MACT
China Huangtai P/S
#7 Unit
300 MW
863XRP
-MRS
42.5
5
Oct. 1987
0.3
86.4
OFA
Maruzumi Paper MFG.
OOE Factory
21Ot/h
623RP
-MRS
9.2
2
Jun. 1988
0
85
A-MACT
Daiseru Aboshi
Co., Ltd.
300t/h
623RP
•MRS
12.9
3
Apr. 1989
—
—
A-MACT
Taio Pulp Ca, Ltd.
Mishima Factory #19 Unit
350t/h
703RP
-MRS
17.6
4
Apr. 1989
—
—
A-MACT
Banyu Shibiyo, Ltd.
140t/h
603RP
-MRS
9.0
2
Oct. 1989
—
—
A-MACT
Hokuriku Electric Power
Co., Ltd. Tsuruga P/S #1 Unit
500 MW
943RP
-MRS
41.7
6
Oct 1991
—
—
A-MACT
Chubu Electric Power Co., Ltd.
Hekinan P/S #1 Unit
700 MW
1103RP
-MRS
82.5
6
Oct. 1991
—
—
A-MACT
Chang Chun
Petrochemical Co., Ltd.
210t/h
623RP
-MRS
14.0
2
Jul. 1989
—
—
A-MACT
Hong Kong Electric Co, Ltd.
Lamina P/S #6 Unit
350 MW
903RP
-MRS
40.6
5
Apr. 1991
—
OFA
Mitsubishi Acetate Co., Ltd.
Toyama Factory
75t/h
563RS
-MRS
7.8
1
Sep. 1985
0
98
A-MACT
Toho Rayon Ca, Ltd.
Tokushima Factory
75t/h
483RS
-MRS
3.2
1
Nov. 1985
0
96.5
A-MACT
Petro.
Sumitomo Chemical Co., Ltd.
CNba Works
305t/h
583RS
-MRS
8.7
2
Dec. 1985
0
98.9
OFA
Coke
Toyo Pulp Co., Ltd.
Kure Factory
80t/h
563RS
-MRS
9.5
1
Jun. 1986
0
96
OFA
Mitsubishi Chemical Co., Ltd.
MizusMma Factory
170t/h
483RP
-MRS
4.8
1
Sep. 1986
0
93
A-MACT
Takeda Chemical
Industries, Ltd.
lOOt/h
563RS
-MRS
6.8
1
Oct. 1986
0
96
A-MACT
Coal/
Petro.
Coke
Mitsubishi Rayon
Co., Ltd.
Ootake Factory
180t/h
583RP
-MRS
11.5/8.2
2
Apr. 1989
—
—
A-MACT
3-92
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
THE XCL BURNER - LATEST DEVELOPMENTS AND OPERATING EXPERIENCE
Albert D. LaRue
Fossil Power Division
Babcock & Wilcox
Barberton, Ohio
ABSTRACT
As reported at the previous symposium in New Orleans, B&W developed the XCL
burner to further reduce NO^ emissions from wall-fired, pulverized coal boilers.
A full complement of XCL burners was installed at Ohio Edison's Edgewater
Station, Unit 4, as a key element of the LIMB demonstration. Optimization in
this application required implementation of corrective measures to other boiler
systems, including the air heater and pulverizer feeder controls. NO^ emissions
well under New Source Performance Standards with high combustion efficiency were
subsequently demonstrated.
B&W's XCL burner is being applied commercially to utility and industrial size
units. A recent contract included the development and use of the XCL Tri-Fuel
burner, capable of firing natural gas, fuel oil, or pulverized coal. The
development included extensive testing on one full-scale prototype burner (120
million Btu/hr) on all three fuels. Fuel-staging technology incorporated in the
XCL resulted in NO^ emissions less than: 0.1 (lb/million Btu) with gas; 0.2 with
No. 6 oil; and 0.35 with high and low volatile bituminous coal. These results
were obtained without use of other NO control measures such as gas recirculation
x
and N0x ports. However, even lower emissions were achieved in combination with
these measures. Consequently, B&W has designed the XCL Oil/Gas burner for use in
low NO^ applications with these fuels. Most recently, B&W developed the XCL-FM
for low NO^ applications in package boilers with unheated combustion air.
Testing in such a boiler at B&W's Alliance Research Center has demonstrated N0x
emissions less than 0.1 for natural gas and distillate oil (burner only) and less
than 0.04 with gas recirculation.
3-93
-------
INTRODUCTION AND BACKGROUND
In 1985 Babcock & Wilcox (B&W) embarked on a program to develop a true second
generation low NO^ burner for firing pulverized coal. The motivation for this
program was to address a changing marketplace in the United States. New utility
and industrial boilers have increasingly been regulated to N0x emission limits
much below federal New Source Performance Standards (NSPS). Furthermore, many
pre-NSPS units are likely to be subject to NO^ limitations in the near future.
These circumstances increase the need for combustion equipment which can
efficiently reduce N0x emissions while being adaptable to existing boilers.
Combustion equipment, being the source of NO^, is most often the first choice for
control of N0x emissions. B&W's Dual Register Burner (DRB) has proven to be very
effective in controlling N0x in new wall-fired utility boilers. All of the 68
units in service with DRB's are satisfying applicable NSPS N0x limitations (1).
This 100 percent record has been achieved without the need for alterations to
reduce NO^ during the commissioning of any of these units. And unlike other
manufacturers, this performance is achieved by virtue of the burner performance
and without augmentation by overfire air ports, underfire air ports, boundary
air, etc. This procedure avoids problems with slagging and corrosion in the
combustion zone associated with air staging.
However, the DRB is difficult to apply to pre-NSPS boilers. One reason for the
DRB's success in control of N0x is its use in combination with compartmented
windboxes (Figure 1). Efficient combustion with low NO^ is dependent on properly
distributed air and fuel for the combustion system. The compartmented windbox
enables reliable control of secondary air to the burners served by each
pulverizer. However, retrofitting compartmented windboxes to existing boilers
can be complicated and expensive, owing to burner layout, space limitations, etc.
Still the need remains for an accurate and reliable means of distributing and
controlling secondary air, which should be addressed by the new burner design.
3-94
-------
In addition, the burner must perform well within the confines of pre-NSPS boiler
designs. N0x reduction has to be achieved in the more compact furnaces
characteristic of these units. This requirement is complicated by the higher
thermal loading to the combustion zone tending to increase NO^, and smaller
furnace dimensions which sometimes constrain flame shape and N0x (air/fuel mixing
restrictions). And, at least equally important, the burner must be mechanically
reliable in the hostile operating conditions for which it is intended.
Mechanical breakdowns not only add to expense in time and material to repair, but
also can impair boiler performance (slagging) and cause load restrictions.
B&W designed a new burner called the XCL (aXial-Controlled-Low NO^ burner) to
address these issues. The XCL (Figure 2) combines features from three
commercially successful burners, drawing on their experience and demonstrated
performance. Then, some features were added to the XCL which uniquely enhanced
its performance. The functional design of the XCL is largely derived from B&W's
DRB, and from Babcock Hitachi's HT-NR burner via licensed technology (2). Fuel
dispersion is achieved in the nozzle by the conical diffuser, resulting in a
fuel-rich ring near the walls of the nozzle and a fuel- lean core. The fuel
exits through a flame stabilizing ring which accelerates ignition and
devolatilization. Reducing species then form by partial oxidation of coal
volatiles from primary air and limited secondary air. The reducing zone in the
core of the flame prevents NC>x formation during devolatilization, and the
reducing species generated serve to decompose N0x as the combustion continues
into the char oxidation phase. Multistage air vanes and the air separation vane
limit burner pressure drop, improve distribution of secondary air within the
burner and, consequently, provide more uniform swirl patterns leaving the throat
and control of near field stoichiometry and downstream mixing. An impeller is
used, when appropriate, to vary flame shape with some trade-off in NO^ emissions.
The mechanical configuration is adapted from B&W's S-type burner. The S-burner's
success commercially has come from demonstrated combustion improvements and
mechanical reliability when retrofitted to pre-NSPS boilers. Like the S-burner,
the XCL uses a slide damper to reliably control secondary air and is equipped
with a pitot grid downstream of the burner entrance to measure secondary air flow
in each burner. These measurements are used during commissioning to identify air
distribution patterns in the windbox to enable balancing air flow to each burner.
The rugged stainless steel construction has solved problems with overheating,
warpage, binding, etc., experienced by other burners.
3-95
-------
Development of the XCL burner began with prototype-scale, 78 million Btu/hr
(MKB) tests to evaluate and tune secondary air flow patterns relative to tests
performed on a DRB and HT-NR. Flow testing was followed by extensive combustion
tests, as reported previously (3). These tests, performed with support by the
EPA, confirmed the XCL burner could meet performance criteria stipulated for the
low NO^ burners to be used on the Limestone Injection Multistage Burner (LIMB)
project. By use of a modified impeller, flame shape could be varied to
accommodate the furnace dimensions. Flame length could be varied from 12 to 22
feet, with NO^ emissions varying from 0.6 to 0.3 lb/MKB (unstaged). Burner
resistance was near 4 inch w.c. and carbon in the ash was 2-3 percent firing
Pittsburgh 8 coal.
XCL BURNER UTILITY BOILER RETROFIT
B&W fabricated and installed 12 XCL burners at Ohio Edison's Edgewater Station,
Unit 4, as a key aspect of the LIMB project. This unit has a B&W natural
circulation boiler rated at 690,000 lb/hr main steam flow at 1480 psig/1000°F
(Figure 3). Prior to XCL burner installation, baseline tests were performed on
the unit with the existing circular burners. After burner tuning, full load N0^
emissions were 0.89 lb/MKB, with 20 ppm CO, 3 percent excess 0£ at the burners,
and averaged 2.2 percent carbon in the ash (Table 1).
The XCL burners were installed during the fall of 1986. As installed, the XCL's
reduced N0^ to 0.53 lb/MKB at full load and 3-3.5 percent excess 0^ at the
burners, with 115 ppm CO and with 2.8 percent carbon in the ash. Flame lengths
were 15-18 feet, well short of the rear wall (22 ft. furnace depth). NO was
x
reduced by 40 percent while maintaining high combustion efficiency.
Burner characterization/optimization testing followed, with plans to reduce N0^
below the 0.5 lb/MKB program goal by making use of flame shaping features of the
XCL. Impeller adjustments reduced NO^ to 0.34 to 0.40 lb/MKB, but CO and
unburned carbon (UBC) increased unexpectedly to unacceptable values. Furnace gas
specie probing at the entrance to the superheater identified zones deficient in
oxygen, which explained the high CO/UBC. Note that air flow readings from the
XCL burner pitot grids did not support the furnace 0^ distribution, and raised
suspicion of fuel distribution among burners. Burner dampers could be tuned to
improve furnace 0^ distribution and sharply reduced CO, e.g. from 2500 to 75 ppm
3-96
-------
average leaving the boiler. Conical diffusers were installed in three of the
burner elevations to retain low NO^ performance, but problems continued with CO
and UBC. Continued furnace probing and investigation showed the O^/CO patterns
leaving the furnace changed considerably from day to day, while burner settings,
load, and excess air were unchanged. The unexplained variations in air/fuel
distribution prevented implementation of "fixes" to balance and tune the system.
Fuel distribution and loading of the pulverizers was identified as the main
problem. The control system for regulating primary air and coal to these
pulverizers has no direct measure of coal flow to the mills. The table feeders
provide neither a volumetric or weight flow indication of coal. In addition, the
pneumatic controllers were subject to maladjustment and loss of calibration. The
combined result was non-repeatable and non-uniform coal feed to the four
pulverizers, and varying coal distribution leaving each of the pulverizers. The
mill serving the second from top burner elevation contributed disproportionately
to the CO/UBC problems so impellers were reinstalled in those burners to reduce
sensitivity to fuel problems. Combustion was also impaired by marginal excess
air. While the XCL pressure drop met the program's 5 inch w.c. limit at MCR, the
burner resistance had increased about 3 inches over the circular burners. This
increase, combined with continuing air heater leakage limited excess air to 3-3.5
percent at full load, or lower, depending on operating conditions. While this
excess air level is reasonable with good air/fuel distribution, it did little to
compensate for fuel flow variations.
Optimization work eventually resulted in N0x being reduced to just under the 0.5
lb/MKB program goal, i.e. to 0.47, with 6 percent carbon in ash and 60 ppm CO
(full load normal excess air). New solid state feeder controls were installed
late in 1988 and are expected to improve fuel control with the table feeder and
reduce carbon in the ash after tuning.
This burner retrofit program showed the importance of accurate fuel/air control
to low NO^ combustion optimization. Inaccuracies in air/fuel distribution and
control may not cause significant combustion inefficiencies with pre-NSPS rapid
mix burners. This circumstance was true for the circular burners at this plant.
In addition, the low N0x burners were able to reduce NO^ considerably (over 40
percent) while maintaining combustion efficiency. However, the XCL burner is
capable of 60 to 70 percent NO^ reduction from uncontrolled levels, which was not
achievable with good combustion efficiency in this unit. In this instance,
further improvement would be necessary to achieve uniform fuel delivery to the
3-97
-------
burners, i.e. gravimetric coal feeders. Gravimetric feeders are the norm for
many larger pre-NSPS units and most post-NSPS units, and are needed for
reliable/repeatable coal regulation. High performance low NO^ burners can reach
their potential only when used in combination with systems which accurately and
consistently regulate their air and fuel supply.
MULTI-FUEL XCL BURNER DEVELOPMENT/COMMERCIALIZATION
Late in 1986, B&W was approached by ENEL, the electric utility of Italy, and
Ansaldo Componenti (ACO) - B&W's licensee in Italy, concerning low NO^ combustion
technology. ENEL was evaluating these technologies world-wide, in preparation
for implementation in Italy. Systems with multi-fuel capability (pulverized
coal, No. 6 oil, natural gas) were needed for use with the many cell
burner-equipped existing units, and also for more conventional wall-fired units.
(As an aside, B&W received a contract to supply 15 multi-fuel Low NO Cells for
one of ENEL's 320 MW units and start-up is scheduled for later this year.) The
focus for the conventional wall-fired units was to perform a state-of-art
demonstration of low NO^ combustion equipment at the Brindisi Sud Power Station.
This new power plant will have four B&W designed/ACO supplied 660 MW,
supercritical units. The units had been designed to fire pulverized coal and No.
6 oil using Dual Register Burners/Compartmented Windboxes. Unit 1 was already
well along in erection and would retain this system, but NO^ ports were added to
further reduce N0x. Unit 2 was at a stage where the combustion system could be
changed but timing was critical.
B&W received the contract to supply 56 multi-fuel XCL burners for Brindisi
Sud 2. The combustion system design would also include addition of gas firing
capability, dual zone N0x ports above the top burner row, and retain gas
recirculation to the compartmented windboxes. The project was contingent upon
demonstrating performance in prototype scale combustion tests firing natural gas,
fuel oil, and coal. B&W designed the XCL multi-fuel burner (Figure 4) (patent
pending) to retain in-burner fuel staging techniques incorporated for coal
firing, and to make use of these principles firing fuel oil and natural gas. All
of the fuel elements for coal, oil, and gas firing are enclosed within the flame
stabilizing ring in the center of the burner. A small quantity of core air is
admitted to the nozzle during gas or oil firing to partially oxidize the fuel and
generate hydrocarbon radicals in the core of the flame. These effectively
3-98
-------
eliminate NO in the near flame and go on to mix with and limit NO formation in
x x
the later stages of combustion (char oxidation). As before, the burner is
configured with multi-stage air vanes to control introduction of secondary air to
the flame and final mixing. A slide damper controls air flow to the burner and a
pitot grid is used to measure air flow.
B&W performed the large scale combustion tests - nominal 120 million Btu/hr - at
Riley Research's tunnel facility in Worcester, Massachusetts. The test furnace,
shown in Figure 5, is 18 ft. wide, 60 ft. long, and 18 ft. tall not including the
hopper bottom. Interior surfaces of the roof, hopper, and sidewalls are
insulated for a distance of 40 ft. from the end wall on which the burner is
mounted. Combustion tests were performed with the burner end wall uninsulated
and insulated to compare thermal effects. Staging ports used in this program
were positioned on the sidewalls 16 ft. from the the burner wall. Secondary air,
primary air, and N0x port air were measured and controlled, as were each of the
fuels. A system was installed for these tests to supply, meter, and regulate
recirculated flue gas (GR) to the secondary air. A B&W CFS gas-fired lighter was
used for ignition, and the facility's flame scanners/flame safety system was used
for these functions. Gases exiting the furnace were continuously monitored for
temperature, 0„, CO, C0„, and NO . The natural gas fired was typically 95
percent methane and 1030 Btu/ft . The No. 6 oil fired was selected to match the
oil specified for Brindisi Sud, and a typical analysis is shown in Table 2. Fuel
nitrogen ranged from 0.30 to 0.41 and averaged 0.37 percent. Two coals were
fired during the tests (Table 2). The "low volatile" (LV) bituminous was similar
to the design coal and the higher volatile (HV) coal was tested for comparison.
The LV coal has a relatively high FC/VM ratio making it less conducive to NOx
control measures, and the HV coal was more typical of bituminous coal used by
U.S. utilities.
XCL PERFORMANCE - NATURAL GAS
Combustion testing took place through fall and winter, 1987. Performance firing
gas and oil were of primary importance (coal performance was already commercially
demonstrated). A total of 215 tests were performed firing natural gas. These
began with evaluations of four different gas spud designs, which included two
standard designs and two new designs to promote fuel staging effects. Figure 6
shows NO^ test results for B&W's standard variable mix spud and the HEMI spud
(patent pending). At high load (100 MKB), without use of GR or staging ports,
and 8 percent excess air, the variable mix spud produced 131 ppm NO^ (all ppm
3-99
-------
figures are referenced to 3% 0^). At the same conditions, the HEMI spud produced
56 ppm N0x with 34 ppm CO. Secondary air temperature was 600°F. Figure 6 also
includes predicted N0x for a circular burner with VM spuds, based on thermal
scaling factors developed in this program and B&W standards. The XCL burner
reduced NO^ 32 percent compared to the circular burner, each using the same gas
spud design. The XCL plus HEMI spud reduced N0x 71 percent, relative to the
circular/VM spud. XCL results are compared to standard N0x predictions for
another low NO burner, the PG-DRB, in Figure 7. The XCL NO emissions are much
x x
lower than the PG-DRB at low GR rates and emissions performance converges at high
levels of GR. Note the PG-DRB is intended for use with GR so the emissions
corresponding to low GR rates do not reflect normal operation. However, the XCL
achieves low N0x emissions without use of GR, for situations which require this;
and produces lower N0x for a given level of GR up to very high GR rates.
Extensive characterization tests were performed on the XCL burner with HEMI
spuds. The end wall, oil which the burner was located, was insulated so that all
furnace surfaces for a distance of 40 ft. were insulated. Increasing firing rate
to 124 MKB (100% load) with the insulated firing wall resulted in 74 ppm N0x with
36 ppm CO. Table 3 shows results for various loads with and without N0x ports
and gas recirculation (GR). Full load N0x emissions were reduced to 50 ppm with
moderate quantities of GR, and down to 33 ppm N0x staged with higher GR.
Excellent emission performance was also achieved at reduced loads, e.g. 56 ppm
NO unstaged/No GR at 32 percent load or 23 ppm NO with GR at the same load.
X X
Flame stability and turndown were exceptional, owing to the placement of the fuel
elements within the flame stabilizing ring. No furnace rumble was encountered in
any of the tests. Stable, well-shaped flames were anchored at the burner from
1.8 MKB (minimum attempted) to 140 MKB (maximum available), representing 78:1
turndown.
XCL PERFORMANCE - NO. 6 OIL
The No. 6 oil combustion tests commenced with evaluation of several Y-Jet and
Racer atomizer sprayer plates. The emphasis was on NOx/CO emissions and smoking
tendencies at high load without GR or staging ports. The facility was not
equipped to measure opacity, but observation ports were sufficient to qualify
furnace clarity. The standard Racer plate, with high oil pressure/low atomizing
steam consumption produced the best results. The flame shape was controllable
and free from smoke over a very wide range of conditions. Characterization tests
3-100
-------
followed with the Racer and XCL burner. Table 4 summarizes some key results of
the 142 tests performed firing No. 6 oil. With the front firing wall insulated
and 600°F secondary air, NO^ emissions were 118 ppm with 38 ppm CO at full load,
unstaged without GR. NC>x was reduced to 88 ppm by staging with moderate GR.
Emission performance was similar at middle loads, but NC>x tended to increase at
lower loads as stoichiometry unavoidably increased. Air staging and GR would
partly offset this tendency, e.g. at 35 percent load, NC>x was limited to 133 ppm
with 48 ppm CO with staging and 17 percent GR.
The XCL burner was reconfigured for a portion of the tests to defeat the NO^
control features and simulate a pre-NSPS circular burner. The flame stabilizing
ring and air separation vane were removed, an impeller was installed, and the
burner vanes were adjusted to produce a short bushy flame characteristic of
circular burners. Note previous lab and field tests had demonstrated the
validity of this approach, which provided a cost-effective means of developing
baseline results for oil and coal, in order to develop thermal scaling factors.
Figure 8 compares performance of the "circular burner" to the XCL burner. The
XCL reduced NO^ 42 percent relative to the circular with zero GR, and by 45
percent with 20 percent GR. The standard prediction for the PG-DRB is also shown
in Figure 8. XCL N0^ emissions are 28 percent lower without any GR, and 24
percent lower with 20 percent GR.
XCL flame stability was good throughout the tests firing No. 6 oil. Turndown,
primarily a factor of atomizer performance, was 10:1.
XCL PERFORMANCE - PULVERIZED COAL
NO^ emissions firing pulverized coal in the multi-fuel XCL in this facility were
similar to results achieved previously in lab and field tests. At conditions of
full load, unstaged, firing the lower volatile coal, NO^ emissions were typically
250 ppm. At the same conditions, the higher volatile coal resulted in NO
emissions of 236 ppm. CO emissions were near 100 ppm for either case. Unburned
carbon loss, as measured by dust samples collected at the furnace exit, was 1.4
percent in this relatively cool, slow-mix furnace. Use of the NO^ ports did
little to impact NO , which is attributed to the proximity of the NO ports to
X X
the burners (4). However, the additional downstream turbulence provided by the
NO^ ports served to reduce unburned carbon loss to 0.3 percent.
3-101
-------
Scale-up of the results from the gas, oil, and coal tests indicated the XCL would
meet or exceed performance expectations. Consequently, B&W manufactured 56 XCL
multi-fuel burners for installation at Brindisi Sud 2, scheduled to start up
later in 1989
XCL OIL/GAS
The exceptional emission performance of the XCL multi-fuel burner firing natural
gas and No. 6 oil led to the decision to commercialize the burner for gas/oil
applications. The XCL oil/gas burner (Figure 9) retains the functional design
parameters of the multi-fuel design, but obviously without the coal components.
The oil atomizer and HEMI gas spuds are enclosed within the flame stabilizing
ring. Secondary air is controlled by a sliding damper, measured by a pitot grid,
and distributed and swirled by multi-stage air vanes (similar to the S-burner).
Provision is made for vane adjustment although the combustion tests identified
optimum settings for the vanes in the inner and outer air zones. These settings
are the same for firing either natural gas or fuel oil. Vane adjustment, if
needed, will be performed during commissioning and then vanes will be locked in
position. Subsequent air control would be provided by the sliding air damper.
The HEMI spuds are designed to permit adjustment or removal from the burner
platform.
The XCL oil/gas can be used in combination with other N0x control measures such
as N0x ports and GR when conditions dictate. The burner is designed for use in
open windboxes for utility and many industrial boiler applications.
XCL-FM
A rather unique industrial boiler type in need of low NO^ combustion equipment is
the package boiler. These boilers are completely shop assembled and shipped,
usually by rail car, for use in a wide variety of applications. These boilers
are characterized by compact furnaces with high volumetric heat release rates,
and are designed to operate with little or no supervision. The vast majority of
the units use unheated combustion air and fire natural gas, No. 2 oil, or No. 6
oil.
B&W developed the XCL-FM for use with package boilers (FM is B&W's designation
for its line of package boilers). The development program centered on an
extensive series of combustion tests performed on the FM boiler (Figure 10) at
our Alliance Research Center, during the spring and summer of 1988 (5). The
3-102
-------
tested conditions included use of natural gas, No. 2 oil, and No. 6 oil (with and
without doping) at various loads and stoichiometries, with and without gas
recirculation to the windbox. Considerable furnace probing was performed to
study temperature, N0x> CO, CC^, and 0^ during flame development. Several
varieties of HEMI spuds and oil atomizers were tested to optimize performance in
this furnace. The XCL-FM burner (Figure 11) was designed for a nominal full load
input of 60 MKB using ambient air. The furnace end of the burner can be seen to
be similar to the XCL oil/gas design, while the upstream portion of the burner
was simplified. The sliding air damper and pitot grid would be superfluous for
these single burner applications.
Some key test results are summarized in Figure 12. NO^ emissions firing natural
gas were very close to results achieved in XCL multi-fuel program. Full load NO^
emissions were 60 to 70 ppm without GR, and 30 ppm with 17 percent GR (maximum
available GR at full load). At 2/3 load, N0^ could be reduced to 19 ppm with 20
percent GR. CO was less than 50 ppm. Emission results with No. 2 oil were
similar to natural gas, with NO^ in the low 60"s ppm without GR. But GR had less
influence on N0x> e.g. 45 ppm NO^ with 17 percent GR. The No. 6 oil tested had
0.2 percent fuel nitrogen inherently, and pyridine was used in selected tests to
raise fuel nitrogen to evaluate the influence on NO^. Figure 12 shows the
influence to be significant and non-linear. As expected, N0x emissions were much
higher with No. 6 oil and some adjustments were necessary to prevent flame
impingement and deposition. While these adjustments were successful in this
regard, they increased N0x> e.g. originally NO^ was typically 155 ppm and
increased to 200 ppm after adjustments (without GR or staging).
A commercial demonstration of the XCL-FM is underway on one of the boilers at the
California Institute of Technology in Pasadena. Burner tuning and performance
tests will take place this spring.
SUMMARY
Babcock & Wilcox has developed a new generation of burners which effectively make
use of fuel staging principles to limit NO^ emissions to very low levels. The
XCL burners benefit from this advanced technology plus B&W's experience in N0x
control over a wide range of applications. The mechanical design is derived from
the S-type burner which has demonstrated reliability and performance improvements
3-103
-------
in numerous coal-fired utility boilers. The various XCL burners are designed to
facilitate use in new or retrofit applications, ranging from multi-fuel (coal,
oil, gas) supercritical steam generators to package boilers. Some applications
are in commercial operation while others are scheduled for demonstration this
year.
REFERENCES
1. LaRue, A. D., Cioffi, P. L., "N0x Control Update - 1989", presented at the
Joint Symposium on Stationary N0x Control, San Francisco, CA. March 1989.
B&W paper - BR-1370.
2. Narita, T., et al, "Operating Experiences of Coal Fired Utility Boilers
using Hitachi NO^ Reduction Burners", presented at the Joint Symposium on
Stationary N0x Control, New Orleans, LA. March 1987.
3. LaRue, A. D., Acree, M. A., "Development Status of B&W's Second Generation
Low NO^ Burner - the XCL Burner", presented at the Joint Symposium on
Stationary NOx Control, New Orleans, LA. March 1987. B&W paper - BR-1315.
4. Wendt, J. 0. L., et al, "Pulverized Coal Combustion: N0x Formation
Mechanisms under Fuel Rich and Staged Combustion Conditions", presented at
the 17th Symposium on Combustion, Pittsburgh, PA. 1978.
5. Eckhart, C. F., "XCL Burner Test Report", B&W in-house report,
December 1988.
3-104
-------
Phase 5 Dual
Figure 1 Dual register burner (Phase V) and compartmentedwindbox.
® •• Combustion Zone of Volatile Matter
Figure 2 XCL Burner.
3-105
-------
Air Control Pitot Fixed
Disk Manifold Vanes
Adjustable
1 3 Vanes
— Spin \
D-Q
Figure 3 XCL Burner Retrofit to Ohio Edison's Edgewater Station.
Figure 4 Multi-fuel XCL burner.
3-106
-------
Tertiary Air Ports
(Unused)
Burner
Opening
Staging Air
Staging Windbox
Burner Windbox
Figure 5 Riley test facility.
250
200-
150-
100-
50-
No Staging, Fuel N2 = 0.40%
"Circular Burner"
XCL Burner
8 12 16 20 24
Gas Recirculation (%)
Figure 8 NO, Emissions Firing 6 Oil.
200
180-
160
140
120-
E 100
a
-2- 80H
o
z 60
40-
20-
0
1 Circular Burner + VM Spuds
3enefit of
XCL Burner Design ,
XCL + VM Spud
Benefit of
Fuel-Staged
Hemi-Spud 0
^ O-vJ
oo o
°
o
o o
XCL + Hemi Spuds
0
6 8 10
Excess Air (%)
12
i
14
Figure 6 XCL Burner - Natural Gas Test Results
No Gas Recirc., No Two-Stage Combustion
No Two-Stage Combustion
Primary Gas - Dual Register Burner
/
4 8 12 16 20 24
Gas Recirculation To Burners (%)
Figure 7 NOxEmissions Firing Natural Gas.
Flame
Pitot Stabilizing
Grid Inner
Vanes
jJb:-
Oil Atomizer
Removable
Gas Spud
T
Sliding Air
Damper
Air
£L Separation
Vane
Dual Stage
Outer Vanes
Figure 9 XCL Oil/Gas Burner.
Sealed Water-Wall
Construction
Figure 10 FM Boiler.
3-107
-------
Figure 11 XCL FM burner.
400-1
300-
200-
100-
Nat. Gas • Full Load • No Staging
Nat. Gas - 2/3 Load - No Staging
2 Oil • Full Load - No Staging
6 Oil • Full Load, 0.9 Stoich
6 Oil • Full Load, Unstages
I
10
Gas Recirculation
i
20
30
Figure 12 XCL FM Test Results.
3-108
-------
Table 2
Prototype Test Fuel Analyses
"Low
High
Volatile"
Volatile
Fuel
#6 Oil
Coal
Coal
Proximate (%)
Moisture
4.3
8.3
Volatile Matter
23.0
29.10
Ash
13.5
10.90
Fixed Carbon
59.2
51.70
FC/VM Ratio
2.6
1.8
HHV (Btu/lb)
18356
12642
12141
Ultimate (%)
Moisture
4.3
8.3
Carbon
86.44
71.0
67.9
Hydrogen
10.71
4.4
4.4
Nitrogen
0.37
1.0
1.5
Oxygen
0.25
3.0
5.4
Sulfur
2.16
2.8
1.6
Ash
0.07
13.5
10.9
Specific Gravity
0.94
API Gravity
12.3
Conradson Carbon
12.1
Asphaltenes
4.4
Table 1
LIMB Retrofit Results*
Circular
XCL Burners-
XCL Burners-
Burners
with Impellers
Tuned
NOx (lb/106 Btu)
0.89
0.53
0.47
CO(ppm @ 3% 02)
20
115
60
Carbon in Flyash
2.2
2.8
6
* ¦ Full load, normal/excess air
Table 3
XCL Burner - Natural Gas Test Results
Load
%
NO*
(ppm @ 3
CO
1% 02)
Burner
Stoich.
Gas
Recirc.
%
Excess
Air
%
100
74
36
1.07
0
7
100
65
46
0.89
0
7
100
50
35
1.07
9
7
100
33
36
0.81
16
8
50
45
35
1.12
0
12
50
29
37
0.85
18
13
32
56
62
1.35
0
35
32
23
39
1.21
20
21
Table 4
XCL Burner - #6 Oil Test Results
Load
%
N0X
(ppm @
CO
3% 0,)
Burner
Stoich.
Gas
Recirc.
%
Excess
Air
%
100
118
38
1.08
0
8
100
95
45
0.90
0
8
100
88
46
0.88
10
7
62
123
36
1.16
0
16
62
83
36
0.80
18
17
35
133
48
0.96
17
37
3-109
-------
(Intentionally Blank)
3-110
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
APPLICATION OF LOW NOx COMBUSTION TECHNOLOGIES
TO A LOW VOLATILE COAL FIRING BOILER
Shigehiro MIYAMAE, Takashi KIGA and Keiji MAKINO
Ishikawajima-Harima Heavy Industries Co., Ltd.
Boilers R&D Dept., Boiler Plant Div.
2-16, Toyosu 3-chome, Koto-ku, Tokyo 135, Japan
and
Kouhei SUZUKI
Ishikawajima-Harima Heavy Industries Co., Ltd.
Advanced Fine Process Dept., Research Institute
1-15, Toyosu 3-chome, Koto-ku, Tokyo 135, Japan
ABSTRACT
Low NOx combustion technologies are introduced on low volatile coal firing with
relation to the application of a INPACT method, developed for low NOx firing which
was proved effective for the reduction of NOx emission on boilers fired steaming
coal(l,2). We fundamentally studied the characteristics of NOx emission for low
volatile coals using a reactor tube and/or a tunnel type test furnace, and these
studies were reflected to design of a 530 t/h boiler fired steaming coal with
blending anthracite or petroleum coke. The field test results showed INPACT method
also contributed both reductions of NOx emission and of unburned loss on low
volatile coal firing by multi-air staging through IAP (Interstage Air Ports) and
OAP (Over-fire Air Ports).
INTRODUCTION
Recently, in Japan, utility boilers and industrial boilers have been become to use
various kinds of coal from foreign countries for the purpose of diversifying fuel
supply source and for political reasons. Especially, industrial users intend to
use anthracite or petroleum cokes for saving fuel cost, by blending of a steaming
coal so as to improve the properties of low volatile coals having less flame
stability, less char oxidation rate and less NOx reduction than higher volatile
coals. For low NOx combustion technologies, IHI already developed INPACT (IHI NOx
Preventing Advanced Combustion Technologies) method, of which firing concepts are
illustrated in Figure 1 in comparison with other low NOx combustion methods, and
demonstrated remarkable NOx reduction of less than 100 ppm for a 600 MWe coal
firing boiler (1). However, blending low volatile coal was roughly estimated
lowering both of NOx reduction and of combustion efficiency, and then fundamental
studies were conducted to reflect to design of a 530 t/h boiler newly installed.
For the evaluations of NOx reduction on low volatile coal firing, we tested various
kinds of coal ranged from 9% to 42% as volatile content in coal using a reactor
tube and/or a tunnel type test furnace. On the fundamental test by the reactor
tube, we investigated both characteristics of coal-N release and of coal-N
3-111
Preceding page blank
-------
conversion to NOx, independently, for five (5) different coals. And then, we
attempted to demonstrate the differences of NOx emission and of the degree of NOx
reduction in fuel rich condition, using the test furnace for four (4) different
coals and three (3) blended coals of high and low volatile coals. For latter
study, we also investigated the influence of coal blending on flame stability in
horizontal firing system conventionally used, by quantitative measurements of
spectra emitted from pulverized coal flame.
Accordingly, we decided an optimum blend ratio of low volatile coal such as
anthracite or petroleum coke taking into consideration of NOx emission and unburned
loss. For minimizing unburned loss, a rotary type classifier was adopted to be
mounted in a roller type IHI-FW pulverizer so as to improve coal fineness. Firing
tests were conducted for 30% of blended coals of anthracite and of petroleum coke,
respectively, and it was demonstrated INPACT method fully available for NOx
reduction on low volatile coal firing, from the operational results of 140 ppm of
NOx emission. Especially, although the reduction of NOx emission was always
recognized to cause the increment of unburned loss, adopting INPACT method was
remarkably resulted to realize both reductions of NOx emission and of unburned
loss.
GENERAL CHARACTERISTICS OF LOW VOLATILE COALS
Generally speaking, in designing a horizontal-firing boiler fired low volatile
coal, it is important to estimate the characteristics of flame stability, of char
oxidation and of NOx emission. We studied such characteristics using a reactor
tube and/or a small size tunnel type test furnace. The fundamental studies were
conducted to clarify the following items.
(1) the degree of NOx reduction and NOx emission for the variation of volatile
content in coal
(2) an allowable blend ratio of low volatile coal and its governing factors, for
blended coal of low volatile coal and of steaming coal
(3) the effect of unburned loss on NOx emission under low NOx combustion
NOx formation on coal firing is well-known to be divided into two phases, that is;
volatile-NOx formed in early combustion stage and char-NOx in post flame. Firing
low volatile coal was roughly estimated to have a property of less NOx reduction
than steaming coals on staged firing as low NOx combustion, due to less volatile-
NOx, which could be easily decomposed into molecular nitrogen in early combustion
stage under fuel rich condition, and higher char-NOx, which had a difficulty of NOx
control by low NOx firing. Therefore, we firstly investigated the effect of
volatile content on coal-N conversion to NOx in fuel rich condition by the reactor
tube. And then, we conducted firing test so as to demonstrate the difference of
NOx reduction for the variation of volatile content in coal. In the latter study,
we also estimated the influences of blending low volatile coal on flame stability
and on unburned loss.
Test Facility
Reactor Tube. The characteristics of coal-N conversion were experimentally studied
in the down flow type electrically heated reactor tube (047 mm in diameter, 1.2m in
length). Pulverized coal particles (70 - 80% pass through a 200 mesh sieve) were
dried and fed in nitrogen stream into the reactor, controlled at constant wall
temperature. Oxygen was diluted with nitrogen and was fed into the top of the
reactor. For simulating staged firing oxygen also fed through alumina probe
3-112
-------
inserted from the bottom of the reactor. Product gas and unreacted char was
sampled though a water-cooled sample probe from the bottom of the reactor.
Nitrogeneous species such as NOx, NH3 and HCN were analyzed by a chemiluminescent
type monitor (HORIBA ENDA-C211) and FTD type gas chromatograph (SHIMADZU GC-7A)
with Porapak Q and Chromsorb 103 as column agents, respectively. Residual char-N
and ash content in unreacted char were also analyzed by an oxygen-circulating
combustion type analyzer (SUMIGRAPH HC-80). Five (5) different coals were examined
and their properties are listed in Table 1.
Test Furnace. Demonstrating combustion performances on low volatile coal firing,
we conducted firing test using the tunnel type test furnace. Single burner was
installed at the front end of the furnace and observation ports were distributed
along with the furnace. OAP ports were also located in a circle to realize stage
firing. Specification of the test furnace is as follows:
Type ; Horizontal, cylindrical and non-pressurized water jacket cooling
Capacity ; 200 kg/h of pulverized coal (equivalent to 1.8 MWt)
Dimension ; 1.3m in inside diameter, 7m in length
Figure 2 shows the general arrangement of the test furnace. The furnace was
divided into two parts by division bricks, they are; combustion chamber and heat
recovery area. Fire proof bricks and castable refractory were installed inside of
the furnace so as to simulate flame temperature of an actual furnace. We tested
four (4) different coals, of which properties are also listed in Table 1, included
anthracite as low volatile coal, and three (3) different blended coals of
anthracite.
Estimating flame stability, an attempt was made to quantitatively measure the
intensity of spectra emitted from coal flame near the ignition field, by means of a
monochromator as spectroscopic analysis. For only checking flame failure, flame
detector can be applicable, however, an analogical output of a conventional flame
detector having tendency indicated saturation value makes no comparison for our
purposes. The system comprises an optical probe, an optical fiber as a transmitter
and a monochromator with a signal analyzer. The optical probe was set on burner
face parallel to burner axis to view the ignition field. We mainly directed our
attention to spectra covering visible band due to minimize the transmission loss of
the intensity through the optical fiber, though pure quarts fiber of large size was
adopted.
Results and Discussions
Characteristics of Coal-N Conversion to NOx. Fundamentally evaluating NOx emission
for low volatile coal, we firstly investigate the effect of volatile content on the
activities of volatile-N release in pyrolysis process and in char oxidation
process, and then estimated the characteristics of coal-N conversion to NOx in fuel
rich condition, using the reactor tube.
Figure 3 shows both of volatile-N release, np^, and of volatile yield, np, plotted
against volatile content, at 1673°K of coal"pyrolysis temperature. Accordingly,
the following empirical relations were obtained.
where, Q and VMp denote the volatile enhancement factor, proposed by Mitchell and
Tarbel (3), and proximate volatile content as dry basis, respectively. Higher
np = Q VMp = 1.4 VMp
np>N = np0.629 = 1.24 (VMp)0.629
at 1673°K
(1)
(2)
at 1673°K
3-113
-------
volatile coal caused more release of volatile-N in the relation to volatile
content, as was expected.
Figure 4 shows the effect of coal pyrolysis temperature on the volatile enhancement
factor as volatile yield and on volatile-N release, for coal E. As the results,
the following empirical relations were also obtained.
Q = 3.4 exp(-2,870/RT) (3)
In (1 - np) = In (Np/No) = -34.6 exp (-12,000/RT) (4)
where, Np, No, R and T represent residual char-N, coal-N in raw coal, gas constant
and reactor temperature, respectively. As reported earlier researchers (3,4), with
the increment of temperature more volatile-N was released. Then, we also
investigated the activities of char-N release in char oxidation process. Char
conversion and char-N release were counted by both data of volatile yield/volatile-
N release in pyrolysis test and of coal conversion/coal-N release in oxidation
test. Figure 5 shows the relation between char conversion, nchar> and char-N
release, nchar.N- In the figure, the results of the char oxidation tests for coal
B and coal E are also plotted. Consequently, following correlations were
experimentally obtained.
nchar.N = 0.875 iichar + 0-05 at nchar < 0.4 (5)
nchar.N = nchar at nchar > 0.4 (6)
Eq. (5) possibly indicated that retained tar-N was released prior to char-N
release. If char retained volatile content, retained volatile-N was also released
prior to char-N release, according to the plotted data of char oxidation test for
coal B, of which char was roughly adjusted to retain 11.6% of ultimate volatile
yield. As the results, it was clarified char-N release approximately coincided
with char conversion.
Next, estimating the differences of NOx reduction for the variation of volatile
content, we investigated the characteristics of coal-N conversion to TFN (Total
Fixed Nitrogen), defined as the sum of NOx, NH3 and HCN, in fuel rich condition
with 1.1 second of residence time. We attempted to make relations between coal-N
conversion to TFN, CRTFN» and the ratio of coal conversion to ultimate volatile
yield, n/np, in first stage combustion. For all coals tested, as plotted in Figure
6, coal-N conversion to TFN could be minimized at the condition of 1.15 of ri/rtp,
becoming less than 5%. Firing low volatile coal in a boiler, consequently, we
could estimate NOx emission becoming higher than the case of steaming coals, at the
condition of same coal conversion in fuel rich region, due to higher value of n/np.
Coal-N conversion to TFN included both TFNs originating from volatile-N and from
char-N. Therefore, an attempt was made to estimate the contributions of conversion
to TFN from both of them, independently. Char-N conversion to TFN was counted from
the char oxidation tests, and the results were also expressed against the parameter
of n/np in Figure 6. Accordingly, in fuel rich condition, higher TFN conversion ir>
low volatile coal oxidation was evaluated to be mainly originated by char-N,
converted to NOx in post flame. This estimation may be consistent with the results
reported by Phong-anant et al (5) and by Chen et al (6).
Evaluation of NOx Emission. Both demonstrating NOx emission for low volatile coal
and estimating NOx level for blended coals of low volatile coal with steaming coal,
we carried out firing tests, using four (4) different coals and three (3) different
blended coals at the tunnel type test furnace. Semi-anthracite, as a low volatile
coal, was selected to attempt to burn independently, and was blended with a
steaming coal, coal F listed in Table 1. Blend ratio of semi-anthracite was
3-114
-------
decided to be ranged from 35% to 75%, taking into consideration of the variation of
volatile content in test coals, and their properties were listed below.
Blend A Blend B Blend C
blend ratio
(%)
35
55
75
volatile matter
(%)
18.1
15.3
12.6
fixed carbon
(%)
64.5
68.4
72.3
ash
(%)
14.8
13.4
12.0
moisture
(%)
2.6
2.9
3.1
coal-N as d.a.f
(%)
1.7
1.8
1.9
Figure 7 represents the relation between NOx emission and volatile content with a
parameter of staging air ratio. Firing low volatile coal showed the results
becoming lower NOx emission without staging combustion, and this was interpreted
NOx emission governed by volatile-NOx in air rich condition. Reducing air
stoichiometry by staged firing, however, low volatile coal was resulted to have the
property of less NOx reduction in comparison with the case of high volatile coal.
That is, NOx reduction was counted as follows;
w/o staging w/30% staging NOx reduction
(ppm)
coal E (31%) 630
coal F (23%) 600
coal G (20%) 480
coal H ( 9%) 550
blend A (18%) 600
blend B (15%) 590
blend C (13%) 500
(ppm)
(%)
150
76
190
68
210
56
350
36
200
67
250
57
260
48
Note: The values in parenthesis above show volatile content in coal.
Accordingly, low volatile coal was resulted to have higher NOx emission in fuel
rich combustion. And, this was recognized to be in good agreement with the results
of preliminary studies mentioned before.
Figure 8(a) shows an alternative expression for the effect of coal blending on NOx
emission, and blending high volatile coal with low volatile coal, NOx emission
could be reduced with almost linear relation to blend ratio, as was expected.
Moreover, for the purpose of checking a possibility of NOx reduction for low
volatile coal, we also estimated the effect of INPACT method on NOx emission, using
a low volatile coal of 18% content. Consequently, further 25% of NOx reduction
could be obtained by INPACT method, while INPACT method gave 35% of NOx reduction
in steaming coals such as coal E, as was already reported in other papers (1,2).
The difference of the effect of volatile content on NOx reduction was estimated due
to that of char-NOx formed in post flame.
Evaluation of Unburned Loss. For blended coal of low volatile coal, low NOx
combustion was estimated to accelerate the increment of unburned loss, due to less
char oxidation rate of low volatile coal (7). The evaluations of unburned loss
were also done with those of NOx emission in firing tests. Figure 9 represents the
relations between NOx emission and unburned loss for blended coals with their
mother coals. In this tests, we adjusted coal fineness to 75 - 80% pass through a
200 mesh sieve for all coals tested. Blending low volatile coal gave the increment
3-115
-------
of unburned loss, of which values were estimated with a linear relation to blend
ratio, as shown in Figure 8 (b).
Evaluation of Flame Stability. We also estimated flame stability for coals blended
semi-anthracite. Firstly, we compared a profile of flame spectra in visible
wavelength band, and then obtained a relation between the intensity of
600 nm spectrum, as a typical spectrum, and a non-dimensional parameter, x/D (x;
ignition distance, D; burner throat diameter), of ignition delay. Ignition
distance was roughly measured by eye through observation ports, the results are
plotted in Figure 10 (a). The ignition delay caused to weaken the intensity of the
spectrum detected parallel to burner axis, due to light scattering through
unreacted pulverized coal dust. In the tests, coal H and Blend C were resulted to
have a big ignition delay even though the air swirl was adjusted, and remarkable
reduction of the intensity was detected. Consequently, it was interpreted that the
spectrum intensity was available for estimating ignition delay, which was in close
relation to flame stability. Figure 10 (b) shows an alternative expression of a
relation between the spectrum intensity and volatile content in coal tested.
Volatile content in coal gave a significant effect on flame stability, and
allowable volatile content was roughly estimated to around 15% in horizontal firing
system conventionally used.
Flame stability was also affected by operational parameters such as the flow ratio
of primary air to coal at coal nozzle, air velocity through burner throat, air
temperature, swirl intensity and so on. Taking into consideration of strict
limitation of NOx emission in Japan, we only investigated the effect of staged
firing, which was in a relation to the second parameter mentioned above, on flame
stability. Reducing air flow through a burner showed to improve spectrum intensity
for low volatile coals, whereas high volatile coals gave no deviations of the
intensity by air staging, as was represented in Figure 11. Accordingly, reducing
air velocity through a burner was experimentally recognized to give better effect
on flame stability.
APPLICATION RESULTS ON A 530 T/H BOILER
Designing of a 530 t/h boiler fired steaming coal with low volatile coal, the
results of fundamental studies were reflected as follows;
(1) adopting INPACT method to realize NOx emission less than 180 ppm for both
blended coals of anthracite and of petroleum coke
(2) adopting a rotary type classifier mounted on a roller type IHI-FW pulverizer
for minimizing the increment of unburned loss by the improvement of coal
finess
(3) adopting BUF (Boost Up Fan) for staging air and adopting the mechanism
adjusting air velocity through 0AP ports for better mixing of staging air and
of unreacted char in complete combustion region of furnace
With regard to blending low volatile coal, we made a decision of an allowable blend
ratio being around 30%, taking into consideration of NOx emission and of the degree
of unburned loss. Figure 12 shows the general arrangement of a 530 t/h boiler, of
which specification is typically listed below.
Type ; IHI-FW natural circulation type
Evaporation ; 530 t/h (147.2 kg/s)
Steam cond. ; 131 kg/cm^G (12.85 MPa) x 541°C at superheater outlet
29.7 kg/cm2G (2.91 MPa) x 541°C at reheater outlet
3-116
-------
Draft system
Fuel
Pulverizer
Burner
Air ports for
staging
balanced system
steaming coal and blended coal with 30% of anthracite or of
petroleum coke
4 sets of IHI-FW MBF pulverizer
type IHI DF (Dual Flow) type burner
16 burners (4 raws x 2 stages, opposed arrangement)
OAP (in each front and rear wall)
SAP (in each front and rear wall)
IAP (in each front and rear wall)
Note SAP; Side-overfire Air Port, IAP; Interstage Air Port
Typical Feature of the Boiler
Adopting INPACT method, staging air through OAP and SAP ports was branched from air
duct connected to air preheater outlet and was boosted up by a BUF, of which
rotation frequency was controlled in relation to flow rate by a hydraulic coupling
connected to a fan for saving of running cost. SAP ports were located at both
front and rear walls between OAP port and each corner, for the purpose of promoting
char oxidation rate in the area near furnace side wall. IAP ports, as typical
ports of staging air in INPACT method, were also located at the elevation between
top raw of burner and OAP ports so as to realize multi-air staging, which was
resulted to be effective for both reduction of NOx emission and of unburned loss.
Staging air through IAP was branched from common air duct of burner windbox and was
controlled by dampers. The locations of OAP, SAP and IAP ports were decided in
consideration of residence time in firing region. Each staging port had a damper
at inlet for adjusting the distribution of staging air over furnace width.
Burner spacing, as a factor of burner arrangement, was also designed taking into
consideration of liberation rate in firing region, according to the fact that
elevating liberation rate gave elevating flame temperature and was resulted to
release more volatile-N and to reduce unreacted char in firing region, as was
interpreted in Figure 4. For adjusting air stoichiometry to opitimum condition for
each burner, IHI has been adopting separate-windbox system, in which air flow was
separately fed into each burner from common air duct. Boundary air system was also
adopted in the boiler to make oxidizing condition near furnace walls, by
introducing fresh air through furnace bottom and through both side walls near
furnace hopper.
We adopted a motor-driven rotary type classifier mounted on a pulverizer instead of
a conventional (adjustable vane) type classifier. A rotary type classifier with
rigid vane has been originally developed for the requirement of producing finer
materials on a cement plant, and their experiences were reflected to apply for
coals. Coal fineness was adjusted by rotation frequency of motor with VVVF
(Variable Voltage and Variable Frequency) control, and a typical result was plotted
in Figure 13.
Results and Discussions
We carried out firing tests for a typical steaming coal and for two blended coals
of anthracite or of petroleum coke. Coal properties are listed in Table 2.
Anthracite and petroleum coke had 11% and 13% of volatile content, respectively,
and this anthracite was strictly classified to semi-anthracite in ASTM coal rank.
