EPA-6OO/7-89-015b
December 1989
FIELD EVALUATION OF LOW-EMISSION COAL BURNER
TECHNOLOGY ON UTILITY BOILERS
VOLUME II
Second. Generation Low-NO Burners
x
A, R. Abele, G. S. Kindt, R. Payne
(Energy and Environmental Research Corporation)
and
P. W. Waaoders
Babeock & Wilcox
20 S. Van Buren Avenue
Barberton, OH 44203
. EPA Contract 68-02-3130
EPA Project Officer: P. Jeff Chappell
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
— TECHNICAL REPORT DATA
(Please read fxislructions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/7-89-015b
3. RECIPIENT'S ACCESSION-NO,
4. TITLE AND SUBTITLE
Field Evaluation of Low-Emission Coal Burner Tech-
nology on Utility Boilers; Volume II. Second Gener-
ation Low-NOx Boilers
S. REPORT DATS
December 1989
6. PERFORMING ORGANIZATION CODE
7' a_uthoh,s,_4! R. Abele, G.S.Kindt, and R. Payne (EERC)
• and. P. W.. W.aanders (Babcock and Wilcox)*
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING OH0ANIZATION NAME AND ADDRESS
Energy and Environmental Research Corporation
18 Mason
Irvine, California 92718
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3130*
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; 9/78-6/86
14. SPONSORING AGENCY CODE
EPA/600/13
15. supplementary notes _^EERL project officer is P. Jeff C
541-3738. (*) Contract with Babcock and Wilcox, P.O.
44203-0351.
lappell, Mail, Drop 63, 919/
Box 351, Barber ton, OH
i6. ABSTHACTv^e report describes tests to evaluate the performance characteristics of
three Second Generation Low-NOx burner designs; the Dual Register faurner~~(DRB),
the Babcock-Hitachi NOx Reducing (HNR) burner, and the XCL burner. The three
represent a progression in development based on the orginal Babcock and Wilcox
DRB. Of particular interest was the identification of burner configurations which
would be suitable for application in the EPA LIMB (Limestone Injection Multistage
Burner) technology demonstration program at Chio Edison's Edgewater Station, Unit
4. The retrofit requirements for this unit were used to establish burner performance
criteria. The testing was conducted with nominal full-scale burner designs, having
a capacity of 78 million Btu/hr (22. 9 MW). Each burner was tested over a wide
range of operating conditions and hardware configurations, and with different coals,,,
With appropriate adjustments, all burners were capable of achieving NOx emissions
below 350 ppm (0% C2, dry).,.. wit^-flam~etl'engtfrs- les^Thah:22"frt6vTTn)T and with
acceptable carbon in ash.-'However, the XCL burner was judged to have the best
overall performance and to meet all the Edgewater boiler retrofit requirements.
Additional brief tests were conducted to evaluate the impact of burner design on
S02 removal by injected sorbent materials.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. cosati Field/Group
Pollution Calcium Carbonates
Coal Calcium Oxides
Combustion Electric Utilities
Nitrogen Oxides
Sulfur Oxides
Burners
Pollution Control
Stationary Sources
Low-NOx Burners
Dual Regis'ter Burners
HNR Burners
XCL Burners
13B
21D
2 IB
07B
13A
IS. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21- NO. OF PAGES
Ik^LJ
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73) 7~2
-------
N0T1CE
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.
-------
ABSTRACT
This report describes a series of tests designed to evaluate the
performance characteristics of three Second Generation Low-NQx burner
designs. These burner designs were, the Dual Register Burner (DRB), the
Babcock-Hitachi N0X Reducing (HNR) burner, and the XCL burner, which
represent a progression in development based upon the original Babcock &
Wilcox DRB. Of particular interest was the identification of burner
configurations which would be suitable for application in the EPA Limestone
Injection Multistage Burner (LIMB) technology demonstration program at Ohio
Edison's Edgewater Station, Unit 4. The retrofit requirements for this unit
were used to establish burner performance criteria.
The testing was conducted with nominal full-scale burner designs, having
a capacity of 78 x 10® Btu/hr (22.9 MW). Each burner was tested over a wide
range of operating conditions and hardware configurations, and with different
coal types. With appropriate adjustments, all burners were capable of
achieving N0X emissions below 350 ppm (0%,02, dry), with flame lengths less
than 22 feet (6.7m), and with acceptable carbon in ash. However, the XCL
burner was judged to have the best overall performance and to meet all the
Edgewater boiler retrofit requirements.
An additional brief series of tests was conducted to evaluate the impact
of burner design on SOg removal by injected sorbent materials. When
limestone and hydrated lime were injected into the upper furnace, remote from
the burners, no impact of burner design was observed. Significant
differences in SO2 removal were measured only when sorbent was injected
through the burner secondary air passage.
This work was carried out by Energy and Environmental Research
Corporation under EPA Contract 68-02-3130, through Babcock and Wilcox
Subcontract 940962 NR.
-------
TABLE OF CONTENTS
Section
Abstract . . . .
Table of Contents
List of Figures,,.
List of Tables .
1.0 SUMMARY. . .
1.1 Test Burners 1-2
1.2 Fuels and Sorbents 1-4
1.3 Burner Performance and N0X Emissions. .......... 1-6
1.4 SOg Reduction Potential 1-11
2.0 INTRODUCTION ........... ...... 2-1
3.0 BURNER DESIGNS AND EXPERIMENTAL SYSTEMS 3-1
3.1 Dual Register Burners 3-2
3.2 Babcock-Hitachi N0X Reducing Burner ........... 3-4
3.3 XCL Burner. . . ................ 3-7
3.4 Burner Hardware Components 3-11
3.5 Fuels and Sorbents. ................... 3-16
3.6 LWS Configuration 3-24
4.0 BURNER PERFORMANCE AND M0X EMISSIONS . 4-1
4.1 Dual Register Burners 4-1
4.2 Evaluation of the Babcock-Hitachi N0X Reducing Burner . . 4-23
4.3 XCL Burner. 4-31
4.4 Comparison of Burner Performance. 4-62
5.0 S02 REDUCTION POTENTIAL WITH SORBENT INJECTION ........ 5-1
5.1 Injection Configurations. ................ 5-1
5.2 Test Results. ........... ... 5-2
6.0 CONCLUSIONS. ....... ............. 6-1
6.1 Burner Performance and N0X Emissions 6-1
6.2 SO2 Reduction Potential 6-5
7.0 REFERENCES ...... .... 7-1
Page
• r,i n
•IE
T. 1 '2 -
. , v i =
:ix;
1-1
Preceding Page Blank
J
-------
LIST OF FIGURES
Figure Page
1-1 Correlation of NQX emissions with flame length . 1-8
3-1 Cross-sectional view of the B&W Dual Register Burner ..... 3-3
3-2 Cross-sectional view of the Babcock-Hitachi N0X Reducing
Burner ........ 3-6
3-3 Babcock-Hitachi NR burner coal swirler . ...... 3-8
3-4 Cross-sectional view of the B&W XCL Burner .......... 3-9
3-5 XCL burner with fixed outer vanes ...... 3-12
3-6 B&W coal pipe venturi . 3-14
3-7 Coal impellers used during second generation low-NOx
burner testing ........ . 3-15
3-8 Flame stabilizing ring 3-17
3-9 Expanded nozzle tip on the XCL burner 1 . . . . . 3-18
3-10 Typical coal particle size distributions . 3-25
3-11 Sorbent particle size distribution ....... 3-28
3-12 Insulation pattern in the LWS for second generation
low-NOx burner tests ........ 3-29
3-13 Typical results from rear wall probing for CO levels .... 3-32
4-1 Sensitivity of low velocity DRB N0X emissions to
burner adjustments . ........ 4-5
4-2 Sensitivity of low velocity DRB with 75° impeller to
burner adjustments , 4-6
4-3 Performance of low velocity DRB with diffuser as a
function of excess air and firing rate ... .... 4-8
4-4 Performance of low-velocity DRB with 75° Impeller as a
function of excess air and firing rate 4-9
4-5 Performance of low-velocity DRB with DeNOx stabilizer
as a function of excess air and firing rate .... 4-10
-------
LIST OF FIGURES (Continued)
Figure Page
4-6 Performance of low-velocity DRB with FSR/ASP/Swirler
Assembly as a function of excess air and firing rate ..... 4-11
4-7 Comparison of performance of low-velocity DRB config-
urations 4-13
4-8 Sensitivity of Phase V DRB configurations to outer
register position 4-18
4-9 Effect of excess air on Phase V DRB configurations'
performance 4-21
4-10 Effect of firing rate on Phase V DRB configurations'
performance ......... 4-22
4-11 Comparison of the Phase V DRB with the low-velocity DRB .... 4-25
4-12 Sensitivity of HNR burner performance to burner adjustments - . 4-28
4-13 Effect of excess air on performance of HNR burner with
coal diffuser at 78 x 10" Btu/hr 4-29
4-14 Effect of excess air and firing rate on performance of
HNR burner with modified outer register 4-30
4-15 Comparison of HNR burner configurations' performance 4-32
4-16 Results of initial screening tests of XCL burner for four
configurations ....... ..... 4-37
4-17 Results of initial screening tests of XCL burner with 30°
impeller ........ 4-38
4-18 Effect of excess air on performance of XCL burner with
coal diffuser 4-39
4-19 Effect of excess air on performance of XCL burner with
30° impeller 4-41
4-20 Comparison of coal di ffuser and 30° impeller XCL burner
configurations .......... 4-43
4-21 Results of final XCL burner screening tests with
standard coal nozzle . ........ 4-46
-------
LIST OF FIGURES (Concluded)
Figure Page
4-22 Results of final SCL burner screening tests with expanded
nozzle tip . 4-48
4-23 Effect of coal nozzle diameter on XCL burner performance . . , 4-50
4-24 Results of final XCL burner screening tests with 30°
impeller and fixed outer vanes ................ 4-52
4-25 Performance characteristics for five XCL burner config-
urations 4-54
4-26 Performance of XCL burner with standard coal nozzle,
diffuser, and fixed outer vanes firing Pittsburgh #8 coal . . 4-56
4-27 Performance of XCL burner with standard coal nozzle,
diffuser, and fixed outer vanes firing Utah coal . 4-57
4-28 Performance of XCL burner with standard coal nozzle,
diffuser, and fixed outer vanes firing lower
Kittanning coal 4-58
4-29 Comparison of performance of XCL burner with standard coal
nozzle diffuser, and fixed outervanes for three different
coals 4-60
4-30 Performance of XCL burner with diffuser and fixed outer
vanes for three coals 4-61
4-31 Comparison of XCL burner performance with diffuser and
30° impeller—fixed outer vane configuration 4-63
4-32 Correlation of N0X emissions with flame length 4-65
5-1 Sorbent injection locations—with coal and upper furnace . . . 5-4
5-2 Burner exit sorbent injection nozzle configuration ...... 5-5
5-3 Summary of SO2 capture with near burner injection of
Vicron 45-3 limestone 5-6
5-4 Summary of SO2 capture with Col ton hydrated lime injection . . 5-8
5-5 Effect of sorbent type on SO2 capture with injection at
2500°F 5-9
f-;- . viii
-------
LIST OF TABLES
Table Page
1-1 Components of Basic Burner Designs ......... 1-3
1—2 Composition of Test Coals 1—5
1-3 Second Generation Low N0X Burners with Flames <22 ft long
(78 x 106 Btu/hr, SRj = 1.20) 1-9
1-4 Optimum Burner Configurations for Edgewater Unit 4
(78 x 106 Btu/hr SRj = 1-20) 1-12
3-1 Dual Register Burner Hardware Modifications 3-5
3—2 Composition of Test Coals 3—19
3-3 Coal Ash Characteristics ......... . 3-20
3-4 Predictions of N0X Emissions Based on Coal Composition .... 3-22
3-5 Daily Coal Variations ........... 3-23
3-6 Coal Fineness Variations—Cumulative Mass Percent Undersize . 3-26
3-7 Physical and Chemical Properties of Tested Sorbent 3-27
3-8 CO Measurements at Rear Wall for Selected Conditions 3-33
4-1 Range of Screening Tests for Low-Velocity Dual Register
Burner 4-3
4-2 Low Velocity DRB Performance Summary ...... 4-15
4-3 Range of Screening Tests for Standard Phase V DRB 4-17
4-4 Selected Optimum Settings for Phase V DRB 4-19
4-5 Summary of Key Phase V DRB Test Conditions . 4-24
4-6 Range of Screening Tests for HNR Burner ...... 4-27
4-7 Summary of Key HNR Burner Test Conditions 4-33
4-8 Range of Initial Screening Tests of XCL Burner 4-35
4-9 Summary of Key Test Conditions from Initial XCL Burner
Screening Tests 4-42
-------
LIST OF TABLES (Continued)
Table Page
4-10 Range of Final Screening Tests of XCL Burner . . , . 4-45
4-11 Summary of Final XCL Burner Screening Test Conditions .... 4-53
5-1 Burner Configurations for Sorbent Injection Tests 5-3
6-1 Optimum Burner Configurations for Edgewater Unit 4
(78 x 1Q6 Btu/hr SRj = 1.20) 6-4
-------
1.0 SUMMARY
The objective of this contract'was to evaluate the performance of low-
emission burner technology, specifically the U.S. Environmental Protection
Agency developed Distributed Mixing Burner (DMB), in a utility boiler
application. The initiation of-the LIMB (Limestone Injection Multistage
Burner) technology demonstration at the Ohio Edison Edgewater Station, Unit
4, provided an opportunity to broaden the overall scope of this project..
The objective of this LIMB program with respect to burner design was to
provide a commercial pulverized coal burner that demonstrates a reduction in
nitrogen oxide (N0X) emissions of at least 50 percent relative to
uncontrolled performance of the original Babcock & Wilcox {B&W) Circular
burners. This performance must be achieved within the following requirements
for the Edgewater boiler:
• 78 x 10® Btu/hr heat input per burner.
• Throat diameter no greater than 35 inches.
• Mechanical reliability meeting commercial standards.
• Flame length less than the firing depth of the boiler, 22 ft 3 in.
• Burner pressure drop within fan limitations nominally 5 in. W.G.
• Acceptable combustion efficiency.
The three B&W 1ow-N0x burner designs considered; the Dual Register
Burner (DRB), Babcock-Hitachi N0X Reducing (HNR) burner, and the XCL burner,
were tested at full scale in the EPA Large Watertube Simulator (LWS) to
determine the optimum design for use at the Edgewater boiler as part of this
contract. Full size 78 x 10® Btu/hr burners can be accommodated by the LWS,
minimizing seale-up questions, and, by coincidence, the LWS has a firing
depth of 22 ft, essentially the same as Edgewater Unit 4. Screening tests of
the three basic burner designs were conducted firing Pittsburgh #8 coal, the
coal to be used during the LIMB demonstration, to determine optimum operating
conditions. In addition to available burner adjustments, a number of burner
hardware components were also evaluated to establish the optimum burner
-------
design. A brief set of sorbent injection tests was completed for a selected
configuration of each basic burner to determine the affect of burner design
on SO2 capture. Following the screening tests of the three burners, selected
XCL burner configurations were character!zed with three additional,
distinctly different coals to broaden the application of this new burner.
1.1 Test Burners
The three Second Generation Low-NOx burner designs considered for this
program represent a progression in development which began with the B&W Dual
Register Burner, The burners are based on the same basic concept, using
multiple air zones to allow controlled, delayed-mixing of the fuel and
combustion air. The Dual Register Burner was the first in the line of
development. Data from tests of two DRB designs are presented: a DRB
designed to fit the same burner exit as the HNR and XCL burners, denoted the
Low-Velocity DRB and a standard Phase V DRB tested under B&W P.O. 635-
0A008408 DM. The Babcock-Hi tachi NR burner, in turn, incorporated
modifications to the basic DRB by varying air flow distribution and
velocities and by adding hardware enhancements. The hardware enhancements
were an extended baffle, or Air Separation Plate, between the inner and outer
secondary air zones and a Flame Stabilizing Ring at the exit of the coal
nozzle. The XCL burner represents the product of B&W development to improve
the HNR burner with enhanced mechanical reliability and reduced pressure
drop, and air measurement capability.
The burners all consist of three concentric passages: a central,
cylindrical coal nozzle surrounded by two annular secondary air passages.
Both secondary air passages for all three incorporate swirl generators,
either axial spin vanes or radial register louvers. Each burner also has
some control, over air distribution between the inner and outer air zones.
The key components of each basic burner design are summarized in Table 1-1.
A number of alternative burner components were evaluated with each of
the three basic burners. Components that can be classified as coal pipe
devices included:
Slr2".
-------
TABLE 1-1. COMPONENTS OF BASIC BURNER DESIGNS
Component
DRB
HNR
XCL
Coal Dispersal
Diffuser
Flame Stability
Ring Swirler
Diffuser
Inner Secondary Swirl
Adjustable Axial
Spin Vanes
Adjustable Axial
Spin Vanes
Adjustable Axial
Spin Vanes
Outer Secondary Swirl
Register of
Radial Louver
High Swirl
Radial Register
Adjustable Axial
Spin Vanes
Inner Secondary Flow
Control
Sliding Sleeve
Damper
Sliding Sleeve
Damper
Sliding Sleeve
Damper
Outer Secondary Flow
Control
Dependant on
Swirl
Dependant on
Swirl
Sliding Sleeve
Damper
Inner/Outer Secondary
Separation
Internal Divider
Extended Air
Separation Plate
Extended Air
Separation Plate
Secondary Air Flow
Measurement
None
None
Inner/Outer
Zone Pi tot Grids
-------
• Coal diffuser--a bluff body dispersal device located at the inlet
of the coal nozzle (tested with DRB, HNR, and XCL).
