EPA-600 '7-90-002
January 1990
A LOW NOX STRATEGY FOR
COMBUSTING HIGH
NITROGEN CONTENT FUELS
By:
Ravi K. Srivastava
Acurex Corporation
Environmental Systems Division
P.O. Box 13109
Research Triangle Park, NC 27709
EPA Contract No. 68-02-3988
EPA Project Officer: James A. Mulholland
U. S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
-------
TECHNICAL REPORT DATA
(Phase read Instructions on the reverse before completingj
1. REPORT NO
EPA-600/7-90-002
2
3. RECIPIENT'S ACCESSION NO.
«. TITLE AND SUBTITLE
A Low NO Strategy for Combusting High Nitrogen
5. REPORT DATE
January 1990
Content Fuels
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS!
8. PERFORMING ORGANIZATION REPORT NO.
Ravi K. Srivastava
8064.210
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Corporation
P. C. Box 13109
Research Triangle Park, North Carolina
10. PROGRAM ELEMENT NO.
27709
11. CONTRACT/GRANT NO.
68-02-3988
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 PERIOD COVERED
Final; 11/84 - 7/87
14. SPONSORING AGENCY CODE
EPA/600/13
is.supplementary notes ^eERL project officer J. A. Mulholland is no Ion
Agency^ For details, contact William P. Linak, Mail Drop 65, 919/
ger with the
541-5792.
16. abstract 11^ report gives results of an evaluation of a multistaged combustion bur-
ner (designed for in-furnace NOx control and high combustion efficiency) for high ni-
trogen content fuel and waste incineration application in a 1. 0 MW package boiler,
simulator. A low NOx precombustion chamber burner has been reduced in size by
about a factor of two (from 600 to 250 ms first-stage residence time) and coupled
with: (l) air staging, resulting in a three-stage configuration; and (2) natural gas
fuel staging, yielding up to four stoichiometric zones. Natural gas, doped with am-
monia to yield a 5.8/0 fuel nitrogen content, and distillate fuel oil, doped with pyri-
dine to yield a 2.0% fuel nitrogen content, were used to simulate high nitrogen con-
tent fuel/waste mixtures. "Minimum nitric oxide (NO) emission levels of 160 and 110
ppm (corrected to 0% oxygen, C2) were achieved for the natural gas and fuel oil
tests, respectively, corresponding to about 85% reduction in NOx emissions com-
pared to uncontrolled emissions from a conventional burner mounted on a 0. 7 MW
commercial package boiler-.^Under the conditions tested, net chemical destruction
of NC via reburning does not seem to be evident. This may be due to the existence
of rather low primary NO concentrations before the application of reburning. How-
ever, a beneficial dilution caused by reburning may provide lower NC emissions=
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. 1DENTIFIE RS/OPEN ENDED TERMS
c. COS ATI Kield/Group
Pollution Wastes
Combustion Simulation
Incinerators
Natural Gas
Fuel Oil
Nitrogen Oxides
Pollution Control
Stationary Sources
Staged Combustion
Reburning
13B 14 G
21B
2 ID
07B
13. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
71
Release to Public
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73J
I
i
-------
ABSTRACT
A multistaged combustion burner designed for in-furnace NOx control and high combustion
efficiency has been evaluated for high nitrogen content fuel and waste incineration application in a
1.0 MW package boiler simulator. A low NOx precombustion chamber burner has been reduced in size
by approximately a lactor of two (from 600-250 ms first-stage residence time) and coupled with 1) air
staging, resulting in a three-stage configuration, and 2) natural gas fuel staging, yielding up to four
stoichiometric zones. Natural gas, doped with ammonia to yield a 5.8 percent fuel nitrogen content, and
distillate fuel oil, doped with pyridine to yield a 2 percent fuel nitrogen content, were used to simulate high
nitrogen content fuel/waste mixtures. Minimum nitric oxide (NO) emission levels of 160 ppm 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 commercial
package boiler. Under the conditions tested, net chemical destruction of NO via rebuming does not seem
to be evident. This may be due to the existence of rather low primary NO concentrations before the
application of rebuming. However, a beneficial dilution caused by reburning, as applied here, may
provide lower NO emissions (on a ppm or lb/10® Btu basis) along with no loss in heat output.
ii
-------
TABLE OF CONTENTS
Section Page
Abstract ii'
List ot Figures iv
List of Tables . ... v:
1 introduction 1
2 Experimental Facility 6
2.1 Precombustion Chamber Burner 6
2.2 Firetube Package Boiler Simulator (PBS) 9
2.3 Combustion Air/Fuel/Dopant Flow Systems 14
2.4 Burner Management/Safety System 19
2.5 Stack Gas Speciation 22
2.6 Temperature Measurements Using A Suction Pyrometer 24
3 Proof-of-Concept Tests 28
3.1 Background 28
3.2 Experiment 30
3.3 Apparatus and Methods 30
3.4 Approach 31
3.5 Baselien Test Results -31
3.6 Rebuming Characterization Test Results 36
3.7 Preliminary Proof-of-Concept Test Results 39
3.8 Discussion and Conclusions 39
4 New Burner Tests 43
4.1 Introduction 43
4.2 Experimental Methods 43
4.3 Burner Baseline Performance 45
4.3.1 Excess Air Variation 48
4.3.2 Load Variation 48
4.3.3 Fuel Nitrogen Variation 48
4.4 Air Staged NOx Controls 52
4.5 Fuel Staged NOx Controls 54
4.5.1 Burner Fuel Staging 54
4.5.2 Boiler Fuel Staging 56
4.6 Discussion and Conclusions 58
5 Quality Control Evaluation Report 61
5.1 Introduction 61
5.2 Discussion of Data Quality 61
6 References 64
iii"
-------
LIST OF FIGURES
Figure Page
2-1 Horizontal low NOx precombustion chamber burner 8
2-2 Burner/boiler transition section 10
2-3 Reduced size low NOx precombustion chamber burner 11
2-4 Vertical, modular, low NOx precombustion chamber burner 12
2-5 Package boiler simulator 13
2-6 Combustion air supply system diagram 15
2-7 Natural gas supply system diagram 16
2-8 Fuel oil supply system diagram 18
2-9 Ammonia supply system diagram 20
2-10 Burner management system logic diagram 21
2-11 Continuous emissions monitoring system (CEMS) diagram 23
2-12 Suction pyrometer design 26
2-13 Suction pyrometer calibration curve 27
3-1 The reburning process 29
3-2 Pilot-scale research facility 32
3-3 Inlluence of first-stage residence time on NO emission for short and long preburner
configurations 33
3-4 Inlluence of fuel nitrogen content on minimum NO emission for short and long preburner
configurations 35
3-5 NO reduction for natural gas reburning application .37
3-6 Contour map of optimum NO reduction by reburning 38
3-7 NO emissions as a function of reburn zone stoichiometry for various reburn fuel injection
locations 40
3-8 NO emissions as a function of reburn zone stoichiometry for various burnout air injection
locations 41
4-1 Pilot-scale combustion research facility. The package boiler simulator has been fitted with a
precombustion chamber burner and air and fuel staging ports 44
4-2 Effect of burner stoichiometry. Shown are results from tests firing a 2.0 percent and
5.8 percent nitrogen gas fuel and a 2 percent nitrogen distillate fuel mixture .46
4-3 Burner temperatures. Shown are measurements firing the low NOx burner on gas and
oil fuels 47
4-4 Excess air variation results. Shown are 2.0 percent and 5.8 percent nitrogen gas and
2 percent nitrogen oil results 49
4-5 Load variation results. Shown are results with a 2.0 percent and 5.8 percent nitrogen gas
fuel and a 2 percent nitrogen oil mixture 50
4-6 Fuel nitrogen variation results. The dashed lines represent data from the North
American (NA) boiler tests 51
4-7 Air staging results. Shown are results from firing a 5.8 percent nitrogen gas fuel and a
2 percent nitrogen oil mixture 53
4-8 Burner fuel staging results. Shown are results from firing a 2.0 percent and 5.8 percent
nitrogen gas fuel and a 2.0 percent nitrogen oil mixture 55
4-9 Boiler Fuel staging results. Shown are results from firing a 5.8 percent nitrogen gas fuel
and a 2.0 percent nitrogen oil mixture 57
iv
-------
LIST OF TABLES
Table Page
1-1 Technical Presentations and Papers 4
2-1 Fuel Analyses 17
2-2 Gaseous Emission Monitor Specifications 25
4-1 Summary of Results 60
4-2 Dilution Corrections 60
5-1 Data Quality Summary 62
-------
SECTION 1
INTRODUCTION
The family of nitrogen oxide compounds, including nitric oxide (NO) and nitrogen dioxide (N02),
is generally referred to as "NOx" These NO/ species are formed during the combustion of coal, oil, and
natural gas by the reduction and oxidation of molecular nitrogen (N2) and nitrogen contained in the fuel.
NO2 is a poisonous gas that the United States Environmental Protection Agency (EPA) has designated
as a criteria pollutant because of its harmful effects to human health.1 In addition, NOx emissions are
known to contribute to the formation of photochemical oxidants and are precursors, along with sulfur
oxides (SOx), of acid precipitation. Two more areas of concern are emerging regarding NOx levels in the
atmosphere. First, forest damage as a result of acid precipitation, reported to be extensive in the Federal
Republic of Germany,2 has been linked with increasing NOx levels. Second, increasing levels of
atmospheric nitrous oxide (N2O) have been measured, levels that are predicted to contribute to both a
decline in the abundance of stratospheric ozone and an increase in climatic warming.3 studies of N2O
and NO concentration in experimental flames and in flue gases indicate that a correlation may exist
between these two gases formed in combustion processes.
The EPA estimates that about 20 million tons (18,000 Gg) of NOx are emitted annually from
stationary and mobile sources in the United States. Unlike SOx emissions, NOx emissions are
increasing.^ Coal- and oil-fired utility and industrial boilers account for over half of these NOx emissions.
Only 15 percent of the stationary NOx sources are regulated by EPA's New Source Performance
Standards (NSPS); the remainder must be addressed with retrofit technologies if significant NOx emission
reduction is to be realized. Another NOx control problem is posed by the potential of incinerating high
nitrogen content wastes in industrial boilers. While incineration of these materials would not constitute a
significant increase in the overall national NOx emission level, individual plant emissions may be sufficient
to cause a local NOx problem that would prevent governmental permitting of on-site incineration. As
-I-
-------
thermal destruction is an attractive alternative to landfill storage of wastes, the need continues for
developing high efficiency, low NOx combustion technologies.