Coal blending was done in the coal conveyance system with running weighers. Prior
to performance tests for NOx reduction, we carried out combustion adjustment in
consideration of air distribution through each burner and staging air port.
3-117
-------
Figure 14 shows the data of NOx emission against BUF rotation as an operational
parameter of staging air flow, for a steaming coal and for a blended coal of
anthracite. In this case, IAP air was adjusted to minimum flow. With the
increment of staging air, NOx emission for coals tested was reduced with a almost
linear relation to BUF rotation, and low volatile coal gave higher NOx emission in
comparison with the case of a steaming coal. The difference of NOx emission
between a steaming coal and a blend coal was 20 - 40 ppm, and was in good agreement
with expected value in accordance with the data of firing tests at the tunnel type
test furnace.
Next, we investigated the effect of multi-air staging through IAP ports on NOx
emission for the purpose of further NOx reduction on blended coal firing. Figure
15 shows the results plotted against damper opening of IAP ports. Increasing IAP
air was resulted to further NOx reduction, as similar as the results of another
boiler (1). NOx emission for blended coal was resulted to be around 140 ppm and
this showed INPACT method was also effective for NOx reduction on low volatile coal
firing.
We also checked unburned loss in a series of the firing tests. For a steaming
coal, unburned loss was only 0.2 - 0.4% even in low NOx condition. For blended
coal of low volatile coal, unburned loss was resulted to become higher. Since
unburned loss had a close relation to NOx emission, we evaluated unburned loss
against NOx emission as plotted in Figure 16. The differences of unburned loss for
both cases with or without multi-air staging adopted in INPACT method were also
expressed in the figure. The degree of IAP air flow rate was represented by that
of damper opening for IAP ports, and we noted that strengthening of multi-air
staging could reduce unburned loss for both blended coals. Consequently, these
results showed INPACT method to give better effect for improving combustion
efficiency under the condition of low NOx firing.
Plant users intend to take special care how reducing condition gives furnace
corrosion in INPACT method or even in conventional staged firing. Generally
speaking, furnace corrosion has a close relation to sulphur content in coal used.
According to our experiences, a high sulphur domestic coal of more than 3% caused
corrosion troubles both for a low NOx firing boiler and for a conventional firing
boiler, however, controlling sulphur content less than 1% by coal blending was
resulted to reduce corrosion troubles remarkably. Fortunately, most of imported
coals are low sulphur coals in Japan, and for the boiler applied INPACT method we
also confirmed that no corrosion phenomena was observed on tube surface in water
wall after one (1) year continuous operation.
CONCLUSION
We clarified both characteristics of NOx emission and of unburned loss for low
volatile coal and then applied INPACT method to a low volatile coal firing boiler.
Following evaluations were summarized from the results in the fundamental tests and
in the field tests of a 530 t/h boiler newly installed.
(1) INPACT method was also fully effective for NOx reduction and for minimizing
the increment of unburned loss on low volatile coal firing.
(2) We clarified the relation between NOx emission and volatile content in coal
included low volatile coal. Consequently, low volatile coal has the
properties of higher NOx emission in comparison with other steaming coals due
to higher char-NOx emission which have a difficulty of NOx control.
3-118
-------
ACKNOWLEDGEMENTS
We would like to express our gratitude to our customer for applying INPACT method
to a boiler newly installed.
REFERENCE
1. T. Kiga, et al. Furnace Design and Application on Low NOx Pulverized Coal
Combustion, Joint Svmp. on Stationary NOx Control, No. 2D, March, 1987
2. S. Miyamae, et al. Low NOx Pulverized Coal Combustion Technology for Large
Utility Thermal Power Plant, ISCC, Sept., 1987
3. J. W. Mitchell et al. A Kinetic Model of Nitric Oxide Formation during
Pulverized Coal Combustion, AIChE, Vol. 28, No. 2, 1982, pp. 302, 311.
4. J. 0. Wendt, et al. Mechanism Governing the Destruction of Nitrogeneous
Species during Fuel Rich Combustion of Pulverized Coal, Proc. 19th Svmp.
(Int.) on Combustion. 1982, pp. 1243, 1251.
5. D. Phong-anant, et al. Nitrogen Oxide Formation from Australian Coals,
Combustion and Flame 62, 1985, pp. 21, 30.
6. S. L. Chen, et al. Influence of Coal Combustion on the Fate of Volatile and
Char Nitrogen during Combustion, Proc. 19th Svmp. (Int.) on Combustion, 1980,
pp. 1271, 1280
7. K. Takahashi, et al. Experimentally Determined Reactivity Index for Assessing
Unreacted Carbon in Pulverized Coal Boilers, ISCC. Sept., 1987.
3-119
-------
Conventional
Air Staging
R.B;Reburning
Figure T. Comparison of Low NOx Firing
-7000 mm-
Flue Gas Exit
[L
Water Jacket
#
c|) (]) c[) c[) (j)i
£
Refractory
Over Air Ports |j
(j) (\) CD (j) c() (j) (j)
Wind Box
F
Burner
4-
Observation Ports,
E-
ifr
Pulverized Coal
Figure 2. General Arangement of Tunnel Type Test Furnace
3-120
-------
<0
0)
TJ r~
r- 0)
0J D£
> z
I
OJ 0)
-p -p
fO
'o o
z
* i
?>
1.0
0.8
0.6
0.4
0.2
^P-N
<
—^
/°
o
10 20 30 40 50
Volatle Content (dry basis) (%)
60
Figure 3. Relation between Volatile Content in Coal and
Both of Volatile Yield and of Volatile-N Release
o
Q.
-3.0
-2.0
-1.0
-0.5
-0.2
T (xlOOO) (°K)
2.2 2.0 1.8 1.6 1.4 1.2
1
1
—I
V
—r . _
*8r
}—°"
s.
1 n(Np/No"
5 6 7 8
10000/T (1/K)
3.0
c
ai
1.0 §
u
c
-------
a>
i/i
to
V
Q)
D£
1.0
0.8
0.6
- 0.4
i
u
at
M
U
O coal A v coa^ E
^ coal B + char of coal B
A coal C ~ char of coal E
0.2
/
/
4
S
/ „
A
/ J
w
0 0.2 0.4 0.6 0.8 1.0
V char , Char Conversion
Figure 5. Relation between Char Conversion and
Char-N Release
1.0 1.2 1.4 1.6 1.8 2.0
7/7P , Ratio of Coal Conversion to Volatile
Yield
Figure 6. Relation between Coal-N Conversion to TFN and
Ratio of Coal Conversion to Volatile Yield
3-122
-------
O w/o air staging
^ w/15% air staging
^ w/30% air staging
Figure 7. Relation between NOx Emission and Volatile
Content in Coal
700
Qi
IS 500
-P
O
d>
u
u
c
u
dP
(n 300
o
100
O
_ O
P
U
L
0 20 40 60 80 100
Blend Ratio of Anthracite (%)
CO -i
1
O w/o air staging
A w/30% air staging
L
;>i
A
k
n-f
^9-
—8
-*"U
0 20 40 60 80 100
Blend Ratio of Anthracite (%)
(a) NOx Emission
(b) Unburned Loss
Figure 8.- Effect of Coal Blending on NOx Emission
and Unburned Loss
3-123
-------
O w/o blending (Coal F)
^ w/35% blending (Blend A)
•A w/55% blending (Blend B)
~ w/75% blending (Blend C)
0 semi-anthracite (Coal H)
NOx (02 6% corrected) (ppm)
Figure 9. Relation between NOx Emission and Unburned Loss
for Blended Coal of Low Volatile Coal
x/D, Parameter of Ignition Volatile Content (%)
Delay
(a) Relation between Flame Intensity (b) Relation between Flame Intensity
and Ignition Delay ^nd Volatile Content in Coal
Figure 10. Evaluation of Flame Stability by Measurement of
Spectrum Intensity Emitted from Coal Flame
3-124
-------
+10
-10
-20
-30
>——
Jr—°—Y' V
Coal E
^Coal G
•
Coal H <
~ X
~
0 10 20
Staging Air Ratio (%)
+10
30
CQ
T3
CD
>
s -10
cn
C
Q)
r0
r-*
Cu
-20
-30
Blend B
r
/
Blend C
f
0 10 20
Staging Air Ratio (%)
(a) Mother Coal (b) Blended Coal
Figure 11. Effect of Air Staging on Flame Stability
30
Burn£I: iff
Burner
Figure 12. General Arrangement of a 530 T/H Boiler
3-125
-------
"J °
o o 60
O fM
(Hz)
Rotation Frequency of Classifier
Figure 13. Relation between Rotation Frequency of Classifier
and Coal Fineness
_ 300
a.
f 250
§ 200
o
<#>
o 150
X
100
400 600 800 1000
BUF Rotation (rpm)
Figure 14. NOx Reduction by Air Staging through 0AP/SAP
Ports, expressed as a parameter of BUF rotation
I
V
... T
1
xl
^blended
coal of
anthracite
v.
i
steaming
coal
i
'x
!
3-126
-------
g, o OAP/SAP damper opening 25/50 %
0 OAP/SAP damper opening 35/60 %
BUF Rotation (rpm) IAP Damper Opening (%)
(a) Blended Coal of Anthracite
Ch
V 250
<1)
200
O 150
100
500
OAP/SAP damoer opening 25/50
OAP/SAP damper opening 40/65
\
°v
?J
—c
600 700 800
BUF Rotation (rpm)
900 0 20 40 60
IAP Damper Opening (\
(b) Blended Coal of Petroleunm Coke
Figure 15. NOx Reduction by Air Staging through IAP Ports
for Multi-air Staging of INPACT Method
3-127
-------
O IAP damper 5%
m IAP damper 30%
_ 3
w
w
a 2
?
U
i 1
3
0
^11
ton
i
100 150 200 250
NOx (02 6% corrected) (ppm)
O IAP damper 15%
0 IAP damper 55%
cn
w
3 2
1
y
% 1
0
100 150 200 250
NOx (02 6% corrected) (ppm)
(a) Blended Coal of Anthracite
(b) Blended Coal of Petroleum
Figure 16. Relation between NOx Emission and Unburned Loss
3-128
-------
Table 1
COAL PROPERTY
(for fundamental study)
Coal
A
B
C
D
E
F
G
H
HEATING VALUE*
(MJ/kg)
27.9
28.3
18.3
20.6
29.1
27.72
31.15
30.5(
Proximate Analysis*
Inherent Moisture
(%)
2.9
3.1
22.8
22.0
3.7
2.1
1.1
3.5
Volatile Matter
(%)
29.5
26.2
34.1
41.0
31.0
22.9
20.1
9.1
Fixed Carbon
(%)
51.6
56.7
32.2
34.7
49.8
57.7
69.0
77.2
Ash
(%)
16.0
14.0
10.9
2.1
15.5
17.3
9.8
10.2
Ultimate Analysis**
Carbon
(%)
69.2
71.7
60.5
67.3
71.0
70.8
80.3
82.1
Hydrogen
{%)
4.5
4.3
4.1
4.3
4.7
3.9
4.7
3.3
Nitrogen
(%)
1.5
1.5
2.0
1.0
1.4
1.3
1.1
1.7
Sulphur
{%)
0.3
0.6
2.1
0.0
0.5
0.3
0.3
0.6
Oxygen
{%)
8.0
7.4
17.2
24.2
6.4
6.1
2.7
1.7
Note: Coal A, B, C, D and E were used for the research using a reactor tube, and
Coal E, F, G and H were burned at a tunnel type test furnace.
* Air dry basis
** Dry basis
Table 2
COAL PROPERTY
(for field test)
Coal or Coke
Heating Value*
Proximate Analysis*
Inherent Moisture
Volatile Matter
Fixed Carbon
Ash
Ultimate Analysis**
Nitrogen
Sulphur
Steaming Coal
(MJ/kg) 29.35
(*)
(%)
(*)
(%)
(%)
(%)
3.8
34.7
51.9
9.6
1.5
0.6
Anthracite
29.31
2.0
11.0
74.0
13.0
1.8
1.9
Petroleum Coke
35.84
0.6
12.8
86.4
0.2
1.0
4.9
* Air dry basis
** Dry basis
3-129
-------
Session 4
MANUFACTURER'S UPDATE
Chairman: G. Offen, EPRI
-------
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
NOx CONTROL:
THE FOSTER WHEELER APPROACH
Joel Vatsky
Director, Combustion and Environmental Engineering
Foster Wheeler Energy Corporation
Perryvilie Corporate Park
Clinton, N.J.
INTRODUCTION
The approach to achieving low NOx emissions, employed by Foster Wheeler Energy Corp., is
based upon utilizing low NOx burners as the primary control method. The Controlled Flow/Split-
flame low NOx burner has been in field service on industrial units since 1976 and in utility service
since 1979. It has been employed as original equipment on new steam generators and has been
retrofitted to several industrial and utility units.
A major advantage of this burner is its flame shape which is similar to that of pre-New Source
Performance Standards turbulent burners. Consequently, it has negligible effect on radiant zone
absorption patterns and furnace exit gas temperature. As such it has been installed on pre-NSPS
units and not degraded boiler performance or efficiency.
The CF/SF burner can be installed on any wall-fired steam generate -hether originally equipped
with circular or cell-type burners. It can be retrofitted on most circular ^-ner equipped units with
no pressure part or structural modifications (eg. as a "plug-in") simply replacing the existing
secondary air register and coal injectors with the new designs.
Typically, NOx can be reduced by 50%, although reductions of up to 60% have been attained,
without simultaneously utilizing supplemental NOx controls such as staging ports. Experience to
date indicates that percent NOx reduction is not strongly dependent on fuel characteristics,
although absolute NOx level is.
Significant additional NOx reductions, up to a total of 80%, can be attained by combining a modern
overfire air system with the CF/SF burner. Field tests in the early 1980's on two utility steam
generators, and subsequent evaluations on our 80 million Btu/Hr Combustion and Environmental
Test Facility, have demonstrated NOx levels below 0.2 lb/million Btu.
4-1
-------
This paper describes the low NOx systems and summarizes the field and developmental results.
CONTROLLED FLOW/SPLIT-FLAME LOW NOx BURNER
Foster Wheeler chose to develop a low NOx burner based on internal staging characteristics (i.e,
fuel and air staged within the burner while retaining historical flame shapes) rather than utilize
delayed combustion (long-flame) or external staging (tertiary air port) burner concepts. The
internally staged CF/SF concept does not adversely affect either heat absorption profiles of the
furnace, the FEGT or the furnace structure as can the delayed combustion concept or the
externally staged concept, respectively.
The Controlled Flow/Split-flame low NOx burner, shown schematically in Figure 1, is a dual register
design with the registers arranged in series. The inner register controls the swirl and stoichiometry
around the coal injector while the outer register affects the flame envelope's shape.
Controlled Flow Split Flame Burner
Figure 1
Surrounding the register assembly is a perforated plate air hood which improves the circumfer-
ential distribution of secondary air entering the outer register. By measuring the pressure drop
across each burner's perforated plate, an index of the secondary air flow can be obtained. This
index aids in balancing burner stoichiometries to minimize excess air requirements. Therefore,
since secondary air distribution within the windbox is optimized, the common windbox can be
retained and there is no need to modulate burner air flow controls.
4-2
-------
The coal injector consists of a tangential scroll inlet, annular fuel barrel and split-flame nozzle with
axially movable inner tip. Adjusting the inner tip will vary the primary air/coal mixture's velocity so
that it can be optimized with respect to the secondary air.
Control of the primary air/coal distribution and velocity is an integral part of achieving low NOx
levels with acceptable flame conditions. The coal is collected into four concentrated streams
which yield substoichiometric flames in the near-throat region while the burner operates at normal
excess air levels.
Advantages of the CF/SF low NOx burner design, which enhance its attractiveness for use on
retrofitted boilers as well as new units, include:
(1) Does not rely on long, narrow delayed mixing flame to achieve low NOx.
(2) Fully compatible with existing burner management, flame scanning and ignition systems.
(3) "Plug-in" retrofitability: No pressure part or structural modifications are required for burner
installation.
(4) Same burner pressure drop as historically used with turbulent burners (2-4" H2O).
(5) Independent control of secondary air flow and swirl at each burner thus allowing air/fuel
ratio to be optimized at each burner without modifying swirl; and without modulating
secondary air flows.
In addition to the low NOx burners, a set of Boundary Air ports, requiring a total of four openings
in the tubewalls for opposed fired units, is offered. Their location is schematically shown in Figure
2. The ports inject windbox air along the sidewalls in order to produce a high excess O2
concentration which will not vary significantly as pulverizers are taken in and out of service over
the boiler's upper load range. In combination with the burner's ability to adjust secondary air flow
distribution both vertically and horizontally within the windbox, the boundary air system is
intended to maximize the overall flexibility to control oxygen distribution during combustion.
EMISSIONS DATA
A substantial body of operational and emissions data for industrial and utility steam generators,
equipped with the CF/SF burner, has been obtained over the past ten years. Some of the more
interesting results are those obtained on steam generators designed for use with turbulent
burners and subsequently retrofitted with the CF/SF low NOx burner.
4-3
-------
Boundary Air System
Figure 2
In 1984 a retrofit was made to a 110,000 Ib/hr 4-burner steam generator at a Du Pont chemical
plant in Martinsville, Virginia; primarily for the purpose of demonstrating SO2 control via sorbent
injection. The demonstration was funded by Consolidation Coal Company; FWEC's part being
the supply of low NOx burners with integral limestone injectors. NOx was reduced 50% without
any measurable change in Furnace Exit Gas Temperature as determined by FWEC's boiler
performance program and in-situ HVT measurements performed by Conoco.
FWEC has also made several retrofits to utility steam generators. Figure 3 summarizes NOx
emission results from three utility units with the data from the 110,000 Ib/hr Du Pont Martinsville
unit included for comparison.
The first unit shown is the 360 MW San Juan #1 of Public Service New Mexico. When sold, after
the 1971 State of New Mexico NOx regulations took effect, it had to meet the new regulation of
0.45 lb/million Btu (in contrast to the Federal EPA limit of 0.7). However, when started up in 1976
FWEC's low NOx burner had not yet been developed. Consequently, the unit was equipped with
turbulent burners with NOx control via overfire air only. Emissions were about 0.95 lb/million Btu
with OFA ports closed (and less than 0.7 with OFA ports open). The State of New Mexico
Regulation (0.45) was not attainable.
4-4
-------
San Juan #1 was retrofitted with Controlled Flow/Split-flame low NOx burners in 1979 with NOx
reduced to about 0.4 lb/million Btu, without overfire air. Subsequent to this success, which was
a true retrofit, the CF/SF burner was installed on several units which were in the construction
phase, in keeping with FWEC's policy of providing our newest technologies where practical. One
of these units, all of which had been designed for turbulent burners, is a 525 M W steam generator.
Data for this unit, with and without overfire air, is also shown on Figure 3. There are several other
steam generators, retrofitted in the early 1980's while being constructed, ranging from 250 MW
to 700 MW which show similar results. Most do not have overfire air ports and those which do
have them do not need them to meet the relevant NSPS requirements.
nob
~'191
•mi
1.0 _
T
04 _
&• .
0.7 .
Ai
IM m
04 .
&4 .
L
L
OJ _
—
—
02 _
0.1 .
OFA
CAPACITY
COAL
GL
HOW
CL
¦28
Of
MO MW
I> HT.
LHT.
Low NOx Burner Retrofits
Figure 3
The most recent retrofit is the 650 MW unit also shown on Figure 3. This is the Pleasants Station
Unit #2 of Allegheny Power Corp. The following section discusses this unit in more detail.
PLEASANTS UNIT #2 LOW NOx BURNER RETROFIT
Allegheny Power System's Pleasants Power Station has two identical Foster Wheeler supercritical
once-through steam generators (Figure 4). These 5,035,000 Ib/hr single reheat units supply
steam to 626 MW (net) Allis-Chalmers turbine-generators and were designed to meet the
requirements of the 1971 new source performance standards, which limit NOx emissions to 0.7
lb/million Btu. They are equipped with wet scrubbers to control sulphur emissions.
4-5
-------
FOSTER 0 WHEELER
Pleasants Unit No. 2
Figure 4
The steam generators were equipped with overfire air ports as the only means of NOx control.
NOx emissions in Unit Nos. 1 and 2 averaged between 0.6 and 0.7 lb/million Btu after start-up in
1978 and 1980, respectively. Unit No. 2 has less margin than Unit No. 1, however, and as a result,
Unit No. 2 experienced periodic derates to avoid exceeding the regulatory limit. It was determined
that the best method of eliminating the derates due to NOx limitations was to install low NOx
burners.
The Controlled Flow/Split-flame low NOx burner was installed in Unit No. 2 during a four week
outage in 1986. In addition to eliminating NOx derates, it was anticipated that the new burners
could help boiler performance in the following areas:
(1) Induced draft fan limitations - The ID fans were designed with insufficient capacity margins
to accommodate some fuel/moisture/load combinations. The low NOx burners operate with
less excess air, thus providing greater capacity margin.
4-6
-------
(2) Furnace sidewall corrosion - The low NOx burners with lower Boundary Air ports admit
more combustion air in the furnace zone by eliminating overfire air thus providing more
oxidizing air for the sidewalls.
(3) Air register maintenance - Reduced maintenance costs would result by not having to
maintain air port and burner registers.
(4) Boiler efficiency - Improved boiler efficiency would result due to reduced dry gas loss re-
sulting from low excess air operations.
The steam generator has 24 burners with 12 on each of the front and rear walls. Each of the six
rows of four burners are supplied by a ball mill pulverizer. The unit has eight overfire air ports
which were used to admit air for NOx control.
The original intervane burners were removed so that the new low NOx burners could be installed.
The intervane burner is a basic design with one air register and the coal nozzle consisting of an
inner and an outer barrel. Coal enters the annulus between the barrels tangentially through the
scroll. The primary air/coal stream mixes with the secondary air in a highly turbulent manner.
The low NOx system installed at Pleasants includes the low NOx burners and boundary air ports
which were installed in the four corners of the furnace above the ash hopper slope as shown in
Figure 2. These air ports inject windbox air along the sidewalls in order to produce high O2 levels
between the wing burners and sidewalls, and is intended to reduce furnace sidewall corrosion
that occurs when a local reducing atmosphere exists in this region.
An important feature of the low NOx system is the ability to reduce overall excess O2 levels. This
can be achieved because of the ability to balance burners in the windbox and by closing the
overfire air ports. In order to be able to determine this minimum excess O2 level, carbon monoxide
(CO) monitors were added to the unit. It was felt that the O2 could be lowered to a point before
CO appeared in unacceptable levels, providing that no other adverse conditions existed (e.g.,
severe slagging, sidewall wastage).
The low NOx burner system was installed during a four-week outage that occurred in February,
1986. The burners were designed so that no pressure part modifications were necessary in the
burner throat area. The existing burners, air registers, and scrolls were removed. The new air
register assemblies, inner and outer barrels were installed and the existing scrolls were reused.
Modifications to the outer windbox wall were completed to accommodate the new manual
actuators, to reposition the main register drive motors. A modification was done to the burner
4-7
-------
refractory that involved removal of some refractory, leaving a leaner profile. This was intended
to reduce the chances of slag accumulations, or "eyebrows." During this outage, the overfire air
ports were repaired and left operable in case overfire air was needed to meet the lower NOx
performance guarantee. However, the overfire air ports have not been used since the installation
of the low NOx burners.
Emission Test Results
The Pleasants No. 2 unit was originally brought into commercial operation in 1981. Installed within
the common windbox, as shown in Figure 4, the overfire air dampers remove secondary air from
the burners in order to decrease burner stoichiometry. This design was set to produce an NOx
margin of about 10% below the guarantee level, based on the contract performance coal.
However, NOx emissions will vary with coal constituents, load, excess air and the number of
pulverizers or burners out of service. With this configuration, burner register position will also
affect NOx, primarily by changing windbox pressure which modifies the amount of overfire air
flow.
Figure 5 compares the NOx emission results as a function of excess O2 for the turbulent intervane
burners with and without overfire air. This data, taken during and shortly after the initial start-up,
Pleasants #2
NOx vs. 02 Intervane Burner
0.4 -
I I I I I I
1-0 u
cxcssa o2 («)
Figure 5
4-8
-------
was typical of subsequent operation. Although there were periodic incidents requiring unit load
to be reduced in order to prevent NOx exceeding the regulatory limit as indicated by the stack
monitor, checks by both FWEC and independent test contractors confirmed the lower curve of
Figure 5.
After installation of the low NOx burners and boundary air ports, during a four week January/Feb-
ruary, 1986 outage, start-up commenced in late February. Initial burner settings (inner and outer
register and movable coal nozzle) were FWEC's standard start-up positions. After reaching full
load with all burners in service and with overfire air and boundary air ports closed, it was found
that the NOx level was in compliance with the revised guarantee, 0.6 lb/106 Btu.
A two-week burner optimization and low NOx monitoring period was undertaken. This produced
the NOx results summarized in Figure 6. With overfire air ports closed, NOx was reduced to about
0.4 lb/106 Btu. With overfire air ports open, the NOx level, during test conditions, was reduced
to about 0.33 lb/106 Btu. Figure 7 summarizes and compares the performance of the intervane
and low NOx burners as a function of overfire air port damper position: NOx is reduced a total of
67% from the completely uncontrolled condition.
NOx
11/10* ITU
&J
onx
i.a
¦i i i
u
¦XCI39 Oj(%)
Pleasants No. 2-NOx vs. 02 Low NOx Burner
Figure 6
Table I summarizes typical flue gas measurements for the two burners. Unburned carbon loss,
being a key measure of efficiency, was measured after the low NOx burners were optimized. As
4-9
-------
OVCH Fine AIR DAMPER POSJTION (% OPeNI
Pleasants No. 2 - NOx vs. OFA
Figure 7
shown on Table I, the unburned carbon level in the flyash was below 2.5% (measured in a loss
on ignition test). For the coals used, this translates to less than 0.35% as efficiency loss, which
is within the normal range for pulverized bituminous coal-fired boilers.
TABLE I.
EMISSION SUMMARY: PLEASANTS UNIT NO. 2
Burner
Intervene
Low NOx
NOx
CO
uc
NOx
CO
uc
Units
lb/106
Ppm
%
OFA: Closed
0.95 _±. 0.05
40
2.5
0.41 + 0.02
40
2.5
OFA: Open
0.65 +. 0.05
40-60
2.5
0.33
100
--
*No data was taken as this was a short-term test condition designed only to observe NOx change.
+ Measured at economizer exit.
The fuel in use at the station is a middle Kittanning high volatile bituminous coal. Table II compares
the contract boiler performance coal with the fuel used during the 1981 start-up and shakedown
and the fuel currently being burned.
4-10
-------
TABLE II.
COAL ANALYSIS: PLEASANTS STATION
Constituent
Performance
1981
1986
FC/VM
1.36
1.31
1.45
Nz
1.0
0.9
1.3
FC
43.6
47.9
50.4
VM
32.1
36.6
34.7
Ash
16.8
13.6
11.7
H2O
7.5
2.3
3.2
C
62.3
69.4
7.0
H
4.1
5.1
5.0
0
5.1
6.4
5.4
S
3.2
2.7
3.4
HHV (Btu/lb)
11,500
13,600
12,700
The most significant differences between the current fuel's characteristics and those of the earlier
fuel are:
• Increase in fixed carbon to volatile matter ratio of about 10% (1.31 to 1.45)
• Increased fuel nitrogen content of about 40% (0.9 to 1.3)
Both of these changes would tend to increase NOx emissions, all other factors being equal.
Operational Results
In parallel with achieving low NOx levels is the importance of not degrading steam generator
performance and efficiency, nor of creating operational difficulties for the station's personnel.
Areas of particular concern are discussed below, along with actual observations:
(1) Burner Pressure Drop: Any increase in burner resistance, either on the primary or second-
ary air side, would result in higher power consumption.
• Primary air fan power consumption is in the same range as with the original
coal injectors, indicating no significant change to nozzle flow resistance.
• Windbox-to-furnace pressure drop is the same as before: at full load with all
burners in service and overfire air ports closed Ap is about 3.0 - 3.5" H2O.
4-11
-------
With OFA open, it is approximately 1.5" H2O, both at normal excess air of 20%
(3.6% O2).
(2) Furnace Exit Gas Temperature: Some types of low NOx systems rely on long, delayed
mixing flames which have the potential for increasing FEGT. If this were to occur on the
Pleasants unit, the resulting increase heat head across the convection pass would cause:
• Higher spray flows to control superheater temperature
• Shifts in reheat control damper positions (the unit uses a parallel pass
convection arrangement for reheat control) to prevent high reheat tempera-
tures. Higher air heater inlet temperature could also result.
None of these potential adverse effects have been noted over the past 14 months of low
NOx burner operation. Consequently, the absorption patterns which can be observed and
measured with normal steam generator performance techniques show no change due to
the low NOx burners.
3. Slagging and Fouling: Since the coal can be classified as having a medium to high slagg-
ing potential, additional ash deposition would be of concern, particularly during low NOx
firing.
• Furnace Slagging: With the original burners, the walls were covered with a
thin, light coating of ash. However, very large "eyebrows," up to 6' in length,
would form over the burner throat openings. In many cases, the flames would
be burning through a "tube" of slag. The burner throat refractory modifications
made during low NOx burner installation have resulted in the eyebrows being
reduced in size to the point where they no longer affect the flame or cause
other adverse impacts.
Ash coverage on the walls is essentially the same as before. The visual
observations made are somewhat subjective since there is no adequate
means of directly quantifying ash extent and quality. However, the thermal
measurements discussed in (2) above also confirm that there has been no
increase in furnace deposition. A significant increase in furnace wall coverage
would reduce absorption in this region thereby causing high FEGT with
previously noted adverse consequences.
• Fouling: Visual observations and evaluation of the thermal measurements
indicate no increase in superheater fouling or convention pass deposition.
Again, this implies that extended flames and upper furnace combustion have
not occurred.
Thus, after three years of operation, there have been no unsatisfactory or adverse side-effects
resulting from the low NOx burner retrofit.
4-12
-------
FIELD DATA COMPARISONS
NOx emissions from coal combustion consist of two components: thermal fixation of atmospheric
nitrogen and conversion of fuel-bound fixed nitrogen to NO. Furnace thermal conditions will affect
thermal NOx generation while burner type will affect both thermal and fuel NO. Foster Wheeler
has determined the characteristic relationships between NOx emissions, furnace configuration
and combustion intensity as represented by the Burner Zone Liberation Rate (BZLR is a measure
of the heat input to the burner zone, the higher this value the higher the furnace temperature).
Figure 8 is a summary of NOx emissions from turbulent and low NOx burners. The upper two
curves represent the variation of NOx with BZLR for single wall and opposed fired units equipped
with intervane burners. The bottom curve is for the CF/SF low NOx burner. Data from industrial
and utility units and FW's test facility (CETF) are shown. There is very good correlation between
data points and predictive curves.
KCV
NO*
1.4
i.a
i.i
i.l
1.0
0.1
o.a
0.7
LB/ioAbtU o.a
M
•.4
o.a
o.a
o.t
UTMJTV
• hi uw niMMni **
© ais uw imqii wmj. men
• aaouw IIIM1* WAU.FMID
• aaa uw orroaao rutin
•INOU WALL FMID UNIT a
MWIWl
q 110.000 lb/w 4 euwea
o na.ooo U/IW 4 aUWMEM
~ CETF
OPPOSED FMED UMTS
'
ao
i i I I I I I
too tao too aao aoo aaa 400 4>o
BURNER ZONE LIBERATION RATE: |10a BTU/HR-FT*)
NOx Reduction Summary
Figure 8
ADVANCED OVERFIRE AIR
Foster Wheeler has been investigating advanced NOx control measures for new unit and retrofit
use. Some technologies which can be relatively easily incorporated into a new steam generator
cannot be economically retrofitted to pre-NSPS units.The following summarizes an advanced
4-13
-------
overfire air technique for achieving lower NOx levels than are achievable with conventional
methods.
Conventional overfire air systems, as presented above, utilize approximately 20% of the total
combustion air flow with ports located above the top burner level. Under these design conditions
NOx reductions of about 30 - 35% are achievable. Although greater reductions are obtainable by
using more overfire air (lowering the burner stoichiometry) the potential adverse side-effects of
such large OFA quantities generally preclude their use.
However, it has been determined that improved NOx reduction can be attained by increasing the
distance between the overfire air level and the top burner level, thereby increasing the residence
time in the stoichiometric flame for reactions which reduce NO to N2 to occur. This concept was
field demonstrated in the early 1980's on two utility steam generators which had been retrofitted
with CF/SF burners during construction. Both units were designed for future lignite firing and, as
such, were equipped with an extra level of burners.
The arrangement of one of these units is schematically shown in Figure 9. The units fired much
higher quality coal when initially started up. When operating with the overfire air ports closed and
the top burner level out of service, at full load, the top burners act as conventional overfire air
urn ustoEitt tik inirnE in
UUISEKIT
<
QVCIIftM AIR
LlVttJ
» UNfT SOUIPPCO WITH 8XTWA LtVO. OP lUANKRt
can ai imso as ovumnc am
Long Resident Time Overfire Air Arrangement
Figure 9
4-14
-------
ports. When the top burner's registers are closed the overfire air ports are used the residence
time is significantly increased over that of the conventional overfire arrangement.
More recently, FW has started up an 80 million Btu/hr Combustion and Environmental Test
Facility. The CETF furnace in equipped with overfire air ports located well above the top burner
level.
Figure 10 summarizes the effect of increased OFA residence time on the two utility boilers (data
from 1981 and 1982) and the CETF (data from 1986). The results are consistent in that the
increased residence time yields a NOx reduction nearly double that of the standard ports: 50%
vs. 30%.
NOx
UTILITY BOILERS
CETF
OPAP t LOWElf CLOSED
limi CLOSED ( (
« TOP BURNER LEVEL USED AS OVERFIRE AIR PORTS
CLOSED OPEN
Effect of OFA Resident Time
Figure 10
These very encouraging results have enabled Foster Wheeler to provide such systems on new
and retrofit applications to utility steam generators.
SUMMARY
Foster Wheeler has obtained considerable experience with commercial low NOx burners on both
new steam generators and retrofitted units. Emission and boiler performance results are
consistent in that NOx is typically 50 - 60% lower than with pre-NSPS turbulent burners and there
4-15
-------
have been no indications of adverse effects on boiler performance or operation due to the low
NOx burners. Of particular significance have been the successful retrofits of industrial and utility
units ranging in size from 110,000 Ib/hr to 650 MW.
Advanced combustion-based NOx controls combining a commercially-proven low NOx burner
with a modern overfire air system can achieve NOx reductions up to 80% from pre-NSPS levels.
4-16
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
N0x CONTROL UPDATE - 1989
Albert D. LaRue
Paul L. Cioffi
Fossil Power Division
Babcock & Wilcox
Barberton, Ohio
ABSTRACT
Babcock & Wilcox is actively expanding its technology base and product lines for
controlling NO^ emissions in the combustion process or, thereafter, in existing
and new boiler applications. Fuel economics and extremely low emission limits in
some locations are causing a resurgence in natural gas and fuel oil low NO^
systems. In retrofit situations, compact furnaces and crowded boiler settings
complicate the addition of N0x control equipment. B&W has recently developed the
XCL Oil/Gas Burner which makes use of air and fuel staging principles in the
burner, resulting in very low NO^ emissions without addition of gas recirculation
systems or NO^ ports. The PG-Dual" Register Burner offers broad commercial
experience in combination with more traditional NO^ control measures for use with
oil or gas. With regards to pulverized coal, the XCL-PC Burner is operating
successfully in its first commercial application in the EPA's LIMB project. The
XCL-PC is capable of even lower NO^ emissions than its predecessor, and flame
shape can be tuned to accommodate furnace dimensions. Development of the Low N0x
Cell-PC Burner is complete and the first demonstration will also include gas and
oil capabilities.
Application of rebuming to a coal-fired cyclone unit proved to be quite
successful at B&W's research facility and negotiations are underway to
commercially demonstrate this technology. Atmospheric and pressurized fluid beds
offer low NO^ emissions for new plants or repowering. An updated overview is
provided here of the systems used by B&W to control N0x emissions in the above
and other situations.
4-17
-------
INTRODUCTION
All new steam generating units in excess of 100 million Btu/hr (MKB) input firing
gas, oil, or coal in the United States are subject to federal New Source
Performance Standards (NSPS) regulating N0x emissions. These units may also face
more stringent N0x requirements enacted by state legislation. In addition, new
units must comply with permitting requirements under provisions of the federal
Clean Air Act, for attainment or non-attainment regions as applicable. Limits
much lower than NSPS are becoming more common for new units. Existing sources
are subject to restrictions in non-attainment regions, either directly or to
provide offsets to permit operation of new units in that region. Still, the
majority of pre-NSPS units are not presently subject to N0^ limitations. Public,
congressional, and international debate over the need for emission limits on
these units seems to be nearing a decision point, with enactment of Acid Rain
Legislation appearing imminent. Acid Rain Legislation will most likely restrict
emissions of N0x and SO^, particularly from units in the central and northeastern
U.S. In summary, new units are increasingly being regulated to N0x emission
limits much below federal NSPS, and many unregulated existing units are likely to
be subject to N0x limitations in the near future.
Babcock & Wilcox (B&W) is a leader in the development and commercial application
of systems to reduce emissions of N0x from industrial and utility steam
generating equipment. B&W's N0x control program began in 1957, with the company
receiving the patent (U.S. Patent No. 3048131) for two-stage combustion shortly
thereafter. The program has continued to grow for over 30 years, in depth of
technology and in diversity of application. B&W low N0x systems are reducing
emissions from over 40,000 MWe of utility steam generators in the U.S., plus a
wide range of industrial units. Technologies range from multi-fuel staged
wall-fired burners for new and existing boilers, to repowering with pressurized
fluid bed boilers.
This paper will describe systems offered by B&W to control N0x emissions from
utility and industrial scale steam generation equipment. The emphasis will be on
4-18
-------
control of N0x during the combustion process but post-combustion N0x control
systems will also be considered.
Some of the systems which will be discussed are being developed to facilitate
significant NO^ reductions on existing units which are not amenable to standard
N0x control measures e.g. the Low N0x Cell Burner and reburning systems for
coal-fired cyclone boilers. Other systems have a broad range of application to
existing or new steam generating equipment. The discussion which follows is
organized by fuel forms—solids, liquids, gas—due to the widening variety of
solid fuels of interest and due to a narrowing of distinctions between systems
once considered either utility or industrial.
SOLID FUELS
PULVERIZED COAL
Dual Register Burner
Pulverized coal remains the dominant fuel for electric power production in the
United States. B&W's first low NO^ PC-fired burner was developed and
commercialized in the early 1970's—the Dual Register Burner (DRB). The DRB
differed from conventional wall-fired circular burners by its axial fuel
injection arrangement and dual concentric registers for increased control of
swirl and air/fuel mixing (Figure 1). The DRB is used in combination with a
compartmented windbox and enlarged combustion zone, and occasionally with NO
ports. This system has successfully met all applicable N0x emission limitations,
ranging from 0.7 to 0.45 lb NO/million Btu (lb/MKB). Actual emissions from
coal-fired units with DRBs range from 0.7 to 0.3 lb/MKB.
Emission levels are a function of coal properties, stoichiometry, and
burner/furnace design for a particular unit. The DRB typically achieves 40 to 50
percent reduction in N0x from uncontrolled levels. This differentiates the DRB
from other low N0x burners which rely on Over Fire Air, Under Fire Air, Boundary
Air, etc. to augment N0x performance. The DRB reduces N0x by virtue of its
air/fuel mixing characteristics, avoiding the expense and slagging/corrosion
problems associated with staging air ports in coal-fired units. All of the 68
units in service with DRBs satisfy applicable NSPS without use of air staging
systems. The adaptability of the design is shown in Table 1, with experience
totaling over 37,000 MW and including lignites, subbituminous, and bituminous
coals.
4-19
-------
The Enhanced Ignition Burner (Figure 2), a modified DRB, is used with low
volatile coals or lower grade lignites. Ignition performance and flame stability
are improved with these "difficult" coals by reduced fuel jet momentum and
increased recirculation of hot gases by the Enhanced Ignition Burner. A
resulting increase in N0x can be offset by NO^ ports when necessary.
XCL Burner
The XCL Burner is the most recent addition to the DRB family. The XCL's
functional performance is an outgrowth of Babcock Hitachi's HT-NR Burner and its
predecessor, the B&W DRB (1). Like the HT-NR, the XCL Burner (Figure 3) reduces
NO^ emissions even more effectively than the DRB, typically 25 percent lower, by
virtue of fuel staging reactions promoted by the design. But the XCL Burner is
designed to facilitate use in retrofit applications by three features. One, the
very low NO emission levels of the burner permit its use without addition of NO
X X
ports, even in the smaller furnaces of pre-NSPS units. Two, the flame shape can
be tuned to fit the furnace by use of impellers as appropriate. For furnaces
with limited depth e.g. single wall-fired units, flame length is reduced to avoid
opposite wall impingement and maintain high combustion efficiency. Larger
opposed-fired units, with hotter combustion zones, will usually not need
impellers and benefit from lower NO^ performance of the XCL. Figure 4 compares
flame length and NO^ versus burner input for B&W circular burners (pre-NSPS/high
N0x) and XCL Burners with and without impellers. Combustion tests reveal the
flame length versus burner input for the XCL with impeller is equivalent to
another manufacturer's low NO^ burner, although N0x is somewhat lower with the
XCL. Three, the XCL is configured for use in wall-fired units without requiring
compartmented windboxes. Air flow control, measurement, and swirl adjustment
functions are separated and provided for each individual burner. This enables
precise tuning during commissioning for higher levels of performance. The
mechanical design is adapted from B&W's S-Burner. The S-Burner solved problems
with overheating, warpage, binding, etc. experienced by other burners, providing
reliability and sustained performance in hostile operating conditions. A full
complement of XCL Burners were retrofitted to Ohio Edison's Edgewater Station,
Unit 4, in September 1986 as part of the Limestone Injection Multistage Burner
(LIMB) demonstration. The unit continues to operate with NO^ emissions of 0.5
lb/MKB and well under one percent unburned carbon loss, on this 30-year-old
single wall-fired boiler.
The XCL-PC Burner has recently been equipped with gas and fuel oil capabilities
(2). Full scale prototype combustion tests (120 MKB/hr) demonstrated advantages
4-20
-------
of in-burner fuel staging techniques with these fuels producing very low NO^
emissions. Refer to the Fuel Oil and Gas Section of this paper (or Reference 2)
for further information.
Low NOx Cell Burner
The Low NO^ Cell Burner was developed by B&W, with EPRI support, to replace
standard two-nozzle cell burners which are in use in over 23,000 MW of capacity
in the U.S. The Low NO^ Cell is designed to fit the wall opening of the existing
standard cell, cutting outage time and expense otherwise required for pressure
part replacement. The Low NO^ Cell is capable of reducing N0x by 50 to 75
percent, as established by pilot and large scale combustion testing (3). This
performance is achieved by virtue of the dedicated overfire air port incorporated
into the upper port of each cell (Figure 5). A "Y-branch" combines the original
two fuel pipes into one larger pipe which feeds the single enlarged nozzle in the
lower throat. This design serves to minimize alterations to the coal piping.
Flame shape is controlled by the impeller at the exit of the fuel nozzle, and by
adjustable vanes in the lower and upper throats.
Future N0x legislation notwithstanding, many utilities equipped with the original
B&W cell type burners have already, or are currently considering, upgrading their
burner hardware to improve mechanical reliability. Reliable mechanical operation
of the burners permits burner adjustments to be made to improve distribution of
secondary air within the wrap-around windbox. Some utilities equipped with cell
burners are experiencing accelerated corrosion of the furnace sidewall tubes due
to poor secondary air distribution. Lack of proper burner register control does
not permit adjustments to correct this problem. The B&W S-type Burner, while not
a low NO burner, has been retrofitted to numerous cell burner units in the past
x r
few years, to solve these problems. This burner is also compatible with the
existing cell burner configuration and has been proven very reliable from a
mechanical standpoint. Hardware features that make the S-type Burner
mechanically superior and commercially attractive were incorporated in the design
of the Low NO^ Cell Burner. In fact, the S-type Burner can be configured
initially to permit future conversion to a Low N0x Cell design with minimal
changes to burner hardware (Figure 6). Utilities can therefore address immediate
concerns regarding mechanical burner upgrades and reduce future costs associated
with converting the unit to low NO^ operation. A full unit retrofit of 80
"convertible" Low NO Cell Burners has been in service since March of 1987, and
x
has resulted in reductions of excess air and unburned carbon in this unit.
4-21
-------
The Low NC>x Cell Burner can also be designed to accommodate oil and/or natural
gas firing. A full unit retrofit of 15 Low N0x Cell Burners capable of firing
coal, oil, and gas will be demonstrated on one of ENEL's 320 MW units within the
next year.
CYCLONE FURNACE
Crushed coal is being fired in cyclones of over 100 boilers, with utility
generating capacity in excess of 26,000 MW. These units represent about 14
percent of pre-NSPS coal-fired generating capacity but produce about 21 percent
of the NO^ emissions. There are currently no commercially demonstrated
combustion technologies for cyclone boilers to reduce NOx> Reburning technology
is a promising alternative for achieving significant NO^ reduction in a
cost-effective manner while maintaining unit integrity. Application of reburning
to a cyclone unit is illustrated in Figure 7. Unlike other reburning systems,
the lower furnace is operated at a stoichiometry over 1.0 to protect the cyclones
and lower furnace from corrosion from the typically fired high sulfur coals.
Other parameters are generally similar to other reburning applications.
Babcock & Wilcox is presently starting the final stage of a three-phase program
to evaluate and commercially demonstrate reburning technology in cyclone-fired
boilers. Phase 1 involved an EPRI/B&W sponsored engineering study which revealed
that the majority of cyclone-equipped boilers could be adapted to this
technology, with particular concern given to adequate residence times for the
process. The study indicated a 50-70 percent NO^ reduction could be expected.
The Gas Research Institute joined EPRI and B&W in Phase 2 of the project, in
which pilot scale tests were performed in the cyclone-fired Small Boiler
Simulator facility at the Alliance Research Center. Figure 8 depicts NO
results as a function of reburning zone stoichiometry for coal, No. 6 oil, and
natural gas reburning fuels (4). Baseline NO^ emissions of 925 ppm were reduced
by 40 percent with coal/55 percent with gas at the higher reburning zone
stoichiometry of 0.95. Increasing the reburning fuel at constant load to reduce
reburning zone stoichiometry to 0.85 resulted in 68 percent N0x reduction for
coal and 75 percent for gas. Additional results of the pilot tests indicated no
major boiler operational problems should inhibit commercial application of
reburning technology to cyclone-fired boilers.
Consequently, Phase 3 of the program is in progress and involves full scale
demonstration of coal reburning technology in Wisconsin Power & Light's 100 MW
Nelson Dewey Station Unit 2. The performance goals are to reduce existing
4-22
-------
uncontrolled NC>x emissions by greater than 50 percent and exhibit no serious
impact on cyclone combustor operation, boiler efficiency, fireside corrosion or
deposition, and ash removal system performance. The contract for this 43 month
project is presently being negotiated under the DOE Clean Coal II solicitation.
The program is being supported by the Department of Energy, Wisconsin Power &
Light, B&W, EPRI, the State of Illinois, and a consortium of utilities. Project
initiation is anticipated for Summer 1989 with system start-up in
Fall 1991.
FLUID BEDS
Atmospheric Fluidized Beds
Fluidized bed boilers can burn a variety of fuels while controlling both SC^ and
NO below federal emission limits. Fluid beds are differentiated as either
x
bubbling bed or circulating (fast) fluid bed, depending on fluidizing velocity.
The near term market for bubbling beds is primarily coal-fired retrofit
applications for both industrial and utility boilers (5) (Figure 9). The driving
forces are to reduce fuel costs (permit use of less desirable, lower cost coal),
while avoiding slagging/fouling problems, and emission control. SO2 is captured
by calcined limestone bed material, with 90 percent capture capability with
proper limestone selection and bed temperature. NO^ is controlled by low bed
temperatures that discourage thermal NOX and reduce reaction rates for fuel N0x-
NO^ depends on the coal devolatilization rate and volatile content, excess air,
bed temperature, CO and SO2 concentrations and bed hydrodynamics. NO^ emissions
of 0.3 to 0.4 lb/MKB are readily achieved without resorting to staging. Unburned
combustibles exceed PC values, but depend on feed system design, coal type, bed
temperature, superficial gas velocity, and recycle rate.
Circulating fluid beds (CFB) (Figure 10) are primarily used in new industrial and
cogeneration units where a wide range of solid fuel firing is desirable (coal
and/or biomass) (6). The combustion process of a CFB is particularly
advantageous for difficult-to-burn fuels, such as coal waste. The CFB will
usually reduce SO2 and N0x emissions and achieve lower unburned combustibles
compared to a bubbling bed. S02 capture is improved by reduced sorbent particle
size and increased sorbent residence time. The effect of these process variables
is to achieve the required SO2 reduction with lower Ca/S stoichiometry. NO is
controlled by use of staging (primary/secondary air split) and low bed
temperatures, to 0.3 to 0.4 lb/MKB typically.
4-23
-------
Pressurized Fluid Beds
Pressurized fluidized bed combustion (PFBC) involves burning crushed coal under
high pressure (150-250 psia) in a bed of limestone (7). The limestone captures
90 percent or more of the sulfur released during combustion. NO^ emissions less
than 0.3 lb/MKB are a consequence of low bed temperatures (less than 1600°F)
limiting formation; and increased bed depths with longer bed residence times,
which serve to reduce N0x from gases leaving the bed. The combined cycle power
plant (Figure 11) is more efficient (10%) since the hot combustion gases drive a
gas turbine which both pressurizes the system and produces electricity.
In 1985, B&W formed a partnership (ASEA Babcock) with ABB Carbon, formerly ASEA
PFBC, of Sweden to supply pressurized fluid bed combustion systems. Through ASEA
Babcock, B&W is presently involved in two major PFBC projects, both involving
American Electric Power (AEP) and partially funded by DOE (Clean Coal Technology
Program). The first contract is to supply equipment for a 70 MWe combined cycle
PFBC demonstration plant to repower AEP's Tidd Plant. Hardware design is
completed and fabrication is in the final stages. The boiler, cyclones, and
other equipment are presently being installed in the 68-ft.-high, 44-ft.-
diameter, 2 7/8-in.-thick combustor vessel. Start-up is scheduled for mid-1990.
The second project is to repower a unit at AEP's Sporn Plant with a commercial
330 MWe PFBC system supplied by ASEA Babcock. DOE has selected the project for
funding under Clean Coal II, and negotiations are in progress. Engineering is
expected to commence later this year.
REFUSE SYSTEMS
At present, less than 10 percent of the refuse generated in the U.S. goes to
refuse plants. New plants are being added at the rate of about 16,000 TPD of
total capacity per year. The two categories of refuse combustion are mass fired
and Refuse Derived Fuel (RDF) fired systems (8). Mass firing has been used
exclusively with new "green field" refuse projects. Mass firing involves using
refuse with minimal fuel preparation i.e. as received with large bulky items
removed. Refuse is fed to a charging hopper and onto a reciprocating stoker
where combustibles are burned off and ash is discharged. Compartmented
undergrate air systems are used on modern units to even air distribution, with an
overfire air system provided to improve combustion and maintain oxidizing
conditions.