• Coal pipe venturi —a coal dispersal device located at the inlet of
the coal nozzle consisting of a venturi to concentrate then
disperse the coal stream (tested only with Phase V DRB).
• DeNOx Stabi1izer--a proprietary B&W device designed for easy
insertion into the coal nozzle. (Tested with Low-Velocity DRB and
XCL burners).
Five coal impellers and swirlers were evaluated:
• 75° included angle impeller (Low-Velocity DRB)
• 20° included angle impeller (XCL)
• 30° included angle impeller (XCL)
• Open impeller (XCL)
• HNR burner swirler
Two coal nozzle exit devices were tested: the Flame Stabilizing Ring, tested
with all three burner designs, and an expanded nozzle tip tried only with the
XCL burner. Modifications to the secondary air zones included modifications
to widen outer air vanes (HNR and XCL burners), addition of an extended Air
Separation Plate (all three burners), and installation of fixed swirl vanes
in the outer zone (XCL burner only).
1.2 Fuels and Sorbents
Four different coals were utilized during the Second Generation Low-N0x
burner tests. Table 1-2 summarizes the composition of the coals. The
Pittsburgh #8 coal was the primary fuel used throughout the burner tests.
Pittsburgh #8 coal is a high-volatile A bituminous coal selected for the LIMB
demonstration project at Edgewater Station Unit 4. Since the ultimate goal
1-4
-------
f)
TABLE$-2. COMPOSITION OF TEST COALS
Coal
1 Pittsburgh #8
Utah
Comanche
. \
Lower
Kittanni ng
Reporti ng
Basis
1 As
! Rec'd
Dry
As
Rec 'd
Dry
As
Rec'd
Dry
As
Rec'd
Dry
Proximate (wt %)
Moisture
3.50
0.00
6.11
0.00
22.44
• 0.00
2.43
0.00
Ash
i 12.92
13.40
8.02
8.55
5.00
6.45
10.19
10.44
Volatile
j 33.75
34.98
41.26
43.96
36.12
. 44.87
23.93
24.52
Fixed C
j 49.83
51.62
44.60
46.73
37.72
48.68
63.45
65.04
Heating Value
Btu/lb
;i2,i77
12,618
12,288
13,088
9,325
12,026
13,551
13,888
MMF Btu/lb
14,876
14,440
12,939
15,701
HAF Btu/lb
14,626
14,311
12,855
15,507
Ultimate, (wt %)
Moi sture
;• 3.50
0.00
6.11
0.00
22.44
0.00
' 2.43
0.00
Carbon
68.13
70.59
- 68.58
71.86
54.25
69.97
76.82
78.73
Hydrogen
4.63
4.79
5.16
5.49
3.80
4.91
4.54
4.65
Nitrogen
: i.2i
1.26
1.28
1.36
0.76
0.98
1.12
1.15
Sulfur
3.22
3.30
0.60
0.64
0.43
0.56
1.13
1.16
Ash
12.92
13.40
8.02
8.55
5.00
6.45
10.19
10.44
Oxygen*
6.41
6.63
10.24
10.91
13.32
17.14
3.77 •.
3.87
Forms of Sulfur
(wt %)
'
Sulfate
0.22
0.23
0.01
0.01
0.02
0.02
0.01
0.01
Pyritic
1.62
1.65
0.13
0.13
0.09
0.12
0.53
0.54
Organic
1.38
1.42
0.46
0.50
0.32
0.42
0.60
0.62
*Oxygen determined by difference.
,1-5
-------
of these burner tests was the selection of the optimum burner for retrofit at
the Edgewater boiler, the use of the same coal enabled direct projection of
expected burner performance. The other three coals represent a wide range of
coal types and were used to characterize the performance of the optimized XCL
burner. Utah coal is a western high-volatile A bituminous coal from the
Starpoint mines in Wattis, Utah. Utah coal has been used at EER as the base
fuel for most of the low-emission, high-efficiency burner development
projects. Use of this coal allows comparison of the Second Generation low-
N0X burner performance with an existing data base of other burners. The
Comanche coal is a subbituminous coal from Wyoming. The Lower Kittanning is
a medium-volatile bituminous coal.
Two sorbents were used during these tests to evaluate SO2 reduction
potential by in-furnace injection, Vicron 45-3 limestone and Colton hydrated
lime. These sorbents have been used at EER as typical limestone and hydrated
lime materials in the development of LIMB technology. Vicron is nominally 99
percent pure CaCC>3 with a mass median diameter of 9.8 /xm. The Colton
hydrated lime is nominally 96 percent Ca(0H)2 with a median particle size of
4.0 /*m.
1.3 Burner Performance and N0X Emissions
Optimization tests of the three basic burner designs screened the
available burner adjustments as well as the various burner component
configurations. The three basic components of each burner; the coal
injector, inner secondary air zone, and outer secondary air zone, were
evaluated in these screening tests. The results from these tests can be
easily generalized for all three low-NOx burners with respect to sensitivity
of performance. In each case, the coal injector was the dominant factor that
determined the key performance characteristies of N0X, flame length, and
carbon burnout. Both the design of the coal injector and the available
adjustments could produce up to 67 percent reduction in NCX emissions. The
outer secondary air zone, the degree of swirl and the air flow rate through
the outer passage, was second in Importance to burner performance. The inner
1-6
-------
air zone parameters of swirl and air flow rate generally had the least effect
on performance.
Consistent and recurring throughout the screening tests of al1 three
burners was the close correlation of N0X emissions with flame length. Data
from tests of the Dual Register Burner, HNR burners, and the initial
screening tests of the XCL burner, summarized in Figure 1-1, clearly shows
this correlation. At 20 percent excess air and full load conditions, this
data indicates that for a flame less than the firing depth of the Edgewater
boiler, N0X emissions in the range of about 300-400 ppm* were achieved by
several burner configurations. With flame length as the most severe
constraint at the Edgewater boiler, only eight of the 20 burner configura-
tions tested in this program and five Phase V DRB configurations achieved
flames less than 22 ft in length. These are listed in Table 1-3.
Dual Register Burner. The Low-Velocity DRB designed to fit within the
same throat as the other two candidate burners, the HNR and XCL, produced
excessively long flames for three of the four configurations. Only an
unoptimized 75° impeller design produced a flame less than the 22 ft furnace
firing depth. Full load N0X emissions for the three configurations which
produced flames over 22 ft long were low, ranging from 264 to 386 ppm. The
75° impeller-equipped configuration produced an average of 732 ppm N0X at
full load with 17-18 ft flames. Available data from the B&W sponsored test
program suggest that the performance of this Low-Velocity DRB is not
representative of the current commercial Phase V DRB. Flames less than 22 ft
long could be achieved with five different configurations of the Phase V DRB
with its slightly higher vel ocities, albeit with si i ghtl y higher N0X
emissions (292-372 ppm). To achieve that performance required tightly closed
burner settings, uncharacteristic of typical DRB operation, which produced
burner pressure drops over 6 in. W.G.
*A11 emission concentrations reported corrected to 0 percent Og on a dry
basis, except where indicated.
1-7
-------
1000
600
C2J 400
liit
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate: 80 x 106 Btu/hr
SRj =1.20
DRB
O Diffuser
A 75° Impeller
~ DNS
O FSR/ASP/Swirler
. HNR
o
0
XCL
Swi rler
Diffuser
Di ffuser
30° Impeller
DNS
FSR
12 14 16 18
Observed Flame Length (feet)
- Figure Correlation of N0X emissions with flame length,
-------
C.TABLE 1-3 . SECOND GENERATION LOW MOx BURNERS WITH. FLAMES^ FT LONG
' " (78 x 1Q6 Btu/hr, SRT = 1.20) _
Burner
Configyration
MQX
0 0% 02
(ppm)
Flame
Length
(ft)
Fly Ash
Carbon
(wt.%/
Burner fiP
(in.
W.G.)
Low-Velocity DRB
75° Impeller
708
18
7.28
6.0
Phase V DRB
Diffuser
372
20-21
6.12
10.8
Vertturi
350
20-21
6.45
11.0
Diffuser, ASP
326
22
3.20
6.4
Diffuser, FSR
292
22
6.96
10.5
Diffuser, FSR, ASP
328
22
5.16
11.0
HNR
Swi rl er
348
18-20
N/A
7.20
Diffuser
289
20
3.34
7.50
XCL
DNS
288
20
N/A
8.20
30° Impeller, Standard
Nozzle
374
20-22
4.42
3.30
30° Impeller, Expanded
Nozzle
546
19-20
1.36
4.30
20° Impeller, Expanded
Nozzle
338
21
4.92*
4.90
30° Impeller, Standard
Nozzle, Fixed Outer
Vanes
420
21-22
3.40
4.60
~Data for SRT =1.16
-------
Babcock-Hitachi NR Burner. The Babcock-Hitachi NR burner relies on
biasing the secondary combustion air to the outer zone coupled with a very
high degree of swirl for flame shaping and N0X control. Minimum N0X
emi ssi ons were 222 ppm with a flame over 22 ft long using burner settings
typical of Babcock-Hitachi practice. The two other configurations evaluated
produced higher NCX emission, 289-348 ppm, but with correspondingly shorter
flames, 18-20 ft long. In each case, however, burner pressure drop was about
7 in. W. G.
XCL Burner. The XCL burner, which represents the latest development in
the B&W Dual Register Burner evolution, was tested in 13 configurations
during 2 series of tests. This burner design demonstrated the most potential
to meet the LIMB demonstration goals because of its inherent flexibility.
N0X emissions ranged from 194-700 ppm with flames from 12 to over 22 ft long.
Only five configurations yielded flames 1 ess than 22 ft long with N0X
emissions from 288 to 546 ppm.., The unique B&W DeN0x Stabilizer achieved the
lowest emissions but required burner settings producing a burner pressure
drop of 8.20 in. W. G. The other configurations were, based on either a 20°
or 30° coal impeller design. The impel 1 er-~equipped XCL configurations could
achieve a.wide range of N0X and flame length by the adjustment of the
impeller position, all with burner pressure drop less than 5 in. W. G. At
optimum conditions, the 20° impeller in an expanded coal nozzle gave 338 ppm
N0X while the 30° impeller in the standard coal nozzle gave 374 ppm.
From these numerous burner configurations, two stand out as suitable for
application for the LIMB demonstration. All configurations tested met the
requirements of a firing capacity of 78 x 106 Btu/hr burner, a throat
diameter no greater than 35 inches, and mechanical reliability meeting
commercial standards. The Edgewater boiler also imposed the constraint on
flame length, 22 ft, and on maximum tolerable burner pressure drop, about
5 in. W.G. In addition, "the burners had to produce a stable flame with low
emissions but high combustion efficiency. The two configurations meeting all
those conditions were:
¦. 1-10-
-------
• XCL burner with 30° impeller in the standard coal nozzle with
appropriate outer vane design.
• XCL burner with 20° impeller in an expanded coal nozzle.
Performance of these two configurations is summarized in Table 1-4. In
addition to meeting all Edgewater boiler requirements, the two impeller-
equipped XCL burner configurations offer a very effective means of optimizing
performance to suit the application. Thi s control mechanism is the
adjustable position of the coal impeller. For both designs, flame length and
N0X emissions can be varied simply by moving the impeller a matter of inches.
The impeller adjustment can thus be used to tune the burner for maximum N0X
reduction within the constraints of available firing depth.
1.4 SOg Reduction Potential
A brief series of sorbent injection tests were performed for a selected
configuration of each burner design; DRB, HNR and XCL burners. Two near
burner locations and two upper furnace locations were evaluated at nominal
full-load conditions. Vicron 45-3 limestone was injected through the three
locations closest the burner and Col ton hydrated lime was injected through
the two upper furnace locations.
As expected, SOg capture was not influenced by burner design when
sorbent was injected through the two upper-furnace locations. At the lower
level, corresponding to a gas temperature of about 2500°F, limestone achieved
33 percent capture at Ca/S molar ratio of 2 while the hydrated lime achieved
36 percent capture. At the uppermost level, corresponding to 2150°F gas
temperature, the hydrated lime achieved 38 percent capture at a Ca/S molar
ratio of 2.
The two near-burner locations considered were injection with the coal
and injection through four high-velocity nozzles in the outer secondary air
passage. Limestone injected with the coal achieved only 28-32 percent
capture at a Ca/S molar ratio of 2 for all three burners. The injection of
,1-11
-------
TABLE 1-4. OPTIMUM BURNER CONFIGURATIONS FOR EDGEWATER
UNIT 4 {78 x 106 Btu/hr SRT = 1.20)
Burner
flOx
? 0% 02
(ppm)"
Flame
Length
(ft)
Fly Ash
Carbon
(wt %)
Burner AP
(in. W.G.)
XCL w/30°Impe11 er,
Standard Coal
Nozzle
374
20-22
4.42
3.30
XCL w/20° Impeller,
Expanded Coal
Nozzle
339
21
4.92
4.90
limestone through the outer secondary air passage yielded higher SOg capture
for the DRB equipped with the 75° impeller (40 percent at CaS molar ratio of
2) than for the HNR and XCL burners (30 percent capture). This difference
in results is not fully understood, but appears to be associated with the
near-burner flow field as suggested by flame shape.
1-12
-------
2.0 INTRODUCTION
In the broad context of the title of this project, the objective of this
program Is the evaluation of low-emission burner technology for utility
boiler application. The particular burner technology of interest was the
Distributed Mixing Burner (DMB), developed by the U.S. Environmental
Protection Agency (EPA) at Energy and Environmental Research Corporation
(EER), The effectiveness of the DMB was to be determined by direct
comparison with the original equipment burners in one representative
operating utility boiler. Difficulties in finding a host boiler to
participate in a demonstration, retrofitting existing burners with the new
DMB technology resulted in delays to the overall program. These delays, in
turn, caused escalating costs for a full utility boiler retrofit with DMBs.
Because of these problems, the program was restructured to achieve its
objective without installation of the DMB in a utility boiler. The approach
taken was extensive testing of DMBs at two scales and two B&W commercial
burner designs in the EPA Large Watertube Simulator (IWS) coupled with field
tests at utility boilers equipped with the two BAW commercial burners. This
approach provided data for burner scaleup, performance characteristics of the
DMB compared to commercial burners, and commercial burner performance in
utility boilers. With this data, the expected performance of DMBs could be
extrapolated to utility boilers with some confidence.
The EPA Limestone Injection Multistage Burner (LIMB) demonstration
provided motivation to further extend the scope of this program. The LIMB
project, being conducted at Ohio Edison's Edgewater Station, Unit No. 4, is a
demonstration of combined N0X and SO2 control with low-NOx burners and in-
furnace sorbent injection for SO2 control. The objective of this LIMB
demonstration with respect to burner design and N0X emissions is to achieve
50 percent N0X reduction compared to uncontrolled baseline performance of the
original burners. This performance must be achieved within the constraints
of the Edgewater boiler. These constraints or requirements include:
• High combustion efficiency.
• 78 x 106 Btu/hr heat input per burner.
2-1
-------
• Throat diameter no greater than 35 inches.
• Flame length less than the firing of the boiler, 22 feet, 3 inches.
• Burner pressure drop commensurate with fan limitations, nominally 5
inches W.G.
• Mechanical reliability meeting commercial standards.
Three B&W 1 ow-NOx burner designs were under consideration for the LIMB
demonstration at the Edgewater boiler: the Dual Register Burner (DRB),
Babcock-Hitachi N0X Reducing burner (HNR), and the XCL burner. The DRB and
the HNR burners have demonstrated low emissions in utility boilers and are
commercially available equipment. The XCL burner was developed based on the
HNR burner concept, but incorporating mechanical enhancements. The Edgewater
boiler design, the shallow firing depth in particular, imposes a severe
constraint on these low-NOx burners. Boilers of its vintage (1950-60) were
desi gned with hi gh swirl, hi gh turbulence Circular burners which produce
short, wide flames conducive to good combustion and high N0X emissions. Low-
NOx burners have longer, narrower flames from the inherent delayed mixing
designs. This constraint was a key motivation to test full size burners
prior to installation at Edgewater to demonstrate compatibility to the boiler
firing depth.
This low-emission burner technology evaluation program provided a unique
opportunity to benefit both this low-emission burner evaluation program and
the LIMB demonstration. This program benefited by broadening the data base
of low-emission burner technology with three additional burners and by
directly participating in the application of low-NOx burners to an operating
utility boiler. The LIMB project, in turn, was provided with an opportunity
to develop and demonstrate a burner that would provide optimum performance
within the constraints of the Edgewater boiler prior to installation. By
coincidence, the test facility used for this program, the LWS, has a firing
depth of 22 feet and thus provided the same constraint as the Edgewater
boiler. Full size, 78 x 106 Btu/hr burners can be accommodated and questions
of scale-up are minimized.
-------
The specific goals for the evaluation of these second generation 1 ow-NOx
burners were to;
1. Evaluate and optimize the performance of three low-emission
burners.
2. Determine the compatibility of each burner to the Edgewater boiler,
3. Project emissions performance of the optimized burner to Edgewater
Unit 4.
These objectives were accomplished in two series of tests in the LWS. The
first series consisted of screening tests to optimize the performance of each
burner In terms of low-NOx emissions, flame length, and combustion effi-
ciency. These screening tests included evaluation of both adjustable burner
parameters and selected hardware modifications over a range of firing rate
and excess air. A brief set of in-furnace sorbent injection tests was
completed for each burner to determine the influence of burner design. The
optimum configuration from these initial screening tests was further refined
in the second test series. The final burner configuration was then
character!'zed with four distinct coal types, including typical eastern
bituminous, subbituminous, western bituminous, and medium-volatile
bituminous. The final burner confi gurati on became the 1ow-N0x burner
selected for the LIMB demonstration.