Most NOx in stack combustion gas is NO. Much is known about the mechanisms of NO
formation in flames, both from molecular nitrogen (source of thermal NO) and from fuel-bound nitrogen
(source of fuel NO). Thermal NO can be reduced by decreasing peak flame temperatures. Fuel NO is
very sensitive to reactant stoichiometry and fuel-rich conditions promote N2 formation over NO formation.
Laboratory studies and field test data have established the importance of fuel NO to the total emission of
NOx from residual fuel oil and coal flames. Therefore, minimizing NOx formation in flames typically
involves controlling air and fuel mixing rates to create fuel-rich reducing zones and extracting heat to
reduce final oxidation temperatures.
To avoid the need for costly post-combustion NOx removal, several in-furnace NOx control
strategies have been developed and applied to boilers. These include reduced air preheat, load
reduction, low excess air, flue gas recirculation, overfire air, deep air staging, fuel staging (or reburning),
and various iow-NOx burner systems. While NOx emissions can be reduced by 20-80 percent using
these technologies, from uncontrolled levels exceeding 1,000 ppm for some high nitrogen content coals,
the application of these combustion modifications can reduce combustion efficiency and increase sooting
and slagging in the boiler. These problems are of particular concern in the boiler cofiring of fuels and
wastes where high waste destruction efficiencies and minimal formation of other incomplete combustion
products are of paramount importance. Furthermore, practical constraints, such as burner and boiler
sizes, limit the effectiveness of NOx control by combustion modification.
EPA is currently involved in the development and field demonstration of two evolving NOx control
technologies: the precombustion chamber burner and reburning (fuel staging). These combustion
modification strategies provide alternatives to expensive post-combustion NOx removal technologies,
such as selective catalytic reduction which is being utilized extensively in Japan and West Germany, for
achieving low NOx emissions when firing high nitrogen content fuels or incinerating highly nitrated
wastes. The goal of this work was to utilize the precombustion chamber burner and air or air/fuel staging
concepts to develop a burner that is practical for both new and retrofit installations and is capable of
2
-------
burning high nitrogen fuel/waste streams with low NOx emissions and high combustion efficiency.
Specifically, a NOx emission of less than 0.2 lb (as N02)/106 Btu (or approximately 175 ppm NOx,
measured dry at 0 percent O2) for firing gaseous and liquid fuels doped with up to 5 percent nitrogen (by
weight) was targeted (1 lb/10® Btu = 0.43 kg/GJ).
This study was carried out in three phases. The first phase of this study dealt with a fundamental
exploration of post-flame combustion technology, known as rebuming. In this phase the fundamentals of
reburning and its suitability to combustion applications were studied in detail. The findings of this study,
which lasted from March 1983 to October 1984, are documented by Mulholland et al.5
The burner used during the first phase of the study was a low NOx precombustion chamber
burner, designed and fabricated under an EPA contract by Energy and Environmental Research
Corporation (EERC), Irvine, California. The burner proved to be a useful research tool, though its
practicality was limited because of its size and cost. As a follow-on of the reburning work, the burner was
reduced in size from 600 ms to 350 ms first-stage residence time. Subsequent proof-of-concept tests
involving the reduced size precombustion chamber low NOx burner and air or air/fuel staging were carried
out. The proof-of-concept tests in this second phase helped to generate the experimental matrix for the
third phase.
During the third phase of the study a vertical downfired combustor was designed, fabricated, and
installed. It was of a modular design to allow residence time variations, and was capable of firing
gaseous or liquid fuels. It had ports for detailed samplings and variable fuels and air injection
locations/methods. Parametric tests were carried out using this new burner to rigorously test and prove
the concepts generated in the second phase.
This report covers the project activities in the period November 1984 to July 1987. Much of the
information in this report is documented in various technical presentations and papers, which are
tabulated in Table 1-1.5
3
-------
TABLE 1-1. TECHNICAL PRESENTATIONS AND PAPERS
Period Description
April 83-April 84 "Application of Reburning for NOx Control to a Firetube Package Boiler,"
J.A. Mulholland and W.S. Lanier, Journal of Engineering lor Gas Turbines
and Power. Vol. 107, July 1985. Presented at the Joint Power Generation
Conference, Toronto, Ontario, Canada. Sept. 30-Oct. 4,1984
May 84-Jan 85 "The Effect of Fuel Nitrogen in Reburning Application to a Firetube Package
Reboiler," J.A. Mulholland and R.E. Hall, 1985 Joint Symposium on
Stationary Combustion NOx Control, Boston, Massachusetts, May 6-9,1985
May 84-Jan 85 "Fuel Oil Reburning Application for NOx Control to Firetube Package
Boilers," J.A. Mulholland and R.E. Hall, Journal of Engineering for Gas
Turbines and Power. Vol. 109, pp. 207-214 (April 1987). Presented at the
Joint Power Generation Conference, Milwaukee, Wisconsin,
Oct. 20-24, 1985.
April 83-April 84 "Reburning Thermal and Chemical Processes in a Two-Dimensional
Pilot-Scale System," W.S. Lanier, J.A. Mulholland, and J.T. Beard,
Twenty-First Symposium (International) on Combustion, Munich, West
Germany, Aug. 3-8,1986
April 83-Aug 85 Reburning Application to Firetube Package Boilers. J.A. Mulholland. E.E.
Stephenson, C. Pendergraph and J.V. Ryan, EPA Final Report,
EPA-600/7-87-011 (NTIS PB87-177515), March 1987
Nov 85-May 86 "A Multistaged Burner Design for In-Furnace NOx Control," J.A. Mulholland
and R.K. Srivastava. Presented at the Joint Power Generation Conference,
Portland, Oregon, Oct. 19-23,1986
June 86-Jan 87 "Pilot-Scale Tests of a Multistaged Burner Designed for Low NOx Emission
and High Combustion Efficiency," J.A. Mulholland and R.K. Srivastava.
Presented at EPA/EPRI Joint NOx Symposium, March 23-26,1987, New
Orleans, Louisiana.
June 86-Jan 87 "A Multistaged Approach for In-Furnace NOx Control," R.K. Srivastava and
J.A. Mulholland. Presented at the AlChE Spring 1987 National Meeting,
Houston, Texas, March 1987.
June 86-Jan 87 "Low NOx, High Efficiency Multistaged Burner: Gaseous Fuel Results,"
R.K. Srivastava and J.A. Mulholland, Environmental Progress. Vol. 7, No. 1,
pp. 63-70 (April 1987).
June 86-Mar 87 "Low NOx, High Efficiency Multistaged Burner: Fuel Oil Results," J.A.
Mulholland and R.K. Srivastava. Presented at the 80th Air Pollution Control
Association Annual Meeting and Exhibition, New York, NY, June 21-26,
1987, and submitted for publication in Journal of the Air Pollution Control
Association.
(continued)
4
-------
TABLE 1-1. TECHNICAL PRESENTATIONS AND PAPERS (concluded)
Period Description
Presentations
Dec 6 and 7,1983 EPA/ASME/APCA Technical Information Exchange, Research Triangle
Park, NC
Jan 25 and 26,1984 EPA Reburning Review and Coordination Panel Meeting, Salt Lake
City, Utah
Mar 28,1985 EPA Reburning Review and Coordination Panel Meeting, Salt Lake
City, Utah
5
-------
SECTION 2
EXPERIMENTAL FACILITY
The major components of Ihe pilot-scale experimental facility used in this study consisted of a
precombustion chamber burner; a package boiler simulator (PBS); air/fuel/dopant flow systems; a burner
management/safety system; and combustion gas temperature and sample measurement systems.
These components, located in Wing G High Bay area of EPA/ERC, are now described in detail.
2.1 PRECOMBUSTION CHAMBER BURNER
The precorrtbustion chamber burner uses a staged combustion technology capable of achieving a
NOx emission of less than 0.1 lb (as NO2)/106 Btu (or approximately 90 ppm NOx), even with high
nitrogen content fuel firing.6 it consists of a primary air and fuel injection system, a large refractory wall
precombustion chamber, and a secondary air injection section. Fuel and primary air are injected, for
rapid mixing, into the nearly adiabatic precombustion chamber to effect a first-stage stoichiometry
between 0.6 and 0.8. A residence time between 0.6 and 1.0 s allows maximum reduction and conversion
of fuel nitrogen species to N2 in the fuel-rich precombustion chamber. First-stage combustion gas
products exit the burner through a convergent section that minimizes both radiative heat loss to the boiler
and backmixing of the secondary air. The transition section between the burner and boiler is
water-cooled to reduce combustion gas temperatures before final air addition.
The precombustion chamber burner has been tested on a full-scale (16 MW) crude-oil-fired
steam generator used for thermally enhanced oil recovery (TEOR) in Kern County, CA.7 A 30-day
continuous monitoring test demonstrated the burner's ability to maintain a nominal NOx emission of
70 ppm and a high combustion efficiency. During burner optimization testing, NOx, CO, and smoke
emissions were measured over a range of first-stage stoichiometries. These data indicate the sensitivity
of NOx emission to first-stage stoichiometry and the good hydrocarbon burnout characteristics of the
6
-------
burner. Earlier pilot-scale tests firing nitrogen-free fuels suggest that the minimum NOx level of about
50 ppm can be attributed to second-stage thermal NOx formation.
A test designed to demonstrate the potential use of this burner for nitrogenous waste incineration
was performed at EPA's Air and Energy Engineering Research Laboratory. A pilot-scale (0.6 MW)
precombustion chamber burner was used to incinerate a surrogate nitrogenous waste mixture of
9.1 percent (by volume) pyridine in fuel oil. A NOx emission of less than 100 ppm was maintained, with
greater than 99.99999 percent destruction of the pyridine. The relative NOx emissions, with and without
pyridine addition, indicated that less than 1 percent of the luel nitrogen was converted to NO. This
preliminary result sparked interest in utilizing the precombustion chamber burner for low NOx, high
efficiency combustion.