4-24
-------
RDF firing is used in new "green field" projects, repowering projects (new boiler
in existing plant), retrofit projects (converting to dedicated RDF boilers), or
as a supplemental fuel. RDF is a higher grade fuel resulting from processing of
refuse to remove and/or recycle a large portion of the non-combustibles. Modern
RDF preparation systems employ conveyance, size separation, shredding, material
recovery (iron, aluminum), density separation, and storage. The combustion
system for a dedicated RDF boiler consists of overbed feeders supplying traveling
grate stokers, equipped with underfire and overfire air systems. B&W's
TM
Controlled Combustion Zone (CCZ ) boiler is used to fire RDF or biomass (Figure
12). The twin furnace arches and overfire air systems provide much better mixing
and flexibility to handle changes in fuel quality.
Refuse-fired boilers are not subject to NSPS limits for NO^ but are subject to
provisions of the Clean Air Act which can require BACT in non-attainment regions,
etc. Concerns about pollutants such as dioxins and furans have led to a strategy
of optimizing combustion conditions to minimize their formation, as there is a
correlation between low CO production and corresponding lower dioxin and furan
emissions. Optimizing the combustion system for low CO has led to a slight
increase in NO^ emissions. Typical refuse-boiler-emission levels are 200 to 350
NO^ and 50 to 100 CO (ppmvd @ 12% CO2). Situations requiring lower N0x emissions
are resulting in use of ammonia injection systems or Selective Catalytic
Reduction (SCR).
LIQUID AND GASEOUS FUELS
Activity levels are increasing for installation of low NO^ burners to fire gas
and oil. Primary applications include main burners and auxiliary burners for new
industrial-scale equipment. However, utility retrofit applications are beginning
to increase, either in response to a variety of regulatory pressures, or to take
advantage of the current attractive prices for these fuels. B&W offers several
systems to reduce N0x on oil- or gas-fired units. These are based on two types
of Low N0x Burners, the PG-DRB or the XCL. These burners can be augmented with
other combustion-related N0x control measures such as recirculated flue gas (GR),
staging ports (NOx ports), or reburning, to further reduce N0x emissions as the
situation may warrant.
4-25
-------
PG-DRB
The Primary Gas-Dual Register Burner (PG-DRB) was developed by Babcock Hitachi
K.K. to improve NC>x reduction capabilities of the standard DRB. Over 11,000 MWe
of generating capacity in Japan is operating with PG-DRBs firing gas or fuel oil.
The distinguishing feature and name-sake of the burner is the use of "Primary
Gas" (Figure 13). This term refers to recirculated flue gas routed directly to
each PG-DRB, and introduced in a dedicated zone which surrounds the primary air
zone in the center of the burner. The primary gas serves to reduce peak flame
temperatures and oxygen concentration in the core of the flame to inhibit
formation of thermal and fuel N0x> The dual air zones surrounding the PG zone
provide control of air/fuel mixing to regulate flame shape and NO^.
The PG-DRB can be applied to new or retrofit situations. The system will usually
include gas recirculation to the burner (PG) and to the windbox, with burners
arranged in compartmented windboxes, and N0x ports installed over the
top burner row. B&W's first U.S. application of the PG-DRB was successfully
commissioned in the summer of '88 at Hawaiian Electric*s Kahe 6 unit. This
retrofit upgrade includes burners, dual zone NO^ ports, GR system modifications,
and controls. The N0x emission guarantee was readily achieved, but opacity and
particulate emissions were initially too high. B&W corrected this by application
of T-Jet atomizers and tuning of the equipment, and thereby satisfied all
guarantees. Los Angeles Department of Water & Power's Hayne Station Unit 4 will
be retrofitted with PG-DRBs, compartmented windboxes, dual zone N0x ports, GR
system modifications, and control upgrades later this year.
XCL Oil/Gas Burner
Oil and gas capabilities were added to the XCL-PC Burner to satisfy a need by
ENEL and Ansaldo Componenti, B&W's licensee in Italy (2). B&W contracted to
supply XCL PC/Oil/Gas Burners for a 660 MW unit in southern Italy, contingent
upon combustion test performance of a full-scale prototype burner (120 MKB). The
multi-fuel XCL (Figure 14), patent pending, is designed with all the gas
elements, coal nozzle, and oil atomizer—all the fuel elements—enclosed in a
flame stabilizing ring in the burner throat. This permits "in-burner" fuel
staging, in addition to controlled air/fuel mixing by the dual air zone
arrangement. The large-scale combustion tests again confirmed XCL performance on
coal, but most importantly revealed extremely favorable emission performance
firing gas and oil. Figure 15 shows the results of XCL Burner design and HEMI
fuel*staged gas spuds (patent pending) relative to a standard circular burner.
4-26
-------
The XCL can reduce NO^ below 0.1 lb/MKB without use of gas recirculation or
staging ports, with CO less than 50 ppm. Flame stability ^nd turndown were
exceptional—80:1. Results firing No. 6 oil indicated the XCL produced
significantly lower NO^ than the circular or PG-DRB, over the wide range of
conditions tested. Consequently B&W manufactured 56 multi-fuel XCL Burners which
are scheduled to start up later this year at ENEL's Brindisi Sud Station, Unit 2.
On the merits of this excellent emission performance, B&W has deleted coal-firing
aspects from the XCL and now also offers it as a gas/oil burner. The XCL 0/G
design (Figure 16) enables its use in an open windbox (compartmented windbox is
unnecessary) as it makes use of the reliable S-Burner configuration. Air flow is
controlled as a sliding air damper, measured by a pitot grid (to facilitate
tuning during commissioning), and swirled by vanes in the dual air zones.
B&W completed combustion tests in the summer of '88 on the XCL-FM. The XCL-FM
(Figure 17) is the designation for a package boiler version of the XCL Burner.
The 50 MKB combustion tests at B&W's Alliance Research Center were performed in
the package boiler on site, and confirmed excellent emission performance, with
ambient air, firing No. 2 oil and natural gas. XCL-FM NO^ emissions were more
than 50 percent lower than the circular or PG burner when firing gas, unstaged
without GR. NO was less than 0.08 lb/MKB with CO less than 50 ppm. A
x
commercial demonstration of the XCL-FM is in progress in one of the package
boilers at Cal Tech.
In-Furnace NOx Reduction
In—Furnace NO^ Reduction (IFNR) provides the means to further reduce NC>x
emissions during combustion. B&W has licensed rights to this reburning
technology, which was developed by Babcock Hitachi and Tokyo Electric Power
Company. Utility applications in Japan operate with N0^ emissions of 10-40 ppm
on natural gas and 40-60 ppm with fuel oil. The IFNR system (Figure 18) consists
of low N0x burners operating below 1.0 stoichiometry in the main combustion zone,
followed by low N0x "planetary" burners operating at much lower stoichiometry to
establish a reburning zone. Overfire air ports are situated beyond the planetary
burners to provide sufficient residence time for NO^ reduction in the reburning
zone and residence time to complete combustion before the gases exit the furnace.
System configuration requirements and residence times are constraints which must
be considered for new or retrofit applications.
4-27
-------
POST-COMBUSTION NO CONTROL
x
A growing number of post-combustion systems are now in demonstration phases or
commercially available to reduce NO^ emissions from boilers. These include
ammonia injection systems situated just downstream of the furnace, in regions
where flue gas temperatures are sufficiently high to avoid the need of catalysts
for the reagent to reduce N0x« Other systems have been demonstrated which use
different reagents that favor somewhat lower temperatures found further
downstream in the convection pass/boiler banks. DOE's Clean Coal program has
spurred demonstration of systems downstream of the boiler which can
simultaneously reduce emissions of N0x and other pollutants. B&W's Puri-Fire™,
the Hot Catalytic Scrubbing Baghouse (patent pending) reduces N0x, S02, and
particulate in one system (Figure 19). Pilot testing indicated 90 percent NO
removal with up to 90 percent SO2 capture and 99 percent plus particulate
removal. B&W Is negotiating with DOE (Clean Coal II) for a 5 MWe slip stream
demonstration of this technology at Ohio Edison's Burger Station. Projected
system start-up is for 1991.
Selective Catalytic Reduction (Figure 20) provides a proven means of reducing N0x
emissions to very low levels. B&W offers SCR systems (9) via our license with
Babcock Hitachi. Economics and/or emission limits determine usage of SCR.
Frequently it is used in combination with low N0x combustion systems to achieve
the best overall system. N0x reduction performance is typically 80 percent, with
installations operating as high as 95 percent removal when conditions dictate.
Ammonia slippage, a concern with such systems, is nominally 10 ppm and can be
reduced at the expense of an additional catalyst.
SUMMARY
Babcock & Wilcox's strategy is to develop and supply advanced combustion systems
and post combustion systems to control N0x emissions. These systems are intended
for new boilers, and also for major categories of existing units which are
expected to face N0x restrictions in the future. The systems range from direct
burner replacements to repowering with pressurized fluid beds. Application of
low N0x systems to new or existing equipment requires careful study and
evaluation to ensure the intended performance objectives can be achieved
compatibly with the balance of plant.
4-28
-------
REFERENCES
1. LaRue, A. D. , Cioffi, P. L. , "NOx Control Update - 1987" presented at the
Joint Symposium on Stationary NO Control, New Orleans, March 1987. B&W
paper - BR-1306.
2. LaRue, A. D., "The XCL Burner - Latest Developments and Operating
Experience" presented at the Joint Symposium on Stationary N0x Control,
San Francisco, March 1989. B&W paper - BR-1371.
3. Clark, M. J. et al, "Large Scale Testing and Development of the B&W Low
NO^ Cell Burner" presented at the Joint Symposium on Stationary NO
Control, New Orleans, March 1987.
4. Farzan, H., et al, "Pilot Evaluation of Reburning for Cyclone Boiler N0x
Control" presented at the Joint Symposium on Stationary NO^ Control, San
Francisco, March 1989.
5. Imsdahl, B., Johnson, H. L., Spada, R. E., "Montana - Dakota Utilities 80
MW AFBC Retrofit Approach to Design and Erection" presented at Energy
Technology Conference, Washington, D. C., April 1987.
6. Belin, F., et al, "Waste Wood Combustion in Circulating Fluidized Bed
Boilers" presented at the 2nd International Conference on CFB, Compiegne,
France, March 1988. B&W paper - BR-1333.
7. Kinsinger, F. L., McDonald, D. K., "Combined Cycle using PFBC" presented
at the American Power Conference, Chicago, 1988. B&W paper - BR-1339.
8. Gibbs, D. R., et al, "Design and Operating Experience with High
Temperature and High Pressure Refuse-Fired Power Boiler" presented at ASME
Waste Processing Conference, Philadelphia, May 1988. B&W paper - BR-1344.
9. Radin, M. G., Boyles, B., "Turbine Exhaust Gas DeNOx using Selective
Catalytic Reduction" presented at the American Power Conference, Chicago,
April 1987. B&W paper - BR-1303.
4-29
-------
Phase 5 Dual
Lighter n Conical Reg^!LlTer
Diffuser
Assembly
Burner Secondary Air
Control Dampers
Burner Secondary
Air Foils
Sliding Air Pitot Grid
1 Damper
&
11 /l\ u
Flame Stabilizing Ring
w
X
Inner Vanes
Dual Stage,
Outer Vanes
Figure 3 XCL Burner.
Figure 1 Dual register burner (Phase V) and compartmented Windbox.
X ¦ Circular
0 • XCL + Impeller
~ - XCL - Impeller
A ¦ Other low N0X burner
Secondary
Variable Input Air
CFS Lighter
Primary Air Slide
+ Damper
Pulverized Coal
Figure 2 Enhanced ignition burner.
Recirculated
Furnace Gases,
Flame
X 0 ~ A
t]
Figure 4 Burner design influences on flame length and NO,.
4-30
-------
ADJUSTABLE AIR ZONE DISK
AIR FLOW MONITOR
JUSSIVE lOUVHiOJMFEH.
„ ADJUSTABLE
LOUVER DAMPERS
ADJUSTABLE
AIR ZONE DISK
OBSERVATION SPIN VANE
PORT ADJUSTMENT
Figure 5 Low N0X Cell Burner.
\T*f
EE
Oversized carrier pipe to allow for larger coal nozzle
Figure 6 Convertible S Burner Low/NOx cell.
• Balance of air )
» 1.15 - 1.20 overalls
stoichiometry j
» 15 • 30% heat input
» 0.2 • 0.5 stoichiometry1
• Flue gas recirculation
(optional)
Overfire
Air Ports
Returning
Burners
• 70 • 85% heat input) i—
(crushed coal) (. Cyclones C
• 1.1 stoichiometry ^ L_
Burnout
Zone
Reburn
Zone
3-4% Excess Op
0.85-0.95
• Stoichiometry
Combustioi
-H Zone
3=
Figure 7 Reburning technology for cyclone-fired boiler.
4-31
-------
600
Cyclone at 10% excess air
10% FGR
6 Million Btu/hr load
• 40
0.95 0.90 0.85
Reburning Zone Stoichiometry
Figure 8 SBS cyclone results with reburning.
60 9-
80 «
100
Steam Drum
Primary
Zone
Fuel & Sorbent Feed
Convection Pass
Particle Separator
- Stand Pipe
L-Valve
Primary Air
Figure 10 Circulating fluid bed.
Air Heater
Secondary
Superheater
Figure 9 Bubbling fluid bed retrofit to utility boiler.
4-32
-------
PFBC Combustor
HP Heater
¦ To Precipitator
and Stack
(""""steam
L Turbine
v s.
55.5MW
:©
( 1—I——h
©
Gross Electrical Output 72.3 MW
Net Electrical Output 70.5 MW
Net Efficiency 34.6%
LP Heaters
Steam ¦
Water -
Air
Gas
Figure 11 TIDD PFBC simplified plant cycle.
fhOE=i,
0
Tertiary Air Inlet
Secondary
Air Inlet
Primary Air Inlet
,Ui—ILJI 11=
Primary Gas Inlet
Figure 13 Primary Gas - Dual Register Burner.
Figure 12 CCZ™ RDF boiler design.
4-33
-------
Nozzle Air Air Separation Vane
Figure 14 Multi-Fuel XCL Burner.
O
z
240
200
160
120
80
40
0
No two-stage combustion
No gas recirculation
Circular burner
with variable mix gas spuds
XCL burner
with variable mix gas spuds
XCL burner
with hemispherical gas spuds
8
—i—i—i—
10 12
14
Excess Air (%)
Figure 15 XCL burner - gas firing.
Figure 17 XCL - FM Burner.
Oil Atomizer
RemovableJ
Gas Spud
r
Sliding Air
Damper
Flame
Pitot Stabilizing
Grid Inner ^'n8
Vanes
5?^
Air
Separation
Vane
Dual Stage
Outer Vanes
Figure 16 XCL oil/gas burner.
Overfire
Air Ports
Low NO„
Planetary
Burners
Low NO,
Main Burners
Complete Combustion
Zone
Reburning
Zone
Main Combustion
Zone
Figure 18 INFR system arrangement.
4-34
-------
Figure 20 SCR reactor module.
Table 1
Dual Register Burner Experience
Units in service
68
Combined capacity (MW's)
37,050
Smallest unit
65 MW (3 mills/6 burners)
Largest units
1300 MW (14 mills/98 burners)
DRB capacity range
Minimum
1.5 MKB (Lab)/98 MKB (Field)
Maximum
218 MKB (Field)
Fuel Ranges - Coal
Minimum Maximum
Moisture
3 35
Volatile Matter
19 40
Ash
4 35
Heating Value
4,900 13,900
- Fuel Oil
- Natural Gas
4-35
-------
(intentionally Blank)
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
1989 UPDATE ON NOx EMISSION CONTROL TECHNOLOGIES
AT
COMBUSTION ENGINEERING
R. E. Donais
M. B. Cohen
M. S. McCartney
Fossil Power Systems
Combustion Engineering, Inc.
Windsor, Connecticut
ABSTRACT
Combustion Engineering, Inc., by itself and in combination with Mitsubishi Heavy
Industries, Ltd. under a licensing agreement, have engineered a complete range
of NOx emission control technology for both new and retrofit applications. This
paper presents a synopsis of each technology along with a brief overview of both
the Kansas Power & Light - P.M. Firing system retrofit demonstration and our
recent domestic SCR experience.
INTRODUCTION
Faced with potential impending acid rain legislation, the owners of domestic -
fossil fuel fired - utility steam generators continue to seek the development
and demonstration of retrofitable NOx emission control technology.
At Combustion Engineering work is continuously underway to both demonstrate our
existing capabilities and develop new technology. As shown in previous joint
NOx control symposia. Combustion Engineering has a diverse array of existing NOx
reduction technologies available. These technologies have been developed by
Combustion Engineering and Mitsubishi Heavy Industries, to address the broad
range of possible reductions mandated by future legislation. A comparison of
these various NOx control techniques for tangential coal fired utility boilers
is shown in Figure 1.
Each of the systems presented in Figure 1 has a maximum percent NOx reduction
which, in the opinion of CE, represents the practical technical limits for that
technology. The degree of NOx control is compared to a base of 100% which would
be typical of pre-New Source Performance Standard (NSPS) tangential fired units
designed in the mid to late 1960s. For retrofit applications, the degree of NOx
control may vary somewhat from the values presented depending upon physical and
operational restraints encountered.
Preceding page blank
-------
This paper will present an updated synopsis of these NOx emission control
technologies Combustion Engineering offers. In addition, this paper will focus
upon the Kansas Power & Light - Pollution Minimum (P.M.) firing system retrofit
demonstration and our recent Selective Catalytic Reduction (SCR) domestic
experience.
TRADITIONAL TANGENTIAL FIRING SYSTEM
Virtually all of the current low NOx burner systems are designed to reduce the
availability of oxygen at some point in the history of the fuel within the
furnace. In the early 1970s, designers of fuel burning equipment pursued the
reduction of high temperature zones within the furnace in an attempt to control
the THERMAL portion of total NOx - e.g., simply stated, NOx = Thermal NOx + Fuel
NOx. Most of the recent fundamental research on the kinetics of fuel-bound
nitrogen focus on the very early devolati1ization process, an initial period of
events in the combustion process during which very large quantities of fuel
nitrogen are released. The potential for nitrogen conversion to NOx is high at
this time. Overfire air (OFA) plays a minor role in controlling NOx formed
early in the combustion process, because the 15% to 20% of the total air that
goes to the OFA is a very small part of the O2 available to the volatile matter
being burned and fuel nitrogen being released. Thus, to control the all-important
fuel nitrogen conversion, large quantities of air must be withheld from the fuel
for the duration of the devolati1ization and small char particle combustion
process. By withholding a sufficient quantity of air from the combustion of
volatile matter, the system is simultaneously deprived of a sufficient quantity
of O2 to oxidize nitrogen to NO. The ideal system of igniting and burning
pulverized coal would be to pre-mix a minimum amount of air with the pulverized
coal, promote strong ignition, and then mix the balance of the required air at
later successive stages during the combustion process.
CE's traditional tangential firing system mimics this ideal arrangement, as
shown in Figure 2. From each corner windbox, separate vertically stacked,
nozzles emit distinct levels of fuel and air, promoting strong ignition near
each corner. The bulk mixing process induced by the rotating fireball pattern
shown in Figure 3, ensures the complete mixing of the fuel and air.
The four distinct, controlled air zones, contained within every CE tangential
windbox are listed below:
1. Primary Air is used to dry and transport the coal, and is injected into the
furnace through the coal nozzle assembly.
All remaining airflow is considered secondary air, and is classified as
follows:
2. Fuel Air is that portion of the windbox air (secondary air) that is injected
into the furnace annularly to the primary air.
3. Auxiliary Air is the balance of the air required for combustion that was
not used by the primary or fuel air or overfire air. Most of the auxiliary
air is injected into the furnace through separate compartments located in
between each fuel compartment.
4. Overfire Air (OFA) is the remaining auxiliary air that is introduced to the
furnace through OFA ports located at the top of (or above and separate
from) the main windbox.
4-38
-------
LOW NOx CONCENTRIC FIRING SYSTEM
In 1980, CE developed a firing system concept referred to as the Low NOx
Concentric Firing System (LNCFS). By design, the LNCFS proportions the
secondary air flow through the windbox so as to effect a decrease in the amount
of fuel air while increasing the amount of auxiliary air. In addition, the
auxiliary air is directed away from the fuel towards the furnace wall in order
to reduce the entrainment of auxiliary air by the expanding primary air/coal jet
(Figure 4). This diversion effectively provides more favorable conditions for
low NOx emissions during the initial combustion process than provided by only
OFA equipped firing systems. While both OFA and LNCFS can be referred to as
"staging" techniques, OFA is a type of "vertical staging" for controlling NOx
emissions, and LNCFS can be thought of as a "horizontal staging" technique -
unique to tangential firing.
As a practical matter, LNCFS affects the "early stoichiometry" for a very
limited amount of time. The cross mixing patterns inherent in tangential firing
are massive and separation of both streams are quickly lost as they penetrate
the furnace. Thus LNCFS requires the use of "flame attachment" nozzle tips to
accelerate the devolati1ization process. Therefore to retrofit LNCFS, all the
air and fuel nozzle tips require replacement (Figure 5). No structural windbox
or pressure part changes are normally required, and after OFA, LNCFS is the next
least costly modification available to reduce NOx. For a unit without
provisions for OFA, LNCFS can be economically installed, in view of periodic
replacement of fuel nozzle tips due to erosive wear.
LNCFS Demonstrations
An extensive field demonstration of the LNCFS concept was conducted at Utah
Power & Light, Hunter #2, under EPA contract. Results from these tests were
presented in previous symposia (Reference 1 and 2) and in the EPA final project
report (Reference 3). More recently, CE's licensee in Great Britain, NEI
International Combustion, Ltd. has completed the demonstration of an LNCFS
system at the Central Electric Generating Board (CEGB) Fiddlers Ferry No. 1
Power Station. Results of NEI's demonstration were presented at the 1987 Joint
NOx Symposium (Reference 4).
Briefly, both demonstrations yielded approximately a 20% NOx emission reduction
beyond the standard tangential firing system equipped with OFA. The tests show
the reductions due to LNCFS are additional to those already achieved with OFA.
This evidence supports the theory LNCFS and OFA retard the conversion of fuel
bound nitrogen at different times during the combustion process.
4-39
-------
CE/MHI PM BURNER
The coal fired PM burner is part of a family of ultra low NOx burners capable of
firing coal, oil, and gas. The PM system was developed by Mitsubishi Heavy
Industries (MHI) and has been licensed to Combustion Engineering for application
in North America. In addition to using a bulk staging combustion technique
(OFA), the PM burner emits low NOx emissions by reducing the local stoichiometry
(available 0?) around the coal particle early in the combustion process. The
extensive laboratory work behind the PM burner has gone further to identify a
relationship between NOx and the pulverized coal/air transport ratio. As seen in
Figure 6, tests show typical transport air/fuel ratios (NOx value at Co) are not
conducive to low NOx. In fact, air to coal ratios that are either higher or
lower than typically found in modern pulverized coal units would result in lower
NOx. With the benefit of this information, the PM burner is quite straightforward.
Since the primary air to fuel ratio from the pulverizer cannot as a practical
matter be changed, the MHI burner separates each fuel line in two streams prior
to the windbox (see Figure 7). In theory, neither the fuel rich (CONC) nor the
air rich (WEAK) injection nozzle has a fuel nitrogen conversion ratio as high as
a single nozzle with the air and fuel combined (see Figure 6). In practice,
this theory has been supported by both large scale laboratory prototype testing
as well as field testing in a large industrial boiler in Japan.
PM Burner Application Study
Under EPRI sponsorship, the coal fired PM prototype was thoroughly evaluated
under a two phase program. This program consisted of PM burner laboratory
combustion testing with U.S. coals, followed by an engineering feasibility and
cost analysis for U.S. applications. The results of this application study were
present in the 1987 Joint NOx symposium and are available from EPRI under
Reference 5.
Briefly, this evaluation indicated NOx emissions in the range of
0.20 - 0.25 lb/10 Btu are both technically and economically feasible.
PM Burner Demonstration
Following up this economic feasibility and technical evaluation analysis, the
next logical step was a full scale utility demonstration of this technology.
This phase was desired since although the system had been demonstrated on new
units, in Japan, the system had never been retrofit.
To serve this purpose, EPRI issued a request for potential PM firing system
candidates. Their project objective was to demonstrate the maximum achievable
NOx reduction with minimal impact on plant performance. After reviewing several
host site candidates, the Kansas Power and Light Gas Service - Lawrence 5 boiler
was selected. In late 1986, KP&L contracted CE to design, fabricate and install
the PM firing system during their upcoming spring 1987 scheduled outage.
This plant was designed by Combustion Engineering in the mid-1960s for a 400 MW
generating capacity, and went into service in 1971. The steam generator utilizes
controlled circulation radiant reheat technology, and at the maximum continuous
rating produces 2,805,000 lb/hr of 1005 F superheated primary steam with a
reheat steam flow of 2,450,000 lb/hr at 1005 F. The boiler is 50'-8" wide by
40'-2-1/8" deep and is equipped with fusion welded waterwalls, a multi-stage
superheater with panels, platens and a finishing pendant, a multi-stage reheater
with a radiant wall and pendant, and economizer (see Table 1).
4-40
-------
The existing firing system was designed for full load operation with either
natural gas, #6 oil or Midwestern Bituminous pulverized coal. Gas and oil can
be fired through 4 elevations of corner located tangential fuel admission assemblies,
and pulverized coal can be fired from the 4 corners through 5 elevations of coal
nozzle assemblies. These coal elevations are supplied fue^ from 5 CE 743 RS
bowl mills. Ignition of the fuel is provided through 2x10 Btu/hr natural gas
fired side mounted ignitors. Since this unit was designed prior to the 1971
NSPS, the existing windboxes were not equipped with OFA.
The new equipment CE supplied as part of the PM firing system included all new
PM burner style coal nozzles including PM fuel piping separators, one new elevation
of gas ignitors, a new secondary air damper control system, and replacing a
portion of the existing pulverized coal fuel piping. These 4 new windboxes each
capable of firing gas, oil or pulverized coal also provided an integral close
coupled OFA arrangement and a separated OFA system arrangement. This separated
OFA system included two individually controlled compartments, with each nozzle
providing both vertical tilt and horizontal yaw adjustment. As designed, the
main windbox and integral OFA arrangement tilt in unison for reheat steam tempera-
ture control and the separated OFA arrangement is independently set for NOx
emission control. A comparison of the taller retrofit PM windbox with the
existing windbox is provided in Figure 8.
A scheduled 8 week outage, starting at the end of April 1987, was set aside for
installing this equipment. This project schedule provided a significant coordin-
ation challenge since only 5 months were available for designing and detailing
the equipment through manufacturing and shipping.
With the firing system design well underway, EPRI contracted the Fossil Energy
Research Corporation (FERCO) to conduct a series of emission tests. Their
objective was to characterize the boiler emissions both prior to and after the
PM firing system retrofit. In addition, the intent was to establish the
baseline test conditions to be used for comparison with future post
modification, and long term tests. At about the same time, EPRI contracted
Combustion Engineering to establish the boiler overall performance during both
the baseline and post retrofit test series.
Following these baseline tests, the scheduled outage began April 26, 1987. The
actual installation went fairly smoothly, with temporary support steel installed
in the furnace bottom through a hole made in the boiler north wall. It was
through this hole that the existing windboxes were slid out after a portion of
the waterwall panel was cut loose around the entire windbox area. Afterwards,
the new windboxes were slid through this opening and lifted up into place (see
Figures 9, 10). Following the waterwall tube fit-up and welding process, the
temporary support steel was removed and the north wall opening was closed. Late
in this outage, a turbine related problem was discovered, forcing a 3 week
outage extension. Although this extension relieved some of the shift time
required for the PM firing system retrofit, the original 8 week schedule could
have been maintained.
The unit was restarted July 16, 1987 and the subsequent start up testing did not
uncover any boiler problems. Following a series of extensive diagnostic testing
and a scheduled fall outage, the unit characterization matrix tests were carried
out. Figures 11-15 summarize these post modification tests.
4-41
-------
The most important program result was the overall boiler performance was only
minimally changed. Comparing the baseline steam side absorption with the normal
PM firing system operation (Figure 11) the economizer sectional absorptions were
essentially identical, and the PM firing waterwall sectional absorption was
slightly higher than the baseline absorption. This is further emphasized by
Figure 12 where the normal PM firing system operation typically yielded a
furnace exit gas temperature approximately 40-50°F lower than the baseline
firing system. For comparison, a 120°F change in FOT was experienced with the
baseline firing system between a clean and dirty furnace.
The NOx emission results, see Figure 13, on the other hand, showed a 50%
reduction at a nominal 320 MW boiler load between the "tuned" baseline firing
system and PM firing system. The "Tuned" baseline performance represents the
"best" normal operation obtained during the baseline characterization matrix
testing, as opposed to the "as found" emissions represented by the discrete data
points. As shown, the PM data represents normal, combined OFA operation of the
PM firing system with both the close coupled and separated OFA compartment in
service providing approximately 25% total OFA. At lower loads, the preliminary
PM firing system results also look encouraging, but at this time, the official
results are not available.
The PM firing system dependence on OFA flow is shown in Figure 14. Based upon
the Lawrence tests, approximately 30% of the NOx reduction can be attributed to
the PM firing system, while approximately 70% of the reduction is attributed to
the OFA air system.
For long term operation, it appears the Lawrence 5 boiler efficiency will be
between 0.2% and 0.25% lower with the PM firing system. This efficiency penalty
is due to increased dry gas losses associated with an approximately 0.5%
increase in boiler 0? required for the 50% NOx reduction/25% OFA operation (see
Figure 15).
CE/MHI MACT
No matter how closely the combustion process is controlled, the prerequisites for
completing the combustion process (temperature, time, and excess oxygen) will be
sufficient to generate at least some NOx. State-of-the-art firing systems such
as the PM burner can do an excellent job of burning pulverized coal as well as
liquid and gaseous fuels, but there is always a minimal value below which
further reduction in NOx is impractical. Thus, for very low NOx requirements,
techniques have been developed to actually destroy NOx already formed by the
combustion process. One such technique has been developed by MHI and
subsequently licensed to CE. This system is called MACT which is an
abbreviation for Mitsubishi Advanced Combustion Technology.
MACT is designed to take advantage of the high temperature kinetic
characteristics of the nitrogen oxide molecule. Nitrogen oxide is relatively
unstable at the elevated temperatures commonly found in the mid furnace sections
of most modern pulverized coal fired boilers. Above 1200 C, the NO molecule can
be "destroyed" by reacting with certain radicals at high temperatures. The high
temperature chemistry is not in itself new, but what is unique to the MHI
process reactions can be represented as follows:
4-42
-------
C„Hm + 02 —*-C„'Hm'* + CO + H20
NO + C„'Hm'*-»-C„"Hm" + N2 + H20 + CO
NO + C„'H„'*-^C„"Hm" + NH, + H20 + CO
The asterisks (*) denote a radical at the initial stage of chemical reaction and
NH.. represents any nitrogen compound.
Conceptually, the MACT process can be seen in Figure 16. Compared to
conventional firing systems, MACT diverts a small percentage of fuel from the
main burner combustion zone and injects the fuel through UFI (Upper Fuel
Injection) ports with an inert propellant, usually flue gas. The fuel injected
into the "De-Noxing Zone" has insufficient oxygen to burn. Upon reaching
sufficient temperature, the fuel generates, for a brief period of time, high
temperature radicals which have such a high affinity for oxygen that they
literally strip the oxygen from the NO molecule. This high temperature thermal
decomposition of NO is called the "De-Noxing" (DN) process.
Since the fuel injected into the mid-furnace area in the DN process must be
burned to completion to salvage the heating value of the fuel and to prevent
hydrocarbon emission, the balance of the required combustion air is injected at
the "AA elevation (Figure 17). The "AA" process is the straightforward
oxidation of the products of incomplete combustion leaving the DN process.
It is interesting to note the last reaction shows the formation of NO. In both
theory and practice, the MACT system does a very effective job of destroying
most of the NO in the DN process but regenerates some of the NO in the "AA"
process. The amount of NOx that is regenerated in the "AA" process is dependent
on the fuel nitrogen content and reactivity of the reburn fuel. Natural gas, or
low nitrogen oil may be used as the deNOx or reburn fuel. This is significant
because it means that MACT can be utilized in conjunction with conventional
firing systems (tangential and wall fired) as well as with low NOx burner
systems such as the PM burner to produce NOx concentrations on new boilers as
low as 70 ppm to 115 ppm (vol. corrected to 3% 0^) firing pulverized coal.
MACT Applications
MHI has had an oil fired PM/MACT combination in commercial service since January
of 1981 in Japan. A 600 MW unit has been in operation since September of 1983
utilizing fuel oil as the "MACT fuel" and oi 1 /pu 1 verized coal in the main
windbox - See Reference 6.
AA process:
C„"Hm" + 02 »-H20 + CO:
CO + o2
NH, + O:
NH, + O;
¦»»co2
*- n2 + h2o
~~NO + H20
4-43
-------
POST COMBUSTION CONTROL OF NOx
The NOx reduction capabilities of in-furnace combustion modifications are
considerable. However, there are minimum levels at which further NOx reduction
by these methods becomes impractical. With the use of dry catalytic tail end
systems, reductions of up to 90% in NOx levels can be achieved.
Due to the stringent NOx regulations in Japan, Mitsubishi Heavy Industries, Ltd.
(MHI) began developing a dry NOx removal process for its steam generating units
in the early seventies. This was a developmental process in which the catalytic
systems were first applied to natural gas fired units, then to low and
high-sulfur oil fired units and finally to coal fired boilers. To complement
existing CE NOx reduction technology, a license agreement was signed with MHI in
April 1980, making the catalytic and low NOx burner technology and products
available to CE for steam generators in the United States and Canada.
SCR DOMESTIC INSTALLATIONS
During the past two years, CE has been involved with the design and installation
of several selective catalytic reduction systems (see Table 2). These projects
have been primarily for gas fired cogeneration heat recovery units and
industrial boilers in the California area. The application of SCR technology
was required for each of these projects in order to meet stringent local NOx
emission standards.
A CE shop assembled boiler and SCR system installed at Chevron's El Segundo
Refinery began operation in February 1988. The SCR reactor at this site is
located adjacent to the steam generator. The flue gas from the boiler flows
vertically downward through the SCR reactor chamber. A single layer of catalyst
modules is installed in the SCR unit to treat the incoming flue gas. Additional
space has been provided at the boiler exit to accommodate the future
installation of a CO catalyst (see Figure 18).
Upstream of the SCR reactor chamber, in a horizontal duct run, the ammonia
injection pipes, nozzles and mixing grid are installed. A diluted mixture of
ammonia gas in air is dispersed into the flue gas stream through orificed
openings in the ammonia injection nozzles. Downstream of the ammonia air
injection point is a fixed gridwork of carbon steel pipes used to achieve a
uniform distribution of ammonia in the flue gas before it enters the SCR reactor
chamber. The gas then exits the SCR chamber, flows through a separately encased
economizer section and finally discharges up the stack.
The steam generator at the Chevron facility is used primarily as a source of
back up power and process steam to an adjacent gas turbine cogeneration plant.
The unit is designed to fire natural gas, refinery gas and future no. 2 oil.
Refinery gas has been the primary fuel since unit start up. The SCR system has
been successfully operating across the full load range of the boiler achieving
greater than 90% NOx removal efficiency. Ammonia slip levels (unreacted
ammonia) have been negligible.
4-44
-------
The first CE installation of a SCR system into a heat recovery boiler is
expected to begin operation during the first quarter of 1989 at Harbor
Cogenerating Plant in Wilmington, California. This unit will provide power
from a GE Frame 7 gas turbine (approximately 75 MW) as well as process steam
(465,0001b/hr) from a waste heat boiler for enhanced oil recovery.
The SCR chamber is located downstream of the first module of heating surface
where suitable SCR operating temperatures are available. The ammonia injection
grid has been designed as an integral part of the first pressure part module.
Upstream of the waste heat boiler is a CO catalyst designed for 90% removal
efficiency.
The SCR system at the Harbor Cogeneration Plant treats the flue gas from a
natural gas and refinery gas fired turbine. The design NOx removal efficiency
is approximately 83%. Unreacted ammonia levels are designed for less than 10
ppmvd at 15% O2. A three (3) year catalyst life guarantee is provided for the
honeycomb ceramic based catalyst.
The SCR units at Chevron's Richmond Refinery have been operating since July
1984. These units are located at the exit of three process heaters. The SCR
systems are designed for 65% NOx removal efficiency. At present, the units are
still performing within design limits and no new replacement catalysts have been
requi red.
SCR technology in the U.S. has been primarily applied to liquid and gaseous
fuels for industrial boilers and gas turbine cogeneration projects. These SCR
applications were required in order to meet strict local emission requirements.
So far, whenever NOx emission requirements have gone below levels practical for
combustion modifications, SCR systems have proven to be a viable option.
CONCLUSION
The technology needed to comply with anticipated future NOx emission regulations
is already commercially available. Furnace combustion modifications, post
combustion catalytic systems or a combination of both methods may be considered
for new or existing units depending upon the level of NOx emissions to be
achieved. As shown in Figure 19, the relative costs of reducing NOx can vary
significantly. Each method must be carefully analyzed to determine the most
appropriate and cost effective technique for the desired NOx emission level.
4-45
-------
REFERENCES
1. Lewis, R.D., et al, "Low NOx Coal Firing System Demonstration Results on a
Tangentially Fired Boiler," presented at 1985 Joint Symposium on Stationary
Combustion NOx Control, May 6-9, 1985, Boston, Massachusetts.
2. Kokkinos, A. and Lewis, R.D., "Field Evaluation of a Low NOx Firing System
for Tangentially Coal-Fired Utility Boilers," Final Project Report, July
1985, EPA Report No. 600/7-85-018.
3. Kokkinos, A., et al, "Low NOx Firing System for Tangentially Coal-Fired
Utility Boilers - Preliminary Testing," proceeding of the 1982 Joint
Symposium on Stationary Combustion NOx Control, Vol. 1, July 1983, EPRI,
Palo Alto, California.
4. Sargeant, M., et al, "Reductions in NOx Emissions from a 500 MW
Cornet—Fired Boiler," presented at the 1987 Joint Symposium on Stationary
Combustion NOx Control.
5. Kokkinos, A. and Donais, R. E., "Evaluation of the PM Burner: A Low-NOx
Pulverized-Coal-Firing system for Tangentially Fired Utility Boilers",
Final Report, February 1987, EPRI Research Projects 1836-1,-2.
6. Murakani, N., et al, "Application of the "MACT" In-Furnace NOx Removal
Process Coupled with a Low-NOx SGR Burner," presented at 1985 Joint
Symposium on Stationary Combustion NOx Control, May 6-9, 1985, Boston,
Massachusetts.
4-46
-------
KEY
T-FIRING TANGENTIAL FIRING
OFA OVERFIRE AIR
LNCFS LOW NOx CONCENTRIC FIRING SYSTEM
*PM POLLUTION MINIMUM BURNER
•MACT MITSUBISHI ADVANCED COMBUSTION
TECHNOLOGY
•SCR SELECTIVE CATALYTIC REDUCTION SYSTEM
•C-E/MHI LICENSE TECHNOLOGY
100
80
60
40
20
T-FIRING
BASE
OFA
LNCFS
W/OFA
PM
W/OFA MACT
W/PM
& OFA
-IN-FURNACE NOx CONTROL-
SCR
-I POST
COMB. NOx
CONTROL
100% BASE NOx LEVELS ARE EQUIVALENT TO PRE-1970
BOILER DESIGNS (TYPICAL NOx RANGE 0.5 - 0.7 LBS/106 BTU)
Figure 1. A Comparision of
Various NOx Reduction Tech-
niques For Tangential Coal
Fired Boilers
Figure 2. Typical Plan
View of Fuel And Auxiliary
Air Streams in Tangential
Firing System
Figure 3. Plan View of Tangentially
Fired Steam Generator Flame Pattern
4-47
-------
Figure 4. Furnace Plan View - LNCFS
FLAME
SCANNER
\
AV
B
X
X
X
[x]
OFA
"AIR (OFFSET)
COAL
AIR (OFFSET)
~COAL IGNITOR
1A AIR (OFFSET)
COAL
AIR (OFFSET)
COAL
WARMUPOIL
"COAL
AIR (OFFSET)
ADJUSTABLE AIR
NOZZLE TIP WITH
FIXED IONING VANE
SCANNER
TILT DRIVE
SIDE ELEVATION
PLAN VIEW
(A—A) OF AUX.
AIR COMPARTMENT
Figure 5. Air Nozzle Tip Design - LNCFS
4-48
-------
STOICHIOMETRIC
STOICHIOMETRIC
RATIO FOR COAL
C-| Co 3~4 C2 7~8
AIR/COAL RATIO (kg/kg COAL)
Figure 6. Concept of Pulverized Coal Fired
Low-NOx C-E/MHI PM Burner
WIDTH = 510mm
FRONT VIEW
OIL
AUX. 2
SGR
CONC
SGR
WEAK
AUX. 1
OIL
CROSS SECTIONAL
SIDE VIEW
SEPARATOR
Figure 7. Structure of Coal Fired Low-NOx
C-E/MHI PM Burner
4-49
-------
GENERATOR RATING
MAXIMUM STEAM FLOW
SUPERHEAT STEAM
TEMPERATURE/PRESSURE
FURNACE WIDTH
FURNACE DEPTH
FIRING SYSTEM
PULVERIZERS
400 MWe
2,805,000 Ib/hr
1005°F/2620 psig
50' - 8"
40' 2-1/8"
TILTING TANGENTIAL
FIVE 743RS (WITH EXHAUSTERS)
FUEL ANALYSIS:
MOISTURE
10.5 %
VOLATILE MATTER
38.0 %
FIXED CARBON
42.9 %
ASH
8.6 %
H
4.50%
C
61.91%
S
0.47%
N
1.27%
O
12.75%
ASH
8.60%
MOISTURE
10.50%
HHV
10,884 Btu/lb
Table 1. Kansas Power And Light - Lawrence 5
Boiler Design Parameters and Actual Coal Analysis
GAS
Alfl/CHL
GAS
GAS
AIR/OIL
GAS
GAS
AlR/OlL
GAS
GAS
AIR/Oll
GAS
SIDE ELEVATION
UNMODIFIED
WINDBOX
JES!
WEAK
COAL
GAS
WEAK
COAL
WEAK
COAL
GAS
WEAK
COAL
SIDE ELEVATION
PM FIRING SYSTEM
MOOlPIED
WINDBOX
Figure 8. Kansas Power And Light - Lawrence 5
Comparison of Baseline And PM Retrofit Windboxes
4-50
-------
Reproduced from
best available copy.
Figure 10. PM Retrofit Windbox Construction
Final Fit-up Into Furnace Corner
4-51
-------
PERCENT OF TOTAL UNIT ABSORPTION
50% |
1S87 BASELINE TEST
ECON WW LTSH SH DIV SH PL HTSH RH RAD HTRH
SECTION ABSORPTION
TEST MO 7B TEST NO 3 TEST NO 10 TEST NO 22
WM BASE LINE ^ NO OFA B EH COMB OF* Hi SEPERATE O
HI CLOSEC OFA
TEST NO 26
ALL TEST HAO 0 TILT, APPRQX 3.1* 02
AND 300 WW LOAD
Figure 11. Steam Side Absorption Comparison of
Baseline vs PM Firing Systems With and Without OFA
2.8
2.7
2.6
2.5
FURNACE OUTLET GAS TEMPERATURE
FULL LOAD - NORMAL OPERATION
2 .4
tf3 ~
2.3
1 1 1—~
3.02 3.06
(Thousands)
NHI - MBTU/HR
—r~
3.1
—i r
3.14
2.9
2.94
2.98
3.18
Figure 12. Furnace Outlet Gas Temperature Comparison
Between The Mean Baseline Temperature And The Normal
PM Operation (25% OFA) Temperatures Over Boiler Load
4-52
-------
N0X EMISSION RESULTS
KANSAS POWER & LIGHT COMPANY
LAWRENCE UNIT NO. 5
700
600
500
LOAD DISPATCH
I • ,
9 m - -
6
£ 400
@
2
Q.
Cl
O
z
300
TUNED PERFORMANCE
100
PRELIMINARY
LOW-NO*
500
300 MW
Burner tilts -14"
OFA tilts ~20°
100 200 300
BOILER LOAD, MW
400
5 10 15 20 25
Overfire Air (%)
Figure 14. PM Firing
System OFA Dependence
Figure 13. NOx Emission
Performance Over Load
o
s?
CO
5
CL
a.
O"
Z
500
400
300
200
100
300 MW
OFA
a 0%
y 7%
<
.15%
"C a 27%
MINIMUM 02
I I l l l I
0 1 2 3 4 5 6 7
O2, %
Figure 15. PM Firing
System Oxygen Requirements
4-53
-------
FURNACE
OUTLET
OFA-
MAIN
BURNER
f COMBUSTION N
I COMPLETION )
zone
/ DENOxING \
ZONE J
/ MAIN \
BURNER \
I COMBUSTION I
\ ZONE /
\
/
FURNACE
OUTLET
AA-
UFI —
OFA-
MAIN r
BURNER 1
f COMBUSTION N
I COMPLETION )
\ ZONE /
( DENOxING \
- / MAIN \
' BURNER \
" ICOMBUSTIONI "
. \ ZONE j -
- \ / .
OFA SYSTEM
MACT IN-FURNACE
NOx REMOVAL
PROCESS
Figure 16. Comparison of MACT In-Furnace NOx
Removal Process With Conventional OFA Method
ADDITIONAL AIR
(AA)C
UPPER FUEL
INJECTOR (UF)
OFA
MAIN BURNER
GAS BOOSTER FAN
Figure 17. Schematic Diagram of MACT
For Steam Generator
4-54
-------
Plant Name
Chevron U.S.A., Inc
Richmond Refinery
Richmond, CA
C-E SCR Experience in U.S.A.
System
Plant Type Supplier Fuel Type
3 Process CE Process Gas
Heaters
Treated Flue Start of
Gas Volume, SCFM Operation
20500/16500/16550 July, 1984
Chevron U.S.A., Inc.
El Segundo Refinery
El Segundo, CA
Package Boiler CE Refinery Gas
Future No. 2 Oil
85000
Feb., 1988
Corona Energy Partners
Corona Cheese Factory
Heat Recovery MHI/CE* Natural Gas
Boiler
294200
April, 1988
Simpson Paper Co.
Heat Recovery MHI/CE* Natural Gas
Boiler
278700
May, 1988
Harbor Cogeneration
Project
Wilmington, CA
Heat Recovery
Boiler
CE Natural Gas
Refinery Gas
528000
1st Quarter
*CE Scope Includes: SCR reactor chamber structure, monorail for catalyst removal, ammonia
injection grid, ammonia supply skid and associated piping and dilution air fan/ductwork.
Table 2. C-E Domestic SCR Systems Experience List
-------
Figure 18. General Arrangement
of a Typical SCR System
KEY
T-FIRING TANGENTIAL FIRING
OFA OVERFIRE AIR
LNCFS LOW NOx CONCENTRIC FIRING SYSTEM
•PM POLLUTION MINIMUM BURNER
•MACT MITSUBISHI ADVANCED COMBUSTION
TECHNOLOGY
*SCR SELECTIVE CATALYTIC REDUCTION SYSTEM
• 20
15
10
O
cc
o.
Q_
<
•C-E/MHI LICENSED TECHNOLOGY
T- I T- I LNCFS PM MACT SCR
FIRING|FIRING|w/OFA w/OFA w/PM
BASE w/OFA & OFA
-IN-FURNACE NOx CONTROL-
-i POST
COMB.
NOx CONTROL
NOTES:
1. BASED ON MATERIAL AND CONSTRUCTION COSTS FOR A
NEW 300 Mw COAL FIRED UNIT
2. SCR SYSTEM COSTS ARE BASED ON 80% NOx REMOVAL
3. SCR SYSTEM COSTS DO NOT INCLUDE NH3 TANK FARM
EQUIPMENT OR EXTERNAL STRUCTURAL SUPPORT STEEL
4. THESE COST APPROXIMATIONS ARE SUBJECT TO CHANGE
DEPENDING ON SPECIFIC UNIT DESIGN REQUIREMENTS
Figure 19. Economic Comparison of Various NOx Reduction
Techniques for Tangential Coal Fired Utility Boilers
4-56
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
STATUS OF N0X CONTROL TECHNOLOGY AT RILEY STOKER
R.A. Lisauskas
E.L. Reicker
T. Davis
Riley Stoker Corporation
Worcester, Massachusetts
ABSTRACT
An overview of on-going efforts at Riley Stoker to develop improved low-NOx
combustion technologies for both new and existing boilers is presented. A new
Controlled Combustion Venturi (CCV) burner design is described along with full-
scale burner test data from the Riley Research coal burner test facility. N0X
reductions of 40 to 60% were achieved with this burner design. In addition,
recent overfire air system operating experience is discussed. Field test data
from a 400MW coal-fired utility boiler retrofitted with new overfire air N0X
controls are presented.
The application of N0X controls to incineration and fluidized bed combustion are
also reviewed. Pilot-scale test results from a municipal solid waste (MSW)
combustor are compared with full-scale field data. Experiments have also been
conducted addressing the effectiveness of gas reburning applied to mass burn
incineration systems. The paper concludes with a brief summary on the status of
N0X controls for circulating fluidized bed boilers.
INTRODUCTION
The average age of the U.S. utility and industrial boiler population continues to
rise, while acid rain legislation becomes increasingly likely. Several states
have, or will soon enact N0X control legislation of their own. N0X control,
therefore, may soon be required on boilers not currently regulated under Federal
New Source Performance Standards (NSPS). N0X control research, therefore, has
remained an important element in Riley Stoker's product development and overall
business strategies.
4-57
-------
This paper provides an overview of recent efforts at Riley Stoker to develop low-
N0X combustion control technologies for both existing and new utility and
industrial boiler systems. Overfire air and low N0X burners continue to represent
two of the most cost effective retrofit low-NOx combustion control technologies
for pre-NSPS boilers. In our update given at the 1987 Joint Symposium we
discussed plans to upgrade our low-NOx coal burner technology, the Controlled
Combustion Venturi (CCV) burner, and to improve overfire air system performance on
an existing utility coal-fired boiler. The results of these activities are
presented here.
We will also discuss the development of N0X controls for several new boiler
technologies. Pilot-scale studies are underway to investigate the effectiveness
of natural gas reburning in reducing N0X emissions from municipal solid waste
combustion. A new combustion test facility has been constructed to support this
research. Circulating fluidized bed boilers have become the combustion technology
of choice in many new industrial coal-fired boiler applications. Low combustion
temperatures combined with staged combustion offer the potential for achieving
significantly lower N0X emission levels than conventional coal-fired systems.
Riley Stoker is currently under contract with the U.S. Department of Energy to
develop an advanced coal-fired fluidized bed combustion system capable of
controlling N0X emissions to 0.1 lb/10® Btu or less.
CCV BURNER DEVELOPMENT
The Riley Controlled Combustion Venturi (CCV) burner (U.S. Patent No. 4,479,442)
was originally developed as a low-N0x coal burner for wall-fired boiler retrofit
applications. Our initial retrofit experience on three utility boilers was first
discussed at the 1982 Joint Symposium^. In order to extend the performance of
this low-N0x burner technology over a wider range of operating conditions
additional design improvements have been developed and implemented. The
performance objectives for this second generation CCV burner design were as
follows:
• Improved mechanical reliability and operability of the secondary air
register system.
• Reduced air side pressure drop requirements for wider retrofit
application.
• Improved low-load operation.
• Maintain low-N0x characteristics.