-------
3.0
BURNER DESIGNS AND EXPERIMENTAL SYSTEMS
The evaluation of B&W second generation 1ow-NOx burners involved testing
three coal burner designs:
• Dual Register Burner
• Babcock Hitachi N0X Reducing Burner
• XCL Burner
These burners evolved from the same basic concept, using multiple air zones
to allow controlled, delayed-mixing of the fuel and combustion air. In fact,
these three burners represent a progression of development which began with
the B&W Dual Register Burner. Babcock-Hitachi, in turn, incorporated
modifications to the basic DRB by varying air flow distribution and
velocities and by adding burner hardware enhancements. The XCL burner
represents the product of B&W development to improve the HNR burner with
enhanced mechanical reliability, air flow measurement capabilities, and
reduced burner pressure drop. The test burners were designed to meet the
requirements at the Edgewater boiler, having a firing capacity of 78 x 10®
Btu/hr and a throat diameter no greater than 35 inches.
The tests were conducted by EER in the EPA Large Watertube Simulator at
EER's El Toro, California, test facility. Parametric screening tests,
including hardware modifications, were conducted for each burner to optimize
their performance for 1 ow-NOx emissions, high combustion efficiency, flame
length less than 22 feet, and burner pressure drop. The coal used for burner
optimization was Pittsburgh No. 8 coal, the high volatile eastern bituminous
coal to be used during the LIMB demonstration at Edgewater Unit 4. The final
optimized burner was evaluated with three alternate coals which represent a
wide range of coal types: Utah western bituminous, Lower Kittaning medium-
volatile bituminous, and Comanche, a western sub-bituminous coal. In addi-
tion to optimizing the burners for low-NOx emissions, a brief series of sorb-
ent injection tests were also conducted for each burner to determine whether
burner design affects the degree of SO2 control. All testing was conducted
in accordance with established Quality Assurance procedures following EPA
f-3-1" '¦
-------
guidelines. Documentation of the Quality Assurance program is in Part V,
Appendix A, of this report.
3.1 Dual Register Burners
The Dual Register burner was developed by B&W to replace the Circular
burner. The DRB has undergone several phases of development, incorporating
modifications to improve operabil ity and combustion efficiency while
achieving low-NOx emissions1-4 . For this program, B&W designed a DRB to fit
the requirements at the Edgewater boiler. The test burner was designed for a
nominal firing rate of 78 x 106 Btu/hr, but was sized to fit the same throat
as the HNR and XCL burners. This was done by B&W to facilitate burner
changes in the LWS and to directly compare the effects of burner hardware.
This resulted in secondary air velocities lower than standard DRBs. While
this Low Velocity DRB was the subject of this program, a 78 x 106 Btu/hr
Phase V DRB with standard design velocities was the subject of B&W sponsored
tests (B&W P.O. 635-OA008408 DM) in the LWS outside this project. A summary
of the standard Phase V DRB tests is included in this report for comparison.
The basic configuration of the Low Velocity DRB and the standard Phase V
DRB is the same. A cross-section of the basic DRB is shown in Figure 3-1.
The burner consists of three concentric passages: a central , cylindrical
coal nozzle surrounded by two annular secondary air passages. Coal, trans-
ported by primary air, enters the coal nozzle through a 90° elbow. In the
basic configuration, a bluff body diffuser is located at the inlet to the
coal nozzle. This diffuser produces a uniform coal distribution across the
coal nozzle without imparting any swirl to the primary air/coal stream. The
secondary air is split between two annular passages. The inner passage is
equipped with an adjustable damper for flow control and a set of adjustable
axial spin vanes for swirl control. The outer secondary air passage utilizes
adjustable radial register vanes, or doors, for both flow and swirl control.
The DRBs, as well as the HNR and XCL burners, utilized a steel throat and
exit which were water-cooled during the tests in the LWS. In actual boiler
installations, the burner exit is generally formed by tube bends in the water
wall covered by a thin refractory layer.
3-2
-------
LO
Primary Air + Coal
Outer Secondary
Air
1 n .^Adjus
_ Regi
Adjustable
ster
Vanes
Inner Secondary
Air
Adjustable
Inner
Damper:
Coal
Diffuser
Adjustable
Spin
Vanes
Inner Secondary
Air
Outer Secondary
Air
Figure 3-1. Cross-sectional view of the B&W Dual Register Burner.
f
-------
In addition to the basic configuration of the Low Velocity DRB and
standard Phase V DRB, several hardware modifications were evaluated to
shorten the flame length and/or enhance flame stability. The modifications
evaluated with each DRB are listed in Table 3-1. These modifications are
described in Section 3.4.
3.2 Babcock-Hitachi N0X Reducing Burner
The Babcock-Hitachi NR burner, shown in Figure 3-2, was developed from
the basic DRB configuration to meet the stringent emissions limits in Japan.
The HNR burner retains the general design features of the DRB, namely a
central coal nozzle surrounded by two concentric annular secondary air
passages; axial spin vanes for the inner zone; a sliding damper to control
air flow to the inner zone, and a register of radial vanes to control the
outer secondary air. From this basic concept, Babcock-Hi tachi developed
modifications to achieve lower emissions than the DRB. The modifications
included air flow distribution between the inner and outer secondary air
passages, secondary air velocities, and burner hardware enhancements to
produce a unique flow pattern.
The burner hardware enhancements that are integral components to the HNR
burner include: a Flame Stabilizing Ring (FSR), Air Separation Plate (ASP),
and an outer secondary air register with a larger number of vanes. The FTame
Stabilizing Ring is located at the exit of the coal nozzle. This device is
designed to produce a stable flame core by creating recirculation eddies at
the coal nozzle exit. The Air Separation Plate is a baffle between the inner
and outer secondary air passages that extends the division of the two air
zones into the exit. This delays mixing between the inner and outer air
zones. The angle of this Air Separation Plate also deflects the outer
secondary air away from the flame core to delay mixing and shape the flame.
Babcock-Hitachi also modified the outer secondary air register to increase
the degree of swirl generated. This was done by increasing the number of
register vanes of the basic DRB outer register assembly.
-------
TABLE 3-1. DUAL REGISTER BURNER HARDWARE,MODIFICATIONS
if
Hardware Modification
Low Velocity
DRB
Phase V
DRB
Coal Diffuser—Current DRB Standard
X
X
Coal Venturi—Original DRB Practice
X
Impeller--10-in. Diameter
75° Included Angle
X
DeNOx Stabilizer
X
Air Separation Plate
X
Flame Stabilizing Ring
X
Air Separation Plate + Flame
Stabilizing Ring
X
X
1 ! 1 •
-------
Outer Secondary
Air
Adjustable
Register
nV.aneSss^
Adjustable Inner
Damper
Adjustable
Spin
Vanes
rn Coal
III Pi ff user
n » I
-F1ame
Stabil-
izing
Air
Separation
Plate
Primary Air + Coal
i
i
\
\ • '
Figure 3-2. Cross-sectional view of the Babcock-Hitachi N0x Reducing Burner.
-------
During the combustion tests of the HNR burner in the IWS, optimization
included evaluating several variations of the basic burner arrangement. The
configurations evaluated were the HNR burner with a coal swirler, the HNR
burner with the B&W coal diffuser, and the HNR burner with a modified outer
register assembly. The coal swirler, shown in Figure 3-3, is located in the
coal pipe just upstream of the Flame Stabilizing Ring. The relatively open
design of the swirler imparts a small amount of spin to the primary coal
stream. The swirler is incorporated in the basic configuration of the HNR
burner. The standard B&W coal diffuser, the current coal nozzle device used
in the standard DRB, was evaluated as an alternative to the swirler in the
HNR burner. The di ffuser is a proven low wear device and low N0X device, and
thus could improve the reliability of the HNR burner. The outer register
assembly was also modified to meet Babcock-Hitachi vane clearance
specifications. The assembly supplied by B&W for the tests apparently had
excessive clearances around the vanes that decreased the effectiveness of
swirl generation by allowing air to leak around the vanes. The vanes were
widened to reduce this leakage.
3.3 XCL Burner
The B&W XCL burner, shown in Figure 3-4, is the most recent development
in the evolution of DRB technology. Building on the HNR burner concept, B&W
incorporated features to improve burner operabi 1 ity by enhancing mechanical
reliability, adding air measurement capabilities, and reducing burner
pressure drop. As the other two burners, the XCL is a multi-passage burner
having a central cylindrical coal nozzle, two concentric annular air
passages, and an identical 90° elbow coal inlet- The basic XCL burner
incorporates the B&W coal di ffuser at the coal nozzle inlet to produce a
uniform coal distribution at the burner exit. The XCL burner also
incorporates an Air Separation Plate similar to the HNR design. The inner
secondary air passage is equipped with a sliding damper for flow control and
adjustable axial spin vanes for swirl control similar to both the DRB and HNR
burner.
-------
u.
CD
O)
2:
%n
"O
C
LlJ
"3
O
U
s-
-------
Outer
Damper
Sleeve
Pi tot Grids
Inner Damper
Sleeve\
^-Air
Separa
Plat
Coal Diffuser
tion
Inner
Spin
Vanes
Outer
ss-Spir
Spin
Vanes
Figure 3-4. Cross-sectional view of the B&W XCL Burner.
-------
The major mechanical changes that make the XCL unique are the outer air
passage assembly and pi tot grids for flow measurement. The outer secondary
register of radial vanes, used in both the DRB and HNR burner, was replaced
with a sliding damper for air flow control and adjustable axial spin vanes
for swirl generation. Thus, the inner and outer secondary air passages are
mechanically the same in the XCL burner. Because of the importance of air
flow distribution between the burner passages and among burners in a multi-
burner boiler installation, B&w incorporated pi tot measurement grids in the
secondary air passages. These pi tot grids will be valuable for burner
optimization in field applications.
Because of the developmental nature of this burner, the XCL was tested
in a number of configurations in its parametric optimization tests. The
focus of the different configurations evaluated were coal nozzle changes. In
addition, two modifications were made to the outer secondary air passage.
The XCL configurations in two series of tests included;
t Flame Stabilizing Ring with coal diffuser.
• Flame Stabilizing Ring with coal diffuser and increased outer spin
vane width.
• DeNOx stabilizer
• 30° included angle coal impeller
• Coal diffuser
• 20° included angle coal impeller
• Open impeller
• Expanded nozzle with coal diffuser
• Expanded nozzle with 20° included angle impeller.
• Expanded nozzle with 30° included angle impeller.
• 30° included angle impeller and fixed outer vane assembly.
• Diffuser and fixed outer vane assembly
The various coal nozzle configurations are described in Section 3.4.
3-10
-------
In addition to the basic outer assembly, two modifications were evalu-
ated. Because of the improvement in HNR burner performance from widening the
outer register vanes thereby decreasing leakage, the outer spin vanes were
widened for the XCL burner. The second modification was the incorporation of
fixed outer vanes. These fixed vanes produce effective swirl generation with
reduced pressure drop. The reduction of pressure drop was achieved by
directional vanes upstream of the fixed 45° angle spin vanes, shown in
Figure 3-5.
3.4 Burner Hardware Components
During the evaluation and optimization of the B&W Second Generation Low-
N0X burners, a number of burner components were incorporated into the
burners. These burner components included coal impellers or similar coal
nozzle devices to disperse the coal, coal nozzle modifications, and secondary
air passage modifications.
3.4.1 Coal Dispersal Designs
The coal dispersal devices evaluated in the three low-NOx burner designs
were of several types: coal pipe devices, coal impellers and swirlers, and
coal nozzle devices.
The coal pipe devices were:
t Coal diffuser/deflector
• Coal pipe venturi
• DeNOx stabilizer
The coal diffuser/deflector assembly is the standard coal dispersal device
utilized by SAW in the DRB. This device has been illustrated in Figures 3-1,
3-2, and 3-4. The coal diffuser/deflector is positioned at the inlet of the
coal pipe just downstream of the coal inlet elbow. It produces a uniform
coal distribution across the coal pipe, breaking up any roping of coal along
the elbow with the deflector and the bluff body di ffuser. The diff user was
3-11
-------
Turning Spin
Vanes Vanes,
XT (45°)
XCL burner with fixed outer vanes.
3-12
-------
utilized in all three burner designs. The coal pipe venturi, shown in Figure
3-6, was the original coal dispersal device developed for the DRB. This
device, also located at the coal nozzle inlet, produced a uniform coal
distribution across the pipe by the acceleration and concentration of the
pulverized coal through the venturi throat followed by an expansion to the
coal pipe diameter. This device was tested only in the standard Phase V DRB
under B&W funding. The DeNOx Stabilizer is a proprietary B&W coal pipe
device developed for potential retrofit for existing DRBs. This device was
designed to produce a stable, fuel-rich flame core expected to also decrease
N0X. The DeNOx stabilizer was evaluated in the low-velocity DRB and XCL
burner.
Five coal impellers or swirlers were evaluated during these tests.
These devices included:
• 75° included angle impeller
• 20° included angle impeller
• 30° included angle impeller
• Open impeller
• HNR burner swirler
The four impeller designs are shown in Figure 3-7. They all share the basic
B&W design concept, utilizing multiple, concentric conical rings to impart a
radial component to the primary air/coal stream at the coal nozzle exit. The
75°, 20°, and 30° impellers are similar in design differing only in the angle
of the conical rings. Each has a center conical bluff body, surrounded by
2 to 4 conical rings. The open impeller was essentially the 30° impeller,
but with the central conical bluff body eliminated. The HNR burner swirler
was described previously in . section 3-2 and was used only with the HNR
burner. The 75° Impeller was tested only with the low-velocity DRB. The
other impellers were tested only with the XCL burner.
Two coal nozzle exit devices were evaluated, the Flame Stabilizing Ring
and a short expanded nozzle tip. The Flame Stabilizing Ring, shown in
3-13
-------
Venturi
4
I
I
Coal/Primary Air
Note: Not to scale.
Figure 3-6. B&W coal pipe venturi.
... . . . , ^
3-14
-------
30° Impeller Open Impeller
Figure 3-7. Coal impellers used during Second Generation low-NCL burner testing.;
I ¦ :
-------
Figure 3-8, is located at the end of the coal nozzle. It was developed by
Babcock-Hi tachi for the HNR burner. The Flare Stabilizing Ring acts as an
orifice, with "teeth" on the circumference of the coal pipe. The FSR then
expands into the inner secondary passage. This construction facilitates a
small recirculation zone at the exit of the coal nozzle and thus enhances
flame stability. This device was tested on the Phase V DRB (under B&W
funding)» the HNR burner (as standard equipment), and the XCL burner. The
expanded nozzle tip, shown in Figure 3-9, was installed on the XCL burner
only, to produce a lower primary velocity at the exit. The tip was short,
5.5 inches long, and increased the coal nozzle diameter by about 2 inches.
3.4.2 Secondary Air Modifications
Besides modifications to widen outer secondary air vanes and increase
swirl effectiveness for the HNR and XCL burners, two other hardware
components were evaluated. The Air Separation Plate (ASP) was tested on all
three burners and a fixed outer secondary vane assembly was tested on the XCL
burner. The Air Separation Plate is standard equipment on the HNR and XCL
burners. The ASP is an extended baffle between the inner and outer air
passages that deflects the outer secondary air radially from the inner zones.
This delays mixing of coal with combustion air and shapes the flame. An
assembly of fixed axial vanes set at about 45° angle was tested on the XCL
burner. This fixed vane assembly, previously shown in position in Figure
3-5, was intended as a mechanical simplification to improve burner
reliability and to reduce burner pressure drop. It was also hoped that the
fixed vanes would improve air distribution around the periphery of the
burner, thereby improving combustion efficiency.
3.5 Fuels and Sorbents
Four different coals were utilized during the Second Generation Low-NOx
burner tests. Tables 3-2 and 3-3 summarizes the composition of the coals and
their corresponding ashes, respectively. The Pittsburgh #8 coal was the pri-
mary fuel used throughout the burner tests. Pittsburgh #8 is a high volatile
A bituminous coal selected for the LIMB demonstration project at Edgewater
3-1,6
-------
Outer Secondary Air
Inner Secondary Air
1
Primary Air + Coal
3"
Flame Stabilizing
Ring
Cross-sectional View
Front Vi
Figure 3-8.
Flame stabilizing ring.
-------
Expanded
Nozzle
Tip
W
Figure 3-9. Expanded nozzle tip on the XCL burner.
3-18
-------
TABLE 3-2. COMPOSITION OF TEST COALS
Coal
Pittsburgh #8
li
Utah
j
Comanche
Lower
Ki ttanni ng
Reporting
8a si s
As
Rec 'd
Dry
As
Rec 'd
Dry
As
Rec 'd
Dry
As
Rec' d
Dry
Proximate (wt %}
Moisture
3.50
! 0.00
6.11
0.00
22.44
0.00
2.43 ¦
0.00
Ash
12.92
13.40
8.02
8.55
5.00
.6.45
10.19
10.44
Volatile
33.75
: 34.98
41.26
43.96
36.12
44.87
23.93
24.52
Fixed C
49.83
I 51.62
44.60
46.73
37.72
48.68
63.45
65.04
Heating Value
¦ Btu/lb ,
12,177
|12,618 '
12,288
13,088
9,325
12,026
13,551
13,888
MMF Btu/lb
114,876
14,440
12,939
15,701
MAF Btu/lb
|14,626. _
14,311
12,855
15,507
Ultimate (wt %)
Moisture
3.50
i 0.00
6ill
0.00
22.44
0.00
2.43
' 0.00
Carbon
68.13
i 70,59
68.58
71.86
54.25
69.97
76.82
78.73
Hydrogen
4.63
; 4.79
5.16
5.49
3.80
4.91
4.54
4.65
Nitrogen
1.21
| 1.26
1.28
1.36
0.76
0.98
1.12
1.15
Sulfur
3.22
; 3.30
0.60
0.64
0.43
0.56
1.13
1.16
Ash
12.92
1 13.40 •
8.02
8.55
5.00
6.45
10.19
10.44
Oxygen*
6.41
6.63 .