However, while the precombustion chamber burner has been demonstrated successfully in
pilot-scale and field tests, its large size makes it impractical for most boiler retrofit applications. Reducing
the burner size results in a first-stage residence time that is insufficient to fully convert the fuel nitrogen to
N2; hence, higher NOx emissions result. More NOx control can be achieved by further staging of the
combustion process in the boiler.
For this study, during Phases 1 and 2 the 0.6 MW precombustion chamber burner, designed by
the Energy and Environmental Research Corporation of Irvine, California, was used. This burner is
shown in Figure 2-1.
The burner consisted of primary fuel and primary air injectors, two 0.51 m diameter by 0.91 m
long spool modules, and one 0.25 m diameter transition section with eight radial ports for secondary air
addition. To achieve rapid mixing in the precombustor, the primary fuel was injected through a divergent
nozzle. The primary air, which was not preheated, passed through fixed swirl vanes. These burner
sections were lined with a thick refractory wall to minimize heat loss and maintain high temperatures that
promote conversion of fuel nitrogen to N2 under fuel-rich stoichiometries. The reduction of burner
diameter to 0.25 m in the transition section minimized radiative loss to the boiler and prevented
backmixing of combustion gases, thus allowing precise control over burner stoichiometry.
7
-------
00
Flame Detector
Castable Refractory
A.P. Green LW-26
Primary
Fuel
Axial 2nd Air
Primary
.""•VVSfi'**
Castable Refractory
A.P. Green LW-33
Pilot
Flame
Port
L—Radial 2nd Air
Block Inaulation-
Figure 2—1. Horizontal low NOjj precombuation chamber burner.
-------
The burner was designed for substoichiometric (fuel-rich) operation, requiring secondary
combustion air for complete burnout of primary fuel. This air was provided through the radial injection
ports in the transition section, shown in Figure 2-2. The transition section was water-cooled so that the
combustion gas was cooled before secondary air addition, minimizing thermal NO generation. Primary
fuel nitrogen was simulated by doping ammonia (NH3) into natural gas or pyridine (C5H5N) into distillate
fuel oil.
For Phase 2 proof-of-concept tests, the precombustion chamber burner was shortened by
removing one of the two spool modules. This shorter burner, shown in Figure 2-3, had a first-stage
combustion gas residence time of 350 ms at a nominal load of 2 x 106 Btu/h (0.6 MW).
For Phase 3 tests, a versatile precombustion chamber burner developed by Acurex Corporation
was used. This burner, shown in Figure 2-4, was vertical, modular (with up to four spool modules), and
had ports for observation, sampling, and air/fuel injections. This burner had all the features of the
Phase 2 precombustion chamber burner. In addition, its modular construction provided variations of
residence time, and it had air/fuel staging, sampling, and observation capabilities implicit in its design.
Specifically, during the Phase 3 testing, the shortest configuration of the burner (including the burner
module and the elbow module) was used, with a first-stage residence time of 250 ms at nominal load.
2.2 FIRETUBE PACKAGE BOILER SIMULATOR (PBS)
The firetube package boiler simulator (PBS) used in this study was manufactured by Dynamic
Sciences, Inc., in 1972. The PBS has a nominal rated capacity of 1.0 MW (3.5 x 106 Btu/h) and can be
fired with natural gas, distillate oil, or residual oil. As seen in Figure 2-5, the radiant section is a
double-walled circular cylinder of 2.3 m (7.5 ft) length and 0.6 m (2.0 ft) internal diameter. A 90° elbow
section directs gas flow to the vertical stack. An induced draft fan can be used to pull gases through a
horizontal convective section (shell and tube heat exchanger) and to return the gases to the stack. The
convective section was not used in this study because all relevant chemistry is effectively frozen at the
radiant section exit temperature.
The radiant section is cooled with Dowtherm G heat transfer fluid, with a heat extraction rate and
wall temperatures similar to those found in commercial firetube package boilers. The Dowtherm system,
9
-------
SECONDARY RADIAL AIR
PBS RADIANT
SECTION
PRECOHBUSTION
CHAMBER
SECONDARY AXIAL AIR
Figure 2-2. Burner/boiler transition section.
-------
Castable Refractory
A.P. Green LW-26
Flame Detector
Primary
Air
Primary
Fuel
1
Flame
Axial 2nd Air
ESS
K
Castable Refractory
A.P. Green LW-33
Block Insulation
— Radial 2nd
Figure 2-3. Reduced size low NOx precombustion chamber burner.
-------
PRIMARY FUEL
BURNER
STAGED
FUEL
•\V-.A
PRIMARY AIR
Figure 2-4. Vertical, modular, low precombuatlon chamber burner.
12
-------
ID Fan
Return
Flue Gas
To ID Fan
Dowtherm Out
A
Convective Section
Heat Exchanger
T
Dowtherm In
Commercial Burner
Radiant Section
Dowtherm Out
Dowtherm
Flow Between
Double Walls
Figure 2-5. Package boiler simulator.
-------
which dumps its heal to a cooling tower through a house industrial water system, precludes problems
associated with raising and disposing of steam. Wall penetrations (originally used in staged combustion
studies) at 0.25 m (10 in), 0.51 m (20 in), 1.0 m (40 in), and 1.3 m (50 in) from the firetube front face were
useful as test ports for the insertion of in-flame sampling probes and reburn fuel injectors. In addition, the
end plate of the boiler has a port for injection of deep-staged air or reburning fuel. The boiler's front face
has eight axial ports for the addition of staged air in Phase 1 tests. For Phase 2 and Phase 3 tests, two of
the eight axial air ports on the PBS were modified to provide ports for staging fuel into the boiler at an
angle of 45°. This design allowed reburning application from the boiler front face, with aerodynamic
separation of the fuel-lean and fuel-rich zones in the boiler.
The combination of secondary air additions through radial and axial ports in the transition section
and the boiler front face, respectively, facilitates the production of a continuous series of secondary
flames, from a short radial flame (all radial secondary air) to a long axial flame (all axial secondary air.
2.3 COMBUSTION A!R/FUEUDOPANT FLOW SYSTEMS
As seen in Figure 2-6, primary and secondary combustion airs were supplied by a single blower
powered by an 11 kW (15 hp) electric motor. A Meriam laminar flow element (LFE) monitored total air
flow. Primary air, secondary radial air, and secondary axial air were monitored by venturi flow meters. All
pressure differentials were measured with fluid manometers. A small amount of the primary combustion
air was bled across the flame detector to prevent condensation on its lens. The inlet static pressures and
temperatures of the flow devices were monitored with Magnehelic pressure gauges and bimetal
thermometers to correct flows to standard conditions (1 atm, 70 °F or 101.3 kPa, 21 °C). All air flow
devices were calibrated against an in-house standard LFE.
Final (burnout) combustion air was supplied by the house compressed air system, as is shown in
Figure 2-6. The compressor output was regulated to a pressure of approximately 90 psig (620.6 kPa)
and the air flow was measured by a venturi meter with a U-tube manometer.
Figure 2-7 shows the natural gas supply system. The main burner natural gas flow was
measured with an orifice plate and an inclined manometer. House gas pressure was regulated to a
nominal 5 psig (34.5 kPa). The gas flow split after the orilice, with one leg going to the main burner and
14
-------
cn
LFE Laninar Flow Element
Venturi
[><3
Flow Valve
Butterfly Valve
Pressure Regulator
Pressure Switch
(^p) Pressure Gauge
(J) Thermometer
PS
Limit Switch
Safety System
PS
-® (i>
f
4-® cl>-|
-®
U-b
Main Air
Btower
Compressor
90psig
Plenum
Ammonia
Primary Air
Secondary Axial Air
Secondary Radial Air
Limit Switch
Safety System
©
PS
1
1
1
1
1
1
9
PS
1
1
1
1
1
,
1
1
1
1
©
LFE
PS
1 9
1
1
1
1
OKF
Tertiary Air
Primary
Atomizing Air
Secondary
Atomizing Air
Purge and Cooling Airs
Figure 2-6. Combustion air supply system diagram.
-------
Orifice Plate
<
Rotameter
(V) Pressure Gauge
(J) Thermometer
M
S
c3Sa
t>o
Motorized Valve (normally closed)
Solenoid Valve (NC-normally closed)
Flow Valve
[XI Gas Cock
Pressure Regulator
CD
Natural
Gas
Supply
5psig
0
-£*3-
©
M
<1
Limit Switch
* Safety System
&<1
M
c5kh
-OKF
Ammonia
Secondary Gas
Limit Switch
Safety System
Pilot Gas
Primary Gas
Figure 2-7. Natural gas supply system diagram.
-------
another to a pilot torch. Normally closed solenoid valves were electrically tied to the burner
management/safety system, allowing fuel flow only after a series of flame safeguards and limit switches
were actuated. The primary burner gas pressure was further regulated before passing through a normally
closed, motorized valve (actuated by the burner management/safety system) and a manual gas cock.
The gas cock and the diameter of the holes in the gas nozzle regulated the gas flow. Six holes were
drilled symmetrically around the gas nozzle centeriine, at a 30° angle to the centerline axis. Several
nozzles with varying diameter holes were used to control gas flow to appropriate levels during burner
operation. Table 2-1 presents the results from fuel analyses performed on this fuel.
The fuel oil system is shown in Figure 2-8. The circuit included solenoid valves which were tied
to the burner managemenVsafety system, allowing oil flow only after a series of flame safeguards and
TABLE 2-1. FUEL ANALYSES
Natural Gas
Analysis Natural Gas
Higher Heating Value
(Btu/cubic ft, @ 60OF, 14.73 psig) 1031.6
(Btu/lbf) 23,563
Encal Chromatograph Analysis (mol %)
Nitrogen 0.35563
Carbon Dioxide 0.91649
Methane 96.11870
Ethane 1.85666
Propane 0.38608
l-Butane 0.09849
N-Butane 0.10139
l-Pentane 0.04272
N-Pentane 0.02865
Hexanes Pius 0.09515
Specific Gravity (@ 60 °F, 14.73 psig, air = 1.0) 0.58480
Theoretical Air
(mol air/mol fuel) 9.69
(lba/lbf or kga/kgf) 16.6
(lba/10,000 Btu) 7.04
a air
j fuel
17
-------
Pressure Relief
To Oil Gun
(Furnace)
Filter
#2 Fuel Oil Supply
{XI-
{Xb
£X!J
#2 Fuel Oil Return
#6 Fuel Oil Supply
#6 Fuel Oil Return
Filter
Flow
Totalizer
Heater
55-Gallon
Drum
Pump
ok]
On/Off Valve
£> Check Valve
tXl Metering Valve
(^) Pressure or Temperature Gauge p>l<^] Solenoid Valve
Figure 2-8. Fuel oil supply system diagram.