4-58
-------
The second-generation CCV burner design is illustrated in Figure 1. N0X control
is still achieved through a patented venturi coal nozzle and low swirl coal
spreader design. Control of fuel and air mixing is achieved by separating the
primary air/coal stream into fuel-rich and fuel-lean layers before mixing into the
secondary air. In addition to the venturi nozzle, the second-generation CCV
burner incorporates a new secondary air register design for swirl and air flow
control. Curved, overlapping air register turning vanes provide improved swirl
control at lower pressure drop. Secondary air flow is controlled by a movable
shroud that slides over the secondary air register entrance. Independent control
over both the shroud and air turning vane positions offers significant flexibility
in controlling combustion even at low load. Inlet secondary air velocity is
controlled at low loads by partially closing the shroud. As as result, swirl and
flame stabilization are maintained even under low load staged combustion
operation. The entire air register mechanism has been redesigned and moved to the
burner front plate away from the hot firing wall for improved mechanical
reliability.
A 100 million Btu/hr prototype of this second-generation CCV burner has been built
and demonstrated in Riley Research's Coal Burner Test Facility (CBTF) located in
Worcester, Massachusetts. This facility is designed to simulate near field
combustion conditions in full-scale furnaces. A more complete description of this
facility was presented at the 1985 Joint Symposium^). Both staged and unstaged
burner tests were conducted. Staged combustion was achieved by introducing a
portion of the combustion air downstream of the burner simulating operation with
overfire air. Results comparing CCV burner emissions with a Riley pre-NSPS Flare
burner are shown in Figure 2. In these tests, primary or burner zone
stoichiometry was varied from 0.8 to 1.23 (unstaged). CCV burner N0X emission^
were 40 to 60 percent lower than emission levels established with the Flare
burner. As found in previous studies, N0X and CO emissions varied with coal
spreader configuration and the amount of overfire air. Spreader design, which
controls flame length, allows the burner to be tuned to various furnace
geometries. Flame lengths from 9-20 feet were observed during the tests. Scale-
up of pilot-scale emission data to field conditions also depends on other
considerations such as fuel type and furnace design parameters^.
The following burner performance characteristics were also established during the
pilot-scale testing:
4-59
-------
• Turndown ratio of 3 to 1
• Combustion efficiency of 99.6 percent
• A reduction in windox furnace pressure drop from 8-10"wc to 4"wc
• Improved mechanical reliability
Second generation CCV burners are now being proposed for a number of utility and
industrial boiler retrofit applications both in the U.S. and Far East.
NEW OVERFIRE AIR APPLICATION
Riley Stoker has retrofitted overfire air systems on both coal and oil/gas fired
boilers. A recent application included field modifications to an existing
overfire air system on a 400MWe coal-fired Turbo furnace. This unit was designed
in 1974 to meet the 1971 NSPS of 0.7 lb/10® Btu. The objective of the field
modifications were: 1) increase boiler firing capacity to 105-110 percent of
maximum continuous rating (MCR), 2) reduce NOx emissions to 0.6 lb/10® Btu at 105
percent MCR, and 3) maintain previous combustion efficiency at lower N0X
operation. An additional objective was to retain the existing burner openings and
avoid extensive pressure part modifications.
The overfire air system design and modified design are illustrated in Figure 3.
The Riley Turbo furnace is characterized by a venturi shaped lower furnace.
Burners are installed in single rows on opposite downward facing arches. This
particular unit is equipped with 24 Riley Directional Flame burners. Overfire air
ports are installed on the vertical furnace wall directly above each burner. In
the original design, overfire air was supplied from the main burner windbox.
Modifications included revision to the air supply duct system and installation of
a separate overfire air plenum or windbox for improved air flow control. The
overfire air ports themselves were redesigned for increased flow capacity based on
guidelines developed under an earlier EPRI Study^^. Overfire air velocity was
increased for better jet penetration across the furnace and to enhance mixing in
the burnout zone. Wing overfire ports between the end burners and furnace walls
were added to improve mixing and burnout in the corners and along the furnace side
walls. Underfire air ports beneath the burners were installed to improve lower
furnace combustion efficiency and to maintain an oxidizing environment along the
lower furnace walls during staged combustion. No additional fan capacity was
required to implement these changes.
4-60
-------
Although the original Directional Flame burners were retained, a number of burner
design modifications were incorporated. Major burner modifications are described
in Figure 4. These modifications were required to increase firing capacity and
improve flame stability under more deeply staged combustion. The Riley Direction
Flame burner utilizes axial and parallel secondary air and primary air/fuel
streams. Air and fuel are introduced through slots formed in the furnace wall.
Adjustable air vanes above and below each pair of coal nozzles direct secondary
air into or away from the primary stream. Burner modifications included the
elimination of the coal spreader and the substitution of a converging coal
nozzle. The coal spreader was removed to eliminate a high wear burner
component. The converging nozzle was added to prevent flame detachment. A
perforated plate was also inserted in the center air slot to maintain secondary
air velocity under more deeply staged conditions. Finally, turning vanes near the
entrance of the coal nozzles were redesigned to provide a more uniform
distribution of primary air and fuel within the coal nozzle. The design of each
of these coal nozzle components was based on the results of cold flow model
studies.
All of the retrofit objectives were met. As shown in Figure 5, the original
overfire air system design, while meeting N0X compliance at 100 percent MCR, did
not provide an adequate amount of air staging to achieve the compliance N0X level
of 0.7 1b/10® Btu at 105 percent MCR. However, field testing demonstrated that
the revised N0X limit emission of 0.6 lb/10® Btu is achieved with the modified
overfire air system at 105 percent MCR. The addition of the redesigned coal pipe
turning vanes, described above, further enhanced overall system N0X control
capabilities. In addition to reducing N0X, these turning vanes also led to lower
furnace exit gas temperature; thereby, eliminating the need for reheat spray above
100% MCR.
Carbon burnout did not deteriorate with the new system even under the higher load
and low-N0x operating conditions. However, in addition to burner modifications,
mill system tuning was required to maintain this level of performance. This
facility is equipped with three Riley ball tube mill coal pulverizers.
This case study emphasizes the importance of a systems approach when evaluating
retrofit low-N0x combustion control options. All aspects of the combustion system
must be considered including the burners, overfire air ports, windbox, combustion
controls, and pulverizer system.
4-61
-------
N0X CONTROLS FOR MUNICIPAL SOLID WASTE COMBUSTION
The Riley Research Center is currently engaged in a research and development
program with the Institute of Gas Technology and the Gas Research Institute to
reduce emissions from mass burn resource recovery boilers. The primary objective
of this program is to determine the effectiveness of an in-furnace control
technology known as reburning in reducing N0X formed during municipal solid waste
(MSW) combustion. In this study natural gas is introduced as a reburning fuel
above a modern mass burn combustion grate. Pilot-scale studies on oil and coal-
fired systems have shown natural gas to be an effective reburning fuel,
particularly at low-N0x emission levels^5).
Riley recently modified its 3 million Btu/hr pilot combustion facility to include
MSW. As shown in Figure 6, the lower furnace has been reconfigured to accommodate
a Riley/Takuma mass burn combustion system. The combustion system consists of a
step grate stoker (Figure 7) equipped with individual undergrate air control for
drying and ignition, combustion and burnout. Overfire air is used to complete
combustion. The pilot facility is capable of burning approximately 450 lb/hr or
5.5 tons/day of shredded refuse. The grate and lower furnace are designed to
simulate conditions at a 100 ton/day Riley/Takuma resource recovery boiler located
in Olmsted County, Minnesota. A comparison of pilot-scale baseline emissions data
with full-scale emission data is given in Table 1.
The test facility also has provisions for introducing natural gas and recirculated
flue gas at various locations above the grate in order to create a fuel rich N0X
reduction zone above the main heat release zone. Up to 40 percent of the total
furnace load can be fired in the reburning zone. Flue gas is introduced with the
natural gas to enhance mixing and reduce excess air requirements. Conventional
MSW systems operate with excess air levels of 80% or more.
Tests have been conducted in this facility to determine the influence of various
MSW combustion and reburning zone parameters on reburning efficiency. Preliminary
test results are shown in Figure 8. These results will be discussed in greater
detail later in this Symposium^. However, a 50 percent reduction in MSW N0X
emissions was achieved with 7 to 15 percent natural gas as a reburning fuel. The
effectiveness of reburning was found to be a function of reburning zone
stoichiometry and residence time, and gas injection location. Future testing will
further explore the impact of these parameters on the reburning process.
4-62
-------
CIRCULATING FLUIDIZED BED COMBUSTION
Considerable experience is now being gained on atmospheric fluidized bed
boilers. In recent years, the development of this technology has focused on the
circulating fluidized bed (CFB) boiler. Since 1983, 66 CFB boilers have been sold
in the U.S. Those units represent over 22 million pounds per hour of steam
capacity. Riley Stoker and its liscensee, Mitsui Engineering and Shipbuilding
Ltd. of Japan, have designed and constructed nine coal-fired CFB boilers varying
in individual capacity from 150,000 to 660,000 lb/hr of steam. The Riley system
is based on a technology known as Multi-solids Fluidized Bed Combustion
(MSFBC)(7).
Riley coal-fired MSFBC boilers use limestone to control SO2 emissions. MSFBC
systems, therefore, are designed to operate at temperatures of 1500 to 1700°F for
efficient limestone utilization and SO2 control. Each MSFB boiler incorporates
staged combustion for enhanced N0X control. As as result, N0X emissions from coal
fired MSFBC units are below current emission limits. Recent field tests in both
the U.S. Japan have demonstrated that MSFBC boilers are capable of operating at
N0X emission levels below 0.3 lb/10® Btu. Emissons vary from unit to unit as a
function of fuel type and system design parameters such as primary zone
stoichiometry, primary zone residence time, and temperature^). In 1987, Riley
Stoker was awarded a U.S. DOE contract to investigate advanced fluidized bed
combustion concepts applicable to small coal-fired industrial and commercial
boilers. One of the goals of this program is to achieve N0X emission levels of
0.1 lb/10® Btu or less with high combustion efficiency.
Riley Stoker is currently installing at its Riley Research Center a 7 million
Btu/hr circulating fluidized bed test facility to study various staged combustion
approaches. The combustor, shown in Figure 9, is over 60 feet tall and capable of
simulating a variety of staged combustion configurations and operating
conditions. Combustion testing will begin this spring.
SUMMARY
Retrofit combustion controls in the form of low-N0x burners and improved overfire
air systems are commercially available. N0X reductions of up to 60 percent are
possible with these technologies. A second-generation CCV burner has been
developed offering greater reliability and extended operating range. Full-scale
operating experience is still required on more advanced combustion techniques,
4-63
-------
such as advanced air staging and reburning, before these technologies achieve
commercial status.
Combustion modification techniques originally applied on oil and coal-fired
systems are now being considered for refuse fired boilers. The application of N0X
combustion controls to incineration systems faces many unique challenges. Refuse,
such as municipal waste, is an extremely heterogeneous and variable fuel. The
combustion history is much less well defined. In comparison with conventional
fuels, combustion conditions are apt to vary considerably over both time and
space. Resource recovery boilers also operate at relatively low temperature and
high excess air levels. A new pilot-scale test facility is being used to improve
our understanding and design of modern refuse fired combustion systems.
Commercial operating experience is being rapidly acquired on coal-fired
circulating fluidized bed boilers. These systems offer greater fuel flexibility
and provide combined NOx and SC>2 control.
4-64
-------
REFERENCES
1. R.A. Lisauskas and A.H. Rawdon, "Status of N0X Controls for Riley Stoker Wall-
Fired and Turbo-Fired Boilers," In: Proceedings of the 1982 Joint Symposium
on Stationary Combustion N0X Control, Vol.1, EPRI CS-3182, July 1983.
2. R. Lisauskas et al., "Experimental Investigation of Retrofit Low-NOx
Combustion Systems," In: Proceedings: 1985 Symposium on Stationary
Combustion N0X Control, Vol. 1, EPRI CS-4360, January 1986.
3. R. Lisauskas, 0. Itse and C. Masser, "Extrapolation of Burner Performance from
Single Burner Tests to Field Operation," In: Proceedings: 1985 Symposium on
Stationary Combustion N0X Control, Vol.1, EPRI CS-4360, January 1986.
4. R. Lisauskas, C. McHale, R.Afonso and D. Eskinazi, "Development of Overfire
Air Design Guidelines for Front-Fired Boilers," In Proceedings: 1987
Symposium Stationary Combustion Nitrogen Oxide Control, Vol.1, EPRI CS-5361,
August 1987.
5. J. McCarthy et al., "Pilot-Scale Studies on the Application of Reburning for
NO Control," In Proceedings: 1987 Symposium Stationary Combustion Nitrogen
Oxide Control, Vol.1, EPRI CS-5361, August 1987.
6. C. Penterson, et al., "Reduction of N0X Emissions from MSW Combustion Using
Gas Reburning," Presented at the 1989 Joint Symposium on Stationary
Combustion N0X Control, San Francisco, March 1989.
7. W. Place and J. Coulthard, "Practical Aspects of Multi-solid Fluid Bed (MSFB)
Systems," Presented at the Second International Conference Circulating
Fluidized Beds, Campienge, France, March 1988.
8. R.W. Breault, "N0X Emissions from a Staged Circulating Fluidized Bed
Combustion Process: The Multi-Solids Fluidized Bed Combustor," Presented at
the ASME Winter Annual Meeting, Boston, December 1987.
4-65
-------
Table 1
COMPARISON OF PILOT-SCALE AND FULL-SCALE MSW TEST
DATA FOR A RILEY/TAKUMA MASS BURN COMBUSTION SYSTEM
Ful1-Scale Pilot-Scale
MSW Heating Value, Btu/lb. 6037 5447
Load, 10® Btu/hr. 37.5 2.36
Excess Air, % 73 70
OFA, % 34 38
Combusti od.Products
02, % U) 9.3 8.7
C02, % 10.1 10.9
CO, ppm (2) 29 27
N0X, ppm ^ 134 142
Furnace Exit, Temp., °F 1620 1570
Residence Time to Furnace Exit, sec. 3.8 2.0
As measured
\ ' Based on 12% 02
4-66
-------
Figure 1. Second Generation CCV Burner with New
Secondary Air Register and Movable Shroud
4-67
-------
o»
00
700
600
o
*
<0
H
<
s
a.
o.
«
z
0
(/>
(/>
1
UJ
400
300
200
K -
C CCV 15 DEG SPREADER -
~ CCV 30 DH-3 SPREADER-
A FLARE
I ¦ ' t ' I ' ' ' '
"o. 7 O.B 0.9 1.0 l.i 1.2 1.3
BURNER ZONE STOICHIOMETRY.SRb
*
o
l-
<
S
a.
o.
«
z
o
CO
CO
£
111
o
a
200
190
160
140
120
100
80
60
40
20
I I < 1 > ! » I ! l I I i I I j t t I i I t i i
0 CCV 15 DEG SPREADER
~ CCV 30 DEG SPREADER
A FLARE
I ' I I I I l ' ' I l ' ' ' 1 i ' l l I i ' i i I
''''
0.7 0.8 0.9 1.0 1.1 1.2 1.3
BURNER ZONE STOICHIOMETRY.SRb
Figure 2. Comparison of 100 Million Btu/hr Pilot-Scale NOx and CO Emissions for
the CCV and Flare Burners (open symbols Pennsylvania bituminous, solid symbols
Illinois bituminous).
-------
ORIGINAL OVERFIRE AIR SYSTEM MODIFIED OVERFIRE AIR SYSTEM
Figure 3. Comparison of Original and Retrofit Overfire Air System Designs
Tested on a 400 MW Coal-Fired Turbo Furnace
-------
ORIGINAL BURNER DESIGN
Figure 4. Original and Modified Direction Flame Coal Burner
4-70
-------
i.l
1.0
~ - ORIGINAL O.F.A. SYSTEM
0 - MODIFIED O.F.A. SYSTEM
— A " MODIFIED O.F.A. SYSTEM
W/ BURNER TURNING VANES
0.9
0.8
52 0.7
0.6
0.5
— A
RILEY 400 MWe TURBO
FIRED UTILITY BOILER
(105% MCR)
0,4 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' ' ' 1 ' '
0.9 1.0 1.1 1.2 1.3
BURNER ZONE STOICHIOMETRY.SRb
Figure 5. The Effect of Overfire Air on
Coal-Fired Turbo Furnace NOx Emissions
4-71
-------
Figure 6. Ri1ey 3 Mi 11ion Btu/hr
Pilot-Scale MSW Combustion Facility
Figure 7. Pilot-Scale Stepped
Combustion Grate
-------
o
o
o
o° %
8 °o °
o
o
o
A
AAA A
O WITHOUT OAS flEOUHNING
A
A WITH OAS RE BUHNINQ
I I I I
0.6 0.8 1.0 1.2 1.4 1.6 1.6
REBURN STOICHIOMETRIC RATIO
Figure 8. The Effect of Natural Gas Reburning
on Pilot-Scale NOx Emissions from Municipal
Solid Waste Combustion
Figure 9. Riley Stoker's 7 million
Btu/hr Pilot-Scale Circulating
Fluidized Bed Test Facility
-------
-------
Session 5A
ADVANCED COMBUSTION TECHNOLOGY
Chairman: N. Holt, EPRI
5A~i
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
THE CONVERSION OF FUEL-NITROGEN TO NOx IN CIRCULATING
FLUIDIZED BED COMBUSTION
YAM Y. LEE
Pyropower Corporation
San Diego, California
and
MATTI HILTUNEN
A. Ahlstrom Corporation
Karhula, Finland
ABSTRACT
Circulating Fluidized Bed (CFB) combustion has become accepted as
a commercial technology for burning coal and low grade fuels in an
economical and environmentally acceptable manner. One of the
foremost advantages of CFB technology is low NOx emission levels.
With staged combustion and a series of NOx formation and
destruction reactions occurring in the combustor, N0X emission from
coal combustion in a CFB is quite low.
This paper discusses the mechanism and the interaction of various
N0X formation and destruction reactions on the conversion of fuel-
nitrogen to NO in a CFB. Experimental data from CFB units will be
presented to illustrate the discussions. The effect of fuel type,
fuel-nitrogen split to volatile-nitrogen and char-nitrogen, and
their conversion efficiency on NOx emission are presented.
Additional NO control methodologies for meeting more stringent
emission standards will also be discussed.
INTRODUCTION
With the growing concern of acid rain and photochemical smog, N0X
and SO, emission from combustion sources have received strong
attention in recent years (1,2,3). In the pursuit of the control
and reduction of NOx and S02 emission, many new technologies which
involve either in-fumace combustion modifications or post-
combustion reduction, have been developed (4). With its many
design advantages, circulating fluidized bed combustion (CFBC) has
been accepted as a commercial technology for burning coal and low
grade fuels in an economical and environmentally acceptable manner.
The specific features and advantages of the AHLSTROM circulating
fluidized bed design will be presented in the next section. The
lower operating temperature and a series of N0X reduction reactions
have resulted in lower NOx emission in both bubbling and
circulating fluidized bed combustors as opposed to that from
pulverized coal furnaces and stokers. Staged combustion, which has
5A-1
-------
been found to be successful in reducing NO emission, is
particularly suitable for circulating fluidized bed combustors.
Data from commercial units have shown that the NOx emission from
CFB is lower than that of bubbling fluidized bed and pulverized
coal furnaces (5). This paper reviews the factors that contribute
to the lower N0X emission in circulating fluidized bed combustors.
The mechanism and interaction of various NO formation and
destruction reactions on the conversion of fuel-nitrogen to N0X in
a CFB will be discussed.
MAIN FEATURES OF AHLSTROM PYROFLOW* CFBC
The primary components of an Ahlstrom circulating fluidized bed
boiler are shown in figure 1. It consists of a combustor chamber,
a cyclone separator, a return leg and loopseal for recirculation
of the bed particles. The combustion chamber is enclosed with
water-cooled tubes and a gas-tight membrane. The lower section of
the combustor is covered with refractory and with openings for
introducing fuel, limestone, secondary air, recycled ash, a gas or
oil burner for startup and bottom ash drain. The majority of the
combustion process occurs in the lower section and heat removal is
achieved mainly by particle convection and radiation in the upper
section of the combustor. The cyclone is refractory-lined and is
designed to separate the entrained solids from the hot flue gas and
return the solids through the return leg and loopseal. The
loopseal provides a seal for backflow of gas and has no movable
mechanical parts. The gas velocity employed in CFB is in the range
of 15-20 ft/sec. The air distribution include primary air,
secondary air, transport air for fuel and limestone feed, air to
loopseal and fluidizing air to ash classifier. The bottom ash
classifier is designed to remove larger bed particles and recycle
small particles back to the combustor for better heat transfer.
The operating bed temperature is usually in the range of 1570-1630
*F. The coal and limestone feed size is normally 1/4'* x 0 and 18
mesh x 0 respectively.
The circulating fluidized bed combustor has the following inherent
advantages over conventional combustion systems and bubbling
fluidized bed system:
- fuel flexibility
- low pollutants emission
- lover NOx emission
- lower limestone consumption for sulfur capture
- higher combustion efficiency
- higher load turndown ratio
- operational simplicity and reliability
5A-2
-------
The above advantages have thrust CFBC into the fore-front of
industrial boilers market in recent years. In addition, the
Colorado-Ute 110 MWe utility project has shown promise in the
scale-up of this technology.
FUEL-NITROGEN CONVERSION PATHWAYS
'Thermal NO1, which is formed from nitrogen in air following
Zeldovich mechanism, contributes to less than 10% of the total NO
emission from fluidized bed combustors because of the lower,
uniform bed temperature (6,7) . It has been proven by studies (8,9)
in which nitrogen in oxygen mixture has been replaced by argon in
oxygen as fluidizing medium. Therefore, NO emission from
fluidized bed combustors is mainly from the oxidation of fuel-
nitrogen.
Figure 2 is a schematic summarizing the pathways in the conversion
of fuel-nitrogen in fluidized bed combustors (10). Upon heating,
fuel particles release volatiles including nitrogenous compounds
such as HON and NH3. The fate of the fuel-nitrogen then depends on
whether it follows the volatile-nitrogen pathway or the char-
nitrogen pathway. In the case of volatile-nitrogen, it can be
assumed that the volatile-nitrogen species are released rapidly,
and will diffuse and react in the gaseous phase outside the
boundary layer of the particle. The conversion efficiency of these
nitrogenous species to NO will depend on the local availability of
oxygen and the temperature. On the other hand, in the char-
nitrogen pathway, nitrogen retained in the char after
devolatilization are converted to NO during char combustion.
However, due to the heterogeneous reactions between NO and char as
well as NO and CO over char, NO formed can be reduced to N02 as it
diffuses out from the pores of the char particle. The NO escaped
from the char particles can also be reduced by volatile nitrogenous
species in the gas phase to N2. Also, the NO formed from volatile-
nitrogen can diffuse to the char surface and be reduced to N2.
Nevertheless, several studies (11,12) have shown that the
conversion efficiency to NO from volatile-nitrogen is generally
higher than that from char-nitrogen.
Even though the pathways for conversion of fuel-nitrogen can be
described adequately by figure 2, the interaction of the fuel-
nitrogen chemistry with the environment in the combustor created
by the design and operating conditions of the circulating fluidized
bed conbustor makes it difficult to predict NOx emission. In order
to determine the relative contribution of volatile-nitrogen and
char-nitrogen to NOx emissions and identify potential strategies
for the control of NOx emissions, Beer et al. (13) and Lee et al.
(14) have developed a mechanistic NOx emission model based on
considerations of the evolution and subsequent transformations of
fuel-nitrogen to NOx and N2 for fluidized bed combustors. Recently,
Johnsson (15) has developed a detailed kinetic model for the
oxidation of volatile-nitrogen and the reduction of NO for
5A-3
-------
fluidized bed combustor. It includes 13 homogeneous and
heterogeneous reactions. A single-phase plug flow reactor model
was used to evaluate the kinetic model. Johnsson has concluded
that more information on the kinetics for the heterogeneous
catalytic reactions is needed before detailed modeling of NO
formation in fluidized bed combustors is possible.
The important reactions in the fuel-nitrogen conversion pathway in
a CFB are summarized in table 1. The main homogeneous reactions
involve the oxidation and reduction reaction of volatile-nitrogen
in the form of NH3. The main heterogeneous reactions involve the
NO reduction by char and the NO + CO reduction reaction over char.
The main catalytic reaction involve the oxidation of volatile-
nitrogen (NH3) over calcined limestones.
An investigation of the factors affecting the pathway on the
conversion of fuel-nitrogen to NO, the effect of the fuel
composition and the operating conditions are discussed below.
Effect Of Fuel-nitroaen Content
An upper bound for fuel-nitrogen contribution to NOx emission can
be obtained from the fuel-nitrogen content. Coal-bound nitrogen
occurs in aromatic ring structures such as pyridine, picoline,
quinoline and nicotines. The nitrogen content of typical American
coals ranges from 0.5 to 2.0 percent. However, in a number of
Ahlstrom pilot plant tests, the total nitrogen content has found
not to be in direct relationship with NOx emission. Even when the
difference in heating value of different fuels is corrected by
investigating the N0X emission as a function of total nitrogen
mg/MJ of fuel, no direct relationship is found. This implies that
burning a high nitrogen content fuel will not necessarily results
in higher NOx emission from CFB. The heating value of the fuel
would affect the total nitrogen feed rate into the combustor for
a given boiler load. This in turn can affect the NO emission ppm
value. Therefore, the conversion efficiency for fuel-nitrogen to
NO would provide better comparison for different fuels than an
absolute ppm value for NOx emission.
CONVERSION EFFICIENCY OF FUEL-NITROGEN TO NO
There are many factors which either directly or indirectly affect
the final conversion of fuel-nitrogen to NO in a CFB. They include
the fuel-nitrogen split, the reactivity of the fuel, the sulfur
content and retention requirement, the reactivity of limestone and
particle size as well as the operating conditions such as bed
temperature, excess air ratio and primary air/secondary air split.
These factors are discussed below.
5A-4
-------
Effect Of Fuel-nitroaen Split
During the thermal decomposition reactions of fuel particles, some
of the nitrogenous species in the fuel matrix are released with the
volatiles. The evolution of the nitrogenous species is important
in determining the final NO emission because of the difference in
conversion efficiency for volatile-nitrogen and char-nitrogen.
They are usually in the form of HON or ammonia. The yield,
relative distribution and release rate of these two nitrogenous
species depend primarily on coal rank, heating rate and final
particle temperature.
Pohl et al. (11) have studied the fuel-nitrogen release as
volatiles by measuring the percentage retention of nitrogen in char
produced by heating samples of coal in crucibles until the weight
approached a constant value. The results have shown that in the
operating temperature range of a CFB, the volatile-nitrogen evolved
is shown to be 18-38% for bituminous coal and 10-22% for lignites.
Solomon (16) has studied 13 different coal types and reported that
coal nitrogen is contained almost entirely in tightly bound rings
which are released without breakage in the tar during the initial
stage of devolatilization and the remainder are released at higher
temperatures when rings are ruptured.
Song et al. (17) have studied the fate of fuel nitrogen during
pyrolysis and oxidation in the temperature range of pulverized coal
combustors using the drop tube furnace. They found that increase
in temperature increases the volatile fraction but decreases the
efficiency for converting to NOx. In addition, oxidation
experiments on chars indicates the non-selectivity between nitrogen
and carbon during oxidation but the char nitrogen may undergo
pyrolysis in parallel with the oxidation. Also, Pohl and Sarofim
(11) have shown using drop tube furnace that the conversion
efficiency of volatile-nitrogen is higher than char-nitrogen.
Chan and Lee (18) have studied the conversion of char-nitrogen to
NO in a 7 cm ID fluidized bed reactor and concluded that the
conversion of char-nitrogen to NO decreases with the increase in
particle size and bed temperature. However, there is a minimum of
conversion near 15% 02 in the range of 3 to 21%. The conversion
efficiency to NO ranges from 5 to 25 % in that study with
Mississippi lignite char.
Since the conversion efficiency of volatile-nitrogen is generally
higher than that of char-nitrogen, it is expected that the volatile
nitrogen content would have an important effect on the N0X
emission. Research studies at the Hans Ahlstrom Laboratory have
investigated the relationship of the N0X emission with the fuel
volatile-nitrogen content for different fuels. The different fuels
were burned in the 1.5 MW circulating fluidized bed combustor and
the N0X emissions were measured. Chemical analysis of the fuels
were obtained and volatile-nitrogen fraction determined from the
subtraction of the remaining char nitrogen from total fuel-nitrogen
5A-5
-------
after the fuel has been devolatilized at 900 °c in an oven. The
chemical analysis results of the fuels are shown in table 2. The
typical operating conditions for these experimental runs were: Vg
= 4.5 - 5.4 m/s, T= 850 - 920 *C, Primary air: 55-65%, Exit 02
concentration (wet): 3-4 %. The results of these studies are
plotted as the conversion efficiency of fuel-nitrogen versus
volatile-nitrogen content in figure 3. It is shown that a good
correlation is obtained. As the volatile-nitrogen content
increases from 11% to 46.67%, the conversion efficiency increases
from 4.7% to 12%. In figure 4, the fuel-nitrogen % is plotted
against volatile % (d.a.f.) and a reasonably good correlation is
obtained. The fuel volatile-nitrogen content increases with the
fuel volatiles content in d.a.f. basis. Therefore, a plot of
conversion efficiency to NO versus volatile % (d.a.f.) is obtained
and shown in figure 5. As the volatile % (d.a.f.) increases, the
conversion efficiency to NO increases in the CFB. Both
correlations in figures 3 and 5 are shown to be reasonably good.
It is, however, simpler to obtain volatile content information.
There are exceptions to these correlations as expected because of
the complicated interactions of different parameters in CFBC.
Effect Of Fuel Reactivity
The reactivity of the fuel depends on the volatile content and the
reactivity of the char. The reactivity of the char depends
primarily on the porous structure and the available active sites.
A higher reactivity of the char would give rise to lower carbon
loading in the combustor. A lower char concentration in the
combustor would reduce the NO reduction reactions with char. This
would result in a higher NO emission. In a related situation, a
higher bed temperature would reduce the char concentration in the
combustor. This results in a higher NOx emission at a higher bed
temperature as is shown in many studies (19,20,21). Indeed, the
higher conversion rate of volatile-nitrogen in the combustor at a
higher bed temperature would also contribute to the higher NO
emission.
Effect Of Sulfur Content And Retention Requirement
The sulfur content of the fuel and the retention requirement
determines the limestone requirement. It has been shown in many
studies by Lee et al. (22,23) , Furusawa et al. (11,24) and Leckner
et al. (25) that calcined limestones catalyze the conversion of
volatile-nitrogen to NO. Therefore, an increase in limestone feed
rate or Ca/S ratio would increase the NO emission. This is
illustrated in figure 6 which shows the effect of Ca/S on fuel-
nitrogen conversion efficiency in a circulating fluidized bed
combustor, where the conversion efficiency increases from 9.68 to
11.48% as the Ca/S molar ratio increases from 1.6 to 4.7. These
are results from burning a low sulfur coal in the Ahlstrom 1.5 MW
pilot plant. The operating conditions were: T = 865 *C, Vg = 5
m/s, Primary air = 65 %, Exit 02 concentration (wet): 3.6 %.
5A-6
-------
Similar trends have been obtained by other studies (21,26).
Considering the reactivity of the limestone and particle size, a
higher reactivity or smaller limestone particle size, but not to
elutriable size from cyclone, would reduce limestone consumption
requirement. Hence, the N0X emission would be reduced.
Effect Of Bed Temperature
The effect of bed temperature on the NO emission has been studied
extensively and reported to have strong influence on NO emission
(19,20,27). In general, higher bed temperature increases NO
emission. In the case of a fuel of low reactivity, a higher bed
temperature is required to provide better combustion efficiency.
This would in turn increase NO emission when burning that low-
reactivity fuel. However, in the case of reduced load, where the
velocity and bed temperature are both reduced, the combined effect
would give rise to a lower NO emission.
Effect Of Excess Air Ratio And PA/SA Split
The availability of oxygen for fuel-nitrogen in the combustor is
very important for the conversion to NO. Therefore, a lower excess
air ratio would reduce NO emission. Staged combustion which is
achieved by splitting the air flow into primary air and secondary
air, is used to reduce the oxygen concentration available to the
volatile-nitrogen in the lower section of the bed. This reduces
the conversion of volatile-nitrogen to NO as well as promoting
reduction between volatile-nitrogen and NO formed. A lower primary
air to secondary air split would give rise to lower NO emission
(20,28). However, the PA/SA split normally would range from 1 to
2.75 because of the need of good fluidization in the lower section
of the bed.
ADDITIONAL NOx CONTROL METHODS
Staged combustion is very effective in reducing NOx emission from
CFBC. However, in places when NO emission control is extremely
stringent, additional NOx control measures will be taken in
addition to staged combustion. This can be accomplished to a
certain extent by flue gas recirculation or injection of ammonia
or urea into the cyclone where reduction of NO by NH, takes place
(28). The efficiency of the ammonia injection method has been
demonstrated both in pilot plant and in commercial boiler
operation. Figure 7 shows the effectiveness of ammonia injection
on NO reduction in AHLSTROM PYR0FL0W* boilers as a plot of NOx
reduction % versus NHj/NOx molar ratio (28). However, due to the
temperature sensitivity of the reaction, under reduced load
conditions, the cyclone temperature may not be high enough to give
good efficiency for N0X reduction.
5A-7
-------
CONCLUSIONS
The conversion efficiency of fuel-nitrogen to NO in CFB as a
function of different fuel characteristics and operating conditions
have been discussed. Simple correlation between conversion
efficiency of fuel-nitrogen to NO and volatile-nitrogen content or
volatile content can be obtained. However, there are exceptions
which illustrate the complexity of the interactions among different
parameters. It is, therefore, necessary to develop a mathematical
model to predict N0X emission from CFB for better understanding of
the processes involved. Nevertheless, experimental studies on the
overall kinetics of heterogeneous catalytic reactions in the fuel-
nitrogen conversion pathways are needed. In addition to staged
combustion, other N0X control measures such as flue gas
recirculation and ammonia or urea injection can reduce N0X further
but limitations exist when following load turndown.
REFERENCES
1. "Acid Rain and Transported Air Pollutants: Implication for
Public Policy" (Washington, D.C.: U.S. Congress, Office of
Technology Assessment, OTA-O-204, June 1984).
2. J.A. Graham, D.E. Gardner and D.B. Menzel, Proceedings of 2nd
N0X Control Technology Seminar, EPRI-FP-1109-SR, Section 2
(1979).
3. L.H. Yaverbaum, "Fluidized-Bed Combustion of Coal and Waste
Materials", Notes Data Corporation (1977).
4. J. Makansi, "Reducing NOx emissions", Power, September 1988.
5. F. Engstrom and J. Sahagian, " Operating Experiences with
Circulating Fluidized Bed Boilers", Circulating Fluidized Bed
Ts p- 309-316.
6. F.J. Pereira, J.M. Beer, B. Gibbs, and A.B. Hedley, "N0X
Emissions from Fluidized-Bed Coal Combustors", 15th Symposium
(International) on Combustion, pp. 1149-1156.
7. T. Furusawa, M. Tsujimura, K. Yasunaga, and T. Kojima, "Fate
of Fuel Bound Nitrogen within Fluidized-Bed Combustor under
Staged Air Firing", Proceeding 8th International Conference
FBC, 3, pp. 1095-1104.
8. A.A. Jonke, et al. "Reduction of Atmospheric Pollution by
Application of Fluidized Bed Combustion," Argonne National
Lab, Annual Report, No. ANL/ES-CEN-1002, July (1969) to July
(1970).
9. T. Hirama, H. Takeuchi, and M. Horio, "Nitric Oxide Emission
from Circulating Fluidized-Bed Coal Combustion", Proceeding
9th International Conference FBC, pp. 898-905.
5A-8
-------
10. J. Tang, and Y.Y. Lee, "Important Aspects of Fuels
Characterization for Circulating Fluidized Bed Boilers",
Presented at the Jt. Power Generation Conference,
Philadelphia, PN, Sept. 25-29, 1988.
11. J.H. Pohl, and A.F. Sarofim, "Devolatilization and Oxidation
of Coal Nitrogen", 16th Symposium (International! on
Combustion Cambridge. MA, 1976.
12. H. Ishizuka, K. Hyvarinen, A. Suzuki, K. Yano, and R. Hirose,
"Experimental Study on NOx Reduction in CFB Coal Combustion",
Second International Conference on Circulating Fluidized Beds,
Proceedings, pp. 147-150.
13. J.M. Beer, A.F. Sarofim, and Y.Y. Lee, "NO Formation and
Reduction in Fluidized Bed Combustion of Coal", Journal of
Institute of Enerctv. 38, March, pp. 38-47.
14. Y.Y. Lee, P.M. Walsh, A. Dutta, J.M. Beer, and A.F. Sarofim,
"Spatial Distributions of Coal Nitrogen and Nitric Oxide in
the Bed of a Fluidized Combustor", Proceedings 7th
International Conference FBC, pp. 264-264.
15. J.E. Johnsson, "Mathematical Modelling of NOx Formation in
FBC", Presented at the IEA AFBC Meeting 26-27 of October 1987
at Siegen University.
16. P.R. Solomon, "The Evolution of Pollutants During the Rapid
Devolatilization of Coal", Report No. R77-952588-3, United
Technologies Research Center, 1977.
17. Y.H. Song, J.M. Beer, and A.F. Sarofim, "Rate of Fuel Nitrogen
During Pyrolsis and Oxidation, Proceedings from the Second
Stationary Source Combustion Symposium, EPA-600/7-77-07ed, 79,
1977.
18. C. Chan, and Y.Y. Lee, "The Conversion of Char-Nitrogen to NO
During Fluidized Bed Coal Combustion", Proceedings of IASTED
international Conference On Energy, Power, and Environmental
Systems, Santa Barbara, CA, May 1985.
19. A. Kullendorf, S. Herstad, and C. Andreasson, "Emission
Control by Combustion in Circulating Fluidized Bed-Operating
Experiences", Circulating Fluidized Bed Technology II. pp.
445-456.
20. N. Berge, "NOx Control in a Circulating Fluidized Bed
CombustorM, Circulating Fluidized Bed Technology II. pp. 421-
428.
21. L.E. Amand, and B. Leckner, "Emissions of Nitrogen Oxide from
a Circulating Fluidized Bed Boiler: the Influence of Design
Parameters", 2nd International Conference on Circulating
Fluidized Beds, Proceedings, pp. 151-154.
5A-9
-------
22. Y.Y. Lee, A. Sekthira, and C.M. Wong, "The Effects of Calcined
Limestones on the NH3-N0-02 Reaction", Proceedings 8th
International Conference FBC, 3, pp. 1208-1218.
23. Y.Y. Lee, M.S. Soares, and A. Sekthira, "The Effects of
Sulfated Limestones on the NH3-N0-02Reaction", Proceeding 9th
International Conference FBC, 2, pp. 1185-1187.
24. T. Furusawa, D. Kunii, M. Tsujimura, and M. Tsunoda, "Calcined
Lime and In Situ Formed Char as Catalysts for NO Reduction by
Reducing Gases", Proceeding 7th International Conference FBC,
pp. 525-534.
25. B. Leckner, and L.E. Amand, "Emissions from a Circulating and
a Stationary Fluidized Bed Boiler: A Comparison", Proceedings
9th International Conference FBC, pp. 891-897.
26. T. Furusawa, and T. Shimizu, "Analysis of Circulating
Fluidized Bed Combustion Technology and Scope for Future
Development", Circulating Fluidized Bed Technology II. pp. 51-
62.
27. T. Furusawa, and T. Shimizu, "Reduction of NO? Emission from
Circulating Fluidized Bed Combustors", Preprint of the 2nd
SCEJ Symposium on Circulating Fluidized Beds, Tokyo, Japan,
June 20-21, 1988.
28. M. Hiltunen and J. Tang, "N0X Abatement in AHLSTROM PYROFLOW*
Circulating Fluidized Bed Boiler", The 2nd International
Conference on Circulating Fluidized Beds, March 14-18, France.
29. G.G. De Soete, "An Overall Mechanism of Nitric Oxide Formation
from Ammonia and Amines added to Premixed Hydrocarbon Flames",
Combustion Institute, European Symposium, pp. 439-444.
30. R.K. Lyon, and J.E. Hardy, "Discovery and Development of the
Thermal DeNO Process", Ind. Eng. Chem. Fundamentals. 25, pp.
19-24.
31. L.K. Chan, "Kinetics of the Nitric Oxide-Carbon Reaction under
Fluidized Bed Combustor Conditions", Ph.D. Thesis, MIT.
32. M. Tsujimura, T. Furusawa, and D. Kunii, "Catalytic Reduction
of Nitric Oxide by Carbon Monoxide over Calcined Limestone",
Journal of Chemical Engineering of Japan. 16, No. 2, pp. 132-
136.
33. M. Tsujimura, T. Furusawa, and D. Kunii, "Catalytic Reduction
of Nitric Oxide by Hydrogen over Calcined Limestone", Journal
of Chem. Eng. of Japan. 16, No. 6, pp. 524-526.
34. L.K. Chan, A.F. Sarofim, and J.M. Beer, "Kinetics of the NO-
Carbon Reaction at Fluidized Bed Combustor Conditions",
Combustion and Flame. 52, pp. 37-45.
35. T. Furusawa, T. Honda, J. Takano, and D. Kunii, "Nitric Oxide
Reduction in an Experimental Fluidized-Bed Coal Combustor",
Fluidization. Cambridge University Press, pp. 314-319.
5A-10
-------
Figure X: CIRCULATING R.UIDIZED. BED BOILER
Vblatil*
Rttngn
Volctil*
MalHdtcegai.
.DmxmI
Hnwyrnout and
* Raducti.cn of NO fey
W3, a*r, 00 and
Otar
¦+»5
Xltro^n
cenfcustion-
->N°x
ri^ira 2: Iha f\Ml-Nitrog«n Conversion fttlMys in Ctewtauator
5A-11
-------
%
2 -
10 20 30 40 50
VOLATILE-NITROGEN %
Coke + Sub. Coal ° App. Coal x Lignite
Culm a Dlatomlte * Sub. Coal
Fig. 3 Fuel-nitrogen conversion versus volatile-nitrogen %
VOLATILE % (da.f.)
• Coke + Sub. Coal D App. Coal x Lignite
0 Culm A Dlatomlte * Sub. Coal
Fig. 4 Volatile-nitrogen % vs. volatile % (daf)
5A-12
-------
4
2
Coke
0 Culm
20 40 60
VOLATILE ( % ) <±a.f.
+ Sub. Coal
& Dlatomlte
a App. Coal
* Sub. Coal
80 100
x Lignite
Fig. 5 Fuel-nitrogen conversion to NO versus volatile % daf
Fig. 6 Effect of Ca/S ratio on fuel-nitrogen conversion
5A-13
-------
NOX
Reduction
%
100
90 -
60 -
70 -
€0 _
0/
50 -
W
6
40 -
/
30 -
mm
• /
20 -
•
1
•70 °C
150 ppa O
• • 670 %
102 ppa
800%
* 110 ppa
Petchora coal
Corrected to 7 I oxygen (dry;
ppa
S$V?}utlcn
i—i—r
110
102
ISO
600
670
670
t
O
O
A
A
A
i—i—r
5 5 7
NHj/NOx molar ratio
1 1 1
Figure 7: NO* Reduction In AHLSTROM PYKOFLOW With
tanonia or Urea* Base HO* Level and
Teaperature at the Injection Location
as Faraaeters
5A-14
-------
TABLE 1 - IMPORTANT OVERALL REACTIONS IN
FUEL-NITROGEN CONVERSION PATHWAYS
REACTIONS REFERF.NCF.S
CaO
4 NHj + 5 0, -»4NO + 6 HjO Furusawa et aL (24), Lee et aL (22)
4 NHj + 5 0, -» 4 NO + 6 H,0 Lyon et aL(30), de Soete(29)
Char
NO + CO -~ * N, + COj Chan et aL(31)
CaO
NO + CO -» * Nj + CO, Tsujimuia et aL(32)
CaO
NO + H, -» * N, + H,0 Tsujimuia et aL(33)
NO + Char -» * N, + CO Chan et aL(34), Furusawa et aL(35)
6 NO + 4 NHj - 5 N, + 6 H,0 Lyon et aL(30), de Soete(29)
TABU 2
FUEL ANALYSIS FOR NO, STUDIES
FUEL DMTOMTB PETROLEUM APPALACHIAN PROCESSED LIGNITE SUB-BITUMINOUS SUB-BITUMINOUS
COKE COAL CULM COAL COAL
Aih Gn diy ntt*) «
7&0
134
110
402
32
203
21.4
C Q* 4gr*oiM0 H
1S3
Mi
743
SU
S0.4
63.4
65.9
H (b diy nttfe) H
23
3.9
4i
13
118
337
4.02
NQtdqr toiidi) H
03
1.1
1.0
0.9
1.17
1.03
1.03
S On dry mUi) H
23
23
1.4
0.4S
336
137
0.6
0 (at jiflaati) H
13
1.9
S3
135
939
933
7.05
HHV MJI/kg
7.18
333
29.4
1937
19.
24.4
25.9
Volatile* NH
47.0
11.0
34.0
2646
39.92
33.04
2335
VoltfOn (daO %
90.91
11.18
3839
933
57.94
36.4
3631
5A-15
-------
(intentionally Blank)
5A-16
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
NOx CONTROL IN COAL GASIFICATION COMBINED CYCLE (IGCC) SYSTEMS
N. A. Holt, E. Clark, and A. Cohn
Electric Power Research Institute
Generation and Storage Division
1989 Joint Symposium on Stationary Combustion NOx Control
San Francisco, California
March 6-9, 1989
ABSTRACT
NOx emissions at the pioneer 120 MW Cool Water Gasification Combined Cycle
plant (located at Southern California Edison's station at Daggett, California)
have been measured at 0.047—.090 lb/million Btu of coal fed. This is equivalent
to 8-15% of the New Source Performance Standards for coal fired power plants.
These low emissions have been obtained on four different coal feedstocks and were
achieved simultaneously with low emissions of SO? (0.024-0.111 lb/million Btu),
particulates (0.01 lb/million Btu) and CO (1-3 ppm).
The Cool Water gas turbine (GE 7000E) has a firing temperature of 1985°F.
Separate combustion studies conducted by manufacturers on various medium Btu gases
indicate that the new higher firing temperature (2300° F) machines, such as the GE
MS 7001F should also be able to meet New Source Performance Standards by an ample
margin using similar techniques to those successfully demonstrated at Cool
Water. Even lower NOx levels should be obtainable with staged combustion designs
that are being developed.
INTRODUCTION
The Acid Rain Issue
Emissions of sulfur dioxide .and nitrogen oxides into the atmosphere increase
the acidity of moisture in the air, resulting in acid rain. Man made sulfur
dioxide and nitrogen oxide emissions In the U.S. are currently estimated at over
40 million tons per year with coal-fired power plants accounting for 55% of the
S0£ released and 27% of the NOx (1).
A Department of Energy (DOE) computer model suggests that sulfur dioxide
emissions in the United States could be reduced by as much as 60% by 2010 through
the broad commercial use of clean coal technologies. Nitrogen oxide emissions
could also be significantly reduced, depending on the technology utilized.
According to DOE, Integrated coal Gasification Combined Cycle (IGCC) would provide
the best NOx reduction; approximately 23%i (2).
There are two sources of NOx when burning fossil fuels. A high proportion of
fuel bound nitrogen 1s converted to NOx during combustion (fuel bound NOx) and
some of the nitrogen from the combustion air is converted to NOx at high
combustion temperatures (thermal NOx).
A generalized Coal Gasification Combined Cycle system 1s shown 1n Figure 1
based on oxygen blowing of the gaslfler. With GCC systems utilizing cold gas
clean up, the production of fuel bound NOx 1s avoided since any fuel bound
nitrogen components are removed in the gas cooling and sulfur removal sections
^Relative to projected total national emissions in 2010 without clean
coal technology.
Preceding page blank
-------
prior to the combustion of the syngas. NOx control in such systems must therefore
primarily rely on control of thermal NOx production during combustion.
FATE OF COAL BASED NITROGEN DURING GASIFICATION
In coal gasification most of the nitrogen found in the coal is converted to
nitrogen gas (N?) in the gasifier. A small fraction of the coal's nitrogen does
form ammonia (NFU) in the gasifier. The amount of NH3 depends on the gasification
temperature and the chemical equilibrium 2 nh^ ^ n. + 3h2.
Higher temperatures result in lower ammonia yields. There are three general
types of gasifiers moving bed, fluidized bed, and entrained flow systems. Moving
bed gasifiers are countercurrent devices with typical outlet temperatures 800-
1200°F. Fluidized bed gasifiers operate at 1500-1900CF and Entrained flow
gasifiers typically operate in the range 2300-2800°F. The percentage of coal
based nitrogen that is converted to NHj is 50-60% with moving beds but typically
only 10-15% with entrained flow gasifiers.
In IGCC systems with cold gas clean up the ammonia is easily removed from the
syngas by condensation and water quenching in the gas cooling train prior to the
acid gas (H2S) removal (see Figure 2). This condensate is then steam stripped and
the overhead fed to the Claus plant where the ammonia is converted to harmless
elemental nitrogen N2. Small traces of hydrogen cyanide are also typically
present in the raw gas which are similarly quenched, condensed, stripped and
converted to elemental nitrogen in the Claus plant.
The US Department of Energy (DOE) has sponsored a considerable amount of work
on the development of air blown hot gas clean up GCC systems. In this kind of
system the control of fuel bound NOx is an additional issue which needs to be
addressed due to the presence of NH3 in the gas going to the gas turbine.
EFFECT OF COAL GAS COMPOSITION ON THERMAL NOx
The gas compositions for various gasification processes are shown in Table
1. Thermal NOx is primarily related to fuel gas flame temperature. The adiabatic
stoichiometric undlsassoclated flame temperatures of various gases are shown in
Table 2. Further discussion of the various effects of gas properties on gas
turbine design have been described in an excellent earlier paper available from
the General Electric Company (Reference 3). An estimate of the effect of flame
temperature on NOx emissions is also shown in Table 2. (taken from Ref. 3)
Coal derived gas fuels are mixtures of hydrogen and carbon monoxide. Methane
may be present 1n the gas from some gasifiers (e.g., BGL and KRW) while carbon
dioxide, nitrogen or water are diluents. For most coal gases considered the ratio
of CO to H2 1s of only minor consequence with regard to flame temperature, however
the effect of diluents 1s of paramount Importance in controlling both flame
temperature and NOx. Most generally considered indices of combustor performance
worsen as volumetric heating value decreases: Fuel/A1r ratio Increases, the amount
of fuel to achieve a given temperature rise Increases and residence time in the
reaction zone decreases. There 1s obviously some limitation on the amount of
diluent that can be added without some sacrifice of combustion efficiency (C0-
burnout) or eventual actual loss of flame. However, the presence of hydrogen
(with its high flame speed) 1n coal derived gases appears to provide an early
ignition and completion of combustion so that low NOx and low CO can be
simultaneously achieved, at least over the range of medium Btu gas compositions
derived from the major candidate coal gasification processes. Of the three major
diluents, C02, H20 and N2, C02 has a greater effect in reduction of flame
5A-18
-------
temperature for a given degradation in heating value. The Texaco/Dow gases and
Shel1/BGL gases present different challenges to the gas turbine combustor
designer. The She 11/BGL gases have lower inerts content and therefore flame
temperatures >100°F higher than Texaco/Dow gases. However, NOx control can be
achieved by adding diluents to the fuel. Generally, H20 is a better choice than
N? unless high pressure N2 is available which will not be used elsewhere in the
cycle. (Ref. 4, 6)
In the highly integrated IGCC configuration described by GKT and Siemens KWU,
air is extracted from the gas turbine compressor to feed a higher than usual
pressure Air Separation Unit (ASU). Approximately 85% oxygen is fed to the
gasifier and the pressurized Nj from the ASU is further compressed and fed to the
gas turbine combustor, thereby assisting in NOx control by dilution of the fuel
gas. (Ref. 7)
Extensive combustion development tests have been conducted in single advanced
burner test stands installed in General Electric1s Gas Turbine test facility in
Schenectady, N.Y. (Ref. 4) Various medium gas compositions representative of
gases derived from the major candidate coal gasification processes were provided
from high pressure tube trailers. NOx measurements showed that addition of
diluents can be used to achieve NOx standards with advanced high temperature gas
turbines (such as the MS7001F) well within NSPS limits for coal fired plants.