10.24
10.91
13.32
17.14
3.77
3.87
Forms of Sulfur
i
(wt %)
Sulfate
0.22
0.23
0.01
0.01
0.02
0.02
0.01
0.01
Pyritic
1..
1.65
0.13 -
0.13
0.09
0.12
0.53
0.54
Organic
1.38
1.42
0.46
0.50
0.32
0.42
0.60 ¦
0.62
*Qxygen determined by difference.
3 -19
-------
TABLE 3-3. COAL ASH CHARACTERISTICS
i
Coal
Pittsburgh #8
Utah
Comanche
Lower
Kittanning
Elemental Ash
(wt %)
Si02
48.67
58.40 '
23.18
49.24
AI2O3
20.19
19.96
, '13.99 '
26.81
Ti02 ,
0.84
0.77
1.04
1.20
Fe203
23.87
4.18
5.07
14.83
CaO
1.60
4.56
28.42
1.67
MgO
0.60
1.05
5.15
0'.87
Na20
2.00 ¦
1.54
1.20
0.28
K20
0.31
1.06
0.29
2.50
P205
0.. 39
0.51
1.41
0.26
S03
1.25
4.77
17.50
•' 1.15
Ash Fusion
Temperatures
(°F)
Oxidizing
IDT
2377
2350
2390
2660
ST
2554
2448
2412,
2700
HT
2580
2546
2425
2700
FT
2616
2653
2451
2700
Reduci ng
IDT
2171
2297
2316
2635
ST
2298
2388
2342
2700
HT
2459
2502
2351
2700
FT ,
2498
2621
2383
2700
> •
- J
, ' j
!
'
3-20
-------
Station Unit 4. Since the ultimate goal of these burner tests was the
selection of the optimum burner for retrofit at the Edgewater boiler, the use
of the same coal eliminated coal composition as a factor in the projection of
expected burner performance. The other three coals represent a wide range of
coal types and were used to characterize the performance of the optimized XCL
burner. Utah coal is a western high volatile B bituminous coal from the
Starpoint mines in Vfattis, Utah. This Utah coal has been used at EER as the
base fuel for most of the low-emission, high-efficiency burner development
projects. Use of this coal allows comparison of the Second Generation low-
N0X burner performance with an existing data base of other burners. The
Comanche coal is a subbituminous coal from Wyoming. The Lower Kittaning is a
medium volatile bituminous coal.
Predictions of expected N0X emissions based on coal composition have
been developed at EER to rank coal types. Empirical correlations have been
developed in a laboratory combustor for N0X emissions based on coal
properties for several distinct combustion conditions.5 Application of these
correlations yielded the results summarized in Table 3-4. The NO predictions
given include theoretical total conversion of fuel nitrogen to NO and NO
emissions predicted for different mixing conditions; a premixed flame, a
radial diffusion flame, and physically staged combustion air conditions for
minimum N0X emissions. The absolute values of the NO predictions are for the
specific laboratory combustor and would be expected to vary with combustor
design and operation. High turbulence, pre-NSPS type burners would be
representative of conditions between premixed and radial diffusion. For the
subject coals, NO emissions would be expected to be higher for the Utah coal
followed by Pittsburgh #8, Comanche, and Lower Kittaning. However, the
correlations predict that the Utah coal is more amenable to staging than
either the Pittsburgh #8 or Lower Kittaning, but the Comanche subbituminous
would be expected to produce the lowest emissions under staged conditions.
As-fired pulverized coal samples were obtained on a daily basis
throughout the testing period. The pulverized coal was sampled downstream of
the primary air exhauster following ASME PTC 4.2 procedures. The objective
of this sampling was to verify the composition and fineness of the coal. The
3-21
-------
TABLE 3-4. PREDICTIONS OF N0X EMISSIONS .BASED ON COAL COMPOSITION
' ' ' * ' 1
Coal
Pittsburgh #8
Utah
Comanche
Lower .
Kittanning
Rank
High Volatile
A Bituminous
High Volatile
B Bituminous
Subbitum-
inous B
Medium Volatile
Bitumi nous
Compost tion
(wt %, daf)
Nitrogen
1,45
1.49
1.05 ,
1.28
Volatile Matter
40.39,
48.07
47.96
27.38
Fixed Carbon
59.61
51.10
52.04
72.62 .
NO Predictions
(ppm © OS O2)
Theoretical
2790
2987 .
2297
2323
Premixed
' 1058
1188
911
832
Radial Diffusion
825
876
694
718,
Minimum Staged
288
275
242
291
i
1.
!
Y
:
j
3-22
-------
TABLE 3-5. DAILY COAL VARIATIONS
!U- 1
; \
; j
Pittsburgh #8
Utah
Comanche
Lower
Kittanning
Coal
Mean
Std.Dev.
Mean
Mean
Mean
Composi tion
(dry, wt %)
Carbon
69.24
1.81
71.92
66.03
76.11
Hydrogen
4.65
0.13
5.29
4.79
4.56
Nitrogen
1.43
0.25
1.45
0.97
1.23
Sulfur
3.30
0.25
0.69
0.47
1.27
Ash
12.01
1.05
7.15
6.60
'11.37
j
1 i I
I
3-23;'
-------
average ultimate composition and relative standard deviation of these daily
samples are summarized in Table 3-5. Only the Pittsburgh #8 was tested and
thus sampled more than one test day. The standard deviation for the
Pittsburgh #8 coal was less than 2 percent for each component, suggesting
very consi stent coal composition. The Utah, Comanche, and Lower Kittaning
coals were only sampled once during the brief test duration, thus standard
deviation could not be determined.
The typical particle size distributions for the four coals are shown in
Figure 3-10. Daily variations of coal fineness are summarized in Table 3-6.
Adjustments to the mill and its classifier achieved similar size
distributions for each coal type. Coal fineness was maintained nominally at
70 percent through 200 mesh (75 fim) with measured values ranging from 66.7
percent for Utah coal to 73.35 percent for Lower Kittaning coal.
Two sorbents were used during these tests to evaluate S0£ reduction
potential by in-furnace injection, Vicron 45-3 limestone and Colton hydrated
lime. These sorbents have been used at EER as typical limestone and hydrated
lime materials in the development of LIMB technology. The physical and
chemical characteristics of these two sorbents are listed in Table 3-7 and
the corresponding size distributions are shown in Figure 3-11. Vicron is
nominally 99 percent pure CaC03 with a mass median diameter of 9.8//m. The
Colton hydrated lime is nominally 96 percent Ca(0H)2 with a median particle
size of 4.0 fim.
3.6 LWS Configuration
The Second Generation Low-N0x burners were tested in the EPA Large
Watertube Simulator shown in Figure 3-12. This facility and test procedures
used, are described in detail in Part I, Sections 4.3 and 4.4 of this report.
Key features of the test facility important to these tests were:
• Furnace dimensions, especially width (16 ft) and depth (22 ft).
The depth of the LWS, coincidentally is essentially the same as
that of the Edgewater boi1er (22 ft, 3 inches ). This feature
3^24
-------
99.99
99.9
99
)
¦ 98
I
I
! 95
>
90
80
70
60
50
40
30
20
10
O Pittsburgh #8
~ Utah
A Comanche
0 Lower Kittarining
A
~ V
J L
_L
J L
10
20
30 40
60 80 100
200 300 400 600 1000$
Particle Size {pim)
Figure 3-10. Typical coal particle size distributions.
~ " 3«25.
-------
TABLE 3-6. COAL FINENESS VARIATIONS—CUMULATIVE MASS PERCENT UNDERSIZE
i J
Coal
Pittsburgh #8
Utah
Comanche
Lower Kittanning
Mean
Std.Dev.
Mean
Mean
Mean
Particle
Size (/xm)
38
48.18
4.01
43.23
50.64
49.26
75
71.83
3.14
66.7
72.5
73.35
150
89.43
3.08
87.09
91.11
88.89
300
97.91
1.04
97.15
99.71
98.58
-------
TABLE 3-7. PHYSICAL AND CHEMICAL PROPERTIES OF TESTED SORBENT
or
Physical Properties
Elemental Ash (wt %)
Sorbent
Theoretical
Characteristics
Median
Diameter
( m)
Density
(gm/cirw)
LOI 0
1000°C
(wt %)
CaO
Fe203
a12°3
Na20
MgO
k2o
Si02
Ti02
P205
Si 03
Vicron 45-3
Hydrated Lime
CaC03
Ca(0H)2
9.8
4.0
2.706
2.279
42.49
22.91
55.64
72.67
0.08
0.15
0.03
0.40
0.01
0.01
0.54
0.42
0.01
0.06
0.20
7.06
0.01
0.02
0.01
0.01
0.02
0.07
-------
i i i i 1 1 1—i—t-lt—i 1 1 1—i 1—i—i 1 r
Colton Hydrated Lime
Vicron 45-3
Limestone
100 80 60 50 40 30
20 10 8 6 5 4 3 2 1
Equivalent Spherical Diameter (/urn)
0.6
_L
0.3 0.2
0.1
Figure 3-11. Sorbent particle size distribution.
-------
19' Level
8' Level
ro
10
TT.m. .**1 (^n. m . 1
m2S>
:w: InsulationWffii
__.5~
xlstlng
Re fmiary
Uninsulated
o
Nountlna ,
¦ Plate
Existing
irebrick
16'
t
Additional
Insulation
Existing
Refractory
it Existing
00 o0y^r_"~_ Refractory —
1 52
SAddi tlpri.a 1 J n$i)
:|:S t) 0 n j
Existing -
Refractory -
Existing
Firebrick
Existing
Firebrick
Rear Wall
Addl tional j:
Insulation?
Existing
Refractory
Existing
tJ>, 'V- _ Refractory
North Side Wall
Front Wall
South Side Ual 1
Figure-3-12. Insulation pattern in the LWS for Second Generation Low NO Burner Tests.
X
-------
facilitated optimization of emission and flame length tradeoffs for
Edgewater with confidence.
• Additional insulation was added to the LWS to produce a thermal
environment more representative of operating boilers. This was
difficult to achieve because of the relatively low firing rate for
the test burns, 78 x 10® Btu/hr. With this additional insulation,
average flue gas exit temperatures were measured to be 1855°F.
• The burner installation onto the LWS was facilitated by B&W design
of the test burners to utilize a common windbox and burner exit.
The common windbox provided air to both the inner and outer
secondary air passages without direct measurement of air flow
distribution. Only the XCL burner was equipped with air
measurement grids.
• A separate oil burner was installed adjacent to the pulverized coal
burners. This burner was supplied with its own oil supply and
metered air stream. This assured a clean burning oil flame during
furnace warmup, startup oxygen balance, as well as ignition of the
test burners.
Standard test procedures, described in• detail in Part I, Section 4.4,
were utilized throughout the project. Because flame length was a critical
performance parameter for these tests, special procedures were used to
determine flame length. Simple visual observation is used routinely;
however, judgement of pulverized coal flame boundaries can be subjective. To
confirm flame length observations and any possible flame impingement on the
rear wall of the LWS, the following measurements were taken at all
appropriate conditions:
• ,C0 levels at the rear of the furnace
t Furnace wall temperatures
• Unburned carbon content (UBC) of fly ash
3-30
-------
The latter two measurements were qualitative and useful only to determine
whether there was flame impingement on the rear wall.
Measurements of CO at the rear of the furnace were more useful in
actually determining and confirming flame length. Figure 3-13 and Table 3-8
present results of these measurements for selected test conditions. A
profile of CO concentrations from the rear wall toward the flame is shown in
Figure 3-13. As the probe approaches the end of the observed flame, the CO
level increases abruptly. CO measurements on the rear wall are listed for a
number of conditions in Table 3-8. For long flames, flames exceeding the
depth of the furnace (> 22 feet), CO levels generally exceeded 10,000 ppm.
Low CO levels, less than 200 ppm, were measured at the rear wall for flames
observed to be shorter than about 20 feet. These measurements confirm the
consistency of the visual flame observations made during the tests.
;3-3l
-------
Observed
F1 ame
Length
§_ 160
EL
3> 80
Distance, from Rear Wall
(ft)
Figure 3-13. Typical results from rear wall probing for CO levels.
3-32
-------
TABLE 3-8. CO MEASUREMENTS AT REAR WALL FOR
SELECTED CONDITIONS
BURNER
TEST HO.
OBSERVED FLAME
LENGTH (FT)
CO
(PPM, MEASURED)
DRB
1.04
>22
11,000-15,000
DRB
1.07
>22
7,000-11,000
DRB
1.13
>22
10,000-11,000
DRB
2.03
>22
9,000-13,000
DRB
2.06
>22
13,000
DRB
2.17
>22
9,000-14,000
DRB
2.28
>22
5,000- 6,000
ORB
2.33
18-20
6,000- 7,000
DRB
4.01
18*-20
80
ORB
4.02
14
55
DRB
4.03
12
40
DRB
4.D4
14
80
DRB
4.05
14
200-1,000
DRB
4.06
16
150-200
DRB
4.07
>22
10,MO
DRB
4.00
20-22
12,000-15,000
DRB
4.09
18-20
60-130
DRB
6.02
>22
32,000
DRB
6.05
>22
35,000-43,000
DRB
7.07
>22
23,000-28,000
DRB
7.12
22
10,000-15,000
DRB
7,13
22
6,000- 9,000
HNR
2.27
>22
18,000-21,000
HHR
4.01
>22
15,000-20,000
XCL
3.22
22 +
9,000-13,000
3-33
-------
-------
4.0 BURNER PERFORMANCE AND N0X EMISSIONS
Three B&W Second Generation Low N0X Burner designs were tested in the
LWS to evaluate their performance and N0X reduction potential. The burner
designs included;
• Dual Register Burner
• Babcock-Hitachi N0X Reducing Burner
• XCL Burner
The tests were conducted in two series. Initially, parametric screening
tests including hardware modifications were conducted. The objective of
these screening tests was to determine the burner configuration which yielded
optimum performance within the constraints imposed by the Edgewater Unit 4
boiler, of which firing depth was most severe. The LWS, coincidentally, has
essentially the same depth as the subject boiler and provided an ideal test
configuration to achieve the objectives. Following the screening tests, the
burner configuration which showed the greatest potential for satisfying the
objectives and constraints for the LIMB demonstration boiler was optimized.
This optimized burner, determined to be a configuration of the XCL burner,
was characterized with three additional, distinctly different coal types.
4.1 Dual Register Burners
Following are the test results of two 78 x 10® Btu/hr Dual Register
Burners; a Low Velocity DRB and a standard Phase V DRB described previously.
The Low Velocity DRB, tested under this program, was designed to fit the same
exit opening as the HNR and XCL burners. This resulted in lower burner
velocities than current B&W standards for DRB designs. In a separate B&W
sponsored LWS test program, a standard Phase V DRB was evaluated. The
results from those tests are summarized for comparison.
-------
4.1.1 Low Velocity Dual Register Burner
The 78 x 1Btu/hr Low Velocity DRB was evaluated in four
confi gurations:
• With standard coal diffuser
• With a 75° included angle impeller
• With DeNOx Stabilizer
• With Air Separation Plate and Flame Stabilizing Ring
Each of these configurations was tested over a range of burner adjustments, "
excess air, and firing rate to optimize and characterize its performance.
The range of test conditions are summarized in Table 4-1. Nominal operating
conditions maintained through screening the adjustable parameters were:
Firing Rate = 78 x 10® Btu/hr
SRp = 0.18
SRt = 1.20
Primary Air Temperature = 150°F
Combustion Air (Windbox) Temp = 550°F
The fuel used during the screening tests was Pittsburgh #8 coal.
/ /
The base confi gurati on of the Low Velocity DRB utilized the diffuser
alone. Initial tests of this base configuration at the above-listed full
load operating conditions produced flames over 22 feet long, that is, flames
exceeding the depth of the LWS. Observed flame impingement on the back wall
was confirmed by measured CO levels over 10,000 ppm. The available burner
adjustments were not effective in shaping the flame within the constraints of
the LWS furnace dimensions for this configuration at full load. Even with
high turbulence, opposed swirl settings of 20° CCW open on the inner spin
vanes and 15° CW open on the outer register, the flame length exceeded 22 ft.
4-2
-------
TABLE 4-1. RANGE OF SCREENING TESTS FOR LOW-VELOCITY
DUAL REGISTER BURNER
Confi guration
Di ffuser
DNS
ASP/FSR
Swirl er
0
75
Impeller
Burner Settings:
Inner Spin Vanes
20oCW-25°CCW
30°CW-20°CCW
20°CW-20°CCW
50°CW-50°CCW
Outer Register
10°-45°CW
15°-55°CH
15°-55°CCW
5-75°CW
Inner Sleeve Damper
25-100% Open
25-50% Open
12-25% Open
12-25% Open
Impeller Position
N/A
N/A
-1 to -10 in.
-6 to +8 in.
Performance Range:
N0X @ 0% O2, Dry (ppm)
188-281
242-336
332 - 420
314 - ;917
Flame Length (ft)
> 22
>22
' ; >22
12 - >22
Burner AP (in W.G.)
3.0-7.6
3.6-10.5
4.6-9.2
3.2-11.8
UCB %
2.5-10.6
5.9
1.8544.44
1.05-1.65
-------
N0X emissions for this setting at full load and 20 percent excess air were
264 ppm,* with fly ash carbon content of 4.8 weight percent.
Sensitivity of the performance for the different Low Velocity DRB
configurations to the available burner adjustments are summarized in Fig-
ures 4-1 and 4-2. The configurations with the diffuser, DeNOx Stabilizer,
and Flame Stabilizing Ring/Air Separation Plate/Swirler, shown in Figure 4-1,
produced flames over 22 feet long for essentially all the conditions tested.