18
-------
limit switches were actuated. The oil flow was metered by the metering valve, located near the 55 gallon
(0.21 m3) drum, and the flow was measured by the flow totalizer. The fuel oil feed circuit included a
recirculation loop.
Ammonia could be supplied to the precombustion chamber burner fuel flow to regulate primary
NOx concentration independent of other primary burner parameters. The supply system is shown in
Figure 2-9. The ammonia flow, which was supplied from cylinders regulated to a nominal pressure of
10 psig, was controlled by a precision needle valve and monitored by a rotameter designed for ammonia
service. For natural gas firing, ammonia was doped into the fuel line sufficiently upstream to ensure
complete mixing prior to injection.
The natural gas secondary fuel supply was tapped from the house natural gas supply system, as
shown in Figure 2-7. The fuel flow was metered by a rotameter designed for natural gas service.
Ammonia could be doped into the secondary natural gas to simulate fuel-bound nitrogen. This flow was
metered, as shown in Figure 2-9.
2.4 BURNER MANAGEMENT/SAFETY SYSTEM
A burner management/safety system, built by the North American Manufacturing Company, was
used to permit safe and consistent operation of the facility. The central feature of the system is a
cam-driven programmer tied to an ultraviolet flame detector. Depending on which fuel is being fired, a
series of safety interlocks must be made to initiate the starter sequence. These include initial and
continuous checks on primary combustion air flow; atomizing air flow (oil fired only); Dowtherm
temperature and flow; and water coolant flow (to burner head, transition section, and PBS end plate). A
control system logic diagram is shown in Figure 2-10. With all interlocks made, the combustion air blower
will prepurge the combustor for 1 min. After this prepurge period, the pilot torch solenoid valve is
actuated for 15 s, during which time the pilot torch is manually lighted, inserted into the combustor, and
detected by the UV flame detector. Then, the main fuel valves are automatically actuated. The pilot
flame remains on for 5 s, during which time the main flame must be lighted. If it is not lighted, the main
fuel valves are de-energized and the combustor is purged for 30 s. The safety system provided
automatic shutdown in the event of a safety check failure.
19
-------
Check Valve
Pressure Gauge
Rotameter
Primary Fuel Ammonia
lOpsig
Secondary Fuel Ammonia
Ammonia Cylinder
Pressure Regulator [\l
©
On/Off Valve
DOd Row Valve V
Figure 2-9. Ammonia supply system diagram.
-------
Manual Rebuming
Enable Switch
Off
UV Primary
Flame Detector
Tertiary Air
Flow Switch
Manual 3-Position
Rebuming Fuel
Selector Switch
Off
Natural
Gas Flow
Dowtherm Motor
Control Relay
Fuel Oil Flow to
Pressure-Atomizing
Delavan Nozzle
Atomizing Air
Flow Switch
Fuel Oil Flow lo
Air-Atomizing
Sonicore Nozzle
Off
Honeywell
Controller
Atomizing Air
Flow Switch
UV Primary
Flame Detector
Primary Air
Flow Switch
Transition Section
Cooling Water Fbw
Switches (2 in series)
Dowtherm Coolant
Over Temperature
Switches (2 in series)
Manual 3-Position
Primary Fuel
Selector Switch
Manual 2-Position
Oil Nozzle
Selector Switch
Natural Fuel Oil
Gas Flow Fbw
Figure 2-10. Burner management system logic diagram.
-------
2.5 STACK GAS SPECIATION
The PBS stack contained both a thermocouple probe and a sampling probe. The sampling probe
drew a sample through a porous metal filter, pulling from the entire cross-section of the stack. The stack
effluent was sampled and analyzed by a Continuous Emissions Monitoring System (CEMS), shown in
Figure 2-11. The combustion gas sample was extracted through a porous metal filter inserted across the
stack cross-section. Transported in stainless steel and Teflon tubing, the sample gas passed through a
Hankerson dryer and a particulate filter. The sample gas was then split, allowing a portion of the flow to
pass through a Dryrite canister for further drying before going to the individual analyzers for oxygen, total
unburned hydrocarbon, carbon dioxide, and carbon monoxide. The rest of the sample went directly to the
NOx analyzer, bypassing the NOx absorbing Dryrite material.
The NOx meter is a Thermo-Electron Model 10A analyzer which operates via
chemiluminescence. A small portion of the sample flow to the instrument is metered into a vacuum
(reaction) chamber where it is allowed contact with an excess of ozone from an integral ozonator. Nitric
oxide (NO) and ozone (O3) react to form N2O, a portion (approximately 10 percent at room temperature)
of which is elevated to an excited state. The excited molecules return to ground state and give off light of
a characteristic frequency. This light is detected by a photomultiplier tube. The output is amplified and
scaled to read directly in parts per million. The NO/NOx analyzer has a range of 0-1,000 ppm. A high
purity N2 and a gas of known NO concentration were used to calibrate this instrument.
For the conditions sampled in the firetube package boiler, nearly all of the NOx was in the form of
NO and, therefore, the NOx analyzer was operated in the NO measurement mode.
The O2 meter was a Beckman Model F3 Oxygen Analyzer, which operates by utilizing the
paramagnetic property of oxygen. Other gases present in significant concentration in the stack effluent
do not exhibit this property. The instrument was calibrated with high purity N2 and room air. Its
measurement range is 0-25 percent.
The CO2 and CO monitors used were Beckman Series 864/865 Infrared Analyzers. They
operate by directing identical infrared beams through a sample cell and a sealed reference cell. The
difference between the strength of the beams, as measured by a detector at the opposite end of the cell,
22
-------
Metal Filter
r
Stack
ro
CO
c
Sampling Probe
Combustor
I I
no
Dryer
Line
Filter
Cal
Gas Rotameter
-€HZ3
Dryrtte -
cr
Sample
F'ump
NOv
Cal
Gas
Cal
Gas
©HH3
Oc
CO
Vent
Vent
o
CC^
CO
__» Vent
Figure 2-11. Continuous emissions monitoring system (CEMS) diagram.
-------
is proportional to the concentration of the target compound in the sample. Calibration was accomplished
with a high purity N2 gas and a known concentration sample of the target gas. The range of the CO2
analyzer is 0-20 percent. The range of the first of two CO monitors, used for stack gas measurement, is
0-2,000 ppm. The second CO monitor, used for in-flame measurements, has a range from 0-5 percent.
The gaseous emission monitor specifications are summarized in Table 2-2.
2.6 TEMPERATURE MEASUREMENTS USING A SUCTION PYROMETER
The suction pyrometer designed for in-flame temperature measurement in the furnace was a
1.0 m (3.3 ft) long probe. The basic design of the suction pyrometer is shown in Figure 2-12. Platinum
and platinum/13 percent rhodium 10 mil (0.254 mm diameter) thermocouple wire was fed through a
two-hole alumina tube and welded to form an R-type thermocouple junction. The thermocouple was
slipped inside a 0.635 cm (0.25 in) diameter alumina tube with a closed end for radiation shielding. To
complete the suction pyrometer, the shielded R-type thermocouple was slipped into the water-cooled
probe assembly. The thermocouple tip extended 15 cm (6 in) past the end ol the water-cooled probe,
inside a close-end alumina nosepiece with a 0.635 cm (0.25 in) hole near its tip (see Figure 2-11).
Suction for the pyrometer was provided by a carbon vane pump; the flow rate was monitored by a
dry gas meter. The pump and the meter were protected from moisture by an EPA Method 5
condensation train consisting of two impingers and a silica gel trap in an ice bath.
Thermocouple output was read by an Omega Model 2166A digital thermometer. When in
operation, hot combustion gases were pulled at a high velocity through the nosepiece port and down the
annulus between the nosepiece and the shielded thermocouple. As the gas flow rate across the shielded
thermocouple increased, convective heat transfer to the thermocouple tip predominated over radiative
transfer to the cold boiler walls, and the actual gas temperature was approached. Figure 2-13 shows that
at flow rates in excess of 110 Umin, asymptotic operation and, therefore, actual gas temperature
indication was approached by the suction pyrometer. The constnjction of the suction pyrometer allowed
temperature measurement up to approximately 2,000 K (3,140 °F).
24
-------
TABLE 2-2. GASEOUS EMISSION MONITOR SPECIFICATIONS
Gas
Sample Type
Instrument Type
Manufacturer/
Model
Range
02
Continuous
Paramagnetic
BeckmarVF3
0 to 5,10, 25%
CO2
Continuous
Infrared
Beckman/864
Oto 5,10, 20%
CO
Continuous
Infrared
Beckman/865
Oto 500,1,000,
2,000 ppm
NO
Continuous
Chemiluminescence
Teco/10A
Oto 2.5, 10, 25,
100, 250, 1,000,
2,500,10,000 ppm
25
-------
ft
1-1/4" Sch. 40j
Blk. Pip*
0
4
h H
fill:
Ceramic
Nose Piece
<&
Type R T/C
H
3/4" O.D. x 0.065W
SS Tubing
Water In ^gvagelok Pitting
=K> 1
Metal Tabs
Ijrf
~
Suction
Water Out
Part B. Water Cooled Probe Aaeenbly
•7/16" SS Tub*
3/8" Single Bore
Ceraalc Tube
e\l 3/16"
^Ceria
O.D. TVo Hole
lc Tube
Part A. Shielded ThetTocouple
Figure 2-12. Suction pyrometer design.
-------
ro
-si
O
o
k_
ID
03
k_
0)
Q_
E
a>
•o
0)
(3
g
T3
C
1200 Asymptotic Temperature = 1192 °C
1150 -
1100 -
1050 -
1000
950
Indicated
Temperature = 1187 °C
Nominal Suction
Flow Rate = 110 L/min
20
40
60
80
100
120
Suction Flow Rate (dry) (Std L/min)
Figure 2-13. Suction pyrometer calibration curve.