Further combustor modifications and if necessary post combustion De-NOx techniques
such as SCR should be able to achieve extremely low NOx (<5 ppm at 15% 02).
PROGRAM DESCRIPTION
Cool Water Coal Gasification Program
The Cool Water Coal Gasification Program has proven that the integration of
coal gasification with combined cycle power production is comrnercially viable and
environmentally superior to conventional coal-fired power plants. The plant has
successfully demonstrated the IGCC technology on four difference bituminous
coals: SUFCo, a low-sulfur coal form Utah; Illinois No. 6, a high sulfur coal
form Illinois; Pittsburgh No. 8, a high sulfur coal from West Virginia; and
Lemington, a low sulfur, high ash fusion temperature coal from Australia.
Although the Lemington was a foreign coal, tested for the Program's Japanese
partners, the run showed that the technology has the ability to utilize the large
reserves of high ash fusion temperature bituminous coals present in the United
States in an environmentally acceptable manner.
The Program's participants (equity owners) are Texaco Inc., Southern
California Edison Company, the Electric Power Research Institute, Bechtel Power
Corporation, General Electric Company and the Japan Cool Water Program
Partnership. The Empire State Electric Energy Research Corporation and SOHIO
Alternate Energy Development Company were contributors to the project.
Process Description
Cool Water is a nominal 120-MW IGCC power plant that uses the Texaco Coal
Gasification Process to produce a med1um-Btu fuel gas. A block flow diagram of
the facility is presented on Figure 2. The gas is combusted in the combined cycle
portion of the plant to produce electricity. The plant was designed to process up
to 1000 tons of coal per day mixed with water, however, rates as high as 1200 tons
per day have been achieved. The coal-water slurry is gasified with oxygen at a
pressure of approximately 600 psig and temperature between 2100 F and 2500 F. The
raw syngas from the gasifier is first cooled in radiant and convective coolers
that generate steam for use in power production. The slag solidifies during its
fall through the radiant cooler and collects in a water sump at the bottom of the
5A-19
-------
cooler, where it is removed periodically using a lockhopper system. After leaving
the radiant and convective syngas exchangers, the cooled synthesis gas enters a
water scrubber, low level heat is recovered from the gas in a series of
exchangers, followed by air and cooling water trim cooling until the gas is
approximately 100 F. Sulfur is removed, primarily in the form of hydrogen
sulfide, in a Selexol absorber. A sulfur-rich acid gas is stripped from the
Selexol solution and routed through a SCOT/Claus sulfur recovery unit to produce
an elemental sulfur product. Following sulfur removal, the gas is reheated and
partially resaturated utilizing hot water available from the raw gas cool down.
This arrangement enhances overall system heat recovery while providing water vapor
to the gas for NOx control. Live steam injection is also available to supplement
the moisture added by saturation. The synthesis gas is fired in a GE model 7000E
combustion turbine to produce electricity. The heat from the gas turbine exhaust
is utilized in the heat recovery steam generator (HRSG): to generate high
pressure saturated steam; to superheat the steam generated in the HRSG, in
addition to the steam from the syngas cooler; and to provide boiler feedwater
heating for both the HRSG and syngas cooler. The cooled gases are then discharged
to the atmosphere through the HRSG stack. Carbon monoxide, NOx, sulfur dioxide,
and carbon dioxide are continuously monitored in the exhaust flow from the HRSG
stack. The first three compounds are monitored per guidelines of the EPA and the
San Bernardino County Air Pollution Control District (SBCAPCD).
NOx Emissions at Cool Water
Cool Water has easily met EPA New Source Performance Standards (NSPS) for NOx
on all of the coals tested. As shown in Table 1, EPA test results for NOx ranged
from 0.047 to 0.090 lb/mm ion Btu coal, compared to a NSPS limit of 0.60
lb/million Btu. Federal requirements for SO2, CO, and particulates were also met
by a wide margin, also shown in Table 3.
Emission Limits
Emissions from the plant are limited by Federal, regional, and local
regulations. National emission limits for coal burning plans were stated in the
New Source Performance Standards (NSPS), enacted in 1979 by the U.S. Environmental
Protection Agency (EPA). Emissions from CWCGP have been well below these
standards, generally 8-1536 of these limits. The permit granted by EPA Region IX
in 1981, prior to plant construction, imposed stringent limits based on design
estimates, and the initial operation at CWCGP demonstrated the ability to meet
those limits. Local permits received in early 1988 from the San Bernardino County
Air Pollution Control District (SBCAPCD) set more stringent NOx limitations. The
actual range of values demonstrated during operations was within each emission
limit in this permit. Compliance with all regulations has been maintained through
careful plant operation.
The various regulations have emission limits with differing bases and
units. The NSPS rules are in terms of percent removal and pounds per million Btu
of coal (lb/MHBtu). The EPA and SBCAPCD permits contain limits in combinations of
pounds per hour and concentration (ppmv dry). The SBCAPCD limit for HRSG NOx
emissions is 80 ppmv corrected to 3% exhaust oxygen; which for a gas turbine with
15% oxygen in the exhaust is the equivalent of 26 ppmv NOx at stack conditions.
Table 1 shows all of the permit limits converted to a lb/MMBtu basis. For each
parameter, the SBCAPCD limits are the most stringent (5).
NOx Control
Cool Water has been able to consistently achieve low levels of NOx for two
5A-20
-------
reasons: (1) No ammonia or organic nitrogen compounds are present in the fuel gas
to form NOx from fuel bound nitrogen, and (2) thermal NOx formation is held to
very low levels by syngas saturation and steam injection.
With most commercial gas turbine combustors, fuel bound nitrogen is converted
almost entirely to NOx during burning. In IGCC systems utilizing cold gas
cleanup, the production of NOx from fuel bound nitrogen is avoided. The majority
of the nitrogen in the coal is converted to harmless nitrogen gas (N2) in the
gasifier. A small fraction of the coal's nitrogen does form ammonia in the
gasifier, but this source of fuel bound nitrogen is removed in the gas cooling and
sulfur removal sections before the syngas is combusted. The ammonia is routed
with the sulfur bearing gases to the sulfur recovery unit where it is converted to
No. Therefore, an IGCC power plant like Cool Water must deal only with thermal
NOx formation.
Thermal NOx production is directly related to a fuel's flame temperature, all
other influences being equal. Cool Water lowers NOx production rates below the
level that would be achieved by burning dry synthesis gas through the use of
syngas saturation and steam injection. By adding a diluent, water vapor, the
reaction zone temperature in the combustor is lowered, reducing NOx production.
The effect of moisture addition on NOx formation is shown on Figure 3. The dry
NOx level of 63 ppmv is reduced to 12 ppmv by syngas saturation. Gas turbines are
limited in the amount of water or steam that can be injected because: (1) the
fire will blow out at some point, (2) prior to the loss of flame, carbon monoxide
emissions will increase and combustion efficiency will decrease, and (3)
combustion components' lives may be decreased by pressure oscillations or noise
generated by the combustion process (4). Cool Water has never experienced a
flameout, even at steam to gas addition ratios as high as 0.38 lb steam/lb dry
gas; where NOx was only 12 ppmv. Carbon monoxide formation was still very low at
this level of saturation; 2 ppmv as shown in Figure 3. The monitoring equipment
on the combustion turbine showed no change in normal vibrations that would have
indicated increasing pressure oscillations or noise from the moisture addition.
These results were very encouraging for future IGCC plants because similarly high
levels of syngas saturation and/or steam injection will be needed to achieve low
NOx levels in the new higher firing temperature gas turbines, such as the GE
MS7001F.
It is also of interest to contrast the medium Btu gas (MBtu) data shown in
Figure 3 with the performance data on natural gas for a typical 7000E combustor.
The dry NOx level is about 140 ppmv which can be reduced to about 40 ppmv with a
steam/fuel mass flow ratio of about 1/1 (Ref. 5, Fig. 1&3). However, this amount
of steam dilution would also result in about 40 ppmv CO (Ref 5, Fig. 1&3).
Furthermore if more steam was added to reduce the NOx with natural gas to about 25
ppmv then the corresponding CO would be >200 ppmv. As mentioned before in this
paper the ability to obtain low NOx (10-20 ppmv) with MBG fuel is attributable to
the hydrogen present in the MBG fuel. It should also be mentioned General
Electric has recently developed a "quiet combustor" or dry low NOx combustor
capable of meeting 25 ppmv NOx at 15% On in MS7001EA fired with natural gas (Ref.
6). Furthermore, the CO emissions can Be reduced by a fairly low cost post
combustion catalytic conversion in the HRSG.
^To date, combustion systems have been developed and field-demonstrated
for fuels that do not contain significant bound nitrogen. A
fundamentally different kind of combustor (i.e., staged rich-quick
quench-lean) is necessary when there is significant bound nitrogen in
the fuel.(4)
5A-21
-------
The Texaco gasifier produces a medium Btu gaseous fuel having a higher
heating value of approximately 265 Btu/SCF. The composition of the synthesis gas
at Cool water, following particulate and sulfur removal, on each of the test coals
are set forth in Table 4.
Changes in gasification section operating parameters can alter the
composition of the gas and subsequently the amount of NOx formation, assuming all
other factors are constant. During operation at high slurry concentration the
plant observed an increase in NOx production. (Ref. 4) The increase in NOx was
expected because operation at high slurry concentration reduces the concentration
of C0£, a diluent, and an increase in flame temperature results. Similarly,
reducing the temperature of the gasifier will lower the CO2 content of the syngas
and raise the NOx production. Adjustments in saturation and steam injection have
allowed the plant to meet NOx permit criteria at all normal gasifier operating
conditions. The data from this testing is plotted in Figure 4. Notice that the
data points for high slurry concentration show higher levels of NOx formation at a
given steam to gas mass ratio than the points for normal slurry concentration.
Figure 4 also shows the effect of reduced load on NOx formation. The trend
line representing 90% load indicates a 8-10 ppmv lower NOx than full load. The
two 65% points show NOx production continuing to drop as load is reduced. The
lower NOx at reduced load is a result of a leaner mixture causing lower
temperatures in the combustor.
In Figure 4, carbon monoxide values are also plotted against steam to gas
ratio. CO is below 4 ppmv at all test conditions except the 65% load points,
where it rises to 6-7 ppmv. The small increase is probably a result of slightly
reduced combustion efficiency at high turndown.
Using the same data, NOx is plotted against fuel lower heating value (wet
basis) in Figure 5. The high slurry concentration points new fall along the same
trend line as the fuel load points which indicates, for the operating range
covered by this data, the turbine NOx production rate at full load is a function
of the fuel's heating value.
NqQ Testing at Cool Water
Nitrous oxide (N2O) is one of the molecules believed to contribute to the
greenhouse effect. Because of increased concern on global warming, the exhaust of
the Cool Water combustion turbine was tested for NoO during 1988. The analytical
work, performed with the gas turbine operating at Base load, showed that the
exhaust levels of nitrous oxide were very low, less than 0.5 ppmv. Two separate
analytical methods were utilized. In the first a flask sample is taken
conditioned and analysed by gas chromatography. In the past this method has given
widely differing results depending on the gas conditioning techniques used.
Recently a new technique using Non-0ispersive Infra Red (NOIR) spectrometry has
been developed at the University of California at Irvine under EPRI sponsorship
for on-line measurement. The results from each of these methods showed N20 only
at about the limits of detection.
Combustor Modifications for NOx Control
The application of fuel saturation and/or combustor steam injection to
further lower NOx emissions in not unlimited. In order to reach levels below 10
ppm, especially with the new higher turbine firing temperatures (2300°F)
combustion turbines now being introduced, new combustor designs will likely have
to be developed. New dry-low NOx combustors are currently being planned for these
5A-22
-------
turbines when burning natural gas fuel. These combustors are designed to allow
less than 25 ppm NOx without any steam or water injection. Besides simplifying
the plumbing and allowing simple cycle operation without a water supply these
combustors result in a somewhat higher combined cycle efficiency. This is because
the steam that would have been injected (or mixed with the fuel) for NOx control
can now instead be expanded in the steam turbine down to high vacuum which
provides more power than would its expansion in the gas turbine down only to
(approximately) atmospheric pressure.
The new dry-low NOx combustors are based on staged combustion and premixing
concepts that keep the fuel-air mixtures as far as possible from the
stoichiometric ratio that produces the highest flame temperature. The NOx
formation rate due to fixing of the air nitrogen by the Zeldovitch Mechanism goes
up exponentially with the flame temperature. The most common new dry-low NOx
combustor designs work at the engine full power point with premix of the natural
gas fuel and air in a lean mixture and then combustion, followed by air dilution
to lower the combustion products to the desired combustor exit temperature. For
starting and partial power, however, premixing is not feasible and the combustors
are designed for staged zones of lean diffusion flames; starting with one zone and
bringing on the second as more power is desired.
Some medium Btu syngas containing mainly CO and as combustibles (e.g.,
Shell and BGL) have a stoichiometric flame temperature about 100°F higher than
natural gas (methane). Thus it is even more important to combust lean away from
the stoichiometric condition. However, premixing is considered impractical with
syngas due to the presence of H2 in the fuel, since the high flame speed of the H2
would tend to cause flashback in the premixing zone. The proposed practical
arrangement would be a staged lean-lean combustor with diffusion flames in both
zones. The use of fuel saturation or steam injection in conjunction with this
lean-lean design would have the promise of bringing NOx emissions to below 10 ppm
without excessive CO formation for syngas fuel.
Leaning out the flame or adding diluents such as water/steam or CO2 is
limited by lack of flame stability and poor combustion efficiency which produces
excessive CO. This basic limit in combustion design and operation seems to be
much less limiting with synfuel than with methane or distillate. The Cool Water
results showing very low CO formation at low NOx levels strongly support this
conclusion. The basic reason appears to be that the H2 in the fuel with its high
flame speed stabilizes the combustion even at very low flame temperature. This is
the basis of the expectation that lean-lean staged combustor can be operated with
syngas in a lean enough manner and even with some steam injection to produce very
low <10 ppm NOx emission with high combustion efficiency.
NOx can also be formed from fuel nitrogen compounds. This direct NOx
formation 1s not a strong function of flame temperature and cannot be supressed by
lean combustion. Fortunately gases from oxygen blown cold clean-up medium Btu
gasification processes do not contain significant amounts of these compounds.
However, fuel bound nitrogen will exist In air blown low Btu syngases with hot gas
clean-up. Combustors with rich-quick quench-lean designs have been partially
developed that show promise of limiting the direct NOx formation. They work on
the principle that in the rich zone the nitrogen compounds are starved for 02 and
in the flame react back to No. Combustion then proceeds 1n the lean zone at a low
enough temperature to limit Zeldovitch NOx formation. The rich-quick quench-lean
combustor is applicable to all fuels. However, it is not as fully developed nor
does it appear to give as low NOx levels as the lean-lean combustor for fuels
without nitrogen compounds.
5A-23
-------
Post Combustion NQx Control (SCR)
The emissions can be further lowered significantly, typically about 90%, by
engine exhaust clean-up with selective catalytic reactors SCR's. Many
installations of SCR's have been made for NOx reduction of natural gas fired
turbines in combined cycle applications. The SCR is installed in the HRSG in the
region where the flue gas temperature is between 550 and 750°F. Ammonia gas NH3
in about the same molar flow as the NOx is introduced. The NOx is reduced to ^
by flowing over the catalyst (usually vanadium pentoxide). Large amounts of
catalyst coated structure are required, about 50 cu. ft. per MW. The catalyst is
sensitive to sulfur and cannot be used with fuels containing more than about .1%
sulfur. All of the syngas systems under consideration meet this criteria. It is
virtually certain that a syngas system would only be used with a combined or other
heat recovery cycle, such as the steam injected gas turbine, rather than the
simple cycle. Thus the 590°-750°F flue temperature requirement would easily be
met. While as yet there has been no significant test on synfuels with SCR, there
is every expectation that in conjunction with the advanced lean-lean combustor and
fuel saturation, the application of SCR would reduce NOx level to below 5 ppm.
This would be at the extra cost, both capital and maintenance of the SCR. The
capital cost may be less significant for the IGCC if its dispatch factor is high
compared to natural gas fueled combined cycles. Conversely, any time required to
replace or maintain the SCR catalyst would impact more on the IGCC.
There are also flue gas CO removal catalytic oxidation systems these are not
temperature sensitive and not as costly as the SCR NOx removal system. Their
drawback, besides the extra cost, is the additional pressure drop they cause.
This would lower the power level and efficiency (as does the SCR system) by about
0.5.1.0%. The CO formation itself is a sensitive indication of poorer combustion
efficiency and thus the heat rate. However, the available use of the CO catalytic
system would allow the application of lean-lean combustors with steam injection or
fuel saturation for very low NOx emission that might otherwise be restricted
because of excessive CO emissions.
Conclusion
NOx emission at Cool Water have been measured at levels only 8-15% of those
required to meet NSPS for coal fired plants. Actual combustion tests have also
been conducted by gas turbine manufacturers on various medium Btu gases which
indicate that GCC systems based on the advanced gas turbines with higher firing
temperatures should also be able to surpass NSPS by ample margins using similar
diluent techniques to those applied at Cool Water. Should still lower NOx levels
be required these should be obtainable by staged combustion designs now under
development. Extremely low levels (e.g., <5 ppm) would probably require use of
the rather cumbersome post combustion treatments such as Selective Catalytic
Reduction (SCR).
REFERENCES
1. Simbeck, Dale R. and Dickenson, Ronald L., "Integrated Gasification Combined
Cycle for Acid Rain Control," Chemical Engineering Progress, Oct. 1986, p. 28.
2. "DOE guesses SO?, NOx reductions," Coal and SynFuels Technology, Vol. 9, No.
40, October 17, 1988, p. 1
3. "Alternate Gaseous Fuels". R. A. Farell, R. L. Gessner and H. von E. Doering
(General Electric Co.) (GER-3092)
5A-24
-------
4. "NOx Emissions from Advanced Gas Turbines Fired on MBTU Gases," - L. B.
Davis, M. B. Hilt, R. B. Schiefer (General Electric); EPRI AP-5343-SR
Proceedings: Sixth Annual EPRI Contractor's Conference on Coal Gasification,
5. Rib, Oavid M., "Cool Water Environmental Performance Utilizing Four Coal
Feedstocks," presented at the Eighth Annual EPRI Coal Gasification Contractors'
Conference, Palo Alto, 1988, pp. 1-4
6. "Gas Turbine Combustion and Emissions",, L. B. Davis (General Electric Co.)
(GER-3568). 32nd GE Turbine State-of-the-Art Technology Seminar
7. "Concepts of PRENFLO-based IGCC Power Plants" - V. H. Buskies (GKT) and J. C.
Gwozdz (Sargent & Lundy); EPRI AP-6008-SR Proceedings; Seventh Annual EPRI
Contractors' Conference on Coal Gasification
5A-25
-------
TABLE 1
GAS COMPOSITIONS FROM VARIOUS COAL GASIFICATION PROCESSES
Clean Dry Basis - Volume Percent
Texaco
Dow
Shell
BGL
KRW (021
KRW (air)
ch4
—
—
—
7
7
2
H2
36
41
31
29
34
12
CO
47
38
62
60
45
21
co2
16
19
2
3
11
13
N2, etc.
1
1
5
1
3
52
TABLE 2
GAS PROPERTIES AND RELATIVE NOx EMISSIONS^1)
LHV Flame ^ Thermal
Gas Btu/SCF Temperature "F Relative NOx
Methane 911 3982 1.00
Carbon Monoxide 321 4536 3.42
Hydrogen 274 4311 2.07
Coal Derived '110 "2850 0.07
Low Btu Gas
(3)
5A-26
-------
TABLE 3
NO.
EPA Test
Results
HRSG ATMOSPHERIC EMISSIONS
(lb/million Btu coal)
Federal
NSPS*
Local Permit
SBCAPCD **
SO,
Utah coal
0.056
0.60
0.07
Illinois No. 6 coal
0.090
0.60
0.07
Pittsburgh No. 8 coal
0.060
0.60
0.07
Utah coal
0.024
0.17
0.033
(.35% sulfur)
Illinois No. 6 coal
0.064
0.60
0.167
(3.1% sulfur)
Pittsburgh No. 8 coal
0.111
0.60
0.167
(2.7% sulfur)
Lemington coal
0.027
0.22
0.033
(.45% sulfur)
CO
Utah coal
Illinois No. 6
Pittsburgh No.
Lemington coal
coal
8 coal
Particulates
Utah coal
Illinois No. 6 coal
Pittsburgh No. 8 coal
Lemington coal
0.002
0.003
0.002
0.001
0.011
0.008
0.009
0.011
.073***
.073***
.073***
.073***
.03
.03
.03
.03
0.015
0.015
0.015
0.015
0.011
0.011
0.011
0.011
New Source Performance Standards for coal fired power plants.
Local permit limitations, received in early 1988 from the San Bernardino
County Air Pollution Control District (SBCAPCD). The Illinois No. 6 and
Pittsburgh No. 8 coals were tested in early 1986 so the local emissions
limits were not 1n effect when these coals were tested. If the Illinois No.
6 or Pittsburgh No. 8 coals were run again at Cool Water, the SBCAPCD
requirements for NOx could be easily met by using additional steam injection
or syngas saturation.
5A-27
-------
Table 4
Clean Syngas Composition and Heating Value
for Each Test Coal Fired at Cool Water
Illinois Pittsburgh
SUFCo No. 6 No. 8 temington
Clean Dry
Composition (vol%)
Carbon Monoxide 43.0
Hydrogen (Ho) 38.1
Carbon Oioxide 18.1
Methane 0.2
Nitrogen (N?), argon
Btu/SCF 264
44.8 44.3 44.2
38.4 39.4 36.9
15.5 15.5 18.0
0.2 0.2 0.1
271 272 262
5A-28
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
DEVELOPMENT OF N0X CONTROL TECHNOLOGIES FOR COAL-FUELED
STATIONARY DIESEL POWER PLANTS
S.A. Johnson
C.L. Senior
C.K. Katz
PSI Technology Company
Research Park, P.O. Box 3100
Andover, MA 01810
R.P. Wilson, Jr.
A.D. Little, Inc.
Cambridge, ma 02138
ABSTRACT
Coal-fueled stationary dlesel engines are being developed to provide electric power
and by-product steam 1n the size range of 2 to 50 MWe. Such engines will be
commercially viable 1f the cost of cleaned coal 1s low relative to dlesel fuel, 1f
engine wear can be minimized, and 1f cost-effective emission control technologies
can be applied. The emission of primary concern 1s N0X.
The paper summarizes NOx emission measurements from a l3-1n. bore, 400 rpm,
single-cylinder test engine modified to f1 re ultra-clean coal-water fuel.
Uncontrolled emissions are a strong function of manifold air temperature (MAT); at
a nominal MAT of 230°F, the uncontrolled N0X emission 1s about 2.8 lb/ma Btu.
The paper also addresses on-going research to reduce N0X emissions to 0.6 lb/mm
Btu. Control techniques tested Include reburning downstream of the engine and
selective catalytic reduction using low-cost catalyst. Prel1a1nary N0X control
experlaents are presented and discussed.
BACKGROUND
Arthur D. Little, Inc. 1s leading a team of subcontractors 1n a OOE-sponsored pro-
gram to develop dlesel-powered cogeneratlon systems using our most plentiful and
5A-29
-------
least costly Indigenous fossil fuel: coal. A key subsystem of such a plant will be
the emission control equipment. The PSI Technology Company 1s responsible for
evaluating, developing, and designing equipment capable of meeting today's EPA New
Source Performance Standards for coal-fired stationary sources. Emissions of
primary concern are nitrogen oxides, sulfur dioxide, and particulate matter. Inte-
gration of processes to control these emissions while maintaining cycle efficiency
and minimizing capital Investment and operating costs 1s the goal of our work.
The program 1s divided into three phases. In the first phase, we have completed
conceptual designs for several candidate N0X control technologies. Since these
technologies have rarely, 1f ever, been applied to stationary engines, these
designs have drawn heavily upon utility and Industrial boiler experience to develop
general arrangement drawings, material lists, and costs. The results of the design
study allowed us to select and prioritize attractive technologies and to direct
Initial research toward key technical Issues that may Impact the potential
application of each technology to dlesel systems.
Laboratory-scale testing of selected NOx-control processes 1s now underway. The
results of these tests will be Integrated Into a final emission-control subsystem
design which 1s the final output of the Phase I effort.
In Phase II, the N0X control equipment selected and designed 1n Phase I will be
built and tested on a Cooper-Bessemer s1x-cy11nder test engine located at Cooper's
research facility 1n Mt. Vernon, Ohio. If successful, an Integrated engine system,
Including optimized fuel, engine, emission control and heat recovery subsystems
will be demonstrated 1n Phase III.
UNCONTROLLED N0X EMISSIONS
JS-1 Engine
Testing to date has been performed at Cooper Bessemer 1n a JS-1 research engine.
This engine has a 13-1n. bore, a 16-in. stroke, and operates 1n the speed range of
300 to 450 rpm. For the tests reported herein, the engine rpm was held constant at
400 and the engine load was a nominal 150 ps1. This engine has been described
previously; (1,2) so this discussion will be limited to the engine features that
may influence N0X emissions.
5A-30
-------
A cross section of the cylinder head is shown on Figure 1. The coal-water fuel
injector 1s mounted 1n the center of the cylinder. There are two other openings
which can be fitted with a diesel pilot injector, a jet cell natural gas ignition,
or optical diagnostic equipment. In the tests reported, diesel fuel was used as a
pilot fuel, although autoignitlon of the coal-water fuel was possible for air
temperatures greater than 250°F. The pilot fuel flow amounted to 7 to 16 percent
of the total heat Input to the cylinder.
A conventional cam-driven jerk pump delivered the coal-water fuel to the injector.
This allowed injection pressure to be varied Independent of total fuel flow. A
separately driven compressor supplies combustion air to the engine, with heat
exchangers to vary the manifold air temperature from 80° to 430°F independent of
pressure or airflow. It will be seen that injection pressure and manifold air
temperature were the key variables influencing N0X emissions.
The coal-water fuel burned during these tests consisted of 52 percent dry coal,
2 percent additives, and 46 percent water. The fuel had mean particle size of
11 microns and a topslze of about 74 microns. The higher heating value of the
as-f1red fuel was about 7550 Btu/lb and the fuel viscosity at 100°F ranged from
180 to 210 centipolse for a shear rate of 100 to 1000 s*1. Fuel-bound nitrogen,
which could contribute to the N0X emission, was 1.6 to 1.8 wt% on a dry basis.
N0y Emission Results
In order to compare N0X levels from different tests, the N0X data must be converted
to a common basis. The N0X concentration of the gas was measured on a wet basis,
while the CO2 and 02 levels are given for a dry gas. Therefore, for each
experimental point, the amount of H2O in the gas was calculated based on the
ultimate analysis of the fuels. Given a value for the HjO content, the O2 content
of the exhaust gas on a wet basis can be calculated. The N0X level 1n the exhaust
gas was then corrected to conditions of 15 percent oxygen.
Table 1 compares N0X emissions for the JS-1 engine operating at elevated manifold
air temperatures during consecutive tests firing diesel fuel and coal-water fuel.
In general, the N0X emissions were similar for each fuel, although considerable
day-to-day variability could occur. We speculate that any decrease in the thermal
N0X formation brought about by quenching peak flame temperature when firing CWF is
approximately balanced by the production of N0X from nitrogen bound 1n the coal.
(Diesel fuel contains negligible amounts of fuel-bound nitrogen.) However, it is
5A-31
-------
expected that the conversion of fuel nitrogen to N0X will be affected by the rate
of coal devolltiUzation, which will depend on the size of the droplets produced
during atomlzation. Atomization changed throughout the test program as injector
design and Injection pressure was varied, and as the injector nozzle wore away due
to the abrasive effect of the coal-water fuel.
NOy emissions were sensitive to both peak injection pressure and manifold air
temperature. Figure 2 shows the N0X concentration as a function of peak injection
pressure for several different values of the intake manifold air temperature. The
N0X concentration that has been used in the preliminary design of the emission con-
trol system (900 ppm at 11* O2) 1s indicated on the figure. The N0X concentration
in the exhaust shows a general trend of increasing with peak Injection pressure.
In Figure 3, data for a narrow range of injection pressures (8,700 to 10,000 psi)
are shown as a function of Intake manifold temperature. As expected, N0X concen-
tration in the exhaust increases with increasing manifold air temperature. At air
temperatures greater than 300°F, the effect on NOx emissions 1s more pronounced.
N0X CONTROL TECHNOLOGIES
It can be seen from the above discussion that uncontrolled N0X emissions from coal-
flred dlesel power systems are likely to be considerably higher than other station-
ary sources. Although the U.S. Environmental Protection Agency has not promulgated
New Source Performance Standards covering this technology, we have examined control
technologies (or combinations of technologies) capable of reducing N0X by 80 per-
cent; i.e., from 2.8 to 0.6 Ib/mmBtu. Modifications to engine design and/or oper-
ation to minimize baseline N0X have not been tested to date. Other than using a
two-chamber engine to achieve rich-lean combustion, any engine modification that
decreases combination temperature or rate of fuel-air mixing 1s expected to also
adversely affect engine efficiency. Therefore, we are limiting our study to post-
combustion N0X control technologies.
At this point, we have completed preliminary designs and cost estimates for the
emission control technologies listed 1n Table 2. The N0X control technologies are
discussed below.
5A-32
-------
Selecti ve Reduction
In selective reduction, a nitrogen-containing compound (typically ammonia, although
urea has also been used) is added to the exhaust gas stream where it reacts with
oxides of nitrogen to form N2. Depending on whether a catalyst 1s employed the
process is called Selective Non-Catalytic Reduction (SNR) or Selective Catalytic
Reduction (SCR). The SNR reactions occur in a narrow temperature window (1600° to
2000°F). Since the maximum exhaust temperature expected from a coal-fueled dlesel
engine 1s about 900°F, any SNR processes are not applicable.
SCR processes employ a catalyst to allow the N0X to react with the nitrogenous
reducing agent at even lower temperatures (300° to 750°F). Commercial applications
on utility boilers employ a variety of catalysts (platinum, vanadium, copper
sulfate, zeolite) coated onto a less expensive substrate and arranged as a
honeycomb or parallel plate to maximize exposure of catalyst surface per unit of
reactor volume.
The major potential operating problems when applying SCR to coal-fueled dlesel
engines are degradation of catalyst performance due to fouling with engine
particulate, ammonia slip, and deposition of ammonium sulfate onto downstream heat
transfer surfaces. These problems appear tractable, but the solutions could be
expensive. Control of ammonia slip requires a sophisticated N0X and/or NH3
monitoring system to assure that the NH3/N0X stoichiometric ratio always remains
less than unity. Catalyst supports must be open or accessible to soot blowers to
prevent pluggage or catalyst blinding by the very fine, carbonaceous flyash
produced by the coal-fueled dlesel. Vendors contacted to date estimate a 2 to
3 year catalyst life under base-loaded operation.
In order to assess whether the N0X reduction requirements of the emission control
system can be net with a less expensive SCR process, two alternative technologies
are being considered. The first alternate SCR process Involves using the
carbon-containing particulate matter from the engine exhaust as a "free" catalyst.
The second Involves an Iron oxide catalyst and non-amnon1a reducing agents that may
be more reactive with N0X.
Recently, the catalytic activity of coke and carbon-containing particulate matter
from utility boilers has been demonstrated on a laboratory scale. (3,4) Using coke
activated with sulfuric acid, a 70 percent reduction of NO by ammonia at 300°F has
been demonstrated. (3) Particulate matter from industrial boilers containing 2 to
5A-33
-------
30 percent carbon by weight has been used as a catalyst for reduction of NO with
ammonia. (4) In the temperature range 570° to 750°F, NO reductions from 40 to
60 percent have been demonstrated. Because of the low ash content of the processed
coal used in the engine, the particulate matter contains a very large amount of
carbon. The carbon content of the ash was measured to be 80 percent at 98 percent
burnout and we calculate 67 percent carbon if combustion efficiency is improved to
99 percent burnout, based on a mass balance using 0.5 percent as the ash content.
Thus, it may be possible to exploit the ash as a reduction catalyst. This process
will be studied further by conducting laboratory-scale feasibility experiments.
Other reducing experts and catalysts for SCR are being pursued in separate
experimental programs at PSIT funded by the EPA and A1r Force under the government
Small Business Innovative Research program.
Reburning
In the reburning process, N0X is reduced to N2 as the exhaust gas from the primary
combustion process passes through a secondary combustion zone. In boiler
applications, fuel-rich reburning conditions allow NO to react with hydrocarbon
radials to form CN/HCN. The kinetic path to N2 is complex, as shown on Figure 4.
H and OH radicals are required to convert CN species to NH^ species (1*0,1,2). N
or NH2 reacts rapidly with NO to form N2; this reaction 1s favored at lower
temperatures, while free radical formation 1s favored by higher temperatures.
Since free radicals are also formed upon the Injection of burnout air downstream of
the reburn zone, further N0X reduction or minimal N0X reformation can be achieved
through the design and location of the burnout air Injectors. (5)
Application of reburning to stationary dlesel power plants 1s different from
boilers in two important regards
0 Diesel engine exhaust contains 10 to 12 percent O2 compared to 2 to
4 percent for boilers
0 The temperature of the exhaust leaving the dlesel engine is about 900°F,
much lower than the reburn gas temperature 1n boiler furnaces.
Fuel-lean reburning has been shown to be an effective N0X control method in some
combustion systems. Yang et al. (6) performed tests relative to process heater
applications 1n which N0X could be reduced by as much as 50 percent. Previously,
various investigators (7-9) noted that N0X formed in the gas turbine portion of a
5A-34
-------
combined cycle power system could be destroyed under certain conditions 1n a duct
burner or exhaust-fired heat recovery boiler flame. This non-opt1m1zed form of
reburning may be very attractive for diesel systems whose very high N0X emissions
provide a large driving force for reburning.
Recently, Brown and Kirby (10) studied reburning for gas turbine and diesel
applications. They concluded that when the exhaust O2 content exceeded 11 percent,
large amounts of reburning fuel (23 to 50 percent of the total heat input to the
system) were required to achieve a 50-percent N0X reduction. However, mixing rate
was difficult to control in their 100,000 Btu/hr test facility.
Our application of this reburning concept is shown on Figure 5. The heart of the
process is a 1ow-N0x reburner located in the gas stream. This reburner consists of
a central fuel injector surrounded by a shroud to control the rate of fuel air
mixing. The shroud 1s perforated to allow a portion of the exhaust to mix rapidly
with the reburning fuel. Ideally this initial mixture would have an air/fuel
stoichiometric ratio of 0.7 to 0.9 to allow stable Ignition. The rest of the
exhaust would pass around the shroud to mix slowly but completely as the excess
reburning fuel (and NO) 1s consumed.
EXPERIMENTAL PROGRAM
As of early January, the PSIT experimental program was just beginning. Preliminary
experiments will examine ammonia (and other reducing agents) Injection upstream of
a baghouse. The baghouse will utilize NOMEX filter material allowing operation up
to a temperature of 450°F or ceramic fibers for higher temperature. The test
program will Involve systematic variations of the following parameters:
• Filter media: clean bag filter, filter plus Hme, filter plus engine
particulate, filter plus Hme and engine particulate, filter plus
activated carbon
• Temperature: 300°, 450°, 600°, 750°F
• Reducing agent: none, ammonia, other.
The key Issues to be resolved during these tests are:
5A-35
-------
• Is engine particulate matter (80 percent carbon, 5 urn mean size, 22 m2/g
surface area) an effective N0X reduction catalyst?
• Under what operating conditions is N0X reduction maximized?
• Can the technology be integrated with a bag filter or will it require a
separate reactor to maximize N0X reduction?
Actual particles collected from the JS-1 engine will be characterized and tested by
PSIT in these experiments.
Reburning experiments will also be conducted. Initial tests will involve
systematic variation of the reburner components to define mixing conditions
conducive to N0X reduction. The bench-scale reburner has been fabricated as shown
on Figure 6. This reburner 1s designed for easy modification to the fuel injector
and the shroud to optimize N0X reduction and burnout.
Once the optimum configuration is established, the following operating variables
will be explored:
•
Inlet N0X concentration:
100 to 1300 ppm
•
Inlet O2 concentration:
8 to 19 percent
•
Inlet gas temperature:
300° to 1800°F
•
Reburn fuel flow:
10 to 40 percent of total heat input
•
Exhaust gas Inlet velocity:
4 to 20 ft/s
•
Reburning fuel additives:
to be determined.
The technical issues to be resolved during these tests Include:
• What is the minimum N0X emission achievable when applying reburning to
coal-fueled dlesel stationary power plants?
• Will the technology be limited by either low temperature or high/low O2
concentration of the Incoming exhaust gas?
• How much reburning fuel will be required to achieve at least a 50 percent
N0X reduction?
5A-36
-------
• What »111 be the temperature and composition of the gas leaving the
reburner reactor, and will this impact the operation/design of downstream
emission control or heat recovery components?
The results of the test, if successful, will result in the design of a reburner for
testing an a 1.8 MWe engine at Cooper-Bessemer.
PRELIMINARY N0X CONTROL SYSTEM DESIGN AND COSTS
The goal of this study is to quantify the cost and probability of success for the
emission control system options. The analysis of the costs takes two forms.
First, the cost of the equipment for the component technologies 1s presented. From
these costs, comparisons are made among the various technologies for controlling
the three major pollutants. Second, the capital and operating costs for integrated
emission controls systems are presented as a function of system size.
Component Technologies
The cost of emission control equipment relative to the power output of the system
(e.g., the cost per kilowatt of electricity generated) decreases as the size of
system increases. In general, the relationship between cost and system capacity is
not linear. Therefore, the cost calculations were performed for five systems that
span the range of Interest for small stationary engines. Table 3 details the
size-dependent operating parameters for the five systems of Interest. Each system
consists of one or more engines (either 6 or 20 cylinder engine). Systems l, 2, 3,
and 5, labeled as cogeneratlon systems, include a waste heat boiler which makes
125 ps1g steam for process applications. System 4, a power generation system, does
not have a waste heat boiler. In this configuration, no heat recovery from the
exhaust gas takes place and a water quench unit 1s employed to reduce gas
temperatures as required by pollution control equipment.
Schematic diagrams and materials lists for the N0X control technologies under
consideration are given on Figures 7 to 9. The materials 11st was used to estimate
process capital costs according to the EPRI methodology. (11) The capital costs in
$/kW presented herein represent Total Capital Investment divided by the net power
production of each system.
5A-37
-------
The operating and maintenance cost is the sum of operating labor, maintenance
(labor and materials), and overhead labor. The annual maintenance costs are
estimated as a percentage of the process capital cost. The EPRI methodology
suggests maintenance cost factors for different technologies. The maintenance cost
is broken down as 40 percent labor and 60 percent materials. The overhead charge
is computed as 30 percent of the operating and maintenance labor and is a charge
for administrative and support labor.
Following the EPRI methodology, the operating and maintenance cost is broken down
into a fixed component and a variable component. The fixed operating and
maintenance cost is the product of the capacity factor (or the utilization) and the
annual operating and maintenance costs; the balance of the cost is assumed to be
variable. The variable operating cost has two components: consumables (power,
chemicals, etc.) and operating and maintenance, as described above.
The total annual operating cost 1s the sum of the fixed operating and maintenance
cost, the variable operating cost, and the cost of capital. The cost of capital 1s
based on a 12-year depredation period at 10 percent Interest. The yearly charge
for the cost of capital is 15 percent of the total capital investment. The
operating cost 1s expressed 1n millions of dollars per year or 1n m11ls/kWh. The
latter unit 1s the annual operating cost (expressed as thousandths of a dollar)
divided by the total hours that the equipment runs (at 65 percent utilization this
is equal to 5694 hr) and the power output of the facility (1n kW).
Table 4 gives the cost of equipment for Individual N0X control technologies
expressed on a per kilowatt basis 1n 1988 dollars. Note first that the commercial
SCR process 1s by far the most expensive option, based on data available from
A.D. Little Inc. for conventional dlesel systems and EPRI data for an 80 MWe coal-
fired boiler (12). The cost of ammonia injection 1s high for the small engine but
lower for larger engines because it 1s dominated by the costs of ammonia storage
and handling, and process controls which are relatively Insensitive to engine size.
The cost for reburnlng 1s the sum of the cost of a duct burner and N0X monitoring
equipment. One duct burner is used per engine. For the 6 cylinder engine, the
burner must supply 4 MMBtu/hr of heat if the reburnlng process represents
20 percent of the total heat input. For the 20 cylinder engine, the burner must
supply 13 MMBtu/hr. Three burner manufacturers were contacted; one quote was used
for the duct burner costs because it was the most complete of the three.
5A-38
-------
Integrated Emission Control Systems
Currently, dlesel-fueled stationary engines are not equipped with the kind of emis-
sion control technology that will be needed to meet NSPS with the coal-fueled
engine. The emission control devices will have to be integrated with the engine
system components. A schematic of the five Integrated systems is illustrated in
Figure 10. Only one engine is shown; in this study, we have made the preliminary
decision that emission control equipment downstream of the turbocharger will be
common to all engines in the system, i.e., exhaust gases from all engines are com-
bined after the turbocharger. Approximate temperatures are Indicated 1n the figure
although the exact temperature will depend on the optimum temperature required for
the individual emission control subsystems and on the engine exhaust conditions.
Figure 11 shows capital cost as a function of power plant size for the five
emission control systems under consideration. Note that the two options pairing
reburnlng or ammonia injection with sorbent injection for SO2 control offer a wide
cost advantage over commercial SCR. Options combining expensive N0X control with
Inexpensive SO2 control (and vice versa) fall 1n the middle. It should be noted
that even 1f reburnlng and ammonia Injection are both needed to meet N0X emission
regulations, the cost 1s likely to be less than an SCR system. ¦ Operating costs,
shown on Figure 12, exhibit similar trends (a catalyst life of 2 years and a bag
life of 1 year have been used to estimate these costs).
Since this study represents a preliminary design, with 20 to 30 percent uncertainty
1n the calculated costs, a formal sensitivity analysis has not been performed.
However, two aspects of the sensitivity of the economic calculations to input data
should be mentioned. First, the engine operating parameters and system config-
uration may change as a result of the on-going JS-1 test program. Second, the
process parameters for control technologies may change as a result of the
laboratory experiments and on more detailed discussions with manufacturers of
commercially available technology. The target date for commercialization of the
technology 1s 1995-2000.
REFERENCES
1. E. N. Balles, K. R. Benedek, R. P. Wilson, and A. K. Rao, "Analysis of
Cylinder Pressure and Combustion Products from an Experimental Coal-fueled
Diesel Engine," presented at the ASME Internal Combustion Engine Division
Conference, New Technology for Coqeneration, ICE, Vol. 2, Kansas City, MO,
October 1987.
5A-39
-------
2. A. K. Rao, C. H. Melcher, R. p. Wilson, E. N. Balles, F. S. Schaub, and
J. A. Klmberly, "Operating Results of the Cooper-Bessemer JS-l Engine on
Coal-Water Slurry," presented at the Energy-Sources and Technology Conference,
New Orleans, LA, 1988.
3. I. Mochida, M. Ogaki, H. Fujitsu, Y. Yoshinobu, and S. Ida, "Catalytic
Activity of Coke Activated with Sulphuric Acid for the Reduction of Nitric
Oxide," Fuel, 62:867, 1983.
4. P. Davinl, "Reduction of Nitrogen Oxides with Ammonia: The Activity of
Certain Soots," Fuel, 67:24, 1988.
5. R. W. Borio, "Application of Reburning on a Cyclone-Fired Boiler," this
Symposium, March 1989.
6. R. J. Yang, J. K. Arand, and F. J. Garcia, "Laboratory Evalutlon of
In-Furnace-NOx-Reduction for Industrial Combustion Applications," ASME Paper
84-JPGC-FU-12, October, 1984.
7. J. K. Arand and L. J. Muzio, "Oklahoma Gas and Electric Combined Cycle N0X
Reduction Laboratory Tests," KVB, Irvine, CA, 1977.
8. S. A. Johnson and A. H. Rawdon, "Control of N0X Emissions from Power Boilers,"
The Institute of Fuel, Adelaide, Australia, November, 1974.
9. M. W. McElroy, "Oklahoma Gas and Electric Full-Scale Demonstration of N0X
Destruction 1n Fuel-Rich Burners," EPRI Program RP-782, Palo Alto, CA, 1977.
10. R. A. Brown and W. C. Kirby, "Application of Reburning for N0X Control in
Cogeneratlon," 1985 Joint EPA/EPRI Symposium on Stationary Combustion N0X
Control, Boston, MA, May 1985.
11. Electric Power Research Institute, Technical Assessment Guide, EPRI Report
No. P-2410-SR, 1982.
12. D. R. Swann and G. D. Drlssell, Feasibility of Retrofitting Catalytic Post-
combustion NOy Control of an 80-MW Coal-F1red Utility Boiler, EPRI Report
No. CS-1372, 1980.
5A-40
-------
NATURAL GAS COAL SLURRY DIESEL PILOT
"JET CELL" INJECTOR INJECTOR
Figure 1. Configuration of Test Engine
600
500
£
q 400
Ui
o
HI
C
|
cT
z
300
200
900 ppm AT
77%o7~ ®"
a
a
~
rj 348'F MAT
105% EXCESS AIR
m 253 *F MAT
60 - 90% EXCESS AIR
9 371'F MAT
_ 120% EXCESS AIR
• 240 T MAT
90 - 95 EXCESS AIR
10 15
PEAKT INJECTION PRESSURE [103 psig]
20
Figure 2. JS-1 Test Results for N0X Versus Peak Injection Pressure
5A-41
-------
INTAKE MANIFOLD TEMPERATURE [-F]
Figure 3. JS-1 Test Results for N0X Versus Intake Manifold Temperature
(Peak Injection Pressure 8,700 - 10,000 psig)
Figure 4. N0X Formation and Destruction Pathways During Natural Gas Reburning
5A-42
-------
Figure 5. Reburn Schematic - Plan View
Photo to come
Figure 6. Laboratory-Scale Reburner
FROM CYCLONE
[880F, 12.0 pttgj —~
->• TO TURBOCHARGER
[1330T.12.0 pstg]
Qj MONITOR
aCH«
CONTROL
VALVE
CONTROL
SYSTEM
CH4
SUPPLY
Figure 7. Reburning
5A-43
-------
Figure 8. Selective Catalytic Reduction
Figure 9. NH3 Injection 1n Baghouse
5A-44
-------
xf* mmtrur* r.r iibttwim
ognnwB
cmtuiaam
IKS
~Q|
- HVBmtEDtliE
s *»*>
«Acn
rM»°.
) *
ammonu iNJgcfio^ucT snaffwr «c
OPTION J
««STf
#M,0^
«M*ft
H«AnRMum
-~•MjO f
iMao MO.l
1
MSwmIOi
~
U««IKUN
Figure 10. Integrated Emission Control Systems Under Consideration
5A-45
-------
1000
1987
BUDGET
ESTIMATE
NET POWER (MWel
Figure 11. Capital Cost for Cogeneration System
100
90
80
I
oi 70
60
PROJECTED COST
OF POWER
I 50
1
OPTION A: COMMERCIAL
TECHNOLOGY
40
30
20 _
10
0
1987
I BUDGET
t ESTIMATE
10 20
NET POWER [MWeJ
30
Figure 12. Annual Operating Cost for Cogeneration Systems
5A-46
-------
Table 1
COMPARISON OF NOx EMISSION FROM JS-1 ENGINE
FOR EACH FUEL TYPE
Manifold
A1 r
Temperature
(°F)
N0X Emission
DF-2 Only
N0X Emission
CWF + DF-2
N0X Emission
CWF Only
ppm 9 15% O2
lb/MBtu
ppm Q 15% O2
lb/MBtu
ppm 9 15% 02
lb/MBtu
340
403
1.69
426
1.79
-
-
440
657
2.52
725
3.04
-
-
434
-
-
1029
4.32
907
3.81
426
864
3.63
696
2.91
Table 2
EMISSION CONTROL SUBSYSTEMS
NOy Control
Reburnlng
SCR: Commercially available unit
SCR: Ammonia Injection over engine particulate catalyst at 450°F
1n baghouse
SCR: Ammonia Injection over CuO catalyst at 750°F 1n granular bed
reactor (Shell/UOP) process
Adsorption on engine particulate bed at 275°F 1n baghouse
5A-47
-------
Table 3
PROCESS PARAMETERS FOR COAL-FUELED DIESEL SYSTEMS
Engine -»
1
LSB 6
2
LSB 6
3
LSVB 20
4
LSVB 20
5
LSVB 20
Application
Co-gen
Co-gen
Co-gen
Power
Co-gen
Number of Engines
1
4
2
2
4
Net power output (MWe)
1.8
7.2
12
12
24
Total horsepower
2,516
10,064
16,772
16,772
33,544
Coal flow (Ib/hr)
1,154
4,616
7,692
7,692
15,384
Heat input (MMBtu/hr)
15.86
63.44
105.66
105.66
211.32
Exhaust gas flow (scfm)
single engine
6,979
6,979
23,260
23,260
23,260
Total (all engines)
6,979
27,916
46,520
46,520
93,040
Exhaust gas flow (klb/hr)
single engine
30.9
30.9
103.0
103.0
10.30
Total (all engines)
30.9
123.7
206.1
206.1
412.2
Emissions (lb/hr)
S02
NOx
Solids
21
45
27
84
178
110
140
297
183
140
297
183
280
594
365
Table 4
EQUIPMENT COSTS FOR NOx CONTROL TECHNOLOGIES
Cost ($/kW)
as a Function of System Size:
Subsystem Option
Installation
Multiplier1
1.8 MW
7.2 MW
12 MW
24 MW
Reburnlng
0.33
18
18
9
9
NH3 Injection 1n
baghouse at 450°F
0.40
47
12
8
5
Commercial SCR
0.37
92
60
53
43
J-To calculate installation cost,
installation cost multiplier
multiply equipment cost by the
5A-48
-------
Session 5B
INCINERATION
Chairman: W. Linak, EPA
5B-i
-------
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
Reduction of NO Emissions
From MSW Combustion Using
Gas Reburning
by
Craig A. Penterson, D. C. Itse
Riley Stoker Corporation, USA
Hamid A. Abbasi
Institute of Gas Technology, USA
Yasujlro Wakamura
Takuma Company Ltd., Japan
David G. L1nz
Gas Research Institute, USA
ABSTRACT
A program has been Initiated to evaluate the potential for reducing air pollutant
emissions from Municipal Solid Waste (MSW) combustion systems using natural gas
reburning. The program 1s being conducted jointly by the Institute of Gas
Technology, the Gas Research Institute, Riley Stoker Corporation and Takuma
Company Ltd.