For this reason, only the sensitivity of N0X to burner adjustments is shown.
For the Low Velocity DRB with the 75° impeller, however, both N0X and flame
length are shown as functions of burner settings in Figure 4-2.
Because of the excessive flame length, most of the screening tests with
the di ffuser were conducted at reduced load, nominally 60 x 10® Btu/hr.
These are the data presented on Figure 4-1 representing the Low Velocity DRB
with di ffuser. The data for the other two configurations in this figure are
at nominal full load operation. From these screening tests, the outer
register vane position proved to be the dominant burner adjustment. The
inner sleeve, which affects air flow distribution between inner and outer
zones, and the swirler position were secondary in effect on emissions. The
inner spin vanes had little or no effect on emissions.
The addition of the 75° impeller to the Low Velocity DRB had a
significant effect on performance, particularly flame length. As shown in
Figure 4-2, the impeller and its position became the dominant factor in
burner performance. In this configuration, flame length could be varied from
12 to over 22 feet simply by moving the axial position of the impeller. In
addition, N0X emissions were inversely proportional to flame length. N0X
emissions as low as 300 ppm could be achieved by retracting the impeller 6
inches into the coal nozzle, at the expense of a flame which exceeded the
furnace depth. The outer register vane position was a secondary factor in
*Note: All emission concentrations reported corrected to 0 percent 02 on a
dry basis, except where indicated.
4-4
-------
Configurations:
Diffuser (63 x 10® Btu/hr)
DNS ) 6
FSR/ASP/Swirler ( 78 x 10 Btu/hr
500
450
o. 400
Q_
£» 350
O
-P»l
-------
800
700-
CL
a.
600-
Q
C
«2!
400-
20-
14-
Li.
Impeller Position Inner Spin Vanes Outer Register
(Inches) (°CW)
Figure 4-2, Sensitivity of low Velocity DR5 with 75°
impeller to burner adjustments.
:4-£
-------
N0X emissions, also demonstrating that N0X emissions decreased with an
increase in flame length. As the degree of swirl decreased by opening the
register vanes, the slower mixing that resulted caused an increase in flame
length and corresponding decrease in N0X.
The data from the screening tests indicated that the following common
burner settings produced optimum performance (with regard to flame length)
for each Low Velocity DRB configuration:
Inner Spin Vanes: 20° CCW open
Outer Register: 15° CW open
Inner Sleeve Damper: 251 open (50% open for 75° impeller)
For the configuration with the 75° impeller, the impeller positioned at the
nozzle exit was determined to be optimum. This position corresponded to a
stable minimum N0X point with a flame shorter than the furnace depth. A coal
swirler position of 5 inches behind the coal nozzle tip was determined to be
optimum for the Low Velocity DRB assembly that incorporated HNR-type
components (Flame Stabilizing Ring, Air Separation Plate, and coal swirler).
Each of these configurations was evaluated over a range of firing rate
and excess air at their selected optimum settings. The results for each
configuration are shown in Figures 4-3 through 4-6. These figures present
N0X and combustion efficiency, as indicated by fly ash carbon content and/or
CO concentration, as a function of excess air and firing rate. The perfor-
mance of the diffuser configuration is shown in Figure 4-3 for three firing
rates. As expected, N0X emissions were highest for the full load, 78 x 10^
Btu/hr, and decreased for progressively lower firing rates. At full load and
an overall stoichiometry (SRj) of 1.20, N0X emissions were 264 ppm. At fir-
ing rates of 60 x 106 Btu/hr and 53 x 106 Btu/hr, N0X emissions were 240 and
200 ppm, respectively. The combustion efficiency is shown as a character-
istic family of fly ash carbon data, with the lowest level of unburned carbon
for the highest firing rate. The data also show the increase in unburned
carbon as excess air is decreased. Fly ash carbon at 20 percent excess air
ranges from 4.80 percent at full load up to 8.53 percent at 53 x 10^ Btu/hr.
4-7
-------
;¦>!
1 I
.001
Nominal Firing Rate:
O 78 x 106 Btu/hr
® 61 x 10® Btu/hr
D 53 x 10® Btu/hr
700
7°
-------
1400
1200 -
1000 -
lO '
Q-
CL
>)
•J-
O
CO
o
o
CBj
X
o
Btu/hr
Btu/hr
Overall Stoichiometry--SR
140
120 -
100 -
£ 80
; >>
S-
o .
c^60
o
o
401—
o
20 -
"i 1 r
CO—Open
Fly Ash Carbon—Solid
1.0
T
A£5^,
o
o^v
1.1 1.2
1 .3
1.4
1.5
Overall Stoichiometry--SRy
Figure 4-4. Performance of low-velocity DRB with 75° impeller as a
function of excess air and firing rate.
-------
600
500 -
i
400 _
s
5
300
D
b
* 200
Nominal Firing Rate:
Oj 80 x 106 Btu/hr
,#! 62 x 106 Btu/hr
~ ' 51 x 106 Btu/hr
i
1.00 1.10 1.20
1.30 1.40
Overall Stoichiometry--SR-p
Q- 80
-------
Nominal Firing Rate:
O 77 x 106 Btu/hr
• j 61 x 106 Btu/hr
~ , 53 x 106 Btu/hr
600
500
Q.
Ol
¦ £ 400
o
C\J
o
o
-------
This family of unburned carbon data indicates the sensitivity of this
configuration in the LWS to excess air and the necessity to operate at higher
excess air levels as load is decreased to maintain carbon burnout.
The performance characteristics of the 75° impeller DRB configuration
are shown in Figure 4-4. These data are for two firing rates, 79 x 10® Btu/
hr and 49 x 10® Btu/hr. Again, N0X emissions were higher and fly ash carbon/
CO lower for full load operation than reduced load. At an overall stoichiom-
etry of 1.20, N0X emissions were 710 and 550 ppm for 79 x 10® and 49 x 10®
Btu/hr, respectively. Corresponding flame lengths were 16-18 feet at full
load and about 12 feet at the lower firing rate. Fly ash carbon content was
also sensitive to excess air level, although the difference between the two
firing rates was not as great as for the diffuser configuration.
Performance characteristics for the DeNOx stabilizer and the assembly of
components from the HNR burner design (FSR, ASP, and swirler) are shown in
Figures 4-5 and 4-6, respectively. For each case, the N0X emissions for the
three firing rates are a family of essentially parallel curves with the N0X
decreasing with decreasing firing rate. N0X emissions for the DNS-equipped
DRB at an overall stoichiometry of 1.20 were 320, 280, and 210 ppm for firing
rates of 80, 62, and 51 x 10® Btu/hr, respectively. The DRB equipped with
the HNR components produced N0X emissions at an overall stoichiometry of 1.20
of 375, 315, and 290 ppm for firing rates of 77, 61, and 53 x 10® Btu/hr,
respectively. Combustion efficiency, indicated by CO levels, was more
sensitive to excess air levels for these two configurations than the coal
impeller equipped DRB. This would be expected for the slower mixing and
longer flames produced by the DNS and FSR/ASP/coal swirler configurations.
Figure 4-7 provides a direct comparison of performance for the four Low
Velocity DRB configurations tested at nominal full load. Highest N0X
emissions were produced with the 75° impeller equipped configuration while
the lowest emissions were produced by the base, diffuser equipped configu-
ration. Combustion efficiency, as indicated by fly ash carbon content was
similar for the four configurations. However, as the key test data in
Table 4-2 shows, only the impeller-equipped DRB produced a flame less than
4-12
1
-------
1400
1200
1000
800
600
400
200
i 1 1 r r
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate = 78.6 x 10® Btu/hr
I
X
_L
_L
'C
• L>}
U
-------
the furnace firing depth (22 feet). Because the conditions selected as
optimum for each configuration were similar, the windbox-to-furnace pressure
differential was al so similar. The tightly closed, hi gh swirl settings
chosen to produce relatively shorter flames resulted in high press differ-
entials, from 6 to 10 in. W.G at full load and 20 percent excess air. In
commercial applications, pressure drop across the DRB is typically 3 to 5"
W.G. (when flame length is not a constraint).
4.1.2 Characterization of Standard Phase V DRB
In a separately funded test program sponsored by B&W, a standard Phase V
DRB was evaluated in the LWS. As described previously, the principal
difference between this Phase V DRB and the Low Velocity DRB that was the
subject of this Second Generation Low Burner evaluation was the exit diam-
eter. The Phase V DRB has a smaller exit, resulting in higher secondary air
velocities. The Phase V DRB was evaluated in five configurations:
• Basic coal diffuser
• Coal venturi
• Diffuser with Air Separation Plate
• Diffuser with Flame Stabilizing Ring
• Diffuser, Flame Stabilizing Ring, and Air Separation Plate
The nominal operating conditions maintained through parametric screening
tests of these configurations were;
Firing Rate = 78 x 106 Btu/hr
SRp =0.18
SRT = 1.20
Primary Air Temperature = 150°F
Combustion Air Temperature = 550°F
These conditions are the same as those for the Low Velocity DRB. The fuel
used was also the same Pittsburgh #8 coal.
4-14
-------
TABLE 4-2. LOW VELOCITY DRB PERFORMANCE SUMMARY
i
I
Confi gura-
tion
Swi rl
Inner
Vanes
Outer
Inner
Sleeve
(%
Open)
Spreader
Position
(Inches)
Fi ri ng
Rate
(106 Btu
/hr)
SRT
Emissions
(0 0% 0?)
Flame
Length
(ft)
Fly Ash
Carbon
(wt.%)
Wi ndbox
Pres.
(in.H20)
N0X
CO
Di ffuser
20°CCW
15°CW
25
N/A
76.9
1.19
264
48
22
4.80
7.10
Di ffuser
20°CCW
15°CW
25
N/A
69.7
1.19
240
72
22
6.52
6.20
Di ffuser
20°CCW
15°CW
25
N/A
61.7
1.21
212
66
22
5.38
4.60
Impeller
20°CCW
15°CW
50
0.0
79.4
1.19
708
24
18
7.28
6.00
Impeller
20°CCW
15°CW
50
0.0
57.5
1.20
579
30
14
5.25
3.30
Impel 1er
20°CCW
15°CW
50
0.0
49.4
1.19
552
36
12
4.11
2.40
DNS
20°CCW
15°CW
25
N/A
79.9
1.19
318
21
22
5.89
10.20
DNS
20°CCW
15°CW
25
N/A
60.7
1.20
278
45
22
6.70
5.90
DNS
20°CCW
15°CW
25
N/A
51.1
1.19
204
36
18-20
6.85
4.20
FSR, ASP,
Swi rler
20°CCW
15°CW
25
-5.0
79.4
1.20
386
24
22
4.44
9.20
FSR, ASP,
Swi rler
20°CCW
15°CW
25
-5.0
60.7
1.20
314
36
20-22
2.77
4.90
FSR, ASP
20°CCW
15°CW
25
-5.0
52.6
1.20
290
30
20
3.63
3.50
-------
The range of test conditions for each configuration of the Phase V ORB
are listed in Table 4-3. A key objective in the evaluation of the.five
configurations was to determine the effect of each subject component on
burner performance and the degree of synergism when several components are
used together. Underlying these tests was a comparison of the Phase V DRB
with the Low Velocity DRB. .To this end, the screening tests had a dual
purpose; to optimize the Phase V DRB configurations and also to evaluate
burner settings which duplicated the previous low Velocity DRB tests.
The results of the screening tests indicated similar trends for the
Phase V DRB and the Low Velocity DRB. The outer register vanes were the
dominant parameter affecting burner performance. A set of data for each
configuration demonstrating this sensitivity is shown in Figure 4-8.
Reducing the degree of swirl generated and also increasing outer zone air
flow by opening the outer register decreased N0X emissions for each
configuration with a correspondi ng increase in flame length. The inner
sleeve damper position, which controls the air flow distribution between the
inner and outer air zones, and the inner spin vanes had a secondary effect on
burner performance.
From the screening tests, two sets of conditions were identified to
characterize all the configurations. These conditions, listed in Table 4-4, ^
can be described as: (1) duplicate of Low Velocity DRB optimum settings, and
(2) nominal, optimum commercial conditions. The Low Velocity DRB duplicate
conditions are characterized as highly turbulent with opposed inner and outer
zone swirl and tightly closed inner and outer vanes for high.swirl genera-
tion. These conditions produce a relatively short flame albeit with higher
N0X emissions. Because of the tightly closed settings, the burner pressure
drop is about 11 in. W.G.; at nominal full load conditions. The other set of
conditions are more representative of typical commercial practice, with more
open swirl vanes and registers and co-swirling inner and outer zones. These
conditions produced a longer flame (22 to over 22 feet) with lower emissions
and also reduced the burner pressure drop to 5-6 in. W.G.
4-16]
-------
TABLE 4-3. RANGE OF SCREENING TESTS FOR STANDARD PHASE V DRB
Confi gurati on
Di ffuser
Venturi
Diffuser +
ASP
Diffuser +
FSR
Di ffuser +
ASP/FSR
Burner Settings:
Inner Spin Vanes
Outer Register
Inner Sleeve Damper
18°CCW-16°CW
15°-45°CW
25-100% Open
18°CCW-26°CW
15°-55°CW
25-100% Open
18°CCW-26°CW
15°-55°CW
25-100% Open
18°CCW-26°CW
15°-45°CW
25-100% Open
18°CCW-7°CW
15°-45°CW
25-100% Open
Performance Range:
N0X @% O2 Dry (ppm)
Flame Length (ft)
Burner P (in. W.G.)
243-372
20->22
4.9-10.8
252-370
20->22
5.0-11.0
314-392
14->22
6.0-11.8
243-298
22->22
4.5-10.5
309-352
22->22
5.7-11.0
-------
300
20C-
T
Nominal Conditions:
Fuel; Pittsburgh #8 K
Firing Rate = 78 x 10° Btu/hr
SRt =1.20
Inner Register = 18 CCW
Sleeve = 251 Open
Configuration
©•, W/Diffuser
|T) W/Venturi
-A- W/ASP/Diffuser
<$> W/FSR/Diffuser
-iL-W/ASP/FSR/Diffuser
10 20 30 40
Outer Register (° Open-CW)
Figure 4-8. Sensitivity of Phase V DRB configurations
to outer register position.
. . 4-18 .
-------
TABLE 4-4. SELECTED OPTIMUM SETTINGS FOR PHASE V DRB
Parameter
Low Velocity
• DRB
Commercial
Practice
Inner Spin Vanes
Outer Register
Inner Sleeve Damper
Nominal .Burner AP
18° CCW ,
'15° CW
251 Open
11 in. W.G.
260 cw
35° cw
100% Open
5-6 in. W.G.
-------
The performance of the five configurations evaluated is compared in
Figures 4-9 and 4-10 as a function of excess air and firing rate, respect-
ively. These data show the effect of each component on burner performance
implicitly. The lowest N0X emissions at full load were produced by the
di ffuser and Flame Stabilizing Ring configurations, both with about 290 ppm
N0X and flames over 22 feet long for the commercial burner settings. The
venturi , Air Separation Plate alone, and the combined Air Separation Plate/
Flame Stabilizing Ring yielded similar, but higher, N0X emissions 330-350
ppm. Only the Air Separation Plate produced a flame observed not to impinge
on the rear wall at the commercial settings. Combustion efficiency, as
indicated by fly ash carbon content, fell within a fairly narrow band for 4
of the 5 configurations. The diffuser had slightly high unburned carbon
levels over the excess air range tested. The effect of firing rate on
performance is not as clear with respect to ranking the emissions performance
of the various configurations over the range tested. Each configuration did
demonstrate a decrease in N0X with load coupled with an increase in unburned
carbon.
Considering the effectiveness of each burner component, the coal
diffuser yielded approximately 12 percent lower N0X emissions, and slightly
longer flames, than did the coal venturi. Each device is positioned at the
coal pipe inlet well upstream of the coal nozzle exit. The Flame Stabilizing
Ring with the coal diffuser achieved performance very similar to that of the
diffuser alone. However, their function is somewhat different. The FSR is
attached at the exit of the coal nozzle producing small stabilizing
recirculation zones near the nozzle and promoting a higher velocity central
coal jet. The small differences with co-swirling inner and outer zones is
again surprising. The Air Separation Plate, whose intended function is to
separate the inner and outer air streams, seemed to do the opposite. Upon
inspection, it appears the angle and outlet velocity may produce an
unintended eddy which actually enhances mixing between the inner and outer
zones. The basic principal of such a baffle is sound, but the data suggests
that the design of this device is critical to the resulting performance. The
performance of the combined ASP/FSR configuration was between that measured
for the two components alone. In those terms, the effect of the combined
.4-20
-------
¦p»
I
ro
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate: 78 x 10® Btu/hr
Configuration:
Burner Settings: O w/Oiffuser
-O
Inner = 26UCW
Outer
35°CW
>> 400
Sleeve = 100%
Open
14
1.1 1.2 1.3 1.4
Overall Stoichiometry--SRy
iA
«c
o
.O
s-
o
*o
QJ
c
s-
3
_o
1.0
D w/Venturi
A w/ASP/Diffusion
O w/FSR/Diffuser
k w/ASP/FSR Diffuser
1.1 1.2 1.3 1.4
Overall Stoichiometry—SRy
Figure 4-9. Effect of excess air on Phase V DRB configurations' performance.
-------
400
300 -
o.
O-
>)
V-
^oo -
o
roj
100 -
o
Nominal Conditions:
Fuel: Ohio-Pittsburgh
SR T = 1.20
35°CW
Inner Register = 26 CW
Outer Register
jister
100% Open
Sleeve
40
50
60
70
80
Firing Rate (10 Btu/hr)
cC
o
X)
i-
o
¦o
-------
components can be thought to be additive, with neither promoting the
effectiveness of the other. Table 4-5 summarizes the key test conditions
with the various Phase V DRB configurations. Primary benefit of the ASP and
the FSR was the reduction in unburned carbon in ash.