-------
SECTION 3
PROOF-OF-CONCEPT TESTS
3.1 BACKGROUND
While the precombustion chamber burner, discussed in Section 2, has been demonstrated to
achieve low NOx levels for high nitrogen fuels, its large size makes it impractical (or many boiler
applications. Reducing the burner size results in incomplete reduction of the fuel nitrogen in the
precombustion chamber and, hence, higher NOx emissions. However, an in-furnace NOx control
technology exists for removing NOx from the combustion gas stream. NOx formed in the primary flame
can be destroyed by fuel staging, or rebuming. By diverting portions of the fuel and combustion air
streams from the main burner(s) for injection into the post-flame gases, rebuming establishes a
three-stage combustion process consisting of a fuel-lean primary zone, a fuel-rich rebuming zone, and a
fuel-lean burnout zone (Figure 3-1). NOx is destroyed by hydrocarbon radical reactions to form
intermediate fixed nitrogen species and, eventually, molecular nitrogen. An overall reaction pathway has
been postulated by Chen et al.8
As many as 30 years ago, experimental results indicated that NO could be destroyed by reaction
with hydrocarbon radicals. This reaction mechanism was commercially utilized in the late 1960's by the
John Zink Corporation, which patented a NOx control process for nitric acid plants based on staged
injection of natural gas.9 Early fundamental studies of NOx destruction by injection of secondary fuel into
the flame zone were performed by Wendt et al, who coined the name "rebuming" to describe the
process.10 Concerted effort to develop rebuming for application to boilers occurred in Japan during the
late 1970's and early 1980's. The first reporting of these efforts was by Takahashi et al, documenting
extensive laboratory-, pilot-, and full-scale evaluations of a fuel staging process they refer to as Mitsubishi
Advanced Control Technology, or MACT.11 MACT results indicated that NOx emissions can be reduced
by at least 50 percent, independent of initial NOx level and fuel type. These promising results have
28
-------
Burnout Zone
Rebuming Zone
Fuel
Air
Primary Zone
Figure 3-1. The reburning process.
29
-------
renewed interest in reburning, and lurther development is being sponsored by the U.S. Environmental
Protection Agency (EPA), the Electric Power Research Institute (EPRI), the Gas Research Institute (GRI),
and the U.S. Department of Energy (DOE).
Since 1982 EPA has carried out both in-house and extramural bench- and pilot-scale reburning
tests. In-house experiments focussed on natural gas and fuel oil reburning. 12-15 Coal reburning tests
have been performed for EPA by Energy and Environmental Research Corporation,8.16 and Acurex
Corporation. 17 Significant findings include the dependence of reburning effectiveness on primary flame
NO level, reburn zone stoichiometry, and reburn fuel nitrogen content. Fifty percent NO reduction is
possible with 10-20 percent of the fuel used for reburning. At low primary NO levels, however, a
nitrogen-free reburning fuel (such as natural gas) must be used to achieve 50 percent reduction. Tests
show that reburning is effective over a wide range of temperatures and with relatively short reburning
zone residence times (less than 0.2 s).
EPA cosponsored with GRI and EPRI a reburning application test on an industrial or small utility
coal-fired boiler, beginning in late 1986 and lasting 3-4 years. A field application test of the
precombustion chamber burner has also been initiated. Thus, independently, these technologies are
nearing the final stage of development.
3.2 EXPERIMENT
The goals of this phase of the study were to develop a multistaged burner design concept that
utilizes a half-sized precombustion chamber burner and natural gas reburning to maintain NOx emissions
of less than 175 ppm for primary fuels with up to 5 percent fuel nitrogen dopant. The design minimized
boiler modification to maximize retrofit potential and commercial applicability.
3.3 APPARATUS AND METHODS
The primary experimental equipment consisted of the pilot-scale package boiler simulator and
precombustion chamber burner (described in Section 2). This facility was fired on natural gas for these
tests. Primary fuel nitrogen was simulated by doping ammonia (NH3) into natural gas.
Secondary natural gas was injected through a water-cooled boom at three axial locations: 107,
30, and 5 cm from the boiler front face. The reburning fuel was injected uniformly into the primary
30
-------
combustion gas stream during reburning characterization tests, using a spoked injector configuration.15
Rebuming natural gas was injected Irom a point source at the boiler centerline lor the optimization tests
described here. Burnout air was injected through eight radial ports in a water-cooled boom, inserted
down the boiler centerline through a reartlange, or through axial ports located on the boiler front face.
When injected through the centerline boom, the location of the burnout air injection was nominally fixed at
the 162 cm axial location.
The experimental facility was designed for independent control and measurement of each fuel,
fuel dopant, air stream, stack and in-flame combustion gas speciation, and temperature. Temperature
was measured using suction pyrometry. The facility shown schematically in Figure 3-2 is described in
Section 2.
3.4 APPROACH
To evaluate this multistaged process for low NOx. high efficiency combustion, a three part
experimental program was devised. First, baseline tests (i.e., no reburning) were performed to evaluate
shortened burner (one-module) performance. First-stage stoichiometry and simulated primary fuel
nitrogen content were varied. Results were compared with long burner (two-module) results. Second,
parametric tests were performed to evaluate rebuming effectiveness under ideal mixing conditions as a
function of primary zone NO (NOpn) level, reburn zone stoichiometry (SRr), and simulated reburn fuel
nitrogen content. These results have been previously reported and are summarized here. 12-15 Third,
preliminary tests were performed with the shortened prechamber burner for optimum reburning
application. These results are discussed and current tests on a new test burner are described.
3.5 BASELINE TEST RESULTS
At a primary burner firing rate of 0.6 MW and a nominal first-stage stoichiometry of 0.7,
shortening the precombustion chamber burner length by removing one of the two modules decreased the
bulk combustion gas first-stage residence time from 600 to 350 ms. In baseline tests (without reburning),
first-stage stoichiometry was varied, with exhaust excess air held constant at about 15 percent. Ammonia
was doped into the natural gas fuel stream, resulting in a fuel nitrogen content of 0.66 percent. In
Figure 3-3 data are shown for both the short and long burner configurations. The minimum NO emission
31
-------
CO
fO
PRIMARY AIR
r- REFRACTORY
RADIAL SECONDARY AIR
STACK
SAMPLE
PORT
PRIMARY FUEL
REDUCED LOW NO BURNER
X
PACKAGE BOILER SIMULATOR
DOWTHERM COOLED
AXIAL SECONDARY AIR
WATER COOLED
Figure 3-2. Pilot-scale research facility.
-------
1000
900
800
700
600
500
o
Shortened
Precombustion
Chamber (lover
residence
\ time)
400
300
Unshortened \
Precombustion >!
Chamber (higher
residence
time
200
100
0
0.7
0.6
0*4
o.t
o.s
0.9
1.0
First-Stage Stoichiometric Ratio
Figure 3-3. Influence of first-stage residence time on NO emission for short and long preburner
configurations.
33
-------
was observed at a first-stage stoichiometry of about 0.7, consistent with England's data at a similar scale
test.7 The sharp minimum in the curve indicated the sensitivity of NO emission to first-stage
stoichiometry. The minimum NO emission increased from 50-75 to 200-250 ppm in going to the
shortened burner. These ranges represented average NO emissions achievable given the small
fluctuations in fuel and air flows.
The amount of NH3 dopant was varied at the optimum first-stage stoichiometry (0.7) for both the
short and long burner configuration. The data in Figure 3-4 show that NO levels in the short burner are
more sensitive to fuel nitrogen content than the long burner. With no NH3 addition, NO emission from the
short burner was 90 ppm; from the long burner the NO emission was 40 ppm. These levels indicated the
thermal NO component of the NO emission, coming from the molecular nitrogen (N2) in the air. During
operation with the short burner, a longer flame was observed in the boiler than during operation with the
long burner. This is because combustion of the hydrocarbon was less advanced at the secondary air
addition location for the short burner. Consequently, the peak temperatures in the boiler burnout zone
were slightly higher than for the long burner. With the addition of NH3 dopant the long burner NO
emissions increased to 100 ppm, or a net 60 ppm contribution from the surrogate fuel nitrogen. The short
burner NO emissions for high NH3 dopant levels in the fuel stream approached 250 ppm, or a 160 ppm
net increase due to fuel nitrogen. The reduced first-stage residence time resulted in less fuel nitrogen
being reduced to N2 in the fuel-rich precombustion chamber. As fuel nitrogen content increased above
2 percent the resulting increase in NO emissions became small.
Thus, an effect of halving the burner size is to increase minimum NO emissions for high nitrogen
fuels from 100-250 ppm. Another effect is to move some of the flame back into the boiler, although still
much of the heat release remains in the prebumer. With a shorter flame length in the boiler with the
prebumer than with conventional burner, reburning, which requires boiler volume, is an Ideal technology
for achieving additional in-furnace NO reduction. Before discussing preliminary tests combining the
reburning and precombustion chamber burner technologies, reburning characterization tests are
discussed.
34
-------
300
250
First Stage
Residence Time:
350 ms
200
E
a
oj
° no
5?
o
Q
0
z
1 100
£
c
S
600 ms
3
0
1
2
ft
Simulated Primary Fuel Nitrogen, percent
Figure 3-4. Influence of fuel nitrogen content on minimum NO emission for short and long preburner
configurations.
35
-------
3.6 REBURNING CHARACTERIZATION TEST RESULTS
Rebuming involves staging a portion of the fuel and air streams downstream of the primary flame
to reduce NO in the primary combustion gases. Parametric tests were performed previously in the
package boiler simulator in which primary NO was varied between 50-500 ppm and reburn zone
stoichiometry was varied from 1.0-0.75."' 5 in these tests primary load was held constant at 0.6 MW,
first-stage stoichiometry was fixed at 0.6 (off-optimum), and primary and burnout zone stoichiometries
were held constant at 1.1. The secondary natural gas was injected at a 1.02 m axial location, just
downstream of the visible primary flame. Using a spoked fuel injector, natural gas was injected to match
the flux of free oxygen from the primary zone, thereby providing radial uniformity in the reburn zone
stoichiometry. Burnout air was added at the 1.62 m axial location. Primary NO was varied by varying the
primary fuel NH3 dopant level. Rebum zone stoichiometry was varied by varying the reburn fuel and air
flowrates. This procedure provided sufficient control to measure the maximum reburning effectiveness in
the package boiler simulator.