Three major tasks are included in the program: (1) acquisition of baseline data
from a commercial operating facility, (2) pilot-scale development and testing of
the reburning technology, and (3) field demonstration testing of the technology.
This paper focuses on the results to date of the first two tasks: baseline data
acquisition and pilot scale testing.
The gas reburning concept being evaluated for MSW combustion Is similar to the
reburning technique being investigated to reduce N0X emissions from fossil fuel
combustion systems. Natural gas 1s Introduced above the main combustion zone to
create a reducing environment where significant amounts of N0X and other fixed
nitrogen species (NH1, HCN), generated 1n the main combustion zone, are reduced to
molecular N2.
Results of the pilot scale testing have demonstrated up to 70* reduction 1n N0X
emissions using gas reburning. The effects of gas Injection location, reburning
zone stolchlometry and residence time, overall boiler excess air and the amount of
natural gas reburning on N0X emissions are characterized. The Impact of gas
reburning on reducing other emissions such as CO, unburned THC and dloxin is also
discussed.
5B-1
-------
INTRODUCTION
With the expanding waste-to-energy market today, the need to understand more about
the combustion characteristics and behavior of municipal solid waste (MSW) and
refuse derived fuel (RDF) has increased dramatically. Of primary concern are the
air pollutant emissions produced during the combustion process and the ability to
control these emissions. Flue gas scrubbers are generally employed in modern
facilities at the boiler exit to control the acid gas emissions such as HC1 and
SOj. Baghouses or electrostatic precipitators are used to remove the particulate
emissions as well as the reacted lime from the scrubber. Other emissions could
potentially be controlled during the combustion process. These include N0X, CO,
unburned THC, polychlorinated dibenzodioxins (PCDD's) and polychlorinated
dibenzofurans (PCDF's).
In April, 1987, the Institute of Gas Technology {IGT), the Gas Research Institute
(GRI), Riley Stoker Corporation (Riley) and Takuma Company Ltd. from Japan
(Takuma) began work on a research program to evaluate the potential for reducing
pollutant emissions from MSW combustion systems using natural gas. The primary
objective of the overall program is to investigate the use of natural gas
reburning in combination with low excess air operation to minimize the production
of NOj. emissions. As conceptually shown 1n Figure 1, natural gas would be
introduced above the main combustion zone 1n a MSW combustor to create a strong
fuel rich or reducing zone where N0X and other fixed nitrogen species (NHi, HCN),
generated in the main combustion zone, would be reduced to harmless molecular
nitrogen. Natural gas reburning might also serve to reduce other pollutant
emissions such as CO, unburned HC, PCDD's and PCDF's by creating a more uniform
and stable thermal environment in the combustor.
Other benefits expected from this type of boiler operation are improved combustion
efficiency, increased boiler efficiency and more stable, reliable operation of the
system. As outlined below, the research program is divided Into three major
tasks. Task 1 was completed in the fall of 1987 while Task 2 is nearly complete.
Task 1 - Acquisition of Baseline Data From a Field Operating Unit.
Task 2 - Pilot Scale Development and Testing of the t
Gas Reburning Technology.
Task 3 - Field Demonstration Testing at a Commercial Installation
This paper discusses the results to date of Tasks 1 and 2. Due to the initial
favorable results incurred during the pilot scale testing at both IGT and Riley,
preliminary plans are currently being formulated for a field demonstration test of
the gas reburning technology.
GAS REBURNING TECHNOLOGY
The gas reburning technology being developed 1n this program 1s similar to the
reburning techniques being Investigated to reduce N0X emissions from coal fired
combustion systems. As shown in Figure 1, gas would be Injected into the furnace
through multiple openings above the main combustion zone at 10-15% of the total
heat Input. Recirculated flue gas, equal to 10-20* of the total boiler flue gas
approaching the take off point, would be used to convey the natural gas and
achieve the proper penetration and mixing required 1n the furnace. Overfire air
(OFA), would then be added above the gas reburning zone to complete the combustion
process.
5B-2
-------
The reburning zone created by the natural gas would provide the H and OH radicals
necessary to react with and reduce the N0X and other fixed nitrogen species,
generated in the main combustion zone to molecular N2. As discussed later in this
paper, the N0X emissions being produced from a typical Riley-Takuma mass burn
system (1) average 150 PPM^1'. Our goal is to reduce the NO emissions to 50 PPM,
comparable to the levels currently attainable only by selective catalytic
reduction (SCR) or urea injection (2, 3).
Another benefit anticipated from using gas reburning is the ability to create a
more uniform and stable thermal environment within the furnace, thus eliminating
localized low temperature zones and pathways. As described later, furnace
temperature fluctuations on the order of ±150°F, over a one minute period, were
measured during the baseline testing at a commercial facility while burning only
MSW. Temperature differences as large as 160°F between the front and rear walls
were also measured. These erratic temperature profiles may contribute to products
of incomplete combustion (PIC) emissions by allowing the reactions to occur at
insufficient temperature. Work by Duvall shows that below approximately 1500°F,
thermal destruction of many PIC's begins to rapidly decrease (4). Gas reburning
should, therefore, enhance the burnout of CO, unburned THC, PCDD's and PCDF's by
creating a more uniform temperature environment in the furnace.
Gas reburning, in combination with low excess air operation, should also improve
boiler efficiency. Reducing the overall excess air requirement from the current
80% to 40% could potentially increase boiler efficiency by 1.5%.
BASELINE FIELD TEST RESULTS
In July 1987, parametric field tests were performed during Task 1 by Riley and IGT
on Unit 1 of the Olmsted County Waste-to-Energy facility in Rochester,
Minnesota. This 100 ton/day unit, shown in Figure 2, was designed and constructed
by Riley to incorporate the Japanese technology for mass burning that Riley
licenses from C. Itoh Takuma Ltd., Japan. As shown in Figure 3, the Takuma stoker
consists of four separate sections: feeder table, drying and ignition grate,
combustion grate and burnout grate.
Between the grate sections are two to three foot drops or steps which reduce top
to bottom fuel stratification and help to break up any large agglomerations of the
fuel. By hydraulically reciprocating every other individual grate row, the fuel
cascades and moves down the various grate sections. Fuel drying, pyrolysis and
ignition occur on the drying and ignition grate. About 95% of the fuel combustion
occurs on and over the combustion grate. The burnout grate provides time for
carbon burnout and reduction of putrisclbles under locally high excess air
conditions. The Riley-Takuma systems are designed to handle refuse throughputs
from 100-1000 tons per day.
During the field tests at Olmsted County, pollutant emissions were measured at the
economizer exit or precipitator inlet as well as at critical locations in the
furnace. These emissions included N0X, SO2, CO, unburned THC and putrlscibles in
the ash. Tests were conducted while varying the amount and location of OFA,
undergrate air (UGA), excess air and boiler load. Using suction pyrometers,
furnace gas temperature measurements were also collected in the upper furnace and
furnace backpass. However, attempts to measure furnace gas temperature in the
lower furnace immediately above the grates were unsuccessful due to probe
plugging.
All emissions have been corrected to 12% 02 unless stated otherwise.
5B-3
-------
Results of the testing showed that operation of the overfire air (OFA) system had
the most significant effect on controlling N0X, CO and unburned THC emissions.
The amount of OFA had the greatest impact. Since excess air is automatically
controlled by the amount of OFA (1), increasing excess air would cause an increase
in the N0X and a decrease in CO emissions, typical of most combustion systems.
This is shown in Figure 4. At design excess air operation, NO emissions entering
the electrostatic precipitator averaged 150 PPM while CO was 28 PPM. Unburned THC
emissions measured at this same location averaged less than 5 PPM. In general,
the CO emissions would only increase at very low levels of excess air (<30%) when
OFA was completely closed. But, the level would still remain below 200 PPM.
Detailed in-furnace gas composition measurements showed the effect of OFA in
controlling emissions. Figure 5 summarizes the gas composition history measured
at various locations throughout the entire unit during full load normal
operation. N0X increases following the addition of OFA while CO and unburned THC
are reduced significantly. Less than 50% of the total N0X emission is formed
immediately above the grate while the remaining N0X is formed in the OFA region.
In addition, since the predicted peak furnace temperature in this area remains
below 2400°F, N0X is primarily formed from the conversion of fuel bound nitrogen.
Figure 6 shows the gas temperature history of the unit based on actual furnace
temperature measurements and calculated flue gas flows. Furnace temperatures are
typically less than 2400°F. However, not shown on the figure, are the significant
temperature fluctuations and stratification measured in the furnace. During
several tests, fluctuations of ±150°F over a one minute period were measured while
the temperature difference between front and rear furnace walls was as large as
160°F. Though the CO and unburned THC emissions remain relatively low with this
type of thermal environment, the benefit of using natural gas reburning to
stabilize this environment may potentially reduce these emissions even further as
well as provide the environment for reductions in PCDO's and PCDF's.
The results of the testing at Olmsted County were then used to form the basis for
the pilot scale testing performed initially at IGT in a furnace simulator and then
at Riley in an actual pilot MSW combustion facility. The purpose of the testing
at IGT was to initially investigate the feasibility of using gas reburning to
reduce N0X emissions and to identify those variables which had the most
significant impact on controlling N0X. The test variables included:
• reburning zone residence time
• reburning zone stolchlometry
• reburning zone temperature
• amount of natural gas reburning fuel
• amount of flue gas recirculation (FGR) for natural gas transport
• overall stolchlometry leaving the furnace
Subsequent testing at Riley Stoker was then performed to further characterize and
define an optimum gas reburning strategy that would be directly transferable to a
commercial installation. The following section briefly describes the major
results of the pilot scale testing at IGT followed by a more detailed discussion
of the initial process development tests conducted at Riley Stoker.
5B-4
-------
IGT PILOT SCALE TESTING (TASK 2)
One of the pilot-scale furnaces, at IGT's Applied Combustion Research Facilities,
was modified to simulate the combustion products measured in the lower furnace
section of the Olmsted County MSW combustor. As shown in Figure 7, combustion
products, that would result from firing MSW at a rate of 1.7 x 10° Btu/hr, were
simulated by firing No. 2 fuel oil into preheated combustion air with subsequent
blending of water, ammonia and oxygen. The hot combustion products were
introduced into one end of the 4.5 ft wide x 3 ft high x 14 ft long furnace
chamber and distributed across the furnace cross section using a refractory
grid. The furnace chamber was refractory lined and equipped with cooling tubes to
simulate the thermal environment measured in the field.
In the combustion chamber, the natural gas for reburning, conveyed by recirculated
flue gas, was introduced through holes in the top and bottom of a 4" diameter
stainless steel pipe. This injector, extending the full width of the combustor,
was placed at the centerllne of the combustion chamber just downstream of the
distribution grid. Secondary air, simulating overfire air, was also introduced
through a similar injector design, downstream of the reburning zone, to complete
the burnout. The furnace is equipped with 32 removable doors along the length
which allowed for significant flexibility in the placement of these injectors and
probes.
The reburn zone stoichiometry and temperature were changed by varying the amount
and temperature of the preheated combustion air, while reburn zone residence time
was varied by placing the secondary air injector at various locations along the
furnace length. Test measurements focused on gas composition (O2, CO, THC, CO2
and N0X) and temperature in both the reburn zone and at the furnace exit.
Table I shows a comparison of the gas composition measured without gas reburning
in the IGT furnace simulator with the gas composition measured in the baseline
test furnace at Olmsted County. The pilot-scale results showed good agreement in
regards to gas composition, furnace temperature and residence time. NO emissions
varied (100 - 300 PPM) primarily depending on the amount of NH3 injected into the
flue gas products.
Testing the affect of gas reburning on reducing N0X emissions showed very
promising results. N0X reductions of up to 70% were measured. The most
significant variable was the stoichiometric ratio of the reburning zone. As shown
in Figure 8, NOx decreased with decreasing stoichiometry. NOx emissions were
reduced significantly down to a reburn zone stoichiometry of approximately 0.6-
0.9, depending on residence time. A stoichiometric ratio of 0.9 utilized 15%
natural gas for reburning in combustion products that originally had a
stoichiometric ratio of 1.1.
Figure 8 also shows the effect of the reburn zone residence time on NO
reduction. The results show that longer residence times are more beneficial in
reducing N0X. The difference between 1.0-1.4 seconds and 4.0-5.2 seconds was 40-
50 PPM more N0X reduction for the longer times. The shorter residence times are,
however, more practical for commercial applications. Residence times on the order
of 2.0-2.5 seconds in the reburn zone of a R1ley-Takuma system are possible. As
shown in Figure 9, at 2.5 seconds residence time, N0X reductions of 50% were
measured.
The furnace temperature in the reburn zone was not a critical factor in reducing
N0X emissions within the temperature range tested, 1950°F-2400°F. Any temperature
effects were dominated by other more significant parameters. In regards to other
emissions, though, CO was reduced from 30-40PPM without gas reburning to <20PPM
5B-5
-------
with 15% gas reburning.
The major findings drawn from this initial testing of the gas reburning concept,
on IGT's Pilot-Scale Furnace Simulator, are summarized below:
• Up to 70% N0X reduction was achieved with 15% gas reburning and
4.0-5.2 seconds residence time in the reburning zone.
*¦ Residence times of approximately 2-2.5 seconds in the reburning
zone appears to be sufficient for 50-60% N0X reduction.
• Reburn zone temperatures are less critical than residence time and
stoichiometry for N0X reduction.
• Reburn zone stoichiometry of 0.9 is effective in controlling N0X
and is attainable in commercial installations.
• There is potential for reducing CO emissions using gas reburning.
RILEY STOKER PILOT SCALE TESTING (TASK 2)
Following the initial pilot scale testing at IGT, process development tests were
conducted at Riley in a 3x10° Btu/hr Pilot MSW Combustion Facility. The pilot
scale step grate stoker design is an actual prototype of a full scale Riley-Takuma
system for mass burning. As conceptually shown 1n Figure 10, the combustor
section is 17'-0" tall x H'-9" long x 3'-0" wide and is designed to burn
processed MSW at a rate of approximately 450 Ib/hr or 5.5 tons/day.
Processed MSW, from an RDF plant in Biddeford, Maine, rather than raw MSW is
burned. The majority of the glass and metals content has already been removed and
the refuse has been reduced in size. This was done to avoid mechanical operating
problems which would result from raw MSW. Table II shows a comparison of the
processed MSW with the MSW from Olmsted County.
The pilot facility includes a 10 ton storage trailer, 45 ft long drag chain
conveyor, refuse feed chute, stoker grate and hydraulic drive system, ash
discharge hopper, overflre air nozzles, natural gas injection nozzles, undergrate
air, a flue gas recirculation system and a startup gas burner. The furnace walls
of the combustor are water cooled with high insulating refractory attached to the
inside surface to produce a thermal environment typical of field operating
conditions.
Figure 11 shows a photograph of the main combustor section. Combustion products
exit the main combustor and pass through heat recovery equipment, a gas analysis
sampling train, baghouse, acid gas scrubber system and finally exhaust out the
stack. The gas analysis train at the furnace exit is used to continuously monitor
flue gas composition Including N0X, 0?, CO, THC, CO2 and SO2. The facility is
currently being equipped to monitor HC1. In-furnace gas composition and
temperatures are measured as desired through various sampling ports located
throughout the combustor. A data acquisition and computer system is used for
direct on-line data analysis of various test conditions.
Similar to the testing performed at Olmsted County, baseline testing was first
conducted in the Pilot MSW Combustion Facility to ensure the results were
comparable to the field data. Normal operating conditions were simulated in
regards to excess air, % OFA flow, OFA and UGA flow distribution and the burning
profile of the processed MSW on the grates.
5B-6
-------
Figure 12 shows the computer output for a typical baseline performance test during
normal or standard operating conditions. Pertinent information regarding load,
flue gas composition, furnace zone stoichiometric ratios (SR), temperatures (T)
and residence times (t) are identified as well as the actual fuel and air flows.
N0X , CO and THC emissions for this one test measured 142, 27 and 0 PPM,
respectively. For all the baseline testing under normal operating conditions the
N0X , CO and THC emissions varied as follows:
NO,. 120 - 165 PPM
CO 10 - 50 PPM
THC 0-2 PPM
The furnace exit temperature indicated in Zone 5 (1570°F) was actually measured
with a high velocity temperature (HVT) probe while all the other furnace
temperatures were calculated based on the amount of heat removed in each furnace
section.
Comparing these results with the field data collected at Olmsted in Table III
showed good agreement in regard to flue gas emissions and furnace temperatures.
The only discrepancy discovered was with furnace residence time. The residence
times at Olmsted were calculated to be approximately twice as long as the pilot
unit because of the physical scale differences. However, based on the pilot
testing performed at IGT, longer residence times are more beneficial for NO
reduction. So, the test results obtained in the Riley pilot unit, when testing
the effects of gas reburnlng, should be conservative.
Additional baseline testing was performed in the Riley Pilot unit to evaluate the
impact of varying OFA injection location on N0X emissions. The normal or
standard configuration is to introduce approximately 35% of the total combustion
air through the lower front (LF), middle front (MF) and lower rear (LR) OFA
nozzles. Refer to Figure 12 for the location of these nozzles. NO emissions
during normal operation with 70% excess air averaged 120-165 PPM. However,
introducing the same amount of OFA through only one location (LF or LR) caused an
increase in N0X to 200-220 PPM. CO emissions remained unchanged. The higher
NO is attributed to greater turbulence and mixing in the lower furnace and a
subsequent increase in fuel N0X conversion.
Conversely, the lowest NOx can be produced by completely closing the OFA and
maintaining 50-60% total excess air. Tests showed that N0„ could be reduced to
109 PPM but CO emissions increased significantly to 150 PPft and THC emissions were
measured to be 6 PPM. Again, similar to the field operating unit, operation of
the OFA system in the pilot unit had a significant effect on controlling N0X, CO
and unburned THC emissions. Variations in excess air, unit load and undergrate
air distribution had no appreciable effect on combustion performance and
emissions.
Tests were then performed to evaluate the Impact of gas reburning on reducing
emissions. The primary variables focused on natural gas and FGR quantity, natural
gas, FGR and OFA Injection location and furnace stoichiometrics. As summarized in
Table IV, the initial results were very encouraging. This data was collected at
2.0-2.5 x 105 Btu/hr unit load with approximately 32-34% OFA. From a baseline N0X
level of 120-165 PPM, N0X could be reduced to approximately 70-80 PPM or nearly
50%. Figure 13 shows the effect of reburn zone stoichiometric ratio on N0X
emissions both with and without gas reburnlng. Typically, this ratio averaged 1.4
without and 0.9 with gas reburning. The scatter in the data is due to variations
in load, overall excess air, grate stoichlometry and gas injection location.
5B-7
-------
Gas reburning was indeed effective in controlling N0X . The low levels of CO and
THC emissions measured during the baseline testing would also remain low using gas
reburning. CO emissions were measured at <25 PPM while THC emissions would remain
at only a trace.
However, these favorable results were very dependent on the injection method and
type of operation being used. The best approach for maximum emission reduction
evaluated during this initial testing was to operate the reburning system as
outlined below, which corresponds to Test 53.
• 17% FGR through LF and LR nozzles
• 13% natural gas through LR nozzles only
• 34% OFA through upper front (UF) and upper rear (UR) nozzles
• 31% overall excess air
All other system operating conditions remained the same as during the baseline
testing. The resulting reburn zone stoichiometry for Test 53 was 0.87 while
reburn zone residence time was calculated to be .91 seconds.
Introducing natural gas through the lower rear nozzles, immediately above the
combustion grate, was also very effective during earlier tests (41-43) in reducing
N0X but CO emissions exceeded 80 PPM. The reason for this is because FGR was only
being introduced through these same nozzles, and the high degree of turbulent
mixing, resulting from introducing FGR through both the LF and LR nozzles, was not
as pronounced.
Test 52 demonstrated that low levels of N0X could still be produced when using
only 7% natural gas for reburning. However, as shown 1n Figure 14, this occurred
only when natural gas was introduced through the LR nozzles. Introducing gas
through the LF nozzles for reburning would create a situation where the degree of
N0X reduction was much more dependent on the amount of natural gas being burned.
Tests 47 and 48 evaluated the effect of introducing OFA through the middle front
(MF) and middle rear (MR) nozzles. This would produce shorter residence times for
N0X reduction reactions to occur 1n the reburning zone. Residence times for these
tests were calculated to be only 0.65 seconds as compared to >0.90 seconds when
introducing OFA through the UF and UR nozzles. During these tests, N0X emissions
were only reduced 28-38% depending on the amount of gas reburning being used.
This showed the importance of reburn zone residence time on N0X reduction.
Figure 15 graphically shows the effect of reburn zone residence time on N0X
emissions from this testing. As discussed later, additional testing, to study the
effect of increasing this residence time to levels approaching 1.5 seconds, will
be performed, since residence times of 2.0-2.5 seconds can be obtained in
commercial installations.
Further analysis of the data to evaluate the effects of grate stoichiometry and
overall excess air 1s currently being performed. Preliminary indications are that
these variables do not have as significant an impact on emissions reduction as
reburn zone stoichiometry, residence time and gas Injection strategy. The data
does show that low N0X levels can be achieved without having to operate the grates
at sub-stoichiometric combustion air levels.
The major findings based on initial testing of the gas reburning concept are as
follows:
5B-8
-------
• 50% NOx reduction was achieved with 7-15% natural gas reburning.
Introducing the gas through the lower rear nozzles was the most
effective.
• Effective burnout of CO and THC emissions can be achieved using gas
reburning.
• The ability to control emissions using gas reburning is very
dependent on injection location, residence time and reburn
stoichiometry.
• A reburn zone stoichiometry of approximately 0.9 is effective in
reducing N0X while residence times >.85 seconds are required for
sufficient reactions to occur, i.e. >40% N0X reduction.
• It appears excess air levels can be reduced from the normal 70% to
30% using gas reburning without adverse effects on emissions.
• No significant increase in furnace exit gas temperature was
measured with gas reburning operating.
• Operating with 7-15% natural gas improved furnace heat input
stability resulting in less variation in the measured O2
concentration at the furnace exit. The variation decreased from
±2% without to ±1% with gas reburning.
FUTURE WORK
Testing in Riley Stoker's Pilot MSW Combustion Facility will continue to further
optimize the gas reburning system. Test variables will focus primarily on
increasing reburn zone residence time and modifying lower furnace mixing in an
effort to obtain > 50% N0X reduction while maintaining low CO and THC emissions.
Modifications will be made to introduce 0FA as high in the furnace as possible in
order to maximize residence time. This will provide the capability of testing
residence times > 1.5 seconds. The best injection strategy determined during the
initial testing will continue to be studied along with variations of this approach
such as reducing the amount of FGR and changing FGR injection locations.
Following determination of the final gas reburning configuration, more detailed
in-furnace testing will be performed to better characterize the gas composition
and temperature history throughout the furnace both with and without gas
reburning. D1ox1n measurements will also be collected.
Assuming success with the final pilot scale testing, plans for a field
demonstration test of this technology in a commercial operating facility will be
finalized during Task 3. A detailed economic analysis of the gas reburning
technology will be conducted to determine the economic advantage of utilizing
natural gas for emissions control. Boiler performance analysis will also be
performed to quantify the Impact of a gas reburning system on overall boiler
performance. If these final analyses are encouraging, a field demonstration
program of the gas reburning technology will be implemented.
5B-9
-------
REFERENCES:
1) Itse, D.C. and Penterson, C.A., "Riley-Takuma Technology for Refuse Combustion
Plants," presented at the AFRC International Symposium on Incineration of
Hazardous, Municipal, and Other Wastes, Palm Springs, California, November,
1987.
2) Yamagishi, M., Yokohama, T., Shibuya, E. and Nippon Kokan K.K., "Air Pollution
Control for Waste-to-Energy Plants," presented at the AFRC International
Symposium on Incineration of Hazardous, Municipal, and Other Wastes, Palm
Springs, California, November, 1987.
3) Fleming, O.K., Khinkis, M.J., Abbasi, H.A., Linz, O.G. and Penterson C.A.,
"Emissions Reduction from MSW Combustion Systems Using Natural Gas," presented
at the Energy from Biomass Conference, New Orleans, Louisiana, February, 1988.
4) Duvall, D.S. and Ruby, W.A., "Laboratory Evaluation of High-Temperature
Destruction of Polychlorinated Biphenyls and Related Compounds", EPA/600/14.
5B-10
-------
Figure 2. Riley-Takuma Waste to Energy
Facility, Olmsted County, Minnesota
5B-11
-------
REFUSE CHARGING
HOPPER AND CHUTE
HYDRAULIC
OtL CYUNOER
PIOOLING hopper.
& CHUTE
AIR PLENUM
COMBUSTION GRATE
FURNACE WATER TU8E PANEL
DOUBLE OUMPER
RIOOLiNG
CONVEYOR
CAST IRON SLOCK OVERLAY
•REFRACTORIES OVERLAY
AUXILIARY
FUEL 3URNER
MAIN ASH CHUTE
Figure 3. Riley-Takuma MSW Stoker Grate
5B-12
-------
100
90
80
70 -
60 -
50 -
40 -
30 -
20 -
10 -
oL
\
OFA
OPEN
CLOSED
*
\
•
CO
o
•
N0X
~
¦
^cxo
_L
20 40 60 80
Excess Air, Percent
250
M
O
200*
150?
O
(J
£
100 a
o»
e
o
50 |
x
O
z
100
120
Figure 4. NOx and CO Emissions as a Function of Excess Air,
Olmsted Waste to Energy Facility - Economizer Exit
Combustion Lower Uppar llafractory Furnnea Dropout Supar Bollar Economlzar
Oral* OFA OFA Exit Exit haatar Dank Exit
E«" Exit Exit
Figure 5. Furnace Gas Composition History Through the
Olmsted Waste to Energy Facility During Normal Full
Load Operation
5B-13
-------
3000
2800
2600
2400
2200
2000
1800
1600
1400
1200
98 Percent Load
75 Percent Excess Air
¦ i Calculated Temperature -
r/
J
I
OFA
0
Temperature >l
X J}
Lower
Overfire
Air
tipper
Overfire
Air
Refractory
Exit
furnace
Exit
I ,
1.0
2.0 3.0 4.0
Residence Time, seconds
Dropout
Hopper
Entrance
I
5.0
Figure 5. Temperature and Products of
Incomplete Combustion Profiles through
Olmsted Waste to Energy Facility
200 E
100
6.0
70*-I?00*F
A IK
UUENCJEK DEMISTCR
Figure 7. Schematic of IGT's
Pilot-Scale Facility
5B-14
-------
300
25C
200
150
100 -
0.6 0.7 0.8 0.9 1.0 I.I 1.2
REBURN STOICHIOMETRIC RATIO
Figure 8. The Effect of Reburn Zone Stoichiometri
Ratio on NOx Emissions in IGT Pilot Unit
Figure 9. The Effect of Reburn Zone Residence
Time on NOx Emissions in IGT Pilot Unit
5B-15
-------
tn
CO
s
j
Air/Flue
Gas
Natural
Gas
o
V-
Air/Flue
Gas
it=n»
11-9*
Startup
Burner
H
Figure 10. Riley Pilot MSW
Combustion Facility
Figure 11. Riley Pilot MSW Combustion
Facility - Combustor Section
-------
(Ul«y Raaaareh Pilot Seal* Coabustion facility
Raducad Data ?«g« i
Data: 11/29/8* Tiaa: 11:00
Taac Coaaant: RE-ENTESED DATA • FSG7 CORR
Fual Coaaant: !tSU
Coaouator Coaaant: NONE
Taat Pt 7 .1 sift
Haat Input <02)t
H3W (02):
Gaa:
2.36 XBtu/Hr
2.36 NBtu/Hr
0 NBtu/Hr
ZONE 6
SR» 1.7
7- 1439 ?
t« .37 a
100 *
0 X
GAS
02 *
CG2 t
--//—> CO ?pa
— KC ppa
SO* soa
433 lb/Hr
0 lb/Hr
Xaaa. §7*02
#12*02
a.7
...
10.9
12.42
7.33
37
42
27
0
0
0
194
221
142
I
I
I
I
I
UFGFA—> I
I
I
SF0FA-->I
LFOFA J
\J
__/
/
ZONE 5 /
S3* 1.7\
7« 1370 ?\
.13 a (
I
ZONE 4 I
$?¦ 1.7 I
7« 1620 F I
t* .2S a I
ZONE 3
SR« 1.7
?• 1711 7
t« .26 a
ZONE 2
SR« 1.42
7« 2181 ?
;* .44 a
. J <—UROFA
. I<—MROFA
I . .
1 \
I i
I DGA IN
ZONE 1
SR* 1.06
T« 2323 i
t* .46 i
I
I
K--L30FA
\
I <—Burner
CCA
I86A I
I Straaa
33U
1 Natural1
Haat t
Air 1
FGR I
1
I Gaa I
Input 1
1
1
i
lb/Hr
1 lb/Hr I
KBtu/Hrf
lb/Hr 1
lb/Hr i
1 raad
433
1 1
2.36 1
1
( 3um«r
1 0
0 (
0 1
t DGA
1 —- 1
1
0 1
( CCA
1 —— 1
—— |
1793 1
1 3GA
1 1
1
26a i
I L7QFA
1 0 I
0 1
376 1
0 1
1 LftQFA
....
1 0 1
0 1
121 1
0 1
1 nfofa
1 —— 1
1
334 1
—— 1
I srgfa
1 1
—1
0 I
1 OFOFA
——
1 -— 1
1
0 1
1
1 toofa
1 1
1
0 (
—— 1
I Total I
433 \
0 I 2.36 t 3313 I
Figure 12. Typical Computer Output from
Riley Pilot Unit - Baseline Testing
5B-17
-------
160
140
120
100
80
60
O
o
o
o
o
o
%
8
o
o
A
&
_J_
O WITHOUT OAS n E ~ UR NIN Q
^ WITH GAS OEBUnNINO
0.6 0.0 1.0 1.2 1.4 1.6
REBURN STOICHIOMETRIC RATIO
1.8
Figure 13. The Effect of Reburn Zone Stoichiometric
Ratio on NOx Emissions in Riley Pilot Unit
1 10
2 100
CL
CM
O
*
CM
O
z
90
80
70
1 1 '—
1 1
INJECTION
1
LOCATION
GAS
OFA
~
LF
MF/MR
/ /
/ ^
~
A
LF
UF/UR
O
LR
UF/UR
A"
o
o
-
1
o
1
O
0
1
i
8 10 12 14
NATURAL GAS, X
16
Figure 14. The Effect of Natural Gas Quantity
on NOx Emissions in Riley Pilot Unit
5B-18
-------
1 10
3
£ ioo
CM*
O
5 90
(-
<
X 80
O
Z
70
0.6 0.7 0.8 0.9 1.0 1.1
RESIDENCE TIME. SEC.
Figure 15. The Effect of Reburn Zone Residence
Time on NOx Emissions in Riley Pilot Unit -
Lower Front Gas Injection
Table I
COMPARISON OF IGT PILOT DATA
WITH FIELD DATA - BASELINE TESTING
Test
Combustion Products
02. *
co2, %
n2, %
H20, %
NOx, PPM
Reburn Zone Temp., °F
Reburn Zone Residence Time, sec.
Baseline
Field Test
Unit
Typical
7.9
10.2
71.2
10.7
130
~2220(2)
1.0-3.0
IGT Furnace
Simulator
Typical
7.9
9.8
70.7
11.6
100-300
1950-2400
1.0-4.5
Wet basis
(2' Estimated from actual measurements of FEGT
5B-19
-------
TABLE II
TYPICAL FUEL ANALYSIS COMPARISON
Moisture, X
Volatile Hatter, X
Fixed Carbon, X
Ash, X
Sulfur, X
Carbon, X
Hydrogen, X
Nitrogen, X
Oxygen, X
HHV, Btu/lb (as rec'd.)
Paper, X
Plastic, X
Miscellaneous, X
Olmsted County
Raw HSW
3.35
33.86
16.22
46.57
0.75
32.00
3.44
0.95
12.94
6037
70.0
15.0
10.0
Riley Unit.
Processed HSW
18.19
57.90
11.70
12.21
0.21
34.49
4.48
0.35
30.07
5447
75.8
6.0
18.2
TABLE III
COMPARISON OF RILEY PILOT DATA
WITH FIELD DATA - BASELINE TESTING
Field Test Riley Pilot
Unit Unit
Test 21 7
Load, MBtu/hr 37.5 2.36
Excess Air, X 73 70
OFA Flow. X 34 38
OFA Configuration Std. Std.
Combustion Products
02, X 9.3 8.7
C02, X 10.1 10.9
CO, PPM 29 27
THC, PPH 0 0
N0X , PPM 134 142
Reburn Zone Temp., °F O 2220 2181
Residence Time to Furnace Exit, sec. ~3.8 -2.0
O Estimated from actual measurements of FEGT.
-------
Table IV
RILEY PILOT MSW COMBUSTION FACILITY
GAS REBURNING TEST RESULTS
Stolchiometry
Injectl
on Location
PPM @ 12%
02
% NOx
Test
Grate
Reburn
Total
% F6R
X NG
OFA
FGR
NG
NOx
CO
THC
Reduction
7(1)
1.06
1.42
1.70
0
0
MF/LF/LR
-
-
142
27
0
-
37
1.26
1.10
1.48
13
13
UF/UR
LF
LF
97
25
2
32
38
1.18
1.01
1.48
15
15
II
II
II
74
60
0
48
39
0.95
0.86
1.27
15
10
M
M
II
56
106
5
60
40
0.97
0.87
1.28
13
10
II
II
M
76
32
1
46
41
1.04
0.90
1.32
15
14
M
LR
LR
76
82
4
46
42
1.05
0.91
1.34
13
14
M
II
M
71
98
50
43
1.05
0.94
1.37
15
11
U
M
II
71
84
1
50
44
1.36
1.15
1.69
13
15
M
LF
LF
88
93
38
45
1.09
0.94
1.37
15
14
II
H
II
83
29
1
42
46
1.10
0.99
1.45
15
10
M
N
H
93
31
1
35
47
0.95
0.86
1.31
15
9
MF/MR
N
N
102
35
28
48
1.09
0.94
1.43
15
14
MF/MR
II
II
89
26
1
37
49
0.97
0.85
1.29
17
13
UF/UR
LF/LR
LF
82
73
4
42
50
1.31
1.07
1.61
18
19
N
M
LF/LR
84
33
1
41
51
1.03
0.90
1.35
17
13
II
U
LF/LR
83
19
1
42
52
1.02
0.95
1.44
17
7
M
II
LR
78
25
1
45
53
0.99
0.87
1.31
17
13
II
II
LR
73
26
1
49
(1) Typical Baseline Test
-------
(intentionally Blank)
5B-22
-------
APPLICATION OF LOW NOx PRECOMBUSTOR TECHNOLOGY TO
THE INCINERATION OF NITROGENATED WASTES
R.K. Srivastava, J.V. Ryan
Acurex Corporation
4915 Prospectus Dr.
Durham, NC 27713
W.P. Linak, R.E. Hall, J.A. McSorley
Combustion Research Branch, MD-65
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
J.A. Mulholland
Department of Chemical Engineering, 66-419
Massachusetts Institute of Technology
Cambridge, MA 02139
ABSTRACT
A 0.6 MW (2,000,000 Btu/hr) precombustion chamber burner/package boiler simulator
facility, designed for in-furnace nitrogen oxide (NOx) control and high combustion
efficiency, has been evaluated for high nitrogen content fuel combustion/waste
incineration application. The 250 to 750 ms residence time precombustion chamber
burner, using air staging and in-furnace natural gas reburning, yields up to four
stoichiometric zones.
During initial facility evaluation, natural gas, doped with ammonia (NH3) to yield a 5.8
percent fuel nitrogen content, and No. 2 distillate fuel oil, doped with pyridine (C5H5N) to
yield a 2.0 percent fuel nitrogen content, were used to simulate high nitrogen content
fuel/waste mixtures. Minimum nitric oxide (NO) emission levels of 160 and 110 ppm
(corrected to 0 percent oxygen, O2) were achieved for the natural gas and fuel oil tests,
respectively. These results correspond to approximately 85 percent reduction in NOx
emissions compared to uncontrolled emissions from a conventional burner mounted on a
0.7 MW (2,500,000 Btu/hr) commercial package boiler. Carbon monoxide (CO) emissions
for both low NOx tests were less than 35 ppm, indicating combustion efficiencies (CEs)
greater than 99 percent. Pyridine destruction efficiency (DE) was greater than 99.99 percent.
Subsequently, an investigation examining the incineration characteristics of a nitrogenated
pesticide, containing dinoseb (2-sec-butyl-4,6 dinitrophenol) in an organic solvent, was
5B-23
Preceding page Wank
-------
conducted on the low NOx facility. The dinoseb formulation contained 6.4 percent
nitrogen. NO emissions without any in-furnace NOx control strategy exceeded 4400 ppm
(0 percent O2). When NOx controls, in the form of air staging and natural gas reburning,
were employed, these emissions were reduced to below 150 ppm. CO emissions were
always less than 25 ppm. Dinoseb DE was greater than 99.99 percent and particulate
emissions were less than the RCRA limit of 180 mg/dsm3. Several common products of
incomplete combustion (PICs) were identified in the exhaust gases at concentrations
typically less than 10 ppb.
INTRODUCTION
Oxides of nitrogen (NOx) are formed during combustion by the oxidation of atmospheric
nitrogen (N2) and nitrogen contained in the fuel. Nitrogen dioxide (NO2) is a toxic gas that
the U.S. Environmental Protection Agency (EPA) has designated as a criteria pollutant
because of its adverse effects on human health.(l) NOx emitted from stationary
combustion sources contribute to the degradation of air quality as well as to acid deposition
and forest damage. (2) Increasing levels of atmospheric nitrous oxide (N2O) have also been
measured. Although the sources of this increase have not been determined, these
increasing levels have been predicted to contribute to both a decline in stratospheric ozone
concentrations and an increase in global climate warming. (3)
Unlike sulfur oxide (SOx) emissions, NOx emissions have been increasing in recent
years.(4) While coal- and oil-fired utility and industrial boilers account for over half of
these NOx emissions, only 15 percent of these stationary sources are regulated by EPA's
New Source Performance Standards (NSPS).(5) The remaining sources must be addressed
with retrofit technologies if significant NOx emission reduction is to be realized. Another
NOx control problem is posed by the incineration of high nitrogen content wastes in
hazardous waste incinerators and industrial boilers. While incineration of these materials
may not constitute a significant increase in the overall national NOx emission levels,
individual plant emissions may be sufficient to result in a local NOx problem thus
preventing on-site thermal destruction from being permitted. With thermal destruction
becoming an increasingly attractive alternative to landfill storage of wastes, there
continues to be a need for developing high efficiency, low NO* combustion technologies.
Additionally, incinerator operating conditions which maximize waste destruction are
typically those that promote the formation of NOx.
NOx species are formed in practical combustion systems by high temperature thermal
fixation of atmospheric nitrogen (N2, source of thermal NOx)(6) and by oxidation of
nitrogen chemically bound in the fuel (source of fuel NOx).(7) Thermal NOx can be
reduced by decreasing peak flame temperatures. Fuel NOx, not as strongly dependent on
temperature, is very sensitive to reactant stoichiometry. Fuel-rich conditions promote N2
formation over nitric oxide (NO) formation. Thus, many of the NOx control technologies
currently in use involve modifications of the combustion process to reduce peak flame
temperatures and create fuel-rich conditions by reducing fuel and oxidizer mixing rates.
These modifications, however, frequently result in reduced combustion efficiency (CE) and
increased vessel wall corrosion and ash deposition. Furthermore, practical constraints,
such as burner and boiler sizes, limit the effectiveness of NOx control by combustion
modification.
5B-24
-------
In-furnace NOx control technologies currently in use, including reduced air preheat, load
reduction, low excess air, flue gas recirculation, overfire air, deep air staging, fuel staging
(or reburning), and various low-NOx burner systems, can reduce NOx emissions by 20 to 80
percent. While these technologies are sufficient to meet current NSPS requirements, new
technologies will be required if more stringent NOx emissions regulations are enacted.
EPA has been involved in the development of two evolving NOx control technologies: the
precombustion chamber burner and reburning. These two combustion modification
strategies provide alternatives to expensive post-combustion NOx control technologies,
such as selective catalytic reduction which is being utilized extensively in Japan and West
Germany. The goal of this experimental program was to utilize the precombustion
chamber burner and reburning concepts in developing a low-NOx, high efficiency
combustion system that is practical for both new and retrofit applications.
BACKGROUND
Precombustion Chamber Burner
The precombustion chamber burner is a staged combustion technology capable of
achieving NOx emissions of less than 0.1 lb (as NC>2)/106 Btu (or approximately 90 ppm
NOx), even with high nitrogen content fuels.(8) The burner, described by Englund et al,(8)
consists of a primary air and fuel injection system, a large refractory lined precombustion
chamber, and a secondary staged air injection section. Fuel and primary air are injected so
as to promote rapid mixing in the nearly adiabatic precombustion chamber with a first
stage stoichiometry of between 0.6 and 0.8. Residence times between 0.6 and 1.0 sec allow
for maximum reduction and conversion of fuel nitrogen species to N2 in the fuel-rich
precombustion chamber. First-stage combustion gases exit the burner through a
convergent transition section which minimizes both radiative heat loss to the boiler and
back-mixing of the secondary air. The transition section between the burner and boiler is
water-cooled to reduce combustion gas temperatures before further air addition and thus
minimize the formation of thermal NOx.
The precombustion chamber burner has been tested on a full-scale (16 MWt) crude-oil-
fired steam generator used for thermally enhanced oil recovery (TEOR).(9) A 30-day
continuous monitoring test using heavy residual oil demonstrated the burner's ability to
maintain a nominal NOx emission of 70 ppm and high combustion efficiency. During
burner optimization testing, NOX/ carbon monoxide (CO), and smoke emissions were
measured over a range of first-stage stoichiometrics. These data indicate the sensitivity of
NOx emission to first-stage stoichiometry and the good hydrocarbon burnout
characteristics of the burner.
A test designed to demonstrate the potential for using this burner for nitrogenated waste
incineration was performed at EPA's Air and Energy Engineering Research Laboratory
(AEERL). A pilot- scale (0.6 MWt) precombustion chamber burner was used to incinerate a
surrogate nitrogenated waste mixture of 9.1 percent (by volume) pyridine (C5H5N) in No. 2
distillate fuel oil. NOx emissions of less than 100 ppm were maintained, with greater than
six nines (99.9999%) destruction of pyridine being achieved. The relative NOx emissions,
with and without pyridine addition, indicate that less than 1 percent of the fuel nitrogen
5B-25
-------
was converted to NO. This preliminary result sparked interest in utilizing the
precombustion chamber burner for low NOx, high efficiency incineration.
While the precombustion chamber burner has been demonstrated successfully in pilot and
full-scale tests, its large size and heavy refractory composition make it impractical for most
boiler retrofit applications. Reducing the burner size results in insufficient first-stage
residence time to fully convert fuel nitrogen to N2. However, further NOx control can be
achieved by further fuel and air staging (reburning) in the boiler. Current research is
investigating the use of light weight refractories to provide improved response to load
changes.
Reburning
Reburning is an in-furnace NOx control technology in which portions of the fuel and
combustion air streams are diverted downstream of the primary flame. In this way, the
combustor has three distinct zones: a fuel-lean primary zone, a fuel-rich reburning zone,
and a fuel-lean burnout zone. NOx reduction via reburning occurs both by combustion of
a portion of the fuel under fuel-rich stoichiometrics (conditions not promoting NOx
formation), and chemical destruction mechanisms involving secondary flame radical
reactions with NO, resulting in N2 formation. An EPA study (10) has shown that 50
percent NOx reduction is possible with 15 to 20 percent of the fuel used for reburning. At
low primary NOx levels, however, a nitrogen-free reburning fuel (such as natural gas)
must be used to achieve 50 percent reduction. Reburning can be used in conjunction with
other NOx control schemes to yield lower NOx emissions. However, application of
reburning often necessitates hardware modifications to the combustor.
LOW NOx EXPERIMENTAL FACILITY
The pilot-scale experimental facility, shown schematically in Figure 1, consists of a vertical
precombustion chamber low-NOx burner and a horizontal package boiler simulator. The
precombustion chamber consists of a primary fuel and air injection module, several 0.25 m
long spool modules with 0.51 m inside diameter (I.D.), and a convergent elbow module.
The modular design of the precombustor permits variation in nominal first-stage
residence time between 250 and 750 ms. The modules have a thick refractory wall lining to
minimize heat loss and, thus, maintain high temperatures that promote conversion of
fuel nitrogen to N2 under fuel-rich stoichiometries. To achieve rapid mixing in the
precombustion chamber, the primary fuel is injected through a divergent nozzle, and the
primary air, which is not preheated, is passed through fixed swirl vanes. The convergent
elbow module minimizes back-mixing of combustion gas and radiation losses to the boiler.
A water-cooled transition module, 0.25 m I.D., cools the combustion gas before secondary
air addition to minimize thermal NOx generation.
The pilot-scale boiler simulator is rated at 0.9 MW thermal input. The boiler's radiant
section is horizontal, 0.6 m I.D., 3.0 m long, and cooled with Dowtherm G heat transfer
fluid. Combustion gas exits the boiler through a vertical stack. The burner/boiler
transition section allows for radial addition of staged air. Additionally, the boiler's front
face has two ports for addition of staged (reburn) fuel at an angle of 45 degrees off axis and
six axial ports for the addition of burnout air. This design allows for reburning application
5B-26
-------
from the boiler front face, with aerodynamic separation of the fuel-lean, fuel-rich, and
fuel-lean zones in the boiler.
The experimental facility is designed for independent control and measurement of each
fuel, fuel dopant (if any), and air stream. Stack gas species are measured by continuous
emissions monitors (CEMs). These monitors include measurement of NO, NOx, 02, CO,
CO2, and total unburned hydrocarbons (THC). All data are reported corrected to 0 percent
O2 dry basis.
BURNER PERFORMANCE EVALUATION
Natural gas, doped with ammonia (NH3) to yield a 5.8 percent fuel nitrogen content, and
No. 2 distillate fuel oil, doped with pyridine (C5H5N) to yield a 2 percent fuel nitrogen
content, were used to simulate high nitrogen content fuel/waste mixtures. In these tests,
the shortest burner configuration, corresponding to a first-stage nominal residence time of
250 ms, was evaluated. A 0.7 MW North American fire-tube package boiler was used to
provide conventional burner results for comparison with the multistaged burner results.
This boiler is a three-pass unit with a continuous service rating of 0.3 kg of steam per
second (2400 lb/hr). Its size and thermal characteristics are similar to those of the package
boiler simulator.(ll)
Burner baseline performance (air staging) and burner operation with reburning (fuel
staging) were evaluated. The parameters affecting the NOx emissions from the facility
without reburning are fuel nitrogen content, combustion gas residence time in the
prechamber, first-stage stoichiometry, and exhaust stoichiometry. The residence time of
combustion gas in the burner depends on precombustion chamber length, load, and
stoichiometry. The nominal load was 0.6 MW. The exhaust stoichiometry was kept at a
nominal value of 15 percent excess air. First-stage stoichiometry was optimized for each
parametric test sequence.
Burner Stoichiometry Variation
First-stage stoichiometry was varied by changing the primary air flow. Secondary radial air
was adjusted to maintain 15 percent excess air. The results are plotted in Figure 2. The
curves indicate a strong sensitivity of stack NO to changes in burner stoichiometry. For
the NH3/natural gas tests, a minimum NO emission of 315 ppm occurred at a burner
stoichiometry of about 0.78. For the CsHsN/fuel oil tests, a minimum of 190 ppm occurred
at a burner stoichiometry of 0.65. By comparison, similar tests using the North American
boiler produced NO emissions of 1000 ppm when firing the 5.8 percent nitrogen
NH3/natural gas fuel at 15 percent excess air. NO emissions of 765 ppm resulted when
firing the 2 percent nitrogen CsHsN/fuel oil mixture at 15 percent excess air. Thus, the low
NOx burner reduced NO emissions by 68 percent for the NH3/natural gas fuel and by 75
percent for the CsHsN/fuel oil mixture.
5B-27
-------
Excess Air Variation
The effect of excess air variation on stack NO is shown in Figure 3. As expected, excess air
had a much stronger effect in the conventional North American boiler tests than in the
low NOx burner tests. While reducing excess air did provide some NO reduction with the
low NOx burner, maintaining an exhaust excess air level of 15 percent is important for
achieving high combustion efficiency.
Load Variation
The effect of load reduction on stack NO is shown in Figure 4. As load was decreased from
a nominal condition of 0.6 MW to 65 percent of nominal, the stack NO concentration
dropped by 20 to 25 percent in the low NOx burner tests. With a conventional burner, NO
emissions are reduced only slightly with burner derating, due to decreased air/fuel mixing
intensity. However, with the low NOx burner, the effect is much greater because load
reduction corresponds to an increase in first-stage residence time and, thus, a decrease in
NO emission. Reducing load by 35 percent increased burner residence time from 250 to 385
ms.
Fuel Nitrogen Variation
In addition to the base case tests, fuel nitrogen content was varied between 0 and 5.8
percent and 0 and 5.0 percent for the NH3/natural gas and CsHsN/fuel oil tests,
respectively. The effect of fuel nitrogen variation is shown in Figure 5. As expected
exhaust NO levels increase with increasing fuel nitrogen content. Results from the
conventional North American boiler were much more sensitive to this effect. These
results demonstrate the precombustion chamber burner's ability to reduce fuel nitrogen to
N2 even at reduced size (250 ms). The full size (600 to 800 ms) precombustion chamber
burner produces NO emissions even less sensitive to fuel nitrogen content.(8)
Reburning Tests
For these tests, total boiler load (0.6 MW, 2,000,000 Btu/hr) was held constant while 35
percent of the fuel was diverted from the primary fuel injector to two secondary injectors.
The reburn fuel (undoped natural gas) was added through two ports in the boiler front face
at an angle of 45 degrees off axis, using water-cooled injectors (see Figure 1). Staged air to
complete primary fuel combustion was added radially into the transition section. Burnout
air for the reburn fuel was added axially near the boiler wall through six axial ports on the
boiler front face. Thus, a four-stage combustion process was established, consisting of a
fuel-rich burner zone and three boiler zones characteristic of reburning (i.e, fuel-lean, fuel-
rich, fuel-lean). The stoichiometry in the third stoichiometric zone (SR3, the fuel-rich
reburning zone in the boiler) is critical in this NOx control process.
The CsHsN/fuel oil and NH3/natural gas results are given in Figure 6. NO emissions
decrease due to both primary NO destruction and primary combustion gas dilution. Two
second-stage stoichiometrics (SR2) were established (1.1 and 1.0), with 35 percent fuel
staging (reburning). The exhaust NO levels were reduced to 120 ppm (from 190 ppm) at a
5B-28
-------
SR2 of 1.1 (and a SR3 of 0.76) and to 110 ppm at a SR2 of 1.0 (and a SR3 of 0.69) for the
CsHsN/fuel oil tests. Similarly, NO emissions were reduced to 195 ppm (from 315 ppm) at
a SR2 of 1.1 and 160 ppm at a SR2 of 1.0 for the NH3/natural gas tests. Note that dilution
accounts for most of the decreases measured. Due to less distinct stoichiometric zones
than typically established in reburning applications, the NO reduction by fuel staging was
not quite as great as that obtainable when the staged fuel is injected farther downstream of
the fuel-lean second-stage zone.(11) However, the configuration used in these tests
requires no additional boiler penetrations other than modifications to the boiler front face.