Direct comparison of the Phase V DRB can be made with the Low Velocity
DRB, given that each was evaluated in the same furnace with the same fuel and
at the same scale. This comparison is shown in Figure 4-11 for the two
burners in their basic configurations, utilizing a coal diffuser alone.
Given the same high turbulence burner settings, the Phase V DRB produced
higher N0X emissions (372 ppm) than did the Low Velocity DRB (264 ppm). The
flame was correspondingly shorter for the Phase V DRB (20-21 ft) than for the
Low Velocity DRB (> 22 ft). Unburned carbon levels, however, were comparable
for the two burners. The Phase V DRB with more conventional burner settings
achieved performance very similar to that of the Low Velocity DRB with its
high turbulence settings, namely N0X emissions of 295 ppm, flame length over
22 feet, and 4.75 percent carbon in ash. This comparison indicates the
importance of burner velocities to resultant performance and confirms that
one of the basic design principal of low-emission burners is lower combustion
air velocities.
4,2 Evaluation of the Babcock-Hitachi NQX Reducing Burner
The Babcock-Hitachi M0x Reducing burner was tested in two basic
configurations. One configuration utilized a coal swirler positioned just
upstream of the Flame Stabilizing Ring and the other replaced this coal
swirler with the standard diffuser at the inlet of the coal nozzle. In
addition, inspection of the HNR burner assembly after initial screening tests
revealed excessive clearance in the outer register vanes. Shims were
installed to reduce this clearance and thus increase swirl generation
effectiveness. This outer register modification constituted a variation of
the HNR with diffuser.
-------
TABLE 4-5. SUMMARY OF KEY PHASE V DRB TEST CONDITIONS
Confi gu-
Regi sters
Sleeve
Firing Rate
SRt
N0X @
CO @
Flame
Length
Fly Ash
Carbon
Burner AP
(In. W.G.)
rati on
Inner
Outer
(%
Open)
(106 Btu/hr)
0% 02
0% 02
(ft)
(wt %)
Diffuser
;18°CCM
15°CW
25
79.3
1.19
372
24
20-21
6.12
10.8
Diffuser
26°CW
35°CW
100
78.7
1.21
293
24
>22
4.75
4.9
Diffuser
26°CW
35°CW
100
61.6
1.19
240
42
>22
7.47
2.9
Venturi
'18°CCW
15°CW
25
80.8
1.18
350
18
20-21
6.45
11.0
Venturi
26°CW*
35°CW
100
81.6
1.22
356
25
>22
2.70
5.7
Venturi
26°CW
35°CW
100
61.6
1.20
253
33
22
7.10
3.0
ASP
18°CCW
15°CW
25
81.5
1.18
392
25
14-15
9.74
11.8
ASP
26°CW
35°CW
100
79.5
1.20
326
25
22
3.20
6.4
ASP
26°CW
35°CW
100
64.4
1.18
237
27
22
5.74
3.8
FSR
18°CCW
15°CW
25
80.6
1.19
292
26
22
6.96
10.5
FSR ,
! 26°CW
35°CW
100
78.9
1.21
285
24
>22
3.29
4.9
FSR
26°CW
35°CW
100
59.4
1.19
240
28
22
6.32
2.7
ASP/FSR
; 18°CCW
15°CW
25
80.4
1.19
328
24
22
5.16
11.0
ASP/FSR
26°CW
35°CW
100
81.6
1.18
309
24
>22
3.02
6.1
ASP/FSR
26°CW
35°CW
100
61.9
1.19
252
18
21-2
2.36
3.7
-------
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate = 78 x 10® Btu/hr
O Low Velocity DRB
~ Phase V DRB
High turbulence Settings
Phase V DRB (Commercial Settings)
\
1.1 1.2 1.3 1.4 1.5 1.0 1.1 1.2 1.3 1.
Overall stoichiometry—SR-j. Overall stoiiohiometry—SRj
Figure 4-11. Comparison of the Phase V DRB with the Low Velocity DRB.
-------
As with all the Second Generation Low N0X burners, screening tests to
identify optimum performance were conducted at the following nominal
operating conditions with the Pittsburgh #8 coal:
Firing Rate = 78 x 10® Btu/hr
SRp = 0.18
SRT = 1.20
Primary Air Temperature = 150°F
Combustion Air Temperature = 550°F
The parametric screening tests were conducted over the range of settings
listed in Table 4-6. For the HNR burner, the swirl direction of the two air
zones was maintained in the same clockwise orientation. No opposed swirl
cases were evaluated.
Results of the screening tests are summarized in Figure 4-12. Perform-
ance of the HNR burner was most sensitive to the outer zone variables. Low
N0X emissions and reasonable flame lengths could be achieved by biasing air
to the outer register by closing the inner damper coupled with the outer
vanes closed to produce a high degree of swirl. The HNR configuration with
the coal swirler produced higher N0X emissions, 348 ppm at 20 percent excess
air, with a correspond!'ngly shorter flame, 18-20 ft long than did the HNR
with coal diffuser, 290 ppm N0X and a 20 ft-long flame. With the modified
outer register, N0X emissions were reduced further down to 222 ppm with a
correspondingly longer flame, over 22 ft. The tight burner settings that
achieved these low emissions with the modified outer register resulted in a
high burner pressure drop, nominally 7 in. W.G.
The performance characteristics, in terms of N0X emissions and unburned
carbon in the fly ash, are summarized for the HNR burner with diffuser and
modified outer register in Figures 4-13 and 4-14, respectively. The N0X
4-26
-------
TABLE 4-6. RANGE OF SCREENING TESTS FOR HNR BURNER
Configuration
Coal
Swirl er
Coal
Diffuser
Burner Settings:
Inner Spin Vanes
Outer Register
Inner Sleeve Damper
Swirler Position
10°-30°CW
15°-45°CW
. 0-50% Open
-5 to -10 .in.
1C°-70°CW
10°-75°CW
0-50% Open
N/A
Performance Range:
• N0X @ 01 02, Dry (ppm)
Flame Length (ft)
Burner AP (in. W.G.)
294-416
13-> 22
4.2-8.9
198-426
18-> 22 •
4.6-13.2
1 • .
1
4-27
-------
Configurations
500
Swirler
450
Diffuser
----- Diffuser
w/Modified Outer
Register Vanes
a- 400
£ 350
CM
° 300
o
250
\
o
200
150
CD
20 40 60 80
20 40 60 80
10 20 30 40
2
4
6
8 10
Inner Spin Vanes (°CW) Outer Register (°CW) Inner Sleeve (% Open) Swirler (in.from Nozzle)
Figure 4-12. Sensitivity of HNR burner performance to burner adjustments.
-------
700
'¦J*
11 '
r-o
;^o
c9^
1.00 1.10 1.20 1.30 1.40 1.50
Overall Stoichiometry--SRy
1.00 1.10 1.20 1.30 1.40 1.50
Overall StoicLhiometry--SRy
Figure 4-13. Effect of excess on performance of HNR burner with
coal diffuser at 78 x 10® Btu/hr.
-------
H*1
i1 ''
o
Q.
Q.
>>
S-
Q
700! r
600 -
500. -
400 -
C\J
o
o
300 -
-------
emissions for the HNR burner with coal diffuser decreased essentially
linearly with excess air, down to less than 200 ppm and an overall stoich-
iometry of 1.07. Fly ash carbon content increased linearly as excess air
decreased, but remained less than 6 percent at very low excess air levels.
The HNR burner equipped with a coal diffuser and the modified outer register
vanes yielded a family of data for three firing rates. Both the N0X emission
and fly ash carbon data formed nearly parallel curves for the three firing
rates over the excess air range tested. As expected, N0X emissions were
highest and fly ash carbon content lowest for full load. Combustion
efficiency for this HNR configuration was very sensitive to excess air and
firing rate.
The performance of the HNR burner with and without the modified outer
register is directly compared in Figure 4-15. The modified outer register
produced the lower N0X emissions with no significant impact on combustion
efficiency. There was a tradeoff, however, in flame length. The lower NQX
emissions with the modified outer register were achieved at the expense of
flame lengths exceeding 22 ft at full load. Table 4-7 summarizes the key
conditions for the HNR burner.
4.3 XCL Burner
The XCL burner was evaluated in two series of tests. The initial series
considered the following configurations to determine the hardware arrangement
for optimum performance within the Edgewater boiler constraints:
• Flame Stabilizing Ring with coal diffuser.
• Flame Stabilizing Ring with coal diffuser and modified outer spin
vanes.
• DeMOx Stabilizer
• 30° coal impeller
t Coal diffuser
The initial tests were conducted with Pittsburgh #8 coal at operating
conditions comparable to the DRB and HNR burner tests:
¦ 4-31\
-------
, to
ro
Q.
CL
>5
s-
~
CM
O
O
C2*
700
600
500
400
300
200
100
I 1 1
i i
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate = 79.35
x 106 Btu/hr
:
? :
i i i
• i
1.00 1.10 1.20 1.30 1.40 1.50
Overall Stoichiometry—SRy
14
12 -
u 10
-------
TABLE 4-7. SUMMARY OF KEY HNR BURNER TEST CONDITIONS
Confi gura-
tion
Registers
Inner
Outer
Inner
Sleeve
Damper
(7o
Open)
Spreader
Position
(in.)
Firing
Rate
(106 Btu
/hr)
SRT
Emi ssions
(0 0% 0?)
N0X CO
F1 ame
Length
(ft)
Fly Ash
Carbon
(wt.%)
Wi ndbox
Pres.
(in.W.G.)
Swirl er
10°CW
15°CW
25
-10
79.0
1.19
348
18
18-20
7.20
Di ffuser
10°CW
20°CW
12
Diffuser
77.7
1.20
289
30
20
3.34
7.50
Out Reg.
Mod.
10°CW
20°CW
25
Di ffuser
79.1
1.19
222
24
22
2.65
6.80
Out Reg.
Mod.
10°CW
20°CW
25
Di ffuser
61.9
1.20
181
42
22
6.86
4.10
Out Reg.
Mod.
10°CW
20°CW
25
Di ffuser
51.0
1.19
144
138
18-19
2.70
i
-------
Firing Rate = 78 x 10® Btu/hr
SRp = 0.18
SRy 1.20
Primary Air Temperature = 150°F
Combustion Air Temperature = 550°F
Based on the initial test series results, the XCL burner was identified as
having the most potential to satisfy the requirements of the LIMB
demonstration. The additional hardware configurations considered in the
second XCL optimization test series included:
• 20° coal impeller
• Open impeller
• Expanded nozzle with coal diffuser
• Expanded nozzle with 20° impeller
• Expanded nozzle with 30° impeller
• 30° impeller and fixed outer vane assembly
• Diffuser and fixed outer vane assembly
The second optimization series was conducted at a lower overall
stoichiometry, 1.16. This condition, specified by 3&W, was representative of
operation in most B&W boilers. At this lower, excess air level, the lower air
flow would result in lower burner pressure drop as well as lower N0X
emissions. Optimizing burner performance at this lower excess air level was
expected to provide data that would fulfill Edgewater boiler constraints.
Following the optimization of XCL burner hardware in this second series, the
final configuration was characterized with three alternate coal's.
4.3.1 Optimization of Burner Hardware
The initial screening tests to determine optimum XCL burner hardware
were conducted over the range of burner variables listed in Table 4-8.
-4-34-
-------
TABLE 4-8. RANGE OF INITIAL SCREENING TESTS OF XCL BURNER
i
Confi gurati on
Di ffuser/
FSR
Diffuser/FSR
w/Mod. Outer
Vanes
DNS
30o
Impel 1er
Di ffuser
Burner Settings:
Inner Spin Vanes
10-30°CW
10-50°CW
10-50°CW
20-70°CW
20-50°CW
Outer Spin Vanes
20-60°CW
10-85°CW
10-50°CW
20-50°CW
20-70°CW
Inner Sleeve Damper
25% Open
0-37.5% Open
25-100%
Open
25-100%
Open
25-75%
Open
Outer Sleeve Damper
85% Open
77% Open
100%
Open
100%
Open
61.5-85%
Open
Impeller Position
N/A
N/A
N/A
-6 to +6
in.
N/A
Performance Range:
N0X @ 0% O2, Dry (ppm)
216-314
211-322
228-306
252-700
229-346
Flame Length (ft)
22->22
21-> 22
19->22
12->22
20->22
Burner P (in. W.G.)
1.2-12.0
0.9-12.5
3.1-13.0
2.5-10.6
2.8-9.5
1
-------
Results of these screening tests are summarized in Figures 4-16 and 4-17.
N0X emissions less than 300 ppm could be achieved with both FSR-equipped XCL
burner configurations, the DeN0x stabilizer, and the standard B8W coal
diffuser for a number of conditions, as shown in Figure 4-16. However, for
most burner settings with these burner components the flame length associated
with these 1 ow-NOx levels exceeded 22 ft. The XCL burner with the DeNOx
stabilizer could achieve low N0X emissions, 288 ppm, and a reasonable flame
length, 20 ft, with tightly closed, high-swirl inner and outer spin vane
positions. These spin vane positions, however, resulted in a high burner
pressure drop, 8.2 in. W.G. As for the DRB and HNR burner, the outer zone
was the dominant burner adjustment affecting burner performance.
The coal diffuser produced low emissions and adjustments could be made
to yield a flame about 22 ft. long with pulses exceeding the furnace depth
{> 22 ft). At optimum settings for the XCL burner with diffuser, N0X
emissions were 276, 205, and 194 ppm at firing rates of 80.8, 59.9, and 50.4
x 10® Btu/hr, respectively. The effect of excess air on the performance of
the XCL burner with the coal diffuser at these three firing rates is sum-
marized in Figure 4-18.
The 30° impeller installed in the XCL burner had a dramatic effect on
performance and burner sensitivity, as shown in Figure 4-17. Position of the
impeller in relation to the coal nozzle was a critical parameter in burner
performance. Inner and outer zone adjustments were of secondary importance,
functioning to fine tune burner performance. Moving the impeller a matter of
inches changed N0X emissions from 600 ppm to 250 ppm with a corresponding
change in flame shape and length, from a widely flaring, short flame to a
long, narrow flame. In addition, a hysteresis was observed during adjustment
of impeller position. Retracting the impeller back into the coal nozzle
{denoted as the negative direction) had little effect on N0X emissions and
flame length until the impeller was 5 inches inside the coal nozzle. At this
point there was a step change in performance, with the flame narrowing and
increasing to over 22 feet in length. N0X emissions dropped correspondingly.
When pushing the impeller back toward the coal nozzle exit (positive
4-36
-------
500
450
400
350
300
250
200
150
>22
22
20
18
16
14
12
10
1 1 1 1
1 1 ¦ 1 1 1 1 1 1
1 1 1 III
: \
\r
-
_ ^
1 1 1 1
/
—i 1—i L i 1 i 1 ¦
.i.i.i.i.
FSR w/Modified
Outer Vanes
FSR
DNS
Dlffuser
10 20 30 40
Inner Spin Vanes
(°CW)
20 40 60 80
Outer Spin Vanes
(°CW)
20 40 60 80
Inner Sleeve Damper
(Percent Open)
4-16. Results of initial screening tests of XCL burner for four configurations
-------
600
550
500
450
400
350
300
250
200
150
>22
22
20
18
16
14
12
10
1 1
1
j
1 > 1 > 1
1 | 1 | • | • | 1
1 | l | . | 1 | r -
¦ i i i ¦ i - i i
<
1 l
>
1
1 1 1 1 1
/
l 1 i 1 > 1 i 1 i
r'
i i—j i— i _ i i —i i
.i.i.i.i.
8 -4 0 4 8 20 40 60 80 20 40 60 80 20 40 60 80
Impeller Position Inner Spin Vanes Outer Spin Vanes Inner Sleeve Damper
(i"-) (°CW) (°CW) (Percent Open)
17. Results of initial screening tests of XCL burner with 30° i
-------
700
600
500
i E
l£
! >,
• s-
, p
400
i *
C\J
'O
%0 d
4-39
300
i x
0
1 z
200
100
1 1 1 1
1
Burner Settings:
Inner Spin Vanes = 30°CW
Outer Spin Vanes = 40°CW
Inner Damper = 75% Open
Outer Damper = 100% Open
—
. O 79 x 106 Btu/hr
. '• 61 x 106 Btu/hr
~ SO x 106 Btu/hr
yO
\
\N
N.
'
6^
l i i i
•
>5
+->
3
(/I
C
o
.a
s-
rtJ
<_>
1.00 1.10 1.20 1.30 1.40 1.50
Overall Stoichiometry—SRy
1.00 1.10 1.20 1.30 1.40 1.50
Overall Stoichiometry--SRj
Figure 4-18. Effect of excess air on performance of XCL burner with coal diffuser.
-------
direction), the flame remained over 22 ft long until the impeller reached its
"0" position, the trailing edge flush with the end of the coal nozzle.
Because of this bistable phenomena, the adjustable burner parameters
were optimized at two different impeller positions representing either side
of the bistable settings, 0 and -5 inches from the nozzle. With the impeller
at the 0 position, N0X emissions were 449 ppm with an 18 ft long flame at
nominal full load conditions (78 x 105 Btu/hr and SRj = 1.20). With the
impeller retracted into the nozzle 5 inches, N0X emissions were 374 ppm with
a 20-22 ft long flame. The XCL burner with the 30° impeller at -5 in. was
characterized over a range of excess air and firing rate, with the results
shown in Figure 4-19. For this configuration, combustion efficiency as fly
ash carbon content was very good down to about 10 percent excess air even at
reduced load. The data indicate that relatively low emissions can be
achieved with good carbon burnout and reasonable flame length.