Results from these natural gas reburning tests are summarized in Figure 3-5. The strong
dependence of NO reduction to primary NO is in agreement with the results reported by Chen8 but in
conflict with the MACT claim. 11 Analysis of the data in Figure 3-5 suggests that the overall NO
destruction rate has a partial order with respect to primary NO of 1.5-1.6. That is,
d[NO]/dt = C [NOpri]1-5-1.6
Additional data were taken to evaluate reburning fuel nitrogen effects.13 From this data a contour
map was drawn depicting the relationship between overall NO reduction by reburning and two controlling
parameters: reburning fuel nitrogen content and primary NO level. In Figure 3-6 it is shown that
50 percent NO reduction is not possible for primary NO levels less than 140 ppm. For NO levels less
than 250 ppm, fuel nitrogen content limits NO reduction. For example, at an initial NO level of 200 ppm,
reburning with fuels exceeding 1 percent in fuel nitrogen results in increased NO emissions.
36
-------
100
c
-------
1.0
Overall NO
Reduction, percent
0.8
U-
O)
-O
0.2
200
250
150
100
Primary NO (Dry, 0% O2), ppm
Figure 3-6. Contour map of optimum NO reduction by rebumirig. From Ref. 5.
38
-------
Based on these results from returning tests performed under idealized mixing conditions, it was
hypothesized that reburning applied in tandem with the precombustion chamber burner could produce NO
emissions of less than 175 ppm for gaseous and liquid primary fuels with up to 5 percent fuel nitrogen.
3.7 PRELIMINARY PROOF-OF-CONCEPT TEST RESULTS
Preliminary tests were performed to evaluate the concept of natural gas reburning of combustion
gases from a half-sized precombustion chamber burner. A primary NO level of 260 ppm was maintained
by operating the burner at an off-optimum stoichiometry of 0.65 to reduce the amount of fixed nitrogen
dopant (NH3) required to achieve this emission. This NO level represents the maximum emission of the
half-sized burner when burning fuels with up to a 4 percent fuel nitrogen. Primary flame zone
stoichiometry leaving the preburner was fixed at 1.1. The locations of staged fuel and air addition were
varied.
Figure 3-7 shows the effect of rebum fuel injection location on NO emissions. Reburning fuel was
added at the 107,30, and 10 cm axial locations in the boiler. Burnout air was added at 168 cm. From
Figure 3-7 it is observed that slightly lower NO emissions were achieved on injecting the rebum fuel
downstream of the primary flame. However, even injecting rebum fuel at the outlet of the precombustion
chamber resulted in significantly reduced NO emissions. The data suggest that rebum fuel can be
injected at the boiler front face and still achieve NO emissions of less than 175 ppm.
Figure 3-8 shows the effect of burnout air injection location on exhaust NO levels. With reburning
fuel injected at 30 cm, air was injected through axial ports on the boiler front face. The results show only
slightly higher NO emissions than with deep-staged (168 cm) burnout air. Thus, injecting burnout air from
the boiler front face results in NO emissions that approach the 175 ppm level.
3.8 DISCUSSION AND CONCLUSIONS
Based on the data presented in Figures 3-4 and 3-6, predictions can be made of NO emissions
as a function of primary and secondary fuel nitrogen content. However, by injecting fuel and air on the
front face of the boiler a reduction in reburning effectiveness was observed (Figures 3-7 and 3-8). With
250 ppm primary NO, a 42 percent reduction was achievable for reburning with a nitrogen-free fuel at the
boiler front face versus 55 percent reburning downstream of the primary flame with uniform injection
39
-------
#2
Primary
Fuel
and Air ,
300
250
200
E
a
o3
O
g
o
ISO
100
so
Burnout Air
\.
0.7
Secondary Air I II
^db~y—
iL- n l
#3
0.8
0.9
SRc
1.0
Reburnirig
Fuel
(#1)
Rebum Fuel /-
injection Location^/^ /
10 cm (#3) Jjt /
¦»
/30 cm /
107 cm
-
(#1)
-
J 1 . l. _
1.1
Figure 3-7. NO emissions as a function of reburn zone stiochiometry for various reburn fuel injection
locations.
40
-------
Secondary Air
#2
Primary
Fuel and
Air
E
~.
a.
OJ
O
SS
Q
o
2
250
200
150
100
50
0.7
Reburning Fuel
t
Burnout Air
(#1)
Burnout Air Injection
Location:
0 cm (#2)
168 cm
(#1)
0.4
0.9
SRr
1.0
1.1
Figure 3-8. NO emissions as a function ol reburn zone stoichiometry for various burnout air injection
locations.
41
-------
(Figure 3-6). Though insufficient data are available to make a new contour map of reburning
effectiveness versus primary NO for reburning application from the boiler front face, it can be
conservatively assumed here that reburning effectiveness is reduced by 10-15 percent from downstream
uniform reburning application. Carbon monoxide (CO) emissions for all the tests performed were lower
than 35 ppm, indicating that combustion efficiencies attained were greater than 99 percent.
42
-------
SECTION 4
NEW BURNER TESTS
4.1 INTRODUCTION
Proof-ol-Concept tests, explained in Section 2, showed that a low NOx strategy using a
precombustion chamber burner and reburning (from the front end of the boiler) was attractive from the
standpoint of NOx reduction and high combustion efficiency. Based on the results of these tests, a new
burner system was designed and installed by Acurex. To increase the burner temperatures by reducing
the radiative heat loss to the boiler and to take advantage of thermal buoyancy effects, the burner was
made vertical. This design also contained four removeable, small modules that vary burner length over a
wide range, allowing residence time studies. The new burner is shown in Figure 2-4.
4.2 EXPERIMENTAL METHODS
The goal of this study was to minimize NOx formation, with an emission target of 175 ppm (dry, at
zero percent O2) or less, and maintain efficient incineration of surrogate fuel/waste mixtures with up to
5 percent fuel nitrogen by using a precombustion chamber burner reduced in size by about half.
The facility shown schematically in Figure 4-1 was described in Section 2. For these tests a
burner length of 1.8 m was used, corresponding to a nominal residence time of 250 ms. A 5 cm diameter
port was available for staging fuel into the burner. Two of the eight axial air ports on the package boiler
simulator front face were modified to provide ports for staging fuel into the boiler at an angle of 45°. This
design allowed reburning application from the front face of the boiler. The end plate of the boiler was
modified to allow the insertion of a water-cooled boom injector for deep staging of air into the boiler.
As explained in Section 2, the experimental facility was designed for independent control and
measurement of each fuel, fuel dopant, and air stream. Stack gas speciation was measured by a
continuous emissions monitoring system. NO and NOx were measured by chemiluminescence.
Reported in this section are NO measurements only, measured on a dry basis and corrected to 0 percent
43
-------
PRIMARY FUEL
BURNER
STAGED
FUEL
PRIMARY AIR
RADIAL SECONDARY AIR
BOILER STAGED FUEL
/
AXIAL STAGED AIR
¦i)
~
s
DEEP STAGED AIR
Figure 4-1.
w h«1beeo°'f;t«rvnrarch f"iiuy- n* »***» *<>"« »i—
•nd fuel sctHnl port,! * <*«->« Ku™r «,
-------
02. Spot-check measurements of NOx indicated that NO emissions accounted for over 95 percent of the
exhaust NOx emissions in these tests.
A 0.7 MW North American (NA) Scotch-type 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 (2,400 Ib/h). Its size and thermal characteristics
are nearly identical to those of the package boiler simulator. 12,14
Evaluated were burner baseline performance and burner operation with various air staged and
fuel staged NOx controls. The parameters affecting the NOx emissions from the lacility with unstaged
controls were 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 depended
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, as per commercial boiler practice.
Nominal fuel nitrogen content for the fuel oil/pyridine mixture was 2 percent by weight; for the natural
gas/ammonia fuel the nominal fuel nitrogen content was 5.8 percent. First-stage stoichiometry was
optimized for each NOx control application tested.
4.3 BUHNER BASELINE PERFORMANCE
During low NOx burner characterization tests, the effects of first-stage stoichiometry, excess air,
load, and fuel nitrogen content on NO emissions were studied.
First-stage stoichiometry was varied by changing the primary air flow. Secondary radial air was
adjusted 1o maintain 15 percent excess air. The results are plotted in Figure 4-2. The curves indicate a
strong sensitivity of stack NO to changes in burner stoichiometry. At nominal fuel nitrogen, for the gas
tests, a minimum NO emission of 315 ppm occurred at a burner stoichiometry of about 0.78; for the oil
tests, a minimum of 190 ppm occurred at 0.65. Thus, additional combustion modifications were
necessary to meet the program goal of less than 175 ppm.
Additional tests were done with gas/ammonia mixtures with 2 percent fuel nitrogen content. The
results are also plotted in Figure 4-2. The shift in optimum burner stoichiometry suggests a variation in
the thermal environment in the precombustion chamber. In Figure 4-3 burner temperatures are shown,
45
-------
900
800
700
600
500
400
300
200
100
0
0
re 4-2.
(5.8% N)
(2% N)
2% N)
.4 0.5 0.6 0.7 0.8 0.9
FIrat-Stage Stoichiometric Ratio
1.
iffect ol burner stoichiometry. Shown are results from tests tiring 2.0 percent
.8 percent nitrogen gas fuel and a 2.0 percent nitrogen distillate fuel oil/pyridii
lixture.
46
-------
2300
2200
2100
* 2000
2 1900
1800
Gat
1700
1600
1500
1.0
0.8
0.4
0.8
•T VtV VtV i •
First-Stage Stoichiometric Ratio
Figure 4-3. Burner temperatures. Shown are measurements firing the
low NOjj burner on gas and oil fuels. The dashed line
represents the adiabatic flame temperature of methane
in air at 1 a tin.
47
-------
measured by suction pyrometry. Radiation and conduction errors are estimated to be less than 30 °C.
The burner temperatures were found to be higher for the oil tests; thus, the shift In optimum
stoichiometry.S
In tests on the North American boiler, a NO emission of 1,000 ppm resulted when firing the
5.8 percent nitrogen gas fuel at 15 percent excess air. A NO emission of 765 ppm resulted when firing
the 2 percent nitrogen oil mixture at 15 percent excess air. Thus, the low NOx burner reduced NO
emissions by 68 percent for the gas fuel and by 75 percent tor the oil mixture.