In addition, complete destruction of the primary fuel/waste stream appears to be ensured
by providing all of the required primary combustion air prior to entry into the boiler.
Burner performance evaluation results indicate that relatively low NOx emission levels
are possible with a shortened precombustion chamber burner provided the first-stage
stoichiometry is optimized and a post-flame NOx control technology like reburning is
used. CO emissions, for all the tests performed during burner performance evaluation,
were less than 35 ppm, indicating that combustion efficiencies attained were greater than
99 percent.
NITROGENATED WASTE INCINERATION
Subsequent to the burner evaluation tests, an investigation to examine the incineration
characteristics of a nitrogenated pesticide, containing dinoseb (2-sec-butyl-4,6
dinitrophenol), in an organic solvent, was conducted on the low NOx burner facility. The
burner was configured with all its modular sections in place so as to maximize the first
stage residence time (750 ms). This dinoseb pesticide product formulation contained 6.4
percent nitrogen. An ultimate analysis and analysis from the material data safety sheet are
shown in Table 1. Incineration conditions with no NOx control, with air staging, and with
air staging and reburning were evaluated using the low NOx burner/package boiler
simulator. The North American boiler was not used for these tests. Natural gas was used
as the reburn fuel and was added in addition to a fixed primary load of 0.6 MW (2,000,000
Btu/hr). This increase in total load (approximately 32%) differs from the procedure used
during the burner evaluation tests where total load was held constant.
A Source Assessment and Sampling System (SASS) train in conjunction with GC/MS
analysis was used to characterize emissions of dinoseb and semi-volatile organic products
of incomplete combustion (PICs). A Volatile Organic Sampling Train (VOST) in
conjunction with GC/MS analysis was used to characterize emissions of volatile PICs.
Dinoseb recovery was characterized by spiking 100 g XAD-2 resin with standard dinoseb
solutions in dichloromethane. All XAD-2 and particulate filter samples were Soxlet
extracted in dichloromethane and methanol, concentrated by Kederna-Danish
evaporation, and analyzed by GC/MS. Prior to the incineration tests, scoping tests were
carried out to determine the maximum dinoseb product concentration in No. 2 distillate
fuel oil for stable operation. During these tests, successively increasing concentrations of
dinoseb product in fuel oil were incinerated until combustion conditions deteriorated.
However, the scoping tests revealed that 100 percent dinoseb product could be stably
combusted. Thereafter, the remaining incineration tests were carried out with 100 percent
dinoseb product. The experimental matrix describing the tests conducted is shown in
Table 2.
5B-29
-------
Gaseous and Particulate Emissions
Table 3 presents the stoichiometric and reburning parameters and the gaseous and
particulate emissions for the dinoseb product incineration tests. All concentrations are
presented dry and corrected to 0 percent O2. Concentrations corrected to 7 percent O2
(RCRA convention) are presented in parentheses. Test 7 indicates that NO emissions
without any combustion modification for NOx control exceeded 4400 ppm. When NOx
controls in the form of air staging with and without natural gas reburning were applied,
these NO emissions were reduced to below 150 ppm. Thus the use of the precombustor
concept reduced the NO emissions by approximately 96 percent. It is interesting to note
that reburning did not offer any further NOx reduction over incineration conditions with
air staging only. This is likely because the configuration of the experimental facility
provided ample first stage residence time for the conversion of fuel nitrogen to N2 and
further allowed secondary air addition at relatively cool temperatures, thus inhibiting the
formation of thermal NO. These effects probably provided a low concentration NO pool
before the addition of reburning fuel and so inhibited reburning efficiency.(lO) It is
interesting to note that the NOx measurements are very similar to those for NO. In fact,
the differences are well within instrument error. This addresses our concern that, because
of the nitrogenated nature of dinoseb, NO2 may be preferentially emitted. We see,
however, that, consistent with other nitrogen containing fuels tested, NO2 is typically less
than 5 percent of the total NOx emitted. It is also important to note that corresponding NO
emissions from uncontrolled dinoseb product combustion in industrial boilers would
likely be somewhat less than that measured in test 7. The high temperature, adiabatic
design of the precombustor promotes thermal NOx formation when operated fuel-lean.
This type of operation, however, is comparable to that used in most hazardous waste
incinerators (high temperatures, adiabatic, high excess air, high residence times) and so is
of practical interest.
The particulate emissions for all tests were below the RCRA limit of 180 mg/dsm3, and
were determined to consist primarily of calcium sulfate and calcium oxide (particulate
analyzed by x-ray diffraction). The particulate emission level for dinoseb product
incineration with reburning was less than that for dinoseb product incineration without
reburning. This may be partially due to increased secondary flame burnout, but is also due
to simple dilution by the addition of natural gas reburn fuel and oxidizer. The CO and
unburned hydrocarbon (THC) emissions were less than 15 ppm and 5 ppm, respectively,
indicating that the combustion efficiencies were in excess of 99 percent.
Dinoseb Destruction Efficiencies and Products of Incomplete Combustion
No dinoseb was detected in any of the concentrated dichloromethane extract combustion
emission samples. These combined XAD-2 and particulate filter extracts were analyzed by
GC/MS. These results and the results of dinoseb recovery tests are shown in Table 4.
These data indicate that, although the extraction method used resulted in low dinoseb
recovery (9.39%), destruction efficiencies (DEs) calculated using measured instrument
detection limits could be determined to be greater than the 99.99 percent RCRA limit. Low
dinoseb recoveries, however, are not unique to these tests. Efforts by Acurex Corporation
on samples from another dinoseb campaign [sponsored by the EPA Risk Reduction
5B-30
-------
Engineering Laboratory (RREL)] and Battelle Columbus Division performing nitrogenated
PIC analyses [for the EPA Health Effects Research Laboratory (HERL)] found similar results.
Qualitative PIC emissions, measured by SASS and VOST sampling in conjunction with
GC/MS analysis, are shown in Table 5. Results indicate that several common combustion
PICs are present and emitted in concentrations typically less than 10 ppb. It is important to
note that these analyses include only qualitative identification of volatile and semi-
volatile compounds. These categories approximately correspond to compounds with
boiling points below 100 °C and those between 100 to 300 °C, respectively. No information
is currently available regarding the potential presence of higher molecular weight
compounds including polycyclic aromatic hydrocarbons (PAHs) and nitrogenated PAHs.
CONCLUSIONS
Results from the burner evaluation experimental program indicate that a precombustion
chamber burner, shortened in length and coupled with reburning, is most effective in
minimizing NOx emissions and is capable of providing an overall NO reduction of
approximately 85 percent. These observations were confirmed in the dinoseb tests, where
NOx reductions of approximately 96 percent were obtained.
The reburning hardware modifications used in this study were designed with retrofit
feasibility and efficient combustion in mind. In the reburning tests performed during the
evaluation phase, staged fuel and air from the front face of the boiler provided a secondary
flame for increased primary fuel/waste destruction.
Finally, the results of this program indicate that the precombustion chamber burner
technology is capable of incinerating highly nitrogenated wastes with DEs in excess of 99.99
percent, maintaining low NOx and PIC emissions, and producing combustion efficiencies
in excess of 99 percent. This technology may, therefore, be an attractive alternative for the
disposal/destruction of highly nitrogenated wastes.
ACKNOWLEDGMENTS
Portions of this work were conducted under EPA contracts 68-02-3988 and 68-02-4701 with
Acurex Corporation. The authors would like to thank C. Pendergraph (Acurex Corp.) and
R.A. Grote (EPA/AEERL) for their help operating the combustion and sampling
equipment, and to J. Lewtas, D.M. DeMarini, R.R. Watts, and L.D. Claxton (EPA/HERL) for
their help with portions of the sample preparation and organic analyses. Additionally, the
authors would like to thank D.A. Oberacker (EPA/RREL) for his helpful advice and the
U.S. EPA Office of Pesticides and Toxic Substances (OPTS) for partial project funding.
DISCLAIMER
The research described in this paper has been reviewed by the Air and Energy Engineering
Research Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect the views and
5B-31
-------
policies of the Agency nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
REFERENCES
1. Clean Air Act, U.S. Environmental Protection agency, Amended in 1977.
2. Keinhorst, H., "Legal Instruments and the State of Technology for Reducing NOx
Emissions in the Federal Republic of Germany," In: Proceedings: 1985 Symposium on
Stationary Combustion NOx Control, Volume 1: Utility Boiler Applications, EPA-
600/9-86-021a (NTIS PB86-225042), July 1986, p. 4-1.
3. Kramlich, J.C., et al., "EPA/NOAA/NASA/USDA N20 Workshop, Volume I," EPA-
600/8-88-079 (NTIS PB88-214911), May 1988.
4. Gerber, C.R., Opening Remarks, In: Proceedings: 1985 Symposium on Stationary
Combustion NOx Control, Volume 1, EPA-600/9-86-021a (NTIS PB86-225042), July
1986, p. 2-1.
5. "Standards of Performance for New Stationary Sources - Volume 1: Summary and
Standards," EPA-340/l-82-005a, June 1982.
6. Zeldovich, Y.B.: Acta Physicochimica U.S.S.R. 21, 577 (1946).
7. Pershing, D.W. and Wendt, J.O.L., Sixteenth Symposium (International) on
Combustion, p. 389, The Combustion Institute, 1977.
8. England, G.C., et al., "Evaluation and Demonstration of Low- NOx Burner Systems for
TEOR Steam Generators - Design Phase Report," EPA-600/7-84-076 (NTIS PB84-
224393), July 1984.
9. England, G.C., et al., "Evaluation and Demonstration of Low- NOx Burner Systems for
TEOR Steam Generators: Final Report - Field Evaluation of Commercial Prototype
Burner," EPA-600/7-85-013 (NTIS PB85-185874), March 1985.
10. Mulholland, J.A., et al., "Reburning Application To Firetube Package Boilers," EPA-
600/7-87-011 (NTIS PB87-177515), March 1987.
11. Mulholland, J.A., and Hall, R.E., "Fuel Oil Reburning Application for NOx Control to
Firetube Package Boilers," Journal of Engineering for Gas Turbines and Power, Vol.
109, ASME, April 1987, pp. 207-214.
5B-32
-------
PRIMARY FUEL
AXIAL REBURN A
a*************************************************
PACKAGE BOILER SIMULATOR
Figure 1. EPA pilot-scale combustion research facility consisting of a refractory-lined
precombustion chamber and package boiler simulator, with modified air and fuel staging
ports.
5B-33
-------
0.4 0.5 0.6 0.7 0.8 0.9 1
First-Stage Stoichiometric Ratio (SRI)
A NH3/natural gas
O C5H5N/fuel oil
Figure 2. Effect of burner stoichiometry on exhaust NO concentration from tests burning a
5.8 percent fuel nitrogen NH3/natural gas and a 2 percent fuel nitrogen C5H5N/N0. 2
distillate fuel oil mixtures.
5B-34
-------
E
a
a.
1400 -
1200
1000 -
cn 800
o
K
O
-o
600 -
400 -
200 -
A NH3/notural gas
O C5H5N/fuel oil
NH3/natural gas N.A.
C5H5N/fuel oil N.A.
10 20 30
Excess Air, percent
40
Figure 3. Effect of excess air on exhaust NO concentration from tests burning a 5.8 percent
fuel nitrogen NH3/natural gas and a 2 percent fuel nitrogen C5H5N/N0. 2 distillate fuel oil
mixtures. Curves with open symbols represent data from the low NOx facility operated at
optimum primary stoichiometrics as shown in Figure 2. Curves with filled symbols
represent corresponding data from the conventional North American (N.A.) boiler.
5B-35
-------
400
300 -
E
a
a
CM
o
K
o
£
¦o
200 -
100 -
A NH3/natural gas
O C5H5N/fuel oil
70 80
Load, percent
100
Figure 4. Effect of load variations on exhaust NO concentration from tests burning a 5.8
percent fuel nitrogen NH3/natural gas and a 2 percent fuel nitrogen C5H5N/N0. 2 distillate
fuel oil mixtures (primary stoichiometrics optimized).
5B-36
-------
1000
900
800
E
700
Ui
a.
M
600
O
K
O
500
£
u
400
O
z
300
200
100
0
A NH3/natural gas
O CSHSN/fuel oil
A NH3/natural gas N.A.
C5H5N/fuel oil N.A.
2 3 4 5
Fuel Nitrogen, percent
Figure 5. Effect of fuel nitrogen on exhaust NO concentration from tests burning
NH3/natural gas and C5H5N/N0. 2 distillate fuel oil mixtures. Curves with open symbols
represent data from the low NOx facility (optimal primary stoichiometrics). Curves with
filled symbols represent corresponding data from the conventional North American
(N.A.) boiler.
5B-37
-------
400
A NH3/not gas SR2-1.0
~ NH3/nat gas SR2-1.1
O C5H5N/oil SR2-1.0
* C5H5N/oil SR2-1.1
10 20 30
Fuel Staging, percent
Figure 6. Effect of boiler fuel staging (reburning) on exhaust NO concentration from tests
burning a 5.8 percent fuel nitrogen NH3/natural gas and a 2 percent fuel nitrogen
C5H5N/N0. 2 distillate fuel oil mixtures. Natural gas data at two second stage
stoichiometrics (primary stoichiometry, SRI = 0.78), and corresponding oil data at two
second stage stoichiometries (primary stoichiometry, SRI = 0.65) are presented.
5B-38
-------
TABLE 1
ANALYSIS OF DINOSEB PRODUCT USED FOR INCINERATION TESTS
Dinoseb in organic solvent (dinoseb product)
Analysis from Material
Safety Sheet Ultimate Analysis
Dinoseb 54.4 % C 65.01 %
Diesel #2 4.04 % H 7.27 %
Xylene Range 32.5% N 6.63%
Aromatic
O* 20.45 %
Inert 9.06 %
S 0.35 %
Ash 0.29 %
Moisture 0.63 %
HHV 13,076 Btu/lb
"Oxygen by difference.
5B-39
-------
TABLE 2
EXPERIMENTAL TEST PLAN
Burner performance evaluation tests
Test Condition NCX Control Analyses
A NH3/natural gas Vary first stage CEM
5.8% fuel nitrogen stoichiometry,
content excess air, load,
fuel nitrogen content, reburning
B C5H5N/N0. 2 distillate Same as above. SASS, CEM
fuel oil 2.0% fuel nitrogen content
Dinoseb recovery tests*
Dinoseb Concentrate Analysis
Spike Volume Concentration
Test
(ue>
(ml)
(ng/ul)
1A
100
10
10
IB
100
1
100
1C
1000
1
1000
Dinoseb product incineration tests
Test Condition NOx Control Analyses
2 Combustion blank Air staging CEM, VOST, SASS
#2 fuel oil
3 Combustion blank Air staging and CEM, VOST, SASS
#2 fuel oil rebuming
4 Dinoseb product Air staging CEM
scoping test (determine
max dinoseb product cone
in No. 2 fuel oil for stable operation)
5 100% dinoseb product Air staging CEM, VOST, SASS
6 100% dinoseb product Air staging and CEM, VOST, SASS
rebuming
7 100% dinoseb product None CEM
"For dinoseb recovery tests, standard dinoseb solutions (in dichloromethane) were spiked on
XAD-2 resin, extracted, and concentrated.
5B-40
-------
TABLE 3
DINOSEB PRODUCT CEM AND PARTICULATE RESULTS*
Test Condition NOv Control SRI** SR2 SR3 SR4
R/P %*
Combustion Air staging
blank
Combustion Air staging
blank and reburning
Dinoseb
product
Dinoseb
product
Dinoseb
product
Air staging
Air staging
and reburning
None
0.72 1.35 1.35 1.35 N/A
0.62 1.14 0.87 1.20 32.5
0.64 1.23 1.23 1.23 N/A
0.61 1.08 0.82 1.17 31.4
1.13 1.13 1.13 1.13 N/A
Test NO NOx CO CO2 THC Particulate Tempi Temp4
(ppm)
(ppm)
(ppm)
(%)# (ppm)
(m^/dsm?)
(°a##
(°Q
2
78
9
4.13
(52)
(6)
(2.75)
3
89
-
13
-
5.02
-
-
(59)
(9)
(3.35)
5
141
128
6
2
151.34
1343
633
(94)
(85)
(4)
(1)
(100.89)
6
138
132
14
1
81.76
1151
652
(92)
(88)
(9)
(0.7)
(54.51)
7
4401
4497
13
1
-
708
606
(2934)
(2998)
(9)
(0.7)
*Stack emissions corrected to 0% oxygen. Emissions corrected to 7% oxygen are given in parentheses.
**SR1, SR2, SR3, SR4 indicate the stoichiometric ratios calculated at the combustor locations shown
in Figure 1.
***R/P indicates gas reburning load/primary load (natural gas added in addition to primary burner
load of 2,000,000 Btu/hr). (CEM concentrations have not been corrected for this dilution.)
#C02 concentrations are suspect and are not included due to instrument malfunction.
##Templ, Temp4 indicate approximate gas temperatures (bare thermocouples, near the wall)
measured at the combustor positions shown in Figure 1.
5B-41
-------
TABLE 4
DINOSEB PRODUCT POHC ANALYSES
Dinoseb recovery tests, instrument sensitivity 10 ng/ul
Test
Analysis
Concentration
("g/uD
Measured
Concentration
("g/uD
Percent
Recovery
1A
IB
1C
10
100
1000
Incineration tests
Test Condition
ND*
ND
93.9
NOv Control
N/A**
N/A
9.39
Dinoseb
Consumed
Sample Vol
as fraction
of Stack Vol
Combustion
blank
Combustion
blank
Dinoseb
product
Dinoseb
product
Air staging
Air staging
and rebuming
Air staging
Air staging
and rebuming
0
0.00705
0.00276
68130.7 0.00238
76023.4 0.00275
Test
Potential Max.
Dinoseb Cone,
in Sample
(g, ng/ul)
Dinoseb Cone.
Measured
in Sample
(ng)
DRE Assuming
10 ng/ul
Instrument
Sensitivity,
100% Recovery
DRE Assuming
10 ng/ul
Instrument
Sensitivity,
9.39% Recovery
2
0
ND
N/A
N/A
3
0
ND
N/A
N/A
5
162.15,4,054,000
ND
> 99.9998
> 99.997
6
209.31,10,466,000
ND
> 99.9999
> 99.999
•Not detected.
**Not applicable.
5B-42
-------
TABLE 5
DINOSEB PRODUCT PIC ANALYSES
VOST samples (GC/MS)
Test
Condition
PICs Identified
mg/snn^
ppb
2A
Comb, blank
w/ air staging
Benzene
2
<0.001
< 1
3A
Comb, blank
Benzene
37
0.002
< 1
w/ air staging
so2
514
0.028
10.7
and rebuming
3B
Benzene
18
0.001
< 1
so2
414
0.022
8.4
5A
Dinoseb product
Benzene
12
0.001
< 1
w/ air staging
Toluene
4
<0.001
< 1
Ethylbenzene
4
< 0.001
< 1
Dodecane
22
0.001
< 1
Naphthalene
1725
0.091
17.4
5B
Benzene
2
< 0.001
< 1
S02
125
0.007
2.7
6
Dinoseb product
Benzene
9
0.001
< 1
w/ air staging
Toluene
10
0.001
< 1
and rebuming
SASS samples (GC/MS)
Test
Condition
PICs Identified
ng/ul
irg/arr^
ppb
1A
Dinoseb spike
Unknown
25
N/A*
N/A
Unknown
16
N/A
N/A
Unknown
7
N/A
N/A
Dichloro-iodo-benzene
12
N/A
N/A
2
Comb, blank
Unknown-phenyl
76
0.141
N/A
w/ air staging
Unknown
185
0.344
N/A
Unknown
14
0.026
N/A
Unknown-ester
6
0.011
N/A
Dichloro-iodo-benzene
63
0.117
10.5
3
Comb, blank
Unknown-ester
43
0.160
N/A
w/ air staging
Dichloro-iodo-benzene
7
0.026
2.3
and rebuming
5
Dinoseb product
Unknown
13
0.141
N/A
w/ air staging
Unknown
10
0.109
N/A
Styrene
8
0.087
20.5
Dichloro-iodo-benzene
5
0.054
4.8
6
Dinoseb product
Unknown-ester
77
0.287
N/A
w/ air staging
Phthalate
21
0.078
N/A
and rebuming
Dichloro-iodo-benzene
5
0.019
1.7
*Not applicable.
5B-43
-------
(Intentionally Blank)
5B-44
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency. The contents do not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
THE EFFECT OF FUEL NITROGEN ON NOx EMISSIONS
FROM A ROTARY-KILN INCINERATOR
By: JoAnn S. Lighty, David L. Gordon, David W. Pershing
Department of Chemical Engineering
University of Utah
Salt Lake City, Utah 84112
Warren D. Owens
Department of Mechanical Engineering
University of Utah
Salt Lake City, Utah 84112
Vic A. Cundy and Christopher N. Leger
Department of Mechanical. Engineering
Louisiana State University
Baton Rouge, Louisiana 70803
ABSTRACT
While thermal and fuel NOx formation has been extensively studied
for fossil fuel combustion, little information is available on NOx
formation from nitrogenous waste constituents. These wastes,
usually destroyed in hazardous-waste incinerators, are prevalent and
exist as solids (plastics, nylons) or liquids (dyes, process waste).
Results are presented from studies conducted with contaminated
solids in a laboratory-scale, batch rotary-kiln simulator.
Constituent parameters, i.e. type and percent fuel nitrogen, were
studied at 1000 K. Sorbent was contaminated with a variety of
compounds ranging in concentrations from 1.0% to 4 6.0% nitrogen by
weight (for 681 g charge).
NOx exhaust-gas concentrations ranged from 50 ppm to 100 ppm for a
base run (no fuel nitrogen) to approximately 1, 000 ppm for a
nitrogenous waste. Results indicated that, at higher fuel-N
concentrations, more NOx was formed. Higher concentrations also
resulted in reduced conversions of fuel nitrogen to NOx* Conversion
Preceding page blank
-------
for different constituents, given the same concentration, varied;
pyridine conversions were slightly higher than aniline, followed by
ethylenediamine.
INTRODUCTION
Nitrogen oxides (NO and N02, commonly referred to as NOx) can be
formed from two sources:
1. nitrogen in the combustion air reacting with oxygen,
defined as thermal NOx, and
2. nitrogen in the fuel reacting with oxygen, or fuel N0X.
While fuel N0X formation has been studied extensively for fossil
fuels, little information is available on the NOx formation for
nitrogenous waste constituents. The maximum fuel nitrogen content
in fossil fuels is of the order of 2.0% by weight. Certain
hazardous wastes can contain as high as 30% by weight nitrogen.
Since hazardous-waste incinerators must be designed to meet
environmental regulations, N0X emissions generated from
incineration of these high-level nitrogenous species must be
understood.
The formation of thermal NO can be described by the Zeldovich
mechanism (1); thermal NO formation is highly temperature dependent;
however, below 1800 K, the mechanism is not favorable and formation
is unlikely. Previous researchers have shown that if nitrogen is
contained in the fuel, increased NO formation results. Pershing and
Wendt (2) found that, in self-sustaining pulverized coal flames,
fuel NO was the primary source of NO emissions. In addition, fuel-
nitrogen oxidation was relatively insensitive to temperature, unlike
thermal NO formation. In general, the research has shown that as
fuel nitrogen content is increased, conversion to NO decreases.
Conversion on the order of 50% can be possible for a fossil fuel
containing 1% nitrogen. As oxygen levels are increased, a larger
fraction of the fuel nitrogen converts to NOx.
Unfortunately, the detailed mechanisms and kinetics involved in
fuel-NO formation are not completely understood; however, the
chemical kinetics of fuel-NO formation are closely related to that
of prompt NO (3-8). Globally, the fuel-NO molecule is hypothesized
5B-46
-------
to pyrolize or react to form an intermediate nitrogen containing
species (HCN, N, HN, NH2, or NH3). The intermediate can then react
with an oxygen containing molecule to form NO, or with NO to form
another intermediate species (HCN, N, HN, NH2, or NH3) or to form
N2 . Unlike thermal-NO, fuel-NO kinetics are not strongly
temperature dependent.
Presently, most hazardous wastes are disposed by incineration or
landfilling. With regulations regarding landfills becoming more
stringent and companies having cradle-to-grave responsibility for
wastes landfilled, permanent alternatives, such as incineration,
must be explored. Little data has been gathered on nitrogen oxide
emissions of high-level nitrogen-containing wastes in the hazardous-
waste incinerator environment. These wastes, solids or liquids, are
the result of spent plastics (nitriles) and nylons (amines),
nitrogroups in dyes, and process-waste streams.
While several types of incinerators are currently commercially
available, rotary-kilns represent a large portion of the systems
used for hazardous-waste incineration. A kiln can be fed a variety
of types of wastes - solids, liquids, or sludges - and the
technology has been proven to destroy most organic wastes. For
these reasons, future advances in rotary-kiln technology are
expected (9) and a rotary kiln was chosen as the primary incinerator
system in this study.
If the high nitrogen waste is burned as a liquid, then advanced,
low-NOx techniques developed for liquid fuels can often be utilized.
England and co-workers (10,11) have shown that extremely low NOx
emission levels can be achieved by providing proper atomization,
very high temperature, and sufficient residence time in an initial,
fuel-rich portion of a two-stage combustion system. Figure 1 shows
typical results from their studies at three different scales
including an 8 million Btu/hr pilot-scale unit. These data indicate
that NOx emission levels of 60 ppm (0% 02) were achieved with a
fuel containing 0.6% nitrogen. Their data also indicated that, with
an optimized system, the minimum N0X was insensitive to fuel
nitrogen content; liquid hydrocarbon wastes with 10% nitrogen would
5B-47
-------
not be expected to produce more than 100 ppm N0X under these
conditions.
The situation with high nitrogen solid wastes is far more complex
and was the focus of this study. The objectives of this research
were to:
1. examine the conversion of fuel-nitrogen to NOx in a waste
material, in this study, a sorbent material contaminated
with a nitrogen species (aniline, pyridine, or
ethylenediamine) and toluene, and
2. determine the influence of operating and material parameters
on the conversion. This paper discusses the influence of
charging rate, nitrogen speciation and fuel-nitrogen
content. Future studies will investigate the effects of
excess oxygen and temperature.
EXPERIMENTAL APPARATUS AND PROCEDURE
To study the transient evolution of contaminants from solids, a
rotary-kiln simulator has been constructed at the University of
Utah. The rotary kiln simulator was designed to study the fate of a
control volume of solids which would pass through a full-scale
rotary kiln. In a full-scale kiln, a control volume of solids
passes through the kiln for a defined residence time which is a
function of kiln length, rotation speed, and slope of the kiln. In
the simulator, as illustrated in Figure 2, a batch of solid is
loaded into the simulator and combusted for a predetermined
residence time. Time is exchanged for distance as the independent
variable. A moveable burner can be positioned to vary the amount of
flame that is directly exposed to the bed. By measuring the gas-
phase component concentrations (02, CO, CO2, NO2, NO, and total
hydrocarbons) and other parameters (temperature, rotation rate), as
a function of time, it is possible to determine the fate of the
hazardous species in a particular control volume.
A. diagram of the rotary-kiln simulator is shown in Figure 3 and
described in detail in Lemieux and Pershing (12). The simulator is
rated at 73 kW and is constructed in three, refractory-lined
5B-48
-------
sections - a burner section, a main section, and an exhaust section.
The main section, where a charge is loaded, has an inner diameter of
0.61-m and is 0.61-m long. A moveable burner, equipped with axial
ana radial gas, is located in the burner section. The burner
section and the main section rotate, while the exhaust section is
stationary and connected to an afterburner system. The charge is
loaded through the stationary exhaust section using a chute which
extends into the main section.
Two windows allow an axial view of the flame and a direct view of
the bed. Cooling coils are cast into the outside layer of
refractory at the shell, and at the interface between the middle and
the outside layers. The kiln is driven by a lhp DC motor using a
200:1 worm reduction gear allowing rotational speeds between 0.2 and
2.0 rpm. Solid samples can be taken at any time during the
experimental run. The afterburner system consists of two natural-
gas fired burners placed in a rectangular combustion chamber
opposing one another. Each burner is rated at 73 kW.
Gas-samples are taken from the flue section, the water in the gas is
condensed, and the gas is then analyzed for C02 and CO, 02, and NOx,
using the instruments listed in Table 1, and total hydrocarbons with
Beckman THC analyzer. The analytical system is calibrated prior to
each run as shown in Figure 4. A data acquisition system, Omegalog,
is used to record the signals every 2.5 seconds for a total of 13
minutes. Data can then be directly transferred into a spreadsheet
for reduction and analysis.
The test matrix for this study is shown in Table 2 . A solid,
sorbent material was contaminated with toluene and a nitrogen-
containing waste (aniline, pyridine, ethylenediamine, or
malononitrile) varying the weight percent nitrogen. The solid was
then loaded into several cardboard containers which were then fed
into the kiln. The total weight was 681.0 g with the exception of
one experiment which was conducted using 340.5 g. This experiment
and several experiments, using only the pure nitrogenous contaminant
in various concentrations, were conducted to examine the effect of
charge size on the conversions. The temperature was constant at
1000 K.
5B-49
-------
RESULTS
Base-Line Definition
Initially a series of experiments was conducted to define the impact
of both the solid charge material and the nitrogen free toluene on
the exhaust N0X emissions. Figure 5, which summarizes these
results, shows the time resolved N0X emissions (normalized to 0%
oxygen). The auxiliary natural gas flame produce approximately 50
ppm NOx which results from thermal fixation of the atmospheric
nitrogen associated with the combustion air. When 680 g of clean
sorbent material in a cardboard container were charged (as indicated
by the dotted line), the NOx increased to approximately 80 ppm for
about six minutes. This increase is probably attributable to a
small amount of nitrogen contained in this commercial sorbent
material and/or the container. Adding 35 g of toluene (nitrogen
free) produced no additional NOx.
These results indicate that, under the conditions of this study,
thermal NOx formation from the combustion of the hydrocarbon
contaminants was not significant. The data further suggest that the
incineration of non-nitrogen containing hydrocarbons in large rotary
kiln systems may not produce major increases in exhaust NOx
emissions. Most kilns are operated with bulk temperatures well
below the thermal NOx formation thresh-hold of 1800°K and the
intensity with which liberated hydrocarbons are burned may not be
sufficient to produce the local temperatures and free radical
concentrations required for significant NOx formation.
Typical Time Resolved Data
Figure 6 shows typical time resolved NOx profiles for separate,
replicate experiments with 650g of sorbent which had been
contaminated with 23.3 g of aniline and 11.7 g of toluene (to
produce a 10wt% nitrogen contaminant). Approximately 100 seconds are
required after the cardboard containers are loaded for the break-up
of the containers and the initial heating of the solid. The
hydrocarbons then began to vaporize and local ignition occurs.
5B-50
-------
This causes a rapid increase in the rate of vaporization and the
NOx emissions increase dramatically. The subsequent decrease, at
approximately 300 seconds, occurs because the bulk of the nitrogen
containing hydrocarbons have been evolved and burned. From 3 50 s to
500 s there is a small amount of NOx formation, likely associated
with the slow evolution of the final aniline molecules. Previous
work with toluene and xylene has shown that the last 5 - 10% of the
contaminant is more difficult to remove than the first 50% .
The time resolved, 10% aniline data are presented on a fractional
conversion basis in Figure 7; exhaust oxygen concentrations are
also shown. Initially, there is a relatively large amount of oxygen
locally available to material being evolved from the bed. However,
as the evolution rate increases and ignition occurs, the local
oxygen is depleted. The overall conversion of the fuel nitrogen to
NOx is relatively low during this regime. Since this occurs at the
point of peak mass evolution, this period of low conversion
fortunately dominates the overall average conversion. During the
final portions of the test (> 300 s), the local oxygen concentration
again increases because the hydrocarbon evolution rate has greatly
slowed; visible bulk combustion is no longer apparent. Nitrogen
conversion increases during this period because of the increased
oxygen availability.
Similar conversion characteristics would be expected in a full-scale
rotary kiln where nitrogen containing solids were being charged in
drums. During the period, of peak burning rate, the instantaneous
fuel nitrogen conversion would be lower than at the beginning or the
end because the hydrocarbon combustion would tend to deplete the
local oxygen .
Feeding Rate Effects
To investigate the effects of charging rate, an experiment was
conducted in which the aniline, toluene, and sorbent material were
all decreased by 50%. These results, shown in Figure 8, suggest
that the integrated N0X emissions are only weakly influenced by the
kiln charging rate. The case with a smaller charge, ignited earlier
and burned out faster as expected. The depression in the exhaust
5B-51
-------
oxygen was approximately the same intensity but of shorter duration,
suggesting similar average local oxygen concentrations through-out
the period of nitrogen evolution. This resulted in an almost
identical overall nitrogen conversion of 19.4 % relative to the base
case nitrogen conversion of 18.7 %. The results are in good
agreement with earlier fossil fuel studies which have indicated the
importance of minimizing local oxygen availability.
Influence of Nitrogen Speciation
Aniline, pyridine, and ethylenediamine were all investigated to
establish the influence of contaminant speciation. Figure 9
summarizes the overall results from a series of experiments where
680 g of total charge was added to the kiln. In these experiments
only the pure nitrogen containing chemicals were used; toluene was
not added to hold the hydrocarbon content of the sample constant.
Rather, an equivalent amount of nitrogen was charged in each case
(0.01 g of nitrogen per g of charge) . These results show an
apparent strong influence of chemical speciation on the overall
conversion of the fuel nitrogen to NOx-
To further investigate this somewhat surprising result, detailed
experiments were conducted with all three compounds at various
concentrations. Figure 10 summarizes typical time resolved
emissions data for the aniline series; in these experiments the
fuel nitrogen content in the hydrocarbon/contaminant mix (aniline
plus toluene) was varied from 1% to 10% while holding the total
hydrocarbon content constant at 35 g and the sorbent mass constant
at 646 g. The time resolved behavior was similar for all three
cases; both the peak and integrated NOx emissions increased with
increasing nitrogen content as expected.
Figure 11 summarizes all of the integrated conversion results
obtained for the three different nitrogen-containing contaminants.
These results are plotted as a function of the nitrogen content in
the hydrocarbon contaminant; previous detailed studies on NOx
emissions from the combustion of liquid fuels, under excess air
conditions (10) have indicated that weight percent nitrogen in the
liquid phase is the primary correlating variable, given constant
5B-52
-------
combustion conditions. Figure 11 indicates that the percentage of
fuel nitrogen to converted NOx decreases as the weight percent
nitrogen increases. This result is consistent with the previous
work of England (11) and kinetic theory. The conversion decreases
with increasing nitrogen availability because the reactions forming
NO are essentially first order with respect to available nitrogen
species while the reactions forming N2 are, of necessity, second
order.
In contrast to the previously available data for liquid fuel
combustion, the results shown in Figure 11 also suggest a secondary
influence of nitrogen speciation. Figure 12 provides a detailed
comparison of the conversion results obtained with the three
compounds at a constant fuel nitrogen percentage (3%). Pyridine,
where the nitrogen is contained in the aromatic ring, produced the
highest relative conversion; ethylenediamine, where the nitrogen is
present as an amine group in the alkane chain, resulted in the
lowest nitrogen conversion. Further investigations are required to
elucidate the mechanistic reason for this difference, but it is
likely that the volatility of the nitrogen species, relative to the
bulk hydrocarbon contaminant, is an important parameter.
CONCLUSIONS
The results of this study indicate that significant quantities of
nitrogen oxides can be formed during the incineration of nitrogen
containing solid waste. In this investigation, thermal NOx
formation was relatively insignificant; however, conversions of the
nitrogen chemically bound in the organic contaminant to N0X ranged
from 2 to 4 6%, depending on the nitrogen speciation and combustion
conditions.
Total N0X emissions were related to the total nitrogen in the
waste, the speciation of the nitrogen, and the nitrogen
concentration in the hydrocarbon phase. N0X emissions increased
with increasing nitrogen content; however, the percentage
conversion decreased in accordance with previous fossil fuel
results. Nitrogen speciation was found to have a measurable but
5B-53
-------
second order effect; conversions with pyridine were higher than
with ethylenediamine.
For the conditions of this study, charging rate did not
significantly influence either the percentage conversion of the
chemical nitrogen, or the NOx emissions per unit mass of
contaminant. The experimental results suggest that the local oxygen
concentration has a major influence on the instantaneous fuel
nitrogen conversion, but additional studies are required to quantify
the effects of excess oxygen and temperature.
ACKNOWLEDGMENTS
This work was sponsored by the Advanced Combustion Engineering
Research Center. Funds for this Center are received from the
National Science Foundation, the State of Utah, 22 industrial
participants, and the U.S. Department of Energy. Funding from the
National Science Foundation and the EPA, in conjunction with
Louisiana State University's Hazardous Waste Research Center, is
also appreciated.
REFERENCES
1. Zeldovich, Y. B., P. Sadovnikov, and D. A. Frank-Kamenetskii.
"Oxidation of Nitrogen in Combustion." Academy of Sciences
U.S.S.R. Institute of Chemical Physics, Moscow-Leningrad,
trans, by M. Shelef, 1947.
2. Pershing, D. W., and J. 0. L. Wendt, Sixteenth Symposium
(International) on Combustion. The Combustion Institute,
Pittsburgh, PA, p. 389-399, 1977.
3. Glassman, I., Combustion. Academic Press, New York, NY, 1977.
4. Haynes, B. S., et al., Alternative Hydrocarbon Fuels:
Combustion and Chemical. Kinetics, editors, C. T. Bowman and
N. Birkland, "Kinetics of Nitric Oxide Formation in
Combustion," Progress in Astronautics and Aeronautics, Vol.
62, pp. 359-94, 1978.
5. Fenimore, C. P., Seventeenth Symposium (International) on
Combustion. The Combustion Institute, Pittsburgh, PA, p. 661,
1979.
5B-54
-------
6. Fenimore, C. P., Thirteenth Symposium (International) on
Combustion. The Combustion Institute, Pittsburgh, PA, p. 373,
1971.
7. Sawyer, R. F., 5., Eighteenth Symposium (International) on
Combustion. The Combustion Institute, Pittsburgh, PA, p. 1,
1981.
8. Glarborg, P., J. A.Miller, and R. J. Kee, Combust. Flame. Vol.
65, 1986, p. 177.
9. Oppelt, E. T. "Incineration of Hazardous Waste: A Critical
Review." JAPCA. Vol. 37, 1987, pp. 558-86.
10. England, G. C., M. P. Heap, D. W. Pershing, R. K. Nihart, and
G. B. Martin, Eighteenth Symposium (International) on
Combustion. The Combustion Institute, Pittsburgh, PA, p. 163-
174, 1981.
11. England, G. C., M. P. Heap, R. Payne, D. W. Pershing, and W.
S. Lanier, "Development of Design Criteria for Low NOx Oil
Burners," paper presented at the 75th Annual APCA Meeting,
New Orleans, LA, June 1982.
12. Lemieux, P. M. and D. W. Pershing, "The Design and
Construction of a Rotary Kiln Simulator for Use in Studying
the Incineration of Hazardous Waste," Review of Scientific
Instruments. submitted, 1989.
5B-55
-------
0.5 0.6 0.7 0.8 0.9 1.0
Stoichiometric Ratio (First Stage)
Figure 1. Influence of First-Stage Stoichiometry on NOx
(Bench-scale and Pilot-Scale) from England (11)
5B-56
-------
• Control Volume Follows Batch of Solid Through Kiln
Full Scale
Simulator
=0£i&>
b
Figure 2. Lab-scale rotary-kiln simulation of a full-scale unit.
-------
Burner Gas Tube
Cooling Coils
Rotating Water Fittings
Rotary Arms
Front Rotary Seal
'n y Chain and Sprocket
¦Ul.'l.q n, i ••••••«
I *
BURNER
SECTION
im
• •
• •
• •
m m
MA
N
SECTI
ON'"
• •
L
200:1
Worm Gear
1 hp
DC Motor
HC/G.C. Sample Probe
Sample
Probe
t
Loading Chute
Oxygen Sample Probe
Figure 3. Rotary-kiln simulator: top view.
5B-58
-------
Calibration Gases
Flue Gas
Sample from
Kiln
To Exhaust
cn
00
cn
10
ue Gas Sample Manifold
Refrigerator with
water traps
C02/C0 Oxygen
Analyzer Analyzer
Figure 4. Sample Analysis System
300
-Q Blank Run
~ Toluene Run
200
400
Time (sec)
600
800
Figure 5. Baseline NOx Formation
(dotted line indicates point of bed charging)
0 200 400 600 800
Time (see)
Figure 6. Typical, Time-Resolved NOx Profile
-------
Time (sec)
Figure 7. Typical Time-Resolved Conversion Data and Oxygen Data
Time (sec)
Figure 8. Influence of Charge Size
-------
8
8
Z
o
Z 6
"8
"o
2 4
0
e
c
<5
1 2
u
*
0
Figure 9. Influence of Compound Type and N Concentration
g 0.01 gN/g charge
r-
% Nitrogen
Time (sec)
Figure 10. Comparison of Nitrogen in Fuel
5B-61
-------
50
O
z
>
s
a
£
0
l . | i
40
q Aniline
^ Pyridine
t Ethylenediamine
30
-
q Malononitrile
-
¦ A
20
¦ s
-
10
~
-
§ 1 *
~ u
$
0
( 0 ,
10
20 30
Weight Percent Nitrogen
40
50
Figure 11. Effect of Nitrogen in Fuel on Nitrogen Conversion to NO
Aniline Pyridine Ethylenediamine
Figure 12. Influence of Compound Type for 3.0 Weight% Nitrogen
5B-62
-------
Table 1
Analytical Instruments
Instrument
Range
Analog Voltage
Output
0 - 10 mV
CO? Ranges 0 - 100 mV
0-6%
0 - 30%
CO Ranges 0 - 10 0 mV
0-1%
0 - 10%
Four range selections 0 - 100 mV
0-5%
0 - 10%
0 - 25%
0 - 50%
Eight range selections
0-2.5 ppm
0-10 ppm
0-25 ppm
0 - 100 ppm
0 - 250 ppm
0 - 10 00 ppm
0 - 2500 ppm
0 - 100 00 ppm
Beckman Four range selections 0 - 100 mV
Model 40 0 0-10 ppm
Total Hydrocarbon Analyzer 0 - 100 ppm
(Flame Ionization Detector) 0 - 1000 ppm
0 - 10000 ppm
Anarad
Model AR-600
Dual Gas Analyzer
(NDIR)
Beckman
Model 755
Paramagnetic
Oxygen Analyzer
ThermoElectric
Series 10
Chemiluminescent
NO - N02 " NOx
5B-63
-------
Table 2
Test Matrix
Constituent g Constituent g Toluene g Sorbent
wt% Nitrogen in
the Total Hydro-
carbon Mixture
none 0.0
(a)
0.0
681.0
0.0
none 0.0
35.0
646.0
0.0
Aniline 2.3
32.7
646.0
1.0
3.5
14.0
323.0
3.0
7 .0
28.0
646.0
3.0
11.6
23.4
646.0
5.0
23.3
(a)
11.7
646.0
10. 0
22.6
0.0
658.4
15.0
45.3
0.0
635.7
15.0
79.2
0.0
601.8
15.0
113.1
0.0
567.9
15.0
Pyridine 8.5
41.5
631.0
3.0
5.9
29.1
646.0
3.0
38.4
0.0
642.6
18.0
115.3
(b)
0.0
565.7
18.0
Malononitrile 56.2
0.0
624.8
42.0
Ethylene- 2.3
32.7
646.0
3.0
diamine 29.1
0.0
651.9
47.0
Notes:
(a) 3 replicate runs
(b) 2 replicate runs
5B-64
-------
Appendix A
EPRI/EPA Symposium on Stationary Combustion NOx Control
03/06/89-03/09/89
San Francisco, CA
List of Attendees
James R. Adams
Project Manager
New York State Electric & Gas
4500 Vestal Parkway, East
Binghamton, NY 13903
607/724-0849
Rui Afonso
Senior Research Engineer
New England Electric System
25 Research Drive
Westborough, MA 01582
508/366-9011
Ken Ahn
N/A
Coen Company
1510 Rollins Road
Burlingame, CA 94010
415/697-0440
Charles D. Allen
Sr. Mechanical Consulting Engr.
Arizona Public Service Company
P.O. Box 53999
Phoenix, AZ 95072-3999
602/371-6750
Sy B. Alpert
EPRI Fellow
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2512
Leonard C. Angello
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2873
James A. Armstrong
Vice President, Operations
ADA Technologies, Inc.
304 Inverness Way South, Suite 480
Englewood, CO 80112
303/792-5615
D. S. Arnold
Engineering Specialist
Kerr-McGee Corporation
P.O. Box 25861
Oklahoma City, OK 73125
405/270-2911
Wayne J. Aronson
Chief, Program Support Section
U.S. EPA
345 Courtland Street
Atlanta, GA 30365
404/347-2864
Patrick F. Aubourg
Supervisor
Owens Corning Fiberglas
Tech. Center, P.O. Box 415
Granville, OH 43023
614/587-7604
A. Baldacci
Senior Engineer
ENEL
Via G. B. Martini 3
00198 Rome, ITALY
6/8509-2898
Lothar Balling
Dipl.-Ing.
Siemens AG, KWU Group
Hammerbacherstrasse 12+14
D-8520 Erlangen
FEDERAL REPUBLIC OF GERMANY
9131/182350
A-l
-------
Joe Barsin
Manager, Industrial Projects
Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44203
216/860-1552
Dwayne Bateson
Executive Assistant
TransAlta Utilities
110 12th Avenue, S.V.
Calgary, Alberta T2P 2M1
CANADA
403/267-7129
Nick Bayard de Volo
President
Energy Technology Consultants
2091 Business Center Drive
Suite 100
Irvine, CA 92715
714/833-2522
Peter R. Beal
Manager, Business Development
NEI International Combustion, Ltd.
Sinfin Lane
Derby DE2 9GJ
UNITED KINGDOM
332/271111
Gordon Beales
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2591
Rob Becker
President
Environmental Catalyst Consultants
P.O. Box 637
Spring House, PA 19477
215/628-4447
Joseph Beer
Department Manager
Siemens AG, KWU Group
Hammerbacherstrasse 12+14
Dept. U3112, 8520 Erlangen
FEDERAL REPUBLIC OF GERMANY
09131/182932
Ted Behrens
Account Executive
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
Gerald R. Bernis
Engineer
California Energy Commission
R&D Office
1516 9th Street, MS-43
Sacramento, CA 95814-5512
916/324-3468
Angelo Benanti
Mgr., Thermal Equipment
ENEL
Via G. B. Martini 3
00198 Rome, ITALY
6/8509-2898
Knud Bendixen
Vice President
Burmeister & Wain Energy
23 Teknikerbyen
Virum, DENMARK 2830
45/2 857100
Mogens Berg
N/A
Elkraft Power Company
5 Lautruphoej
DK-2750 Ballerup, DENMARK
45/2 660022
Leif Bernergurd
Technical Officer
Environment Protection Board
Box 1302
S-17125 Solna, SWEDEN
8/799-1119
A-2
-------
Sisto Bertacchi
Senior Researcher
ENEL
Via A. Pisano 120
Pisa, ITALY 56100
050/535693
Gary L. Bisonett
Senior Engineer
Pacific Gas & Electric
245 Market Street, Room 434A
San Francisco, CA 94106
415/973-6950
Kerry Blaraire
Generation Studies
Nova Scotia Power Corporation
P.O. Box 910
Halifax, Nova Scotia
CANADA B3J 2W5
902/428-6655
Verle Bland
Manager, Applied Technology
KVB, Inc.
9342 Jeronimo
Irvine, CA 92718
714/587-2326
Richard Boardman
Student
Brigham Young University
184 CB
Provo, UT 84602
801/378-2076
Richard T. Bobick
Senior Process Engineer
Unocal Corporation
1201 West 5th Street
P.O. Box 7600
Los Angeles, CA 90051
213/977-6435
Dieter Boekenbrink
Director
RWE
5180 Eschweiler, Box 1448
WEST GERMANY
02403-732112
Diego Bonolis
N/A
ENEL
Via G.B. Martini 3
00198 Rome, ITALY
06/85095192
Peter J. Booras
President
Yankee Energy Corporation
80 Boylston Street
Boston, MA 02116
617/542-0550
Richard W. Borio
Principal Consulting Engineer
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-2229
Kerry W. Bowers
Senior Research Engineer
Southern Company Services
P.O. Box 2625
Birmingham, AL 53202
205/870-6825
Wallace H. Bradley
V.P., Engineering
Austell Box Board Corporation
P.O. Box 157
Washington Street
Austell, GA 30001
404/948-3100
Jan Brandin
N/A
University of Lund
Chemical Engineering II
P.O. Box 124
S-221 00 Lund, SWEDEN
046/108284
Ron L. Bredehoft
Staff Planning Analyst
Chevron U.S.A.
324 W. El Segundo Boulevard
El Segundo, CA 90245
213/615-5000
A-3
-------
Bernard P. Breen
President
Energy Systems Associates
1840 Gateway Three
Pittsburgh, PA 15222
412/392-2380
Brent Brigham
Project Engineer
Unocal Corporation
P.O. Box 758
Wilmington, CA 90748
213/513-7600
R. G. Broderick
V.P..Engineering
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
W. J. Brooks
Senior Engineer
CEGB
Barnett Way, Barnwood
Gloucester GL4 7RS
UNITED KINGDOM
0452/652-306
Bert Brown
Applications Manager
Joy Technologies, Inc.
404 E. Huntington Drive
Monrovia, CA 91016-3633
818/301-1172
Douglas R. Brown
Engineer
Applied Utility Systems
1140 E. Chestnut Avenue
Santa Ana, CA 92701
714/953-9922
Gordon Brown
Staff Engineer
Exxon Chemical Americas
5000 Bayway Drive
Baytown, TX 77522
713/425-5033
A-4
C. P. Brundrett
Manager
W. R. Grace & Company
10 E. Baltimore Street
Baltimore, MD 21202
301/659-9125
Mark R. Burkhardt
Research Chemist
ADA Technologies, Inc.
304 Inverness Way South, Suite 480
Englewood, CO 80112
303/792-5615
Kenneth R. Burns
Product Manager
Engelhard Corporation
Edison, NJ 08818
201/632-6640
Tony A. Burns
V.P. & Mgr., Env. Technology Group
S-Cubed, Division of Maxwell Labs
P.O. Box 1620
La Jolla, CA 92038
619/453-0060
John W. Byrne
Senior Research Chemist
Engelhard Corporation
Menlo Park
Edison, NJ 08818
201/321-5153
Scott Cameron
Sales Engineer
Babcock & Wilcox
17172 Abalone Lane, #207
Huntington Beach, CA 92649
714/846-0817
Gary A. Camody
Product Services Manager
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-5039
-------
E. J. Campobenedetto
Marketing Manager
Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44203
216/860-6762
Giovanni Caprioglio
Consultant
General Atomics
P.O. Box 85608
San Diego, CA 92138-5608
619/455-2918
E. J. Capriotti
Vice President, Sales
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
James Carnevale
Supervising Mechanical Engineer
Los Angeles Dept. of Water & Power
P.O. Box 111, Room 604
Los Angeles, CA 90051
213/481-4895
Baldwin K. Chan
Sr. Mechanical Engineer
Pacific Gas & Electric
77 Beale St., Room 2523
San Francisco, CA 94106
415/972-5236
Charles Chang
Assistant Supervisor, Air Quality
Los Angeles Dept. of Water & Power
111 N. Hope Street
Los Angeles, CA 90012
213/481-3235
Shih-Ger Chang
Senior Scientist
Lawrence Berkeley Laboratory
University of California
Berkeley, CA 94720
415/486-5125
Wei-Yin Chen
Assistant Professor, Research
Louisiana State University
Department of Chemical Engineering
Baton Rouge, LA 70803
504/388-3059
Lisa J. Chrisman
Engineer
Fossil Energy Research Corporation
821 Hamilton Drive
Pleasant Hill, CA 94523
415/937-9007
Roger C. Christman
Program Manager
Radian Corporation
13595 Dulles Technology Drive
Suite 200
Herndon, VA 22071
703/834-1500
Ed Cichanowicz
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2374
Leonard Clayton
Field Engineering
Bay Area Air Quality Mgmnt. Dist.