These initial screening tests indicated that the XCL burner had poten-
tial to meet the LIMB demonstration goals at Edgewater. Of the
configurations and conditions evaluated, listed in Table 4-9, the coal
diffuser and coal impeller were most promising. Although the coal diffuser
produced a flame slightly longer than the furnace depth, N0X emissions were
very low and warranted consideration. The 30° impeller could achieve
relatively low emissions and acceptable flame lengths and, more importantly,
provided a wide range of adjustment to performance. The performance of the
XCL burner with these two components is compared in Figure 4-20. The N0X
emissions from these two configurations followed the same trends over a range
of excess air, with diffuser producing about 100 ppm less N0X. The 30°
impeller, however, had significantly better carbon burnout permitting
operation at lower excess air levels.
The encouraging results from this initial series of screening tests
provided direction to the second test series. The second series of tests
focused on additional development of the coal nozzle/impeller design.
Further refinement of the coal nozzle/impeller was expected to achieve an
optimum compromise between the very low N0X produced by the diffuser and the
4-40
-------
I-P»l
i-e*!
Q.
Q.
700
600 -
500 -
>>
° 400
CM
O
O
<2>
X
o
T
O 82 x 106 Btu/hr
•I 50 x 106 Btu/hr
300 -
200
100 -
Burner Settings:
Impeller Position = -5 in.
Inner Spin Vanes = 30°CW
Outer Spin Vanes=60°CW
Inner Damper = 75% Opeq.
Outer Damper = 100% Open
I I I I :
-------
TABLE 4-9. SUMMARY OF KEY TEST
CONDITIONS FROM INITIAL XCL BURNER SCREENING
TESTS
Sleeves
Fi ri ng
j
Confi gura-
tion
Regi sters
Inner
(%
Open)
Outer
Impeller
Rate
(106 Btu
/hr)
SRj
Emi s;
;ions
Flame
Fly Ash
Windbox
|
Inner
Outer
(S
Open)
Position
(Inches)
@ 0%
N0X
0 0%
CO
Length
(ft)
Carbon
(wt.%)
j Pres.
jtin.W.G.
FSR, O.R.MOD
30°CW
40°CW
25
77
N/A
78.0
1.19
222
36
>22
3.10
FSR
20°CW
40°CW
25
85
N/A
78.4
1.19
246
36
>22
3.20
Diffuser
30°CW
40°CW
75
100
N/A
80.8
1.19
276
48
22+
6.97
5.10
Di ffuser
30°CW
40°CVJ
75
100
N/A
59.9
1.20
205
42
22+
1.98
2.50
Diffuser
30°CW
40°CW
75
100
N/A
50.4
1.21
194
64
22+
9.20
1.50
DNS
20°CW
30°CW
75
100
N/A
77.8
1.19
288
36
20
8.20
30° Impeller
30°CW
60°CW
75
100
0
77.7
1.21
449
14
18
3.50
30° Impeller
30°CW
60°CW
75
100
-5
79.9
1.20
374
24
20-22
4.42
3.30
30° Impeller
30°CW
60°CW
75
100
-5
59.6
1.19
306
26
20-22
5.18
1.60
30° Impeller
30°CW
60°CW
75
100
-5
49.8
1.20
301
30
19-22
4.44
[ 1.00
r- ¦ !
i i.
-------
300-
20C-
t 1 r
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate = 80.6 x 10° Btu/hr
_L
X
±
1.00 1.10 1.20 1.30 1.40 1.50
Overall Stoi.cbiometry—SRj
Figure 4-20. Comparison of coal diffuser and 30°
,Oi Diffuser
^>30° Impeller
1.10 1.20 1.30 1.40 1.50
1.00
Overall Stoichiometry—SRj
impeller XCL burner configurations.<
-------
effective flame shaping control arid exceptional burnout achieved with the 30°
impeller. In addition, the potential to minimize burner pressure drop and
enhance mechanical reliability was evaluated using fixed position outer swirl
vanes.
The range of these final screening tests for the XCL burner is listed in
Table 4-10. As described previously, these final tests were conducted at *'
lower excess air, an SRy = 1.16, than the previous tests, SRy = 1.20. These
tests can be grouped in three sets:
• Impeller evaluation with standard coal nozzle
• Evaluation of expanded coal nozzle tip
• Evaluation of fixed outer swirl vanes
Results/of screening tests of three impeller designs with the standard
coal, nozzle are summarized in Figure 4-21. The impellers evaluated included
the previously tested 30° impeller, 20° included angle impeller, and an open-
design 30° impeller. Because there was a span of time between the initial
XCL screening tests and the final series, repeat of the 30° impeller served
to verify similarity between tests and provided a basis with which to compare
the new designs. General trends of performance with burner adjustments were
similar for the initial and final screening tests of the 30° impeller
configuration. * However, the hysteresis in performance related to impeller
position observed during the initial test series did not occur. Because the
impeller affects the aerodynamics near the burner, the change in performance
between the two test series is probably due to differences in burner
conditions, namely excess air level and spin vane position. The lower flow
rate of air through the burner lowers both the axial and tangential momentum
of the burner. During this final series, N0X emissions ranged from 310 to
467 ppm with flames from 16 to 22 ft long.
The 20° impeller was intended to be a compromise design between the 30°
impeller and the coal diffuser. The 20° included angle impeller was designed
to impart a radial component less than that from the 30° impeller but more
than the purely axial flow, from the diffuser. Burner performance with the
4-44
-------
TABLE 4-10. RANGE OF FINAL SCREENING TESTS OF X
^ • 1 -i
CL BURNER
Configuration
20°
Impeller
30°
Impeller
Open
Impeller
Diffuser w/
Expanded
Nozzle
20O
Impeller
w/Expanded
Nozzle
300
Impeller
w/Expanded
- Nozzle
30°
Impeller
w/F i xed
Vanes
Burner Settings:
Inner Spin Vanes
Outer Spin Vanes
Inner Sleeve Damper
Outer Sleeve Damper
Impeller Position
20-40°CW
30-50°CW
75% Open
100% Open
-6 to +3
Inches
20-40°CW
30-60°CW
12.5-100%
Open
100% Open
-6 to +3
Inches
30°CW
30-50°CW
37.5-75%
Open
100% Open
-3 to +3
Inches
30°CW
10-40°CW
12.5-75%
Open
100% Open
N/A
20-35°CW
30-50°CW
37.5-75%
Open
100% Open
-5 to 3
Inches
20-35OCW
30-60°CW
37.5-75%
Open
100% Open
-5.5 to
3.0 in.
20-40°C
45°CW
12.5-75
Open
42-100%
Open
-3.0 to
+3.0 in.
Performance Range:
N0X @ 0% O2, Dry
(ppm)
Flame Length (ft)
Burner P (in.W.G.)
252-397
20->22
4.1-7.0
310-467
16-22
3.4-6.7
233-310
>22
4.5-8.0
173-291
>22
6.0-10.6
196-618
15-> 22
4.2-6.9,
397-560
14-20
3.0-7.1
280-557
5->22
3.5-6.2
-------
-Pa
i I
i-P*i
•>
*-
O
400
CM
O
350
o
(St
300
X
250
200
150
+j
at
ai
22
22
20
18
16
14
12
10
i 1 i ' i '—r
J i—i I I i L
-L
_L
-4-
_L
~i 1 1 r
_L
_1 L
_L
20° Impeller
¦ — 30° Impeller
* ••• Open Impeller
-6 -4 -2
Impeller Position
(in.)
+2 10 20 30 40
Inner Spin Vanes
(5CW)
20 30 40 50
Outer Spin Vanes
(bCW)
20 40 60 80
Inner Sleeve Damper
(Percent Open)
Figure 4-21. Results of final XCL burner screening tests with standard coalnozzle.
-------
20° impel 1 er was most sensitive to impeller position and outer spin vane
position. As for all other configurations, N0X emissions and flame length
were closely related. N0X emissions ranged from 252 to 397 ppm with
relatively narrow flames 20 to over 22 ft long.
The open impeller consisted of a large central opening surrounded by two
conical rings with a 30° included angle. The principle behind the design was
to strip a fraction of the coal from an axial flow and divert it directly
into the swirling secondary air to provide stability. The remaining coal
would be introduced through the open center in an axial jet. It was hoped
that this configuration would produce a compromise between the 1ow-NOx, long
flames from the diffuser and the hi gh-NOx, short flames from the 30°
impeller. The actual performance was comparable to the conical diffuser,
with low N0X emissions (233 to 310 ppm) and long flames (over 22 feet).
Apparently the outer rings did not deflect much of the coal into the
secondary air stream. The bulk of the coal was introduced axially through a
smaller, higher momentum jet, thus resulting in long flames.
The expanded nozzle consisted of a short, abrupt expansion replacing the
end of the straight coal nozzle. This larger diameter was installed to
evaluate the effect of lower coal velocities on burner performance. Although
primary air flow could be varied, decreasing the primary air flowrate not
only decreased velocity but also lowered the stoichiometry in this primary
air zone. By making the physical change, the aerodynamics would be evaluated
independently of combustion stoichi ometry. This expanded nozzle tip was
tested with the standard coal diffuser, a slightly modified 20° impeller, and
the 30° impeller. The results of the screening tests with the expanded
nozzle are shown in Figure 4-22.
The performance of the XCL burner with the coal diffuser, located far
upstream of the coal nozzle tip, was not significantly affected by the
expanded nozzle tip. The resulting emissions were 173 to 198 ppm with flames
over 22 feet long. The short nozzle tip, less than 0.5 coal pipe diameters,
probably did not allow the coal jet to fully expand to the larger diameter
4-47
-------
-e»
i
-c*
CO I
600
550
500
3 450
»>
O 400
° 350
V!
°x 300
250
200
150
>22
22
20
18
16
14
CT>
C
01
Q)
£
12
10
I ' I ¦ I ¦ I ' I
x
x *.
X .
X
•x
• X
• X
s
N
: --r'"
j i i ¦ ' ¦ i
_L
_L
~i r
_L
_L
_L
_L
1 1 1 r
_L
_L_
_L_
-6 -4 -2
Impeller Position
(in.)
+2 10 20 30 40
Inner Spin Vanes
(°CW)
20 30 40 50
Outer Spin Vanes
(°CW)
20 40 60 80
Inner Sleeve Damper
(Percent Open)
Diffuser
30° Impeller
20° Impeller
Figure 4-221. Results of final SCL burner screening tests with expanded nozzle tip.
-------
before exiting the coal pipe. Thus, the effective coal jet velocity was
probably not significantly different than the standard nozzle.
The 20° impeller was slightly modified to fit the expanded nozzle. An
additional conical ring was added to enlarge the diameter of the impeller and
more fully fill the expanded nozzle. The N0X emissions achieved ranged from
196 to 618 ppm with flames from 15 to 22 feet long. Again there was a very
close correlation of N0X with flame length. This configuration demonstrated
effective burner performance control with impeller position. N0X emissions
decreased essentially linearly from 618 ppm to 196 ppm (flame length from
15.16 to over 22 feet) as the impeller was moved from -3 to +3 inches. In
the short test period, there did not appear to be any hysteresis in per-
formance. As with the other impeller configurations, outer register vane
position was also an effective control. Apparently, the low velocity from
the expanded tip could be taken advantage of with a device that filled the
entire large diameter.
The 30° impeller, with no modifications to accommodate the larger
diameter nozzle tip, was also evaluated with this expanded nozzle configura-
tion. Again there was a nearly linear response in N0X emissions as the
impeller was moved. There was, however, an indication that some hysteresis
might be present upon startup at the impeller "0" position. With N0X
emissions ranging from 397 to 560 ppm and, corresponding 20 to 14 foot long
flames, further evaluation of this "hysteresis" was not warranted.
The effect of the expanded coal nozzle tip is shown explicitly in Fig-
ure 4-23. Data with the standard and expanded nozzles are compared for the
20° impeller and 30° impeller, respectively. The 30° impeller, which was not
modified and thus had the same outside diameter, yielded similar performance
in both coal nozzles. The performance of the XCL burner with the 20°
impeller, however, was significantly different with the expanded nozzle tip.
The lower velocity at the nozzle tip coupled with the enlarged 20° impeller
achieved N0X emissions, which could be varied from 600 ppm to 200 ppm simply
by moving the impeller 6 inches. The N0X performance was coupled to flame
length ranging from about 16 ft to over 22 ft.
' 4-49;
-------
>, 500
-400 "
i 1 1 1 r
CD
E
ro
20 Impeller
^—
A
'A
o—o—6*\-o—6
0 A
\ 1 1 1 \
¦5 -4 -3 -2 -1 0 +1 + 2 +3
Impeller Position (Inches)
800
700
E
i
Q-
O.
i 600
>1
'500
1
¦ S-
o
CO
o
0
0
>3-
'
o
<2j
300
X
o
sr
200
100
0
, 22
t i T r
O Standard Nozzle
A Expanded Nozzle
t—i—r
30° Impeller
1 1 1 1 1 1 1"
C7>
C
-------
The final XCL burner component evaluated was a set of fixed position
outer secondary air axial swirl blades. These fixed outer vanes, set at a
45° angle, represent a mechanical simplification and, with appropriate
design, allow efficient swirl generation with low pressure drop. Both of
these factors are important in commercial application. The fixed vanes were
tested with the 30° impeller and briefly with the coal diffuser in the
standard coal nozzle. The screening test results with the 30° impeller are
shown in Figure 4-24. Again burner performance was most sensitive to
impeller position. N0X emissions ranged from 280 to 557 ppm with correspond-
ing flames from over 22 ft to 15 ft long. With the diffuser, the XCL burner
and fixed outer vanes yielded 287 ppm N0X with a flame length over 22 ft.
The fixed outer vane design gained a modest decrease in burner pressure
drop--about 0.5 in W.G., over the adjustable vane design.
Key conditions from these parametric optimization tests are summarized
in Table 4-11 for all eight configurations. Five of the configurations were
also evaluated over a range of excess air at full load. These results are
shown in Figure 4-25. Lowest N0X emissions with flames less than 22 ft long
over the range of excess air evaluated were obtained with the 30° impeller
and the 20° impeller/expanded nozzle. The performance for these two configu-
rations was essentially the same with the 20° impeller achieving somewhat
better combustion efficiency.
4.3.2 Characterization of XCL Burner Performance
The XCL burner design represents an advancement of B&W technology, but
with specific operational experience limited to this test program. To gain
further experience with the XCL burner and to broaden its application, two
configurations were selected for further evaluation. The two configurations
selected were the coal diffuser in the standard coal nozzle with fixed outer
vanes and the 30° impeller in the standard coal nozzle with fixed outer
vanes. The coal diffuser configuration represents minimum N0X emissions in a
practical burner design. Its use would be restricted to applications whose
firing depth can accommodate long flames. The 30° impeller arrangement is
suitable for tight boilers, such as Edgewater Unit 4, with its ability to
' 4-51
-------
650
600
550
500
450
400
350
300
250
200
150
22
20
18
16
14
12
10
i—i—i—i—r
' ' ' » I I L
X
X
X
X
X
X
-3 -2 -1 0 +1 +2 +3
Impeller Position
(in.)
10 20 30 40
Inner Spin Vanes
(%)
20 40 60 80
Inner Sleeve Damper
(Percent Open)
ire 4-24. Results of final XCL burner screening tests with
30° impeller and fixed outer vanes.
4-52
-------
TABLE 4-11. SUMMARY OF FINAL XCL BURNER SCREENING TEST CONDITIONS
Sleeves
Fi ri ng
Configura-
tion
Registers
Inner
(%
Open)
Outer
(%
Open)
Spreader
Posi tion
(Inches)
Rate
(106 Btu
/hr)
srt
N0X
0 0%
02
CO
@ 0%
02
Flame
Length
(ft)
Fly Ash
Carbon
(wt.%)
Wlndbox
Pres.
(in.H20)
Coal
Inner
Outer
20° Impeller
30°CW
40°CW
75
100
0.0
80.0
1.16
315
29
22+
-
5.30
Pitt. #8
30° Impeller
30°CW
50°CW
38
100
0.0
76.9
1.17
317
33
21
6.11*
4.30
Pitt. #8
Open Impeller
30°CW
40°CW
75
100
0.0
79.1
1.16
268
43
>22
-
5.10
Pitt. #8
-p»i
1 i
Expanded
Nozzle
w/Diffuser
30°CW
50°CW
75
100
NA
59.7
1.19
198
90
>22
-
-
Pitt. #8
:22
- "
4.30
Pitt. #8
*UBC Data Listed for SRy = 1.19
-------
800
O' 30° Impeller ^
iD];30° Impeller and Expanded Nozzle' 1
A | 20° Impeller and Expanded Nozzle
jOI 30° Impeller and Fixed Outer Vanes
Q Diffuser and Fixed Outer Vanes
Nominal Conditions:
Fuel: Pittsburgh #8
700
80 x 10° Btu/hr
I
600
CL
g 500
Ll_
o
300
o
200
100
1.10
1.30
1.40
1.50
1.00
1.20
1.30
1.40
1.10
1.20
1.00
Overall Stoichiometry—SR
Overall Stoichiometry—SRj
Figure 4-25. Performance characteristics for five XCL burner configurations.
-------
adjust performance using impeller position. The additional testing consisted
of evaluating the XCL burner with three additional coals at nominal full-load
operating conditions over a range of excess air. The alternate coals were a
Utah high-vol ati1e bituminous coal, Lower Kittanning medium volatile
bituminous coal, and a Wyoming subbituminous coal used at Colorado Public
Service Comanche station.