4.3.1 Excess Air Variation
The effect of excess air variation on stack NO is shown in Figure 4-4. Excess air had a much
stronger effect in the conventional North American burner tests than in the low NOx burner tests, as
expected. While reducing excess air provided some NO reduction with the low NOx burner, maintaining
an exhaust excess air level of 15 percent was important for achieving high combustion efficiency.
4 3.2 Load Variation
The effect of load reduction on stack NO is shown in Figure 4-5. As load was decreased from a
nominal condition of 0.6 MW to 65 percent of nominal, the stack NO concentration dropped by
20-25 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 250-385 ms. However, boiler steam requirements make load reduction an
impractical means of NOx control.
4.3.3 Fuel Nitrogen Variation
The effect of fuel nitrogen variation is shown in Figure 4-6. Exhaust NO level increased with
increasing fuel nitrogen content in the fuel/waste stream, as expected, with a much greater sensitivity
observed in the conventional North American burner tests. These results demonstrate the precombustion
chamber burner's ability to reduce fuel nitrogen to molecular nitrogen, even with its reduced size
48
-------
1400
/ Ga«,
/ N.A.
1200
a 1000
/ on
800
600
400
"Q-
40
Excess Air, percent
Rgtrg-M. Excess a;f variation results. Shew.-, are 2 0 percent and 5.8 perceni r.irossn oil res,j!!S The
dasMd lines represint resits feem tne Nonn Amer-caa (N A,;. &oi!«r tests.
4y
-------
400
300
200
100
GAS C5.*%m
OIL (2-<
gas r 2.0 y.m
l
i
50
60 70 80
Load, percent
90 100
Figure 4-5,
d variation results. Shewn are results with
a*2 0 np6nt andJ5,8 Percen* nitrogen gas fuel and
a 2.0 percent nitrogen oil mixture.
50
-------
1000
A
/ Gat,
_ 600
12 3 4
Fuel Nitrogen, percent
Figure 4-6. Fuel olcrogen variation results. The dashed lines
data from the North American (N.A.) boiler tests.
51
-------
(250 ms). The full size (600-800 ms) precombustion chamber burner produces NO emissions even less
sensitive to fuel nitrogen content.6
4.4 AIR STAGED NOx CONTROLS
During air staged NOx control tests, the effect of moving the location of secondary air addition
from the radial ports in the transition section to the axial and deep staged ports in the boiler was
examined. The secondary air was distributed between radial and axial locations in one series of staging
tests, and between radial and deep staged ports (located 132 cm from the boiler front face in a second
series of tests. This created a three-stage combustion process: a fuel-rich burner zone, a less fuel-rich
(SR2) zone after radial secondary air injection in the transition section, and a fuel-lean burnout zone after
final air addition. A water-cooled boom, aligned with the centerline axis of the boiler, supplied the deep
staged air.
Fuel oil/pyridine and natural gas/ammonia test results for the two cases of secondary air staging
in the boiler are given in Figure 4-7. The axial air staging resulted in a drop in minimum NO concentration
from 190 ppm, with no staging, to 150 ppm for the fuel oil/pyridine burning, and from 315 ppm to 220 ppm
for the natural gas/ammonia burning with 100 percent of the secondary combustion air moved from the
radial injectors to the axial injectors. The reduced NO levels for air staging were caused by lower
combustion gas temperatures and delayed mixing of the secondary air and combustion gas. Hence, less
thermal NO was generated and less fuel nitrogen species fragments were oxidized with axial air addition
into the boiler than with radial air addition into the transition section. Similarly, lower temperatures in the
boiler at the deep staged air location and longer fuel-rich zone residence time resulted in a lower NO
emission when the secondary air was deep staged. With all of the secondary air added from the deep
staged location, the minimum NO level was 130 ppm for the fuel oil/pyridine case and 160 ppm for natural
gas/ammonia case.
From the test results just described, deep staging of secondary air is shown to be an effective
means of minimizing NO emissions. However, such staging may lead to burnout problems, which are of
great concern when cofiring waste. Extending the fuel-rich zone from the precombustion chamber into
the boiler exposes the reducing combustion gases to cool boiler walls, which may result in slagging
52
-------
PRIMARY FUEL
( PRIMARY AIR
—I RADIAL SECONDARY AIRI
$
axial staged AIR DEEP staged air .
E
Q.
Q.
"c\J
o
~w
c
»
2,
O
-4
loo
40 60
Percent Air Staging
loo
Figure 4-7. Air staging results. Shown are results Irom firing a 5.8 percent nitrogen gas fuel and a
2.0 percent nitrogen oil mixture.
53
-------
and/or sooting. Further, the lower combustion gas temperatures in the boiler slow the tuel-rich nitrogen
Kinetics.
4.5 FUEL STAGED NOx CONTROLS
During fuel staged NOx control tests, the effect of diverting part of the primary fuel to staged
locations in the burner and boiler was studied. The staged fuel was natural gas, with no ammonia dopant.
In the case of fuel staging in the burner, the secondary natural gas was added, using a water-cooled
injector, down the centerline axis of the boiler; secondary air was added axially into the boiler to allow
sufficient fuel-rich zone residence time. In the case of fuel staging in the boiler, the secondary natural gas
was added at an angle of 45°, using water-cooled injectors, from two ports in the boiler front face.
Secondary air to complete primary fuel combustion was added radially into the transition section, while
burnout air for secondary fuel combustion was added axially near the boiler wall, using six axial ports on
the boiler front face.
4.5.1 Burner Fuel Staging
Holding total load constant, fuel was diverted from the primary fuel injector to a secondary injector
near the precombustion chamber exit (see Figure 4-1), thus creating two stoichiometric zones in the
burner in addition to the burnout zone in the boiler. The staged fuel was undoped (nitrogen-free) natural
gas. Secondary air was injected through the axial injectors. Two effects may contribute to a decrease in
NO emissions with this fuel staging scheme: 1) possible acceleration of fuel nitrogen reduction
mechanisms due to increased hydrocarbon radical concentrations, and 2) dilution of primary combustion
gases by secondary combustion gases.
Fuel oil/pyridine and natural gas/ammonia results are given in Figure 4-8. With optimized burner
stoichiometric ratios (0.7 for natural gas/ammonia and 0.65 for distillate fuel oil/pyridine) and 35 percent
fuel staging, the decreases in NO levels were from 220 ppm to 190 ppm and from 150 ppm to 110 ppm
for the cases of 5.8 percent fuel nitrogen gas/ammonia and 2.0 percent fuel nitrogen distillate fuel
oil/pyridine mixtures. It is interesting to note that with no fuel staging, the NO levels are 220 ppm and
150 ppm for the (5.8 percent N) gas and (2.0 percent N) oil tests respectively with this test configuration
where secondary air is injected axially, as opposed to 315 ppm and 190 ppm (Figure 4-5) obtained in
54
-------
Primary Fuel
Burner
Staged Fuel
Primary Air
\
Axial Staged Air
a
a
a.
fN
O
c
4)
U
U
®
a.
o
u
«
N
Ss
U
¦V
V_X
o
z
Gas (5.8% N)
Gas (2%N) Q
150
100
-A.
10 20 30 "40
Percent Fuel Staging
Figure 4-8. Burner fuel staging results. Shown are results from firing
2.0 percent and 5.8 percent nitrogen gas fuel and a
2.0 percent nitrogen oil mixture.
55
-------
baseline tests where secondary air was injected radially. This difference can be attributed to a delayed
combustion gas/secondary air mixing in the case of axially injected secondary air. The results with
2.0 percent N gas/ammonia mixture need more data for explanation.
Thus with this scheme of fuel staging, under the experimental conditions examined, the NO
emissions for 5.8 percent N gaseous fuel firing were not lower than the project goal of 0.2 lb (as
NC>2)/106 Btu (or approximately 175 ppm at 0 percent 02).
4.5.2 Boiler Fuel Staging
As in the tests just described, total boiler load was held constant while fuel was diverted from the
primary fuel injector to two secondary injectors. In this case, the staged fuel (undoped natural gas) was
injected into the boiler downstream of the secondary radial air addition, 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 stage (SR3, the fuel-rich
reburning zone in the boiler) is critical in this NOx control process. As already described, all of the staged
air and fuel flows were injected from the front face of the boiler at various angles, resulting in the three
boiler stoichiometric zones.
Fuel oit/pyridine and natural gas/ammonia results are given in Figure 4-9. Two second-stage
stoichiometries (SR2) were established: 1.1 and 1.0. The NO emissions under no staging conditions for
the cases of 5.8 percent N gaseous fuel firing and 2.0 percent N liquid fuel firing were 315 ppm and
190 ppm. With 35 percent fuel staging and 5.8 percent N gaseous fuel firing, the NO emissions
decreased to 195 ppm at a SR2 of 1.1 (aid a SR3 of 0.72) and to 160 ppm at a SR2 of 1.0 (and a SR3 of
0.65), Again, with 35 percent fuel staging and 2.0 percent N liquid fuel firing, the NO emissions
decreased to 120 ppm at a 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). Due to less distinct stoichiometric zones than typically established in reburning application, the
decrease in NO levels by fuel staging was not quite as great as that obtained when the staged fuel was
injected farther downstream of the fuel-lean primary combustion zone;12,16 however, the configuration
56
-------
Primary Fuel
fH
Primary Air
lladial Secondary Air
Boiler Staged Fue]
Axial Staged Air
325i r
a
a
a.
fM
o
c
«
U
w
V
a
N
u
•o
o
z
Figure 4-9.
100
0 10 20 30 40
Percent Fuel Staging
Boiler fuel staging results, shorn are results from flrin.
oll'mixture?^ nl"°8e" 8" fuel «"• » 2-° P"=enc nitrogen
57
-------
used in these tests requires no boiler penetrations. 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.
4.6 DISCUSSION AND CONCLUSIONS
Table 4-1 summarizes results from the various combustion modification schemes for the
reduction of NO in the oil and gas tests. These results were obtained at optimal first-stage
stoichiometries. From the table it can be seen that deep air staging and fuel staging in the boiler
(reburning) seem to be most effective in minimizing NO emission, resulting in an overall NO reduction
(ppm levels) of about 85 percent. Both configurations—shortened precombustion chamber burner/deep
a;r staging and shortened precombustion chamber burner/ooilerfuel staging—met the program goal of
attaining NO emissions of approximately 175 ppm from firing gaseous and liquid fuels doped with up to
5 percent nitrogen.