939 Ellis Street
San Francisco, CA 94109
415/771-6000
Robert J. Collette
Contract Manager
Combustion Engineering
1000 Prospect Hill Road, 5034-GC25
Windsor, CT 06095-0500
203/285-5687
Roger S. Cook
Environmental Engineer
Dept. for Environmental Protection
Kentucky Div. for Air Quality
18 Reilly Road
Frankfurt, KY 40601
502/564-3382
A-5
-------
Frederick M. Coppersmith
Manager, Research & Development
Con. Edison Co. of New York
4 Irving Place
New York, NY 10003
212/460-3098
David A. Cowdrick
Senior Engineer
Tampa Electric Company
P.O. Box 111
Tampa, FL 33601
813/671-3361
Ed Cowle
Project Engineer
Bechtel Power Corporation
12440 E. Imperial Highway
Norwalk, CA 90650
213/807-2423
Anthony Cowley
Mechanical Design Supervisor
B. C. Hydro
970 Burrard Street
Vancouver, British Columbia
V6Z 1Y3
604/663-2849
William R. Cress
Manager, Engineering Studies
Allegheny Power Service Corporation
Cabin Hill Drive
Greensburg, PA 15601
412/838-6721
Diane V. Croson
Mgr., LLW/NOx Process Dev.
Westinghouse Idaho Nuclear Company
P.O. Box 4000
Idaho Falls, ID 83403
208/526-3402
Michael Davidson
Mgr., New Product Development
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-9005
Charles B. Davis
Senior Staff Engineer
Virginia Power
5000 Dominion Boulevard
Glen Allen, VA 23060
804/273-2619
Gerard G. De Soete
Dr. es Sciences
Institut Francais du Petrole
1-4, Avenue de Bois-Preau
Box N311, 92506 Rueil-Mailmaison
Cedex, FRANCE
331/47526145
Dennis D. Delaney
Senior Research Chemist
Unocal Corporation
376 S. Valencia Avenue
Brea, CA 92621
714/528-7201 X2184
Mukesh S. Desai
Engineering Specialist
Bechtel Power Corporation
15740 Shady Grove Road
Location 2A-12
Gaithersburg, MD 20877-1454
301/258-3150
Dilip Deshpende
Supervising Engineer
Alberta Power Company
10035 105th Street
Edmonton, Alberta
CANADA D5G 2V6
403/420-7177
William DePriest
Project Engineer
Sargent & Lundy
55 E. Monroe Street
Chicago, IL 60603
312/269-6678
A-6
-------
Bill Dick
Design Specialist
Ontario Hydro
700 University Avenue
Toronto, Ontario
CANADA M5G 1X6
416/592-8593
Maurizio Didio
N/A
Snamprogetti Spa
c/o Snamprogetti USA, Inc.
666 5th Floor
New York, NY 10103
212/399-1090
Richard Diffenbach
N/A
U.S. Department of Energy
P.O. Box 10940
Pittsburgh, PA 15236
412/892-6090
David J. Dodd
Mgr., Chemical Research
Ontario Hydro
800 Kipling Avenue
Toronto, Ontario
CANADA M8Z 5S4
416/231-4111 X6519
Roger Dodds
Air Quality Engineer
Wisconsin Electric Power Company
333 W. Everett Street
Milwaukee, WI 53201
414/221-2174
Zambelli Donato
Engineer
Azienda Servizi Municipal
Via. Lamarmora N. 230
Brescia, ITALY
030/3311-291
Yogesh P. Doshi
Senior Environmental Engineer
New Jersey Dept. of Env. Protection
401 E. State Street, 2nd Floor
Trenton, NJ 08625
609/633-7249
Dennis C. Drehmel
Technical Manager
U.S. EPA
AEERL -- MD-04
Research Triangle Park, NC 27711
919/541-7505
Joseph R. Dulovich
Plant Superintendent
Ohio Edison Company
1047 Belmont Avenue
Niles, OH 44446
216/544-7327
Michael K. Dunbar
Mechanical Engineer
Pacific Gas & Electric
77 Beale Street, Room 2547
San Francisco, CA 94106
415 972-0606
Richard A. Dye
Fossil Energy Department
U.S. Department of Energy
FE-4, Forrestal
Washington, DC 20585
202/252-6499
Owen W. Dykema
President
Dykema Engineering
5850 Canoga Avenue, Suite 400
Woodland Hills, CA 91367
818/712-0070
William C. Eddins
Dir., Division of Air Quality
Dept. for Environmental Protection
18 Reilly Road
Frankfurt, KY 40601
502/564-3382
Greg Eirschele
Environmental Engineer
Wisconsin Power & Light Company
222 W. Washington Avenue
Madison, WI 53701
608/252-3084
A-7
-------
John W. Eldridge
Professor of Chemical Engrng.
University of Massachusetts
39 Kendrick Place
Amherst, MA 01003
413/253-5991
Tom E. Emmel
Mgr., Combustion Engineering Dept.
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
919/541-9100
Robert Epperly
Executive Vice President
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906'
203/359-1320
David Eskinazi
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2918
Tony Facchiano
N/A
Coen Company
1510 Rollins Road
Burlingame, CA 94010
415/697-0440
Hamid Farzan
Senior Research Engineer
Babcock & Wilcox
1562 Beeson Street
Alliance, OH 44601
216/821-9110
Howard Feibus
Dir., Office of Coal Combustion
U.S. Department of Energy
F#-232, GTN
Washington, DC 20545
301/353-4348
W. D. Fellows
Senior Staff Engineer
Exxon Research & Engineering Co.
P.O. Box 101
Florham Park, NJ 07932
201/765-1250
P. G. Finlay
Head, Electric Power Section
Environment Canada
351 St. Joseph Blvd., 13th Floor
Ottawa, Ontario K1A 0H3
CANADA
819/953-1126
Dave Finnegan
Counsel
Committee on Energy and Commerce
2125 Rayburn Building
Washington, DC 20515
202/225-2927
Kelly A. Fortin
Environmental Engineer
U.S. EPA -- Region 9
2921 Santos Lane, Apt. 2126
Walnut Creek, CA 94596
415/974-7043
Bill Fraser
Senior Vice President
TransAlta Utilities
P.O. Box 1900
Calgary, Alberta
CANADA T2P 2M1
403/267-7482
Steve Frey
N/A
U.S. EPA -- Region 9
215 Fremont Street
San Francisco, CA 94105
415/974-8071
Yuan Fu
Chemical Engineer
U.S. Department of Energy
P.O. Box 10940
Pittsburgh, PA 15236
412/892-4841
A-8
-------
Rajendra P. Gaikwad
Engineer
Ontario Hydro
800 Kipling Avenue
Toronto, Ontario
CANADA M5L 2Z6
416/231-4111
Michael Gamburg
Regional Sales Manager
Fuel Tech
750 17th Ave., Suite 102
San Francisco, CA 94121
415 221-6177
Gerald M. Gardetta
Environmental Affairs Admin.
Southern California Gas Company
810 S. Flower Street
Los Angeles, CA 90017
213/689-3365
James S. Geier
Public Health Engineer
Colorado Department of Health
4210 East 11th Avenue
Denver, CO 80220
303/331-8500
John Gerdes
Design Engineer
John Zink Company
P.O. Box 702220
Tulsa, OK 74170
918/592-4810
Marco Ghiringhelli
Liaison Engineer
Ansaldo -- Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44203
216/860-6029
Robert L. Gilbert
Program Manager
MK-Ferguson Company
One Erieview Plaza
Cleveland, OH 44114
216/523-6618
Alan F. Gillespie
Manager, Engineering Services
Foster Wheeler, Ltd.
P.O. Box 3007
St. Catharines, Ontario
L2R 7B7, CANADA
416/688-4434
Tom P. Gilmore
Product Line Manager
KTI Corporation
1333 S. Mayflower Avenue
Monrovia, CA 91016-4099
818/303-4711
Dan V. Giovanni
Consultant
Electric Power Technologies, Inc.
P.O. Box 5560
Berkeley, Ca 94705
415 653-6422
Lisa Glatch
Process Engineer
Fluor Daniel
3333 Michelson Drive, MC-B1B
Irvine, CA 92730
714/975-3047
Frans Goudriaan
Senior Process Technologist
Kon./Shell Laboratory
Badhuisweg 3, Dept. HCP
1031 CM Amsterdam
HOLLAND
31/20 303957
Toby R. Gouker
Mgr., Stationary Emission Control
W. R. Grace & Company
7379 Route 32
Columbia, MD 21044
301/531-4131
John T. Graves
Environmental Superintendent
Minnkota Power Cooperative
P.O. Box 127
Center, ND 58530
701/794-8711
A-9
-------
Francois Grimard
N/A
Fuel Tech Europe, Ltd.
28a Cadogan Square
London, SWIX OJH
UNITED KINGDOM
01/581-2051
Michael Grimsberg
M. Sc.
University of Lund
Dept. Chem. Engineering
P.O. Box 124
S-221 00 Lund, SWEDEN
46/46 10 82 76
Charles W. Grinnell
V.P. and General Counsel
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
Robert R. Grinstead
Associate Scientist
Dow Chemical USA
2800 Mitchell Drive
Walnut Creek, CA 94598
415/944-2077
P. A. Gristwood
Head Combustion Technology Group
Babcock Energy, Ltd.
165 Great Dover Street
London SE1 4YB
ENGLAND
01/407 8383 X2272
George Gunter
Corporate Engineering Specialist
New Brunswick Power
P.O. Box 2000
Fredericton, New Brunswick
CANADA E3B 4X1
506/458-4381
Vir V. Gupta
EPE IV
U.S. EPA
200 Churchill Road
P.O. Box 19276
Springfield, IL 62794-9276
217/782-2113
Jim Guthrie
Mechanical Engineer
California Air Resources Board
P.O. Box 2815
Sacramento, CA 95812
916/327-1508
Donald K. Hagar
President
Eagleair, Inc.
1150 Mauch Chunk Road "
Bethlehem, PA 18018
215/868-1616
Leo E. Hakka
New Venture Manager
Union Carbide Canada, Ltd.
P.O. Box 700
Pointe-Aux-Trembles, Que
CANADA H1B 5KE
514/493-2617
Bob Hall
Chief, Combustion Research Branch
U.S. EPA
AEERL -- MD-65
Research Triangle Park, NC 27711
919/541-2477
David Ham
Vice President
PSI Technology Company
P.O. Box 3100
Andover, MA 01810
508/475-9030
Richard Haman
Principal Engineer
Detroit Edison
2000 Second Avenue
Detroit, MI 48226
313/897-0208
A-10
-------
Peter J. Handy
Engineering Sales Director
Rolls Royce Canada
P.O. Box 544, Montreal AMF
Montreal, Quebec
CANADA H4Y 1B3
514/631-3541
Thomas Hansen
Chief, Mobile Source Section
U.S. EPA
345 Courtland Street
Atlanta, GA 30365
404/347-2864
Stan Harding
Research Group Leader
Consolidation Coal Company
4000 Brownsville Road
Library, PA 15129
412/854-6735
Peter J. Hart
Supervisor, Emissions Comp.
Allison Gas Turbine Div. of GM
P.O. Box 420, SC T-01
Indianapolis, IN 46206
317 230-4186
Scott Hassett
Principal Engineer
Utah Power & Light Company
168 North, 1950 West
Salt Lake City, UT 84140
801/220-4839
Greg Haussmann
Research Assistant
Stanford School of Engineering
Blgd. 520, Dept. of Mech. Engrng.
Stanford, CA 94305
415/723-1823
Edward Healy
Project Planning Engineer
Southern Company Services
P.O. Box 2625
Birmingham, AL 35202
205/868-5212
Dietmar Hein
Professor
Siemens AG, KWU Group
Hammerbacherstrasse 12+14
D-8520 Erlangen
FEDERAL REPUBLIC OF GERMANY
9131/182350
Klaus Hein
Prof. Dr-Ing.
RWE AG, BV Fortuna
Postfach 1461
5010 Bergheim 4
FEDERAL REPUBLIC OF GERMANY
02271/584-2234
Todd He11ewe11
Principal Engineer
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-4919
Eric Hennen
Environmental Engineer
Dairyland Power Company
2615 East Avenue, So.
Lacrosse, WI 54601
608/787-1371
Bo Herrlander
Business Area Manager
Flakt Industres AB
S-35187 Vaxjo
SWEDEN
01146/470-87400
Steven A. Hickerson
President
Emcotek Corporation
8220 Doe Avenue
Visalia, CA 93291
209/651-2000
David Hickman
Senior Scientist
Corning Glass Works
SP-FR-6-1
Corning, NY 14831
607/974-3715
A-ll
-------
M. B. Hilt
Mgr., Externally Funded Programs
General Electric
1 River Road
Schenectady, NY 12345
518/385-9191
Anna Hinderson
Engineer
ABB Carbon
Box S61220
SWEDEN
N/A
Anna-Karin E. Hjalmarsson
Environment Group
IEA Coal Research
14-15 Lower Grosvenor Place
London SW1W OEX ENGLAND
01/828-4661
Lennart Hjalmarsson
Engineer
ABB Carbon
Box S61220
SWEDEN
N/A
Patrick A. Ho
Principal Engineer
Tampa Electric Company
702 N. Franklin Street
Tampa, FL 33602
813/228-4844
John E. Hofraann
Vice President, Development
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
Bill Holt
Manager, Business Development
Allied Signal
P.O. Box 580970
Tulsa, OK 74158-0970
918/266-1400
Neville Holt
Senior Program Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, Ca 94034
415/855-2503
Richard G. Hooper
Vice President
NaTec, Ltd.
216 Sixteenth Street Mall
Suite 1500
Denver, CO 80202-5129
303/820-5200
Hann S. Huang
Chemical Engineer
Argonne National Laboratory
9700 S. Cass Avenue
Argonne, IL 60439
312/972-3125
Allen Hubbard
Environmental Engineer
Wis. Dept. of Natural Resources
P.O. Box 7921
Madison, WI 53707
608/266-3450
Terry Hunt
Mechanical Engineer
Public Service Co. of Colorado
5900 East 39th Avenue
Denver, CO 80207
303/329-1113
Adrian Hyde
Engineer
UK Department of Energy
Thames House South, Millbank
London, SW11 4QT
UNITED KINGDOM
01/211-4547
Ivan V. Insua
Staff Engineer
Salt River Project
P.O. Box 52025
Phoenix, AZ 85072-2025
608/236-5240
A-12
-------
Paul Ireland
Project Engineering Manager
United Engineers & Constructors
P.O. Box 5888
Denver, CO 80217
303/692-3420
Reda S. Iskandar
Project Manager
Corning Glass Works
MP-WX-01-5
Corning, NY 14831
607/974-7306
Sven Jaras
Engineer
Eka Novel AB
S-445 01 Surte
SWEDEN
+46/303 98377
Jim Jarvis
Senior Engineer
Radian Corporation
8501 Mo-Pac
Austin, TX 78720
512/454-4797
Torgny Johansson
Manager Process Technology
Flakt Industres AB
S-35187 Vaxso
SWEDEN
01146 470-8700 0
David L. Johnson
Mgr., Production, Engrng, & Const.
Otter Tail Power Company
215 S. Cascade Street
Fergus Falls, MN 56537
218/739-8399
Keith Johnson
N/A
Foster Wheeler Power Products
Olympic Office Center
No. 8 Fulton Road
Wembly, Middlesex HA9 0JH, UK
01/900-2533
Kevin L. Johnson
Project Manager
Radian Corporation
3200 E. Chapel Hill Road
Research Triangle Park, NC 27709
919/541-9100
Steve Johnson
Mgr., Applied Combustion Tech.
PSI Technology Company
P.O. Box 3100
Andover, MA 01810
508/475-9030
Wayne L. Johnson
Sr. Process Engineering Specialist
ESSO Resources Canada, Ltd.
237 Fourth Avenue, S.W.
Calgary, Alberta
CANADA T2P 0N6
403/237-2721
Dale G. Jones
Vice President
Emcotek Corporation
8220 Doe Avenue
Visalia, CA 92391
209/651-2000
Gary D. Jones
Program Manager
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
919 541-9100
Thomas E. Kaforey
Generating Plant Engineer
Ohio Edison Company
76 South Main Street
Akron, OH 44308
216/384-5974
Marie Kalinowski
Advanced Scientist
Owens Corning Fiberglas
Route 16
Granville, OH 43023
614/587-7620
A-13
-------
Dale A. Kanary
Engineer
Ohio Edison Company
76 South Main Street
Akron, OH 44308
216/384-5744
Hans T. Karlsson
Associate Professor
University of Lund
Dept. Chem. Engineering Box 124
S-221 00 Lund, SWEDEN
46 46 10 82 44
Ari Karppinen
Research Scientist
Finnish Meteorological Institute
Sahaajankatu 22E
SF 00810 Helsinki
FINLAND
358/0 7581345
Robert F. Kaupp
Marketing Manager, Projects
Johnson Matthey
436 Devon Park Drive
Wayne, PA 19087
215/341-8502
Martin Kay
Supervising Air Quality Engineer
So. Coast Air Quality Mgmnt. Dist.
9150 Flair Drive
El Monte, CA 91731
818/572-6258
Robert T. Keen
Principal Engineer
Fichtner USA, Inc.
Overlook 1, Suite 360
2849 Paces Ferry Road, N.W.
Atlanta, GA 30339
404/432-6983
Robert L. Keller
Corporate Engineer
Dayton Power & Light Company
9200 Chautauqua Road
Miamisburgh, OH 45342
513/865-6234
Stephen E. Kerho
Consultant
Electric Power Technologies, Inc.
25108 Marguerite Pkwy., Ste 8239
Mission Viejo, CA 92692
714 380-7316
Fred Kessler
Project Manager
KTI Corporation
1333 S. Mayflower Avenue
Monrovia, CA 91016
818/303-4711
Lawrence P. King
Senior Marketing Specialist
Babcock & Wilcox
1562 Beeson Street
Alliance, OH 44601
216/829-7576
Masaaki Kinoshita
Assistant Manager
Mitsubishi Heavy Industries
1-1, Akunoura-Machi
Nagasaki, JAPAN
0958/28-6431
Larry H. Kirby
Assoc. Development Scientist
Dow Chemical USA
Texas Operations, B-1605 Building
Freeport, TX 77541
409/238-7721
Thomas Kirkey
Account Manager
Norton Company
2500 E. Ball Road
Anaheim, CA 92806
714/635-4051
John B. Kitto, Jr.
Program Manager
Babcock & Wilcox
Research & Development Division
1562 Beeson Street
Alliance, OH 44601
216/829-7710
A-14
-------
J. S. Klingspor
Senior Engineer
Combustion Engineering
31 Inverness Center Parkway-
Birmingham, AL 35243
205/991-2832
Jeff Klueger
Associate Mechanical Engineer
Los Angeles Dept. of Water & Power
111 N. Hope Street, Room 661
Los Angeles, CA 90051
213/481-4647
Kevin Knill
Research Engineer
Int'l Flame Research Foundation
P.O. Box 10000
1970 CA Ijmuiden
THE NETHERLANDS
31/2510-9364
Cathy L. Knowles
Fibers Department
E.I. du Pont de Nemours & Co., Inc.
Laurel Run Building
P.O. Box 80, 705
Wilmington, DE 19880-0705
302/999-3905
Kris W. Knudsen
Supervising Scientist
Duke Power Company
P.O. Box 33189
Charlotte, NC 28242
704/373-5104
Bernie Koch
Director, Project Development
Consolidation Coal Company
4000 Brownsville Road
Library, PA 15129
412/854-6612
A. Kok
Engineer
N0VEM
P.O. Box 17
6130 AA Sittard
THE NETHERLANDS
4490/95308
Angelos Kokkinos
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2494
Johanna Koppius-Odink
Engineer
EPON
P.O. Box 10087
8000 GB Zwolle
THE NETHERLANDS
038/272900
Richard S. Kostkas
Dir., Special Marketing Projects
The Peoples Natural Gas Company
625 Liberty Avenue
Pittsburgh, PA 15222
412/553-6608
Toshio Koyanagi
SCR Engineer
Mitsubishi Heavy Industry
2 Houston Center
Houston, TX 77010
713/652-9269
Donald E. Krider
Mgr., Thermal De-NOx Sales
Exxon Research & Engineering Co.
P.O. Box 390
Florham Park, NJ 07932
201/765-2339
Hal M. Kruger
Engineer
Centerior Energy
P.O. Box 94661
Cleveland, OH 44101-4661
216/447-2233
H. Kulper
Manager, Fuel & Burner Department
Royal Schelde
P.O. Box 16, 4380AA
Vlissingen, NETHERLANDS
311184 82583
A-15
-------
Robin L. Kump
Environmental Consultant
E. I. DuPont
P.O. Box 1089
Orange, TX 77631-1089
409/886-6239
Alfred Kurzenhauser
Technical Support Supervisor
LFC Power Systems Corporation
Building One, Suite 255
4000 Kruse Way Place
Lake Oswego, OR 97035
503/697-0259
H. K. Kwee
Engineer
Stork Boilers B.V.
P.O. Box 20
7550 GB Hengelo
THE NETHERLANDS
31/74 401857
D. C. Langley
Regional Service Manager
Babcock & Wilcox
7401 W. Mansfield Avenue, #410
Lakewood, CO 80235
303/988-8203
Robert Langthorne
Manager
Canadian Energy Services, Ltd.
401 Salter Street
New Westminster, B.C.
CANADA V3M 5Y1
604/521-3322
Dorothy Lau
Design Specialist
Ontario Hydro
700 University Avenue, H10 F2
Toronto, Ontario M5G 1X6
CANADA
416/592-7306
Dennis Laudal
Research Engineer
Energy & Mineral Research Center
15 North 23rd Street
P.O. Box 8213
Grand Forks, ND 58201-8213
701/777-5138
Nick J. Lavingia
Senior Project Engineer
Chevron U.S.A.
324 W. El Segundo Boulevard
El Segundo, CA 90245
213/615-5784
Peter Lawson
Supervising Engineer
Ontario Hydro
700 University Avenue
Toronto, Ontario
CANADA M5G 1X6
416/592-5393
Albert D. LaRue
Principal Engineer
Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH
216/860-1493
Douglas J. Leadenham
Analyst
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2954
G. K. Lee
Manager, Combustion Research
Energy Resources Canada
555 Booth Street
Ottawa, Ontario K2G 0N3
613/996-3179
George C. Lee
Engineering Specialist
Bechtel National, Inc.
50 Beale Street
San Francisco, CA 94105
415/768-1216
A-16
-------
Yam Y. Lee
Research Specialist
Pyropower Corporation
5120 Shoreham Place
San Diego, CA 92122
619/552-2312
George R. Lester
Senior Research Scientist
Allied Signal
P.O. Box 5016
Des Plaines, IL 60017
312/391-3314
Robert D. Lewis
Principal Engineer
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-5968
F, Y. Liao
Senior Associate Engineer
Mobil R&D
P.O. Box 1026
Princeton, NJ 08540
609/737-4955
Vladimir Lifshits
N/A
Goen Company
1510 Rollins Road
Burlingame, CA 94010
415/697-0440
Sergio Ligasacchi
N/A
ENEL
Via A. Pisano 120
Pisa, ITALY 56100
050/535622
JoAnn S. Lighty
Assistant Professor
University of Utah
2254 MEB
Salt Lake City, UT 84112
801/581-5763
William P. Linak
Project Officer
U.S. EPA
AEERL -- MD-65
Research Triangle Park, NC 27711
919/541-5792
David G. Linz
Mgr., Land and Water Quality Res.
Gas Research Institute
8600 West Bryn Mawr Avenue
Chicago, IL 60631
312/399-8198
Robert A. Lisauskas
Director, Research & Development
Riley Stoker Corporation
Riley Research Center
45 McKeon Road
Worcester, MA 01610
508/792-4801
Steve Londerville
N/A
Coen Company
1510 Rollins Road
Burlingame, CA 94010
415/697-0440
Phillip A. Lowe
Pres ident
Intech, Inc.
11316 Roven Drive
Potomac, MD 20054
301/983-9367
Francis C. Luck
Research Engineer
Rhone-Poulenc France
14 Rue des Gardinoux
93308 Aubervilliers, FRANCE
33/147-396-361
Richard K. Lyon
Senior Scientist
Energy & Environmental Research
P.O. Box 189
Whitehouse, NJ 08888
201/534-5833
A-17
-------
0. L. Maaskant
N/A
Shell International Chemical Co.
P.O. Box 162
2501 AN The Hague
HOLLAND
070/773741
Denis Maftei
Sr. Process Engineer
Environment Ontario
880 Bay Street, 4th Floor
Toronto, Ontario
CANADA M5S 1Z8
416/965-5776
Mehdi Maibodi
Engineer
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
919/541-9100
Jason Makansi
Senior Technical Editor
Power Magazine
11 W. 19th Street, 2nd Floor
New York, NY 10011
212/337-4074
James W. Malone
Sales Engineer
Babcock & Wilcox
3734 Gundry Avenue
Long Beach, CA 90807
213/595-8357
Aaron Mann
N/A
Sierra Pacific Power Company
P.O. Box 10100
Reno, NV 89520
702/689-4011
M. N. Mansour
President
Applied Utility Systems
1140 E. Chestnut Avenue
Santa Ana, CA 92701
714/953-9922
Gerald J. Maringo
Development Engineer
Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44203
216/860-6321
John L. Marion
Consulting Engineer
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-4539
Bob Markoja
Manager, Environmental Services
Sierra Environmental Engineering
3505 Cadillac Avenue, N-3
Costa Mesa, CA 92626
714/432-0330
Gregory A. Martel
Senior Engineer
Northeast Utilities
1866 River Road
Middletown, CT 06457
203/638-3175
Blair Martin
Deputy Director, AEERL
U.S. EPA
AEERL -- MD-60
Research Triangle Park, NC 27711
919/541-7504
Gianni Mascalze
Doctor
Ansaldo Componenti Spa
V.Le Sarca 336
Milano, ITALY 20126
2/6445 2570
Howard Mason
Mgr., Energy Engineering Department
Acurex Corporation
485 Clyde Avenue
Mountain View, CA 94039
415/961-5700
A-l 8
-------
M. Massoudi
Project Manager
Ebasco Services, Inc.
One Parket Plaza, Spear Bldg.
Suite 600
San Francisco, CA 94105
415/777-3000
Mike Maxwell
Chief, Technology Dev. Branch
U.S. EPA
AEERL -- MD-04
Research Triangle Park, NC 27711
919/541-3091
Phil May
Project Director
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
919/541-9100
Michael McCartney
Mgr., Firing Systems Engineering
Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-4677
Michael W. McElroy
Consultant
Electric Power Technologies, Inc.
P.O. Box 307
Menlo Park, CA 94026
415 322-0843
Marilyn Mcllvane
Managing Editor
Mcllvane Company
2970 Maria Avenue
Northbrook, IL 60062
312/272-0010
Robert Mcllvaine
President
Mcllvane Company
2970 Maria Avenue
Northbrook, IL 60062
312/272-0010
William M. McKinney
V.P., New Business Development
United Catalysts, Inc.
P.O. Box 32370
Louisville, KY 40232
502/634-7218
Ron Melton
Section Manager
Battelle Northwest
P.O. Box 999
Richland, WA 99352
509/375-2932
Jim Merriam
Sr. Environmental Licensing Engr.
Jersey Central Power & Light Co.
310 Madison Avenue
Morristown, NJ 07960
201/455-8563
Jean Mevel
Test Manager, Research & Dev.
Stein Industries
1921 Morane Saulnier Avenue
Velizy, Vellacoublay
FRANCE
1/34 65 44 48
James E. Meyers
Chemical Equipment Engineer
Detroit Edison
2000 Second Avenue
Detroit, MI 48226
313/897-0806
Paul G. Mikolaj
Consulting Engineer
Tosco Corporation
Avon Refinery
Martinez, CA 94553
415/228-1880
Thomas Mileus
Sales Manager
Flakt Industres AB
S-351 87 Vaxjo
SWEDEN
01146 470-87000
A-19
-------
Scott A. Miller
Staff Engineer
Dayton Power & Light Company
P.O. Box 468
Aberdeen, OH 45101
513/549-2641 X5127
Kenneth J. Mills
Market Manager
Norton Company
P.O. Box 350
Akron, OH 44309
216/673-5860
Ed Milone
Engineer
Con. Edison Co. of New York
4 Irving Place
New York, NY 10003
212/460-4887
Shigehiro Miyamae
Senior Engineer
Ishikawajima-Harima Heavy Ind.
2-16, Toyosu 3-chome, Koto-ku
Tokyo 135 JAPAN
03/534-4335
Cal Mock
District Manager
Babcock & Wilcox
1990 N. California Boulevard
Walnut Creek, CA 94596
415/947-1100
Hukam C. Mongia
Chief, Combustors R&D
Allison Gas Turbine Div. of GM
P.O. Box 420, Speed Code T14
Indianapolis, IN 46206-0420
317 230-5945
Curtis A. Moore
Counsel
U.S. Senate
Washington, DC 20510
202/224-5762
Dave Moore
Supervisor
Dayton Power & Light Company
Killen Station -- P.O. Box 147
Manchester, OH 45144
513/549-3911
K. A. Moore
N/A
Dykema Engineering
5850 Canoga Avenue, Suite 400
Woodland Hills, CA 91367
818/712-0070
Tom Morasky
Tech. Transfer Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94034
415/855-2468
Akira Mori
Senior Engineer
Hitachi, Ltd.
c/o Hitachi America. Ltd.
950 Elm Avenue, Suite 195
San Bruno, CA 94066-3094
415/872-1902 X8131
Shigeki Morita
Senior Engineer, Kure Works
Babcock-Hitachi K. K.
c/o Hitachi America, Ltd.
950 Elm Avenue, Suite 195
San Bruno, CA 94066-3094
415/872-1902 X8131
Karen D. Morris
Mechanical Engineer
Pacific Gas & Electric
77 Beale Street, Room 2547
San Francisco, CA 94106
415 972-3384
J. K. Mueller
Project Manager
Unocal Corporation
1201 West 5th Street
Los Angeles, CA 90051
213/977-6419
A-20
-------
T. Murataka
Senior Engineer
Babcock-Hitachi K. K.
c/o Hitachi America, Ltd.
950 Elm Avenue, Suite 195
San Bruno, CA 94066-3094
415/872-1902
Hitoshi Murayama
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2777
Paul Musser
Program Manager
U.S. Department of Energy
Germantown, MD 20874
301/353-2846
Lawrence J. Muzio
Vice President
Fossil Energy Research Corporation
23342 C South Pointe
Laguna Hills, CA 92653
714/859-4466
Peter Necker
Prokurist
Neckarwerke AG
Kuferstrasse 2
7300 Esslingen am Neckar
FEDERAL REPUBLIC OF GERMANY
0711/3190-3001
Stefan Negrea
N/A
KRC/Noell GnbH
Alfred, Nobel Strasse 20
D-8700 Wurgburg
WEST GERMANY
N/A
Mike Nelson
Senior Engineer
Southern Company Services
P.O. Box 2625
Birmingham, AL 35202
205/870-6518
Kenneth M. Nichols
Asst. Prof., Engineering & Research
The Institute of Paper Chemistry
P.O. Box 1039
Appleton, WI 54912
414/738-3385
Per R. Nielsen
Senior Manager
Nordic Gas Technology Center
Dr. Neergaardsvej 5A
DK 2970 Hoersholm
DENMARK
45/276 6995
Ralph M. Nigro
Project Engineer
Delmarva Power
800 King Street
Wilmington, DE 19899
302/454-4872
Michael Novak
Engineer
Verbundkraft Elek.
AM H0F 6 A
Vienna A-1011 Wein
AUSTRIA
222/531 13 2590
James H. Nylander
Senior Engineer
San Diego Gas & Electric Co.
P.O. Box 1831
San Diego, CA 92112
619/931-7294
George Offen
Program Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-8942
Yoshiya Ogasawara
Asst. Manager, Int'l Sales Div.
Hitachi, Ltd.
c/o Hitachi America, Ltd.
950 Elm Avenue, Suite 195
San Bruno, CA 94066-3094
415/872-1902 X8131
A-21
-------
Delbert M. Ottmers, Jr.
Vice President
Radian Corporation
P.O. Box 201088
Austin, TX 78720-1088
512/454-4797
Joseph H. Oxley
Program Manager
Battelle
505 King Avenue
Columbus, OH 43201-2693
614/424-7885
Y. S. Pan
Project Coordinator
U.S. Department of Energy
P.O. Box 10940
Pittsburgh, PA 15236
412/892-5727
Andy S. Paranj ape
Senior Project Engineer
Kerr-McGee Corporation
P.O. Box 25861
Oklahoma City, OK 73125
405/270-2906
Rasik Patel
Manager, Thermal Design
KTI Corporation
1333 S. Mayflower Avenue
Monrovia, CA 91016-4099
818/303-4711
Jarl Pedersen
Mgr., Combustion Systems
Burmeister & Wain Energy
23 Teknikerbyen
Virum, DENMARK, 2830
45/285-2100
Craig Penterson
Staff Engineer, R&D
Riley Stoker Corporation
5 Neponset Street
Worcester, MA 01590
508/792-4829
Ralph Perhac
Department Director
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2572
Michael Perlsweig
Flue Gas Cleanup Program Manager
U.S. Department of Energy
Fossil Energy, FE-23
Washington, DC 20545
301/353-4399
David W. Pershing
Dean, College of Engineering
University of Utah
2202 MEB
Salt Lake City, UT 84112
801/581-5057
Robert D. Petersen
Senior Mechanical Engineer
Burns & McDonnell
4800 East 63rd Street
Kansas City, MO 64141
816/333-4375
Ellen Petrill
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-8989
James D. Phillip
Plant Chemist
Tenneco Minerals Company
P.O. Box 1167
Green River, WY 82935
307/872-6519
Larry A. Pierson
Project Manager
Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44203
216/860-1103
A-22
-------
Everett L. Plyler
Dir., Combustion & Indoor Air Div.
U.S. EPA
AEERL -- MD-54
Research Triangle Park, NC 27711
919/541-2918
Gary Poe
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-8969
John H. Pohl
Dir., Research & Development
Energy Systems Associates
15991 Red Hill
Tustin, CA 92680
714/259-9520
David Price
Project Engineer
Unocal Corporation
P.O. Box 758
Wilmington, CA 90748
213/513-7600
W. L. Prins
Manager, Process Technology
Tests Engineers & Consultants
P.O. Box 10.000
1970 CA Ijmuiden
THE NETHERLANDS
+312510 98603/92209
Walter Puille
Technology Transfer Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2470
Suresh K. Punatar
Engineering Supervisor
Bechtel Power Corporation
50 Beale Street
San Francisco, CA 94119
415/768-0495
Brian Quil
Mechanical Engineer
Naval Energy & Environmental Suppt.
NEESA 11A
Port Hueneme, CA 93043
805/982-4748
Les Radak
Senior Research Engineer
Southern California Edison
2244 Walnut Grove Avenue
Rosemead, CA 91770
818/302-9746
Paul Radcliffe
Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2720
James M. Rames
Sr. Mechanical Engineer
Pacific Gas & Electric
77 Beale Street, Room 2553
San Francisco, CA 94106
415 972-2735
Jay Ratafia-Brown
Dir., Fossil Energy/Env. Projects
SAIC
1710 Goodridge Drive
P.O. Box 1303
McLean, VA 22102
703/448-6343
James L. Reese
Consulting Engineer
Applied Utility Systems
1140 E. Chestnut Avenue
Santa Ana, CA 92701
714/953-9922
Joseph N. Reeves
Research Manager
Southern California Edison
2244 Walnut Grove Avenue
Rosemead, CA 91770
818/302-9515
A-23
-------
Amir Rehmat
Mgr., Utility Gas Development
Institute of Gas Technology
3424 South State Street
Chicago, IL 60616-3896
312/567-5899
Ernst Ringle
Delegate
Noell
199 Wells Avenue
Newton, MA 02159
617/332-6510
Alan P. Risch
Manager, Technical Marketing
UOP
39 Old Ridgebury Road
Danbury, CT 06817
208/794-5542
R. C. Rittenhouse
Senior Editor
Power Engineering Magazine
1250 S. Grove Avenue, Suite 302
Barrington, IL 60010
312/382-2450
Roger Robb
Senior Engineer
Northern Indiana Public Service Co.
5265 Hohman Avenue
Hammond, IN 46320
219/853-5821
Dick Robertson
Manager, Air Quality
TU Electric
400 North Olive, LB-81
Dallas, TX 75201
214/812-8416
Chris P. Robie
Process Engineer
United Engineers & Constructors
700 South Ash Street
P.O. Box 5888
Denver, CO 80217
303/692-2803
Tracy M. Robinson
Marketing Programs Manager
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
William Rogers
Sales Engineer
Engelhard Corporation
2000 Powell Street
Emeryville, CA 94608
415/596-1703
Bill Rovesti
Senior Project Manager
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2519
Charlie Rowden
Lead Process Engineer
Fluor Daniel
3333 Michelson Drive
Irvine, CA 92730
714/975-3047
Richard E. Rozelle
Project Manager
Dow Chemical USA
Texas Operations, B-1605 Bldg.
Freeport, TX 77541
409/238-9846
Hans-Georg Rych
Electric Power Systems Division
EVN
Johann-Steinbock-Strasse 1
A-2344 Maria Enzersdorf am Gebirge
AUSTRIA
02236/83 611 526
N. C. Samish
Staff Research Engineer
Shell Development Company
P.O. Box 1380
Houston, TX 77251
713/493-7944
A-24
-------
Henry Sanematsu
Mechanical Engineer
Los Angeles Dept. of Water & Power
111 N. Hope Street
Los Angeles, CA 90012
213/481-4906
Charles E. Schlittler
Program Manager
Kerr-McGee Corporation
P.O. Box 25861
Oklahoma City, OK 73125
405/270-3188
Herbert Schuster
Head, Combustion R&D Dept.
Deutsche Babcock Werke AG
Duisburger Str. 375
D-4200 Oberhausen 1
FEDERAL REPUBLIC OF GERMANY
0208/833-2853
Charlie Sedman
Project Officer
U.S. EPA
AEERL -- MD-62
Research Triangle Park, NC 27711
919/541-7700
Noriyuki Shimizu
Assistant Manager
Electric Power Development Co.
15-1, Ginza, 6-chome, Chuo-ku
Tokyo, JAPAN 104
3/546-9403
Gary H. Shiomoto
Engineer
Fossil Energy Research Corporation
23342 C South Pointe
Laguna Hills, CA 92653
714/859-4466
Dale E. Shore
Manager, Western Region
Energy Systems Associates
15991 Red Hill Avenue, Suite 110
Tustin, CA 92680
714/259-9520
Stefan P. Shoup
Senior Staff Engineer
Inland Steel Company
3210 Watling Street
East Chicago, IL 46312
219/399-5733
George Shulof
President & CEO
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
Sandy Simko
Project Manager
Radian Corporation
P.O. Box 201088
Austin, TX 78720-1088
512/454-4797 X5301
Robert E. Sistek
Environmental Met. Engineer
LTV Steel Company
3100 E. 45th Street
Cleveland, OH 44127
216/429-6477
Lowell L. Smith
Vice President
ETEC
2091 Business Center Drive #100
Irvine, CA 92715
714/833-2523
Richard C. Smith
Supervising Engineer
Union Electric Company
P.O. Box 149, MC-450
St. Louis, M0 63166
314/554-3529
Douglas Smoot
Dean, College of Engrng. & Tech.
Brigham Young University
270 Clyde Building
Provo, UT 84602
801/378-4327
A-25
-------
Gerry C. Snow
Senior Engineer
Radian Corporation
P.O. Box 13000
Research Triangle Park, NC 27709
919/541-9100
David W. South
Economist
Argonne National Laboratory
370 L'Enfant Promenade, S.W.
Suite 702
Washington, DC 20024
202/586-4758
Ruggero Spagliarisi
Engineer
Aeru -- Milano
Via S. Sisto
8-20123
Milano, ITALY
02/808966
Barry Speronello
Principal Scientist
Engelhard Corporation
Menlo Park
Edison, NJ 08818
201/321-5155
Dan Spisak
Project Engineer
Bechtel National,
50 Beale Street
San Francisco, CA
415/768-1645
Richard T. Squires
Research Officer
CEGB
Marchwood Labs
Marchwood, Southampton
UNITED KINGDOM
0703/663232
Ravi K. Srivastava
Project Engineer
Acurex Corporation
P.O. Box 13109
Research Triangle Park, NC 27709
919/541-2692
Susan J. Stamey
Mechanical Engineer
Tennessee Valley Authority
MR3S92B
Chattanooga, TN 37402
615/751-5324
Mark T. Staniulis
Senior Staff Chemist
UOP
Old Sawmill River Road
Tarrytown, NY 10591
914/789-3865
David L. Stavenger
N/A
Dow Chemical USA
2020 Willard H. Dow Center
Midland, MI 48674
517/636-5236
Harry Stecyk
Sr. Engineering Specialist
ESSO Resources Canada, Ltd.
237 Fourth Avenue, S.W.
Calgary, Alberta
CANADA T2P 0H6
403/237-2721
Harvey Stenger
Professor
Lehigh University
Building A, MTC
Dept. of Chemical Engrng.
Bethlehem, PA 18015
215/758-4791
Donald J. Stepson
Marketing Director
Rolls-Royce, Inc.
16800 Greenspoint Park Drive
Suite 155S
Houston, TX 77060
713/875-2100
Dick D. Stern
Chief, Technology Applications Br.
U.S. EPA
AEERL -- MD-63
Research Triangle Park, NC 27711
919/541-2973
Inc.
94105
A-26
-------
Jim E. Stresewski
Engineer
Black & Veatch
1500 Meadow Lake Parkway
Kansas City, MO 64114
913/339-2000
Y. P. Su
Principal Environmental Engineer
Stone & Webster Engineering Corp.
330 Barker Cypress
Houston, TX 77094
713/492-4334
Keith Summons
Senior Design Specialist
Ontario Hydro
700 University Avenue, H13 C10
Toronto, Ontario
CANADA M5G 1X6
416/592-5935
Kohei Suyama
Manager
Mitsubishi Heavy Ind. America
c/o Combustion Engineering
1000 Prospect Hill Road
Windsor, CT 06095
203/285-2491
Markus Tahtinen
Research Scientist
Technical Research Ctr. of Finland
P.O. Box 169
SF-00181 Helsinki
FINLAND
358/0 648931
Katherine M. Tait
Engineer
MONENCO
400 Monenco Place, 801-6th Ave. SW
Calgary/Alta
CANADA T3C 1N1
403/298-4161
Yojiro Takahashi
Manager
Nippon Shokubai America, Inc.
101 East 52nd Street, 14th Floor
New York, NY 10022
212/759-7890
Harry S. Tang
Senior Research Engineer
Shell Development Company
P.O. Box 1380
Houston, TX 77251-1380
713/493-8424
Donald P. Teixeira
Technical Manager Fossil R&D
Pacific Gas & Electric
3400 Crow Canyon Road
San Ramon, CA 94583
415/866-5531
Robert Telesz
Marketing Manager
Babcock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44230
216/860-1103
Preston E. Tempero
Plant Manager
KPL Gas Service
Lawrence Energy Center
P.O. Box 249
Lawrence, KS 66044
913/843-8118
Jean Thomas Rose
Engineering Manager
Betz Industrial
Somerton Road
Trevose, PA 19047
215/355-3300 X2365
T. R. Thomas
N/A
Westinghouse Idaho Nuclear Company
P.O. Box 4000
Idaho Falls, ID 83403
208/526-3086
A-27
-------
Bruce Thompson
District Manager
Nalco Chemical Company
4310 North 75th Street, Suite A
Scottsdale, AZ 85251
602/996-2238
Paul F. Thompson
President
Tenerex Corporation
303 Laurel
Friendswood, TX 77546
713/482-5801
Richard E. Thompson
President
Fossil Energy Research Corporation
23342 C South Pointe
Laguna Hills, CA 92653
714/859-4466
Richard S. Topor
Int'l Regional Manager, Licensing
Combustion Engineering
1000 Prospect Hill Road, 7110-1923
Windsor, CT 06095
203/285-9934
Maj ed A. Toqan
Program Manager, CRF
Massachusetts Inst, of Technology
60 Vassar Street, Bldg. 31-261
Cambridge, MA 02139
617/253-1721
Paul J. Torpey
Executive Director
Empire State Elec. Energy Res. Corp
1155 Avenue of the Americas
New York, NY 10036
212/302-1212
Ian Torrens
Department Director
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, CA 94304
415/855-2422
Sanh V. Tram
Senior Service Engineer
Babcock & Wilcox
12552 Tours Avenue
Garden Grove, CA
714/554-3844
Jay R. Turner
Research Assistant
Washington University
Chemical Engineering Department
Campus Box 1198, One Brookings Dr.
St. Louis, MO 63130
314-889-6042
James M. Valentine
V.P. Marketing & Business Dev.
Fuel Tech
61 Taylor Reed Place
Stamford, CT 06906
203/359-1320
Don Van Buren
Field Engineering
Bay Area Air Quality Mgmnt. Dist.
939 Ellis Street
San Francisco, CA 94109
415/771-6000 X165
Simon Vanderheijden
Chief Chemist
Saskatchewan Power
2025 Victoria Avenue
Regina, Saskatchewan
CANADA S4P 0S1
306/566-3073
Galliano Variali
Professor
University of Rome, Italy
Via Eudossiana 18
Rome, ITALY 00184
396/4687321
Joel Vatsky
Dir., Combustion & Env. Systems
Foster Wheeler Energy Corporation
Perryville Corporate Park
Clinton, NJ 08809-4000
201/730-5450
A-28
-------
Gary R. Veerkamp
Sr. Mechanical Engineer
Pacific Gas & Electric
77 Beale Street, Room 2555
San Francisco, CA 94106
415/972-1576
Jan L. Vernon
Head, Environment Group
IEA Coal Research
14-15 Lower Grosvenor Place
London SW1W OEX ENGLAND
01/828-4661
John Verurink
Senior Combustion Engineer
Stork Boilers B.V.
P.O. Box 20
7550 BD, Hengelo
THE NETHERLANDS
074/401370
Denise M. Viola
Sales Manager
Engelhard Corporation
Menlo Park CN28
Edison, NJ 08818
201/321-5039
J. Von Bergmann
General Manager
Fuel Tech GmbH
Frankfurter Strasse 33-35
Eschborn 6236
FEDERAL REPUBLIC OF GERMANY
49/6196-48613
Paul Wagner
Project Engineer
Delmarva Power
1-95 & Route 273
P.O. Box 9239
Newark, DE 19714
302/454-4844
Ali A. Wali
Planning Analyst
Wisconsin Public Service Comm.
4802 Sheboygan Avenue
P.O. Box 7854
Madison, WI 53707
608/267-3595
Mats A. Wallin
M. Sc.
University of Lund
Dept. Chera. Engineering Box 124
S-221 00 Lund, SWEDEN
46 46-108244
Ed Warchol
Engineering Specialist
ESSO Resources Canada, Ltd.
237 Fourth Avenue, S.W.
Calgary, Alberta
CANADA T2P 0N6
403/237-2721
Wendell Warnacut
Mech. Engineer
Tennessee Valley Authority
1101 Market St. MS-3S103C
Missionary Ridge Place
Chattanooga, TN 37402
615/751-6708
Andy M. Warren
Resident Service Engineer
Babcock & Wilcox
1990 N. California Blvd., Suite 400
Walnut Creek, CA 94596
415/947-1100
Larry Waterland
Program Manager
Acurex Corporation
485 Clyde Avenue
Mountain View, CA 94039
415/961-5700
A-29
-------
Greg Weber
Research Supervisor
Energy & Mineral Research Center
15 North 23rd Street
P.O. Box 8213
Grand Forks, ND 58202-8213
701/777-5222
W. L. Weier
N/A
NV KEMA
P.O. Box 9035
6800 ET Arnhem
THE NETHERLANDS
085/566023
Roy J. Weiskircher
Sr. Env. Control Engineer
USS a Div. of USX Corporation
4000 Tech Center Drive
Monroeville, PA 15146
412/825-2824
Jost O.L. Wendt
Professor
University of Arizona
Dept. of Chemical Engineering
Building 11
Tucson, AZ 85721
602/621-6050
Arthur W. Wesa
Principal Engineer
The Detroit Edison Company
2000 Second Avenue
Detroit, MI 48226
313/237-5117
Phil Westin
Sr. Environmental Engineer
Arco Products Company
1900 W. Crescent Avenue
Anaheim, CA 92803
714/491-6854
Ed Wiken
Chief
Environment Canada
2067 Fairbanks Avenue
Ottawa, Ontario
CANADA
819/997-1090
Dale V. Wilhelm
Environmental Engineer
Tennessee Valley Authority
Environmental Quality Staff
Knoxville, TN 37902
615/632-6695
Ronald Wilkniss
Staff Engineer
Mobil Oil Corporation
3700 W. 190th Street
Torrance, CA 90509
213/212-4587
Steve M. Wilson
Research Engineer
Southern Company Services
P.O. Box 2625
Birmingham, AL 35202
205/877-7835
Phil Winegar
Senior Engineer
New York Power Authority
1633 Broadway
New York, NY 10019
212/468-6729
Johan Witkamp
Engineer
NV KEMA
Utzechtseweg 310
6900 ET Arnhem
THE NETHERLANDS
085/563625
Evan W. Wong
Engineer
Air Resources Board
P.O. Box 2815
Sacramento, CA 95812
916/445-5980
A-30
-------
Bernhard J. Wood
Senior Chemist
SRI International
333 Ravenswood Avenue
Menlo Park, CA 94025
415/859-2305
Ralph T. Yang
Professor
State University of New York
Dept. of Chemical Engineering
Buffalo, NY 14260
716/636-2519
Jerry L. Yee
Senior Mechanical Engineer
Hawaiian Electric Company
P.O. Box 2750
Honolulu, HI 96840
808/543-7708
Nancy Yee
Field Engineering
Bay Area Air Quality Mgmnt. Dist.
939 Ellis Street
San Francisco, CA 94109
415/771-6000
Cherif F. Youssef
Research Project manager
Southern California Gas Company
Box 3249 Terminal Annex, M.L. 73ID
Los Angeles, CA 90051
818/307-2695
Kenneth P. Zak
Marketing Research & Business Dev.
W. R. Grace & Company
7379 Route 32
Columbia, MD 21044
301/531-4383
Jiansheng Zhao
M. Sc.
University of British Columbia
Chemical Engineering Department
2216 Mail Mall
Vancouver, B.C. V6T lWSmbia
604/224-8579
Melvin L. Zwillenberg
Principal Staff Engineer, R&D
Public Service Electric & Gas Co.
MC-T16H, 80 Park Plaza
Newark, NJ 07101
201/430-6636
A- 31
-------
|