In the case of deep air staging, the decrease in NO emissions is due to a longer fuel rich zone
residence time (leading to greater fuel N to N2 conversion) and adding burnout air at a relatively cooler
deep staged air location (thereby generating less thermal NO). However, as mentioned before, this
approach does expose boiler walls to a reducing fuel rich zone and may cause sooting or slagging on
boiler walls.
In the case of reburning (as applied here;, the net decrease in NO emissions seems to be due to
a dilution of primary combustion gases by secondary combustion gases. This can be seen in Table 4-2
where NO concentrations obtained after adding appropriate dilution to base NO concentrations are very
close to those obtained during corresponding reburning applications. Thus, net chemical destruction of
NO during reburning under these experimental conditions does not seem to be evident. This may be
because of the existence of rather low primary NO concentrations before applying reburning. However, a
beneficial dilution caused by reburning, as applied here, may provide lower NO emissions (on a ppm or
lb/106 Btu basis) along with no loss in heat output. This beneficial dilution aspect of reburning application
can be seen by comparing the results of Figures 4-6 and 4-9. Substituting 35 percent of a 5.8 percent N
gaseous fuel with a nitrogen-free one would yield a primary N content of 3.77 percent, and from Figure 4-6,
58
-------
firing this fuel would result in approximately 290 ppm NO. However, for a corresponding case in
reburning, NO emission can be as iow as 160 ppm.
59
-------
TABLE 4-1. SUMMARY OF RESULTS
NO, ppm (percent reduction)
Control Modification Amount Natural Gas/Ammonia Fuel Oil/Pyridine
Burner Baseline
• N.A. Burner — 950 765
• Low NOx Burner 250 ms 315(67) 190(75)
Burner Characterization
• Excess Air 5% 290 (69) 175 (77)
•Load 65% 260(73) 140(82)
• Fuel Nitrogen 1% 200(79) 150(80)
Air Staging
•Axial 100% 220(77) 150(80)
•Deep 100% 160(83) 130(83)
Fuel Staging
•Burner 35% 190(80) 110(86)
— Boiler 35%
— SR2 = 1.1 195(79) 120(84)
— SR2 = 1.0 160(83) 110(86)
TABLE 4-2. DILUTION CORRECTIONS
Configuration
Dilution Added
Two Stage To Two Stage Reburn
Air Staging
Reference Figure
Waste Load
Fuel
No. 2 Fuel Oil/Pyridine
(2% N)
Natural Gas/Ammonia
(2% N)
Natural Gas/Ammonia
(5.8% N)
4-5 : "7 4-9
100% 65% 65% 65%
(SR2 = 10)
190 135 88* 110
185
315
130
260
169
160
4-7
100% 100%
Axial Deep
150 130
220
160
* Calculations showing the addition of dilution:
Natural gas/ammonia (5.8 percent N): 0.65 x 260 ppm = 169 ppm
No. 21uel oil/pyridine (2.0 percent N): 0.65 x 135 ppm = 88 ppm
60
-------
SECTION 5
QUALITY CONTROL EVALUATION REPORT
5.1 INTRODUCTION
The quality assurance objectives of this project were to assure that 1) the combustion conditions
in the furnace were representative of the intended experimental design, and that 2) the collected data
were accurate and useful. The QA/QC procedures documented in the Quality Assurance Project Plan
required that all continuous monitors and other measurement instruments, where possible, be carefully
calibrated using manufacturer's procedures or other published methods. Zero and span checks were
performed each test day on the continuous monitors. Annual calibrations of air flow and gas flow devices
were also performed. These data are documented.
5.2 DISCUSSION OF DATA QUALITY
A summary of the precision, accuracy, and completeness achieved in the relevant measurements
is listed in Table 5-1, along with the original goals for each type of measurement. In all cases, the
completeness, accuracy, and precision of measurements met or exceeded the Quality Assurance
objectives listed in the Quality Assurance Project Plan (QAPP) developed for the project.
Primary, secondary, and tertiary airflows on the Low NOx facility were measured by venturi
arrangements. Precision of these Venturis was checked annually during the test program. The precision
check procedure for these Venturis consisted of the following steps:
1. Close the damper on the stack of the low NOx facility operating with this causes the
in-lurnace pressures to be positive and eliminates in-leakage.
2. Vary each of the airs (primary, secondary, and tertiary) individually and read excess air
levels against a previously calibrated oxygen monitor. Now knowing that 9.69 mol of air are
required per mole of natural gas (refer to house natural gas analysis5), curves can be plotted
relating pressure drop across the respective Venturis to respective air flows in scfm.
61
-------
TABLE 5-1. DATA QUALITY SUMMARY
Completeness (%) Precision (%) Accuracy (%)
Parameter
(Method)
Planned
Actual
Planned
Actual
Planned
Actual
NO (Chemiluminescence)
a
>90
a
4.407
2.5
1.460
CO (NDIR)
a
>90
a
1.907
5
0.547
CO2 (NDIR)
a
>90
a
2.482
5
-2.722
O2 (Paramagnetic)
a
- >90
a
2.218
2.5
-1.504
Primary Air Flows (b)
a
>90
a
0.084 (% RSD)
a
C
Secondary Axial Air Flows (b)
a
>90
a
0.047 (% RSD)
a
C
Secondary Radial Air Flows (b)
a
>90
a
0.049 (% RSD)
a
C
Tertiary Air Flows (b)
a
>90
a
0.084 (% RSD)
a
C
Primary Gas Flows (b)
a
>90
a
c
5
C
Secondary Gas Flows (b)
a
>90
a
0.10 (% RSD)
5
c
a. Objectives were not defined in the QAPP
b. See Section 5.1 tor explanation
c. Not done during the program
-------
The primary natural gas flow used in the above calibration was measured by an orifice which had been
calibrated against a dry gas meter. Prior to air calibration, the natural gas orifice was inspected and
recalibrated if necessary.
Secondary gas flows on the low NOx facility were measured by a rotameter. These flows were
also calibrated using the scheme outlined above for the various air flows.
An overall facility precision check was performed each test day by repeating an arbitrarily
selected "baseline" data point at the end of a test sequence. If the NO and O2 monitor readings were
within 10 percent precision for the two baseline measurements, then overall facility operation was
considered to be consistent.
NO2, NOa, N20, NH3, HCN, and combustion gas velocity measurements were not made during
this program. Hence, Quality Assurance objectives for these parameters were not completed.
Primary natural gas flow venturi, thermocouples measuring process temperatures, Bourdon
gauges measuring process pressures, and suction pyrometer measuring combustion gas temperatures
were inspected periodically and found to be in good working order. Hence, accuracy of these elements
was not checked.
In summary, the collected data are sufficient for the original intentions of the project.
63
-------
SECTION 6
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
NO* Control, Volume 1: Utility Boiler Applications, EPA-600/9-86-02la {NTIS PB86-225042),
Environmental Protection Agency, July 1986, p. 4-1.
3. Lanier, W.S., and S.B. Robinson. EPA Workshop on N2O Emission from Combustion..
EPA-600/B-86-035 (NTIS PB87-113742), Environmental Protection Agency, September 1986.
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), Environmental Protection
Agency, July 1986, p. 2-1.
5. Mulholland, J.A., et al. Reburning Application to Firetube Package Boilers. EPA-6Q0/7-87-011
(NTIS PB87-177515}, Environmental Protection Agency, March 1987.
6. 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),
Environmental Protection Agency, July 1984.
7. 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-60CV7-85-013 (NTIS PB85-185874), Environmental Protection Agency, March 1985.
8. Chen, S.L., et al. NOx Reduction by Reburning with Gas and Coal-Bench Scale Studies. In:
Proceedings of the 1982 Joint Symposium on Stationary Combustion NOx Control, Vol. I,
EPA-600/9-85-022a (NTIS No. PB85-235604), Environmental Protection Agency, July 1985,
p. 16-1.
64
-------
9. Reed, R.D. Process for the Disposal of Nitrogen Oxide. John Zink Company, U.S. Patent
#1274637, 1969.
10. Wendt, J.O.L., et al. Reduction of Sulfur Trioxide and Nitrogen Oxides by Secondary Fuel
Injection. In: Proceedings of the Fourteenth Symposium (International) on Combustion, The
Combustion Institute, 1973, p. 897.
11. Takahasi, Y., et al. Development of Mitsubishi 'MACT In-Fumace NOx—Removal Process.
Presented at the U.S.^Japan NOx Information Exchange, Tokyo, Japan, May 25-30,1981.
Published in Mitsubishi Heavy Industries, Ltd.: Technical Review, 18(2).
12. Mulholland, J.A. and W.S. Lanier. Application of Reburning for NOx Control to a Firetube
Package Boiler. Journal of Engineering for Gas Turbines and Power, 107(3):739,1985.
13. Mulholland, J.A. and R.E. Hall. The Effect of Fuel Nitrogen in Reburning Application to a Firetube
Package Boiler. In Proceedings: 1985 Symposium on Stationary Combustion NOx Control,
Vol. 1, EPA-600/9-86-021a (NTIS PB86-225042), Environmental Protection Agency, July 1986,
p. 22-1.
14. Mulholland, J.A. and R.E. Hall. Fuel Oil Reburning Application for NOx Control to Firetube
Package Boilers. Journal of Engineering for Gas Turbines and Power, 109(2):207-214,1987.
15. Lanier, W.S., et al. Reburning Thermal and Chemical Processes in a Two-Dimensional
Pilot-Scale System. In: Proceedings: Twenty-first Symposium (International) on Combustion,
The Combustion Institute, Munich, W. Germany, August 13-18,1986, pp. 1171-1179.
16. Overmoe, B.J., et al. Pilot Scale Evaluation of NOx Control from Pulverized Coal Combustion by
Reburning. In Proceedings: 1985 Symposium on Stationary Combustion NOx Control, Vol. 1,
EPA-600/9-86-021 a (NTIS PB86-225042), Environmental Protection Agency, July 1986, p. 21-1.
17. Kelly, J.T., et al. Fuel Staging for Pulverized Coal Furnace NOx Control. In: Proceedings: 1982
Joint Symposium on Stationary Combustion NOx Control, Vol. I, EPA-600/9-85-022a (NTIS No.
PB85-235604), Environmental Protection Agency, July 1985, p. 17-1.
65s
/
------- |