The burner settings for the characterization tests of the impeller
equipped XCL burner were:
Impeller position = +1.5 in. ¦
Inner Spin Vanes = 40°'CW
Fixed Outer Vanes = 45° CW
Inner Damper = 381 Open
Outer Damper = 1002. Open
The performance results for this configuration are shown in Figures 4-26, 27,
and 28 for Pittsburgh #8, Utah, and Lower Kittanning coals, respectively.
The general trends for each coal are the same, with higher N0X emissions and
lower unburned carbon levels (UBC) for higher firing rates and excess air
levels. For an overall stoichiometry of 1.20, the results for each coal
were:
Nominal
Firing Rate
(10® Btu/hr)
Pittsburgh #8
Utah
Lower Kittanning 1
N0X
UBC
N0X
UBC
N0X
UBC
80
420
3.4
414
2.9
573
6.5
62
306
4.9
276
5.9
390
6.5
49
229
6.5
187
10.6
210
22.0 |
4-55
-------
800
700 -
600 -
500 "
CVI
O
--o 400
-p»i
• i
:«1
!CT> X
Jo
300 -
200 -
100
1 r
O 78.4 x 106 Btu/hr
~ 62.2 x 106 Btu/hr
A 48.4 x 106 Btu/hr
1.0 1.1 1.2 1.3
Overall Stoichiometry--SRj
1.4
9. 10
\
1.1 1.2 1.3
Overall Stoich-iometry—SRj
Figure 4-26. Performance of XCL burner with standard coal nozzle, diffuser, and
fixed outer vanes firing Pittsburgh #8 coal. '
-------
o
I
81.8 x
1
106 Btu/hr
1
_ ~
63.3 x
106 Btu/hr
A
51.7 x
106 Btu/hr
-
O
O
/
~
/ /
-
A
O
j/ /
/
~'
i
»
«
1.0 1.1 1.2 1.3
Overall Stoichiometry—SRy
1.4
\
9i 10
1.1 1.2 1.3
Overall Stoichiometry—SRj
Figure 4-27. Performance of XCL burner with standard coal nozzle, diffuser, and
fixed outer vanes firing Utah coal.
-------
T
T
O 79.6 x 10b Btu/hr
~ 62.0 x 106 Btu/hr
A 46 x 106 Btu/hr
A"
1.0 1.1 1.2 1.3
Overall Stoichiometry--SRy
1.4
1.1 1.2 1.3
Overall Stoich'iometry—SRj
Figure 4-28. Performance of XCL burner with standard coal nozzle, diffuser, and fixed
outer vanes firing lower Kittanning coal.
-------
These results are generally consistent with the rankings expected based on
the correlation of fuel composition with N0X discussed in Section 3.5 of this
report. Given the level of N0X, this configuration may be considered
somewhere between a staged and radial diffusion flame. Direct comparison of
the performance for each fuel is shown in Figure 4-29. The two high-volatile
bituminous coals yielded essentially the same N0X emissions. Much higher N0X
emissions were measured for the medium volatile Lower Kittanning coal.
Combustion efficiency, as indicated by fly ash carbon content, was also very
similar for the Pittsburgh #8 and Utah coals. The Lower Kittanning coal,
however, yielded the highest fly ash carbon levels.
The XCL configuration with the diffuser was characterized using the
following burner settings:
Inner Spin Vanes = 30° CW
Fixed Outer Vanes = 45° CW
Inner Damper =75% Open
Outer Damper = 100% Open
This configuration was characterized for three coals at nominal full load;
Pittsburgh #8, Lower Kittanning and Comanche. The results are summarized in
Figure 4-30. The lowest N0X emissions were achieved with the subbituminous
Comanche coal as expected. The emissions for the other two coals were higher
than the Comanche, but very similar to each other. With this very low N0X
configuration, which can be considered aerodynamically staged, the fuel
composition-NOy correlation indicates that the Lower Kittanning and Pitts-
burgh #8 coals could be expected to produce similar results. Combustion
efficiency was substantially different, with the subbituminous coal yielding
the lowest unburned carbon levels and the medium volatile bituminous Lower
Kittanning coal producing the highest fly ash carbon levels. Flame length
for the Pittsburgh #8 and Comanche coals exceeded 22 ft, while the Lower
Kittanning coal yielded flames 21-22 ft long.
4-59
-------
800
700
600
E '
Q."
Q.
>—'
1
>,
i
S-
j
Q
«\
CVJ
o
-p*
1
o
Ol
o
<&
X
o
0~1 Pittsburgh #8 Coal
q j Lower Kittanning Coal
- , A Utah Coal
400
300
200
100
X
_L
X
X
1.00 1.10 1.20 1.30
Overall Stoichiometry—SRj
1.40
14
12
10
d) I
)
1 1 1 1-
Nominal Conditions:
Firing Rate = 80 x 10® Btu/he
1.00
1.10
1,20
1.30
Overall Stoichiometry--SRy
Figure 4-29, Comparison of performance of XCL burner with Standard Coal Nozzle Diffuser,
and fixed outervanes for three different coals.
1.40 <
-------
1700
'600
500
iji
' i
en
Dl
cl
>>
i-
o
<\J
o
o
-------
The performance of the two XCL burner configurations are directly
compared in Figure 4-31 for two common coals, Pittsburgh #8 and lower
Kittanning. At nominal full-load conditions with 20 percent excess air, N0X
emissions with the Pittsburgh #8 coal were 420 ppm and 304 ppm for the 30°
impeller and diffuser, respectively. Under similar conditions the difference
in N0X emissions was even greater with the Lower Kittanning coal, namely 573
ppm for the 30° impeller and 300 ppm with the diffuser.
4.4 Comparison of Burner Performance
Twenty configurations of three basic burner concepts, the Dual Register
Burner, Babcock-Hitachi NR. burner, and XCL burner, were evaluated as part of
this program. In addition, data from five configurations of another ORB
design, the Phase V DRB, was available from an analogous test program
sponsored by B&W. To facilitate comparison of such a large number of
burners, the performance of the burners should be screened with respect to
this program's ultimate objective—selection of the second generation low-NOx
burner for retrofit at the EPA limb demonstration site, Edgewater Unit 4.
This imposes the following constraints-that the selected burner must meet:
• Nominal 78 x 106 Btu/hr heat input per burner.
• Flame length less than the firing depth of the boiler, 22 ft 3 in.
• Burner pressure drop within fan limitations, nominally 5 in. W.G.
• Throat diameter no greater than 35 inches.
• Low N0X emissions, at least 50 percent less than baseline burners.
t High combustion efficiency.
• Mechanical reliability meeting commercial standards.
Each of the burner configurations were designed for firing at 78 x 10® Btu/hr
with a throat diameter no greater than 35 inches. In addition, B&W
incorporated its standard commercial hardware designs thereby ensuring that
mechanical reliability will be at least equivalent if not improved, to its
4-62
-------
Pittsburgh #8 Coal
Lower Kittanning Coal
800
700
600
6'
Q-
Q-'
Ts 500
u
o
CVJ
o
£ g 400
U)
CO,
300
200
100
T T
0 Diffuser
30° Impeller
T
1.00
1.10
1.20
1.30
Overall Stoichiometry--SRy
1.40
/
>> 500
Overall Stoichiometry--SRy
CE
V
Figure 4-31. Comparison of XCL burner performance with diffuser and 30° impeller-
fixed outer vane configuration.
1.40l
-------
commercial burners. The performance of the burners is therefore the key to
proper burner selection.
Of the remaining requirements, flame length imposes the most severe
constraint. Data from tests of the Low Velocity ORB, HNR burner, and the
initial screening of the XCL burner, presented in Figure 4-32, clearly
demonstrate the correlation of N0X emissions with flame length. At
20 percent excess air and full load conditions, the test data indicate that
for a flame less than the firing depth of the LW5, or the Edgewater boiler,
N0X emissions in the range of about 300-400 ppm were achieved by several
burner configurations. Using flame length as the initial selection criteria
yields 13 candidate burners listed in Table 4-12. From this list, the
potential of each burner design for application to the Edgewater boiler can
be evaluated.
Only one of the four Low Velocity Dual Register Burner configurations
tested met the flame length criteria. The configuration with the 75°
impeller did indeed shorten the flame to 18 ft, but resulted in N0X emissions
of 708 ppm. The 75° impeller is a design from a high turbulence B&W Circular
burner which was not optimized for low emission performance. The impact of
this impeller design demonstrates the importance of coal injector design in
that, even though the Low Velocity ORB incorporated low-NOx design character-
istics (low combustion air velocity and multiple air zones), the overall
burner performance was dominated by the impeller.
Data more representative of DRB performance was collected during the B&W
sponsored testing of a Phase V DRB. All five configurations tested could
produce a flame 22 ft or less in length. In addition, each configuration
achieved emissions less than 372 ppm. However, this could only be achieved
with tightly closed inner and outer vanes to produce highly swirling flow.
Applying a second requirement, burner pressure drop no more than 5 in. W.G.,
eliminates all five Phase V DRB configurations from application at Edgewater
Unit 4.
4-64
-------
900
800
700
600
500
400
300
200
100
o'
£
i—i—i—i—i—i—i—i—i—i—i—i—i—r
A
AA
A
I -
4 * A
o- o.
~ A- *
A 0
2 K
1s»ti
Nominal Conditions:
Fuel: Pittsburgh #8
Firing Rate: 80 x 106 Btu/hr
SRj =1.20
DRB ¦
O Diffuser j
A 1 75° Impeller
~ DNS
O FSR/ASP/Swirler
r HNR
O Swirler
0 Diffuser
XCL "
• Diffuser
A 30° Impeller
¦ ;dns
* FSR
J 1 1—J 1 I 1 I l i i ' i '
10 12 14 16 18 20 22
Observed Flame Length (feet)
Figure 4-32. Correlation of N0y emissions with flame length.
-------
TABLE 4-12$ ' SECOND GENERATION LOW NOx BURNERS WITH FLAMES <22 FT LONG
{78 x 1Q6 Btu/hr, SRj = 1.20)
Burner
Conf f guratiori
N0X
@ 0"? 02
ippm)
Flame
Length
(ft)
Fly Ash
Carbon
(wt.%)
Burner AP
(in.
W.G.)
Low-Velocity DRB
75° Impeller
708
18
7.28
6.0
Phase V DRB
Oiffuser
372
20-21
6.12
10.8
Venturi
350
20-21
6.45
11.0
Diffuser, ASP
326
22
3.20
6.4
Diffuser, FSR
292
22
6.96
10.5
Diffuser, FSR, ASP
328
22
5.16
11.0
HNR
Swirl er
348
18-20
N/A
7.20
Diffuser
289
20
3.34
7.50
XCL
DNS
288
20
N/A
8.2n
30° Impeller, Standard
Nozzle
374
20-22
4.42
3.30
30° Impeller, Expanded
Nozzle
546
19-20
1.36
4.30
20° Impeller, Expanded
Nozzle
338
21
4.92*
4.90
30° Impeller, Standard
Nozzle, Fixed Outer
Vanes
420
21-22
3.40
4.60
*Data for SRj = 1.16
'"f-66 :
-------
The Babcock-Hitachi MR burner had two configurations meeting the flame
length requirement while producing NOx emissions less than 350 ppm. To
achieve this performance, however, the HNR burner relies on tight burner
settings to produce high swirl to shape the flame. These tightly closed
burner settings resulted in a burner pressure drop over 7.0 in. W.G., again
over the limitation at the LIMB demonstration site.
From the 13 XCL burner configurations tested, five achieved the flame
length limitation. The proprietary B&W coal nozzle insert, the DeNOx
Stabi 1 i zer, achieved the lowest N0X emissions of all the configurations with
flames less than 22 ft long. However, this performance could only be
achieved with a high burner pressure drop, the result of tightly closed, high
swirl burner settings. The other four XCL configurations met both the flame
length and burner pressure drop constraints. Three of those configurations
utilized the 30° impeller design. The lowest emissions with the 30° impeller
were achieved in the standard coal nozzle, 374 ppm, and the highest emission
In the expanded coal nozzle, 546 ppm. An intermediate level of N0X, 420 ppm,
was achieved with the 30° impeller in the standard coal nozzle and the fixed,
45° angle outer swirl vanes. The lower emissions for the variable outer spin
vane assembly was the result of a lesser degree of swirl at its optimum
setting of 50°. It is likely that with appropriate design, fixed outer vanes
should achieve identical performance as that from the variable spin vanes.
The final XCL configuration meeting both the flame length requirement and the
burner pressure drop limitation utilized the 20° impeller in the expanded
nozzle. This configuration achieved the lowest emissions, 338 ppm, of the
four configurations which met both the flame length and pressure drop
constraints.
From the thirteen burner configurations meeting the Edgewater boiler
flame length requirement, two configurations stand out as the most likely
candidates for application: (1) the 30° impeller in the standard nozzle with
appropriate outer vane design, and 12) the 20° impeller with the expanded
coal nozzle tip. Each offer reasonably low emissions, good combustion
efficiency, and acceptable flame length and burner pressure drop. In addi-
tion, these two configurations offer a very effective "handle" with which to
4-67"
-------
adjust flame length to suit the application—the variable position of the
coal impeller. For both impeller designs, flame length and N0X emissions can
be varied simply by moving the impeller a matter of inches. This adjustment
can be done on-line, while the boiler is in operation. Since the impeller is
a coal nozzle device, it does not affect the pressure drop across the burner
registers and spin vanes. The impeller adjustment can thus be used to tune
the burner for maximum N0X reduction within the constraints of available
firing depth.
4-68
-------
5.0
S02 REDUCTION POTENTIAL WITH SORBENT INJECTION
The technology demonstration at Edgewater Unit 4 will utilize the
injection of dry sorbents into the furnace for SO2 control in addition to the
installation of low-N0x burners, hence the term) "LIMB" from Limestone
Injection Multistage Burner. As part of the initial screening tests, a brief
series of sorbent injection tests were conducted for the Low Velocity DRB,
Babcock-Hitachi NR burner, and XCL burner to determine the effect of burner
design on potential for SO? reduction. The operating variables evaluated
included:
• Sorbent composition
• Sorbent feed rate
t Injection location
These tests were not intended to be comprehensive, only screening tests to
evaluate possible burner differences with established injection locations.
5.1 Injection Configurations
The three burner configurations evaluated with sorbent injection are
listed in Table 5-1. Nominal test conditions during all the sorbent
injection tests were:
Firing Rate = 78 x 106 Btu/hr
SRp = 0.18
SRT = 1.20
Primary Air Temperature = 150°F
Combustion Air Temperature = 550°F
Four different injection locations were considered:
• With the pulverized coal
5-1
-------
• Two upper furnace locations
• Around the periphery of the burner exit
Injection of sorbent with the coal and through upper furnace penetrations is
done as standard procedure to screen for SO2 control in the LWS facility,
including the tests described in Part I, Section 6 of this report. These two
locations are illustrated in Figure 5-1. Sorbent is injected into the
pulverized c.oal stream, just downstream of the pulverizer, allowing thorough
mixing through the length of transport pipe to the burner. For these tests,
only the Vicron 45-3 limestone was injected with the coal.
The two locations in the upper furnace evaluated were 8 ft and 19 ft
above burner center!ine. The 8 ft level, which corresponds to a furnace.gas
temperature of 2500°F, had two 3-in. injection nozzles. Velocity of the
sorbent jet at the nozzle exit at this level was nominally 45 ft/s. Both the
Vicron 45-3 limestone and Col ton hydrated lime were injected at the 8-ft
level. Four 2-in. ID ports were used at the 19 ft level, corresponding to a
gas temperature of about 2150°F. Nominal sorbent jet velocity through these
four ports was 50 ft/s. Only the Colton Hydrated lime was injected at this
uppermost level.
Based on general interest in near-burner injection alternatives, Vicron
45-3 limestone was also injected through four 1-in. ID nozzles equally spaced
in the burner exit. This configuration is illustrated in Figure 5-2. The
nozzles were positioned to follow the divergence of the burner exit with the
nozzle tip flush with the burner exit. The nominal sorbent jet velocity
through these burner nozzles was 200 ft/s.
5.2 Test Results
Figure 5-3 compares the results of all three burners tested with "near"
burner injection of the Vicron 45-3 limestone. There was little difference
in SO2 capture with the limestone injected with the coal for the three
burners, with 28-32 percent capture at Ca/S molar ratio of 2. There was a
1-2.
-------
TABLE 5-1. . BURNER CONFIGURATIONS FOR SORBENT INJECTION TESTS
Burner
Configuration \ .
Impeller
Position
(in.)
Inner .
Vanes
Outer
Vanes
Inner
Damper
Outer
Damper
Low Velocity
DR8~
75° Impeller
0.0
20°CCW
15°CW
50% Open
N/A
HNR
01ffuser, Modified
Outer Vanes
f/A
u
, 10°CW
20°CW
251 Open
N/A
XCL
Diffuser, Standard
Coal Nozzle
N/A
30°CW
40°CW
75% Open •
100% Open
1
! 1
1
-------
cn
; 'I
Sorbent
— from
Feeder
Coal Feeder
m.
Metered
Hot
Air
Burner
Pulverizer
Exhauster
Sorbent Injection with Coal
2-in. ID
19'
8'
3-in. ID
/o
*
19'
/I
Burner
i^r
2150°F
2500°F
Upper Furnace Locations
Figure 5-1. Sorbent injection locations--with coal and upper furance.
-------
orbent
Trans-
port
Air
Nozzles
Spaced
90° Apart
Sorbent
Transport
Air
Figure 5-2. Burner exit sorbent injection nozzle configuration.
-------
With Coal
1
1 1 1
Nominal Conditions:
Fuel:
Pittsburgh No. 8
Firing Rate - 80 x 10 Btu/hr
SRj =
1.20
-
# -
/°
c/
/
_
1
l 1 1 ...
5j«l
70
60
50
40
at
I .
CO
o
tn
20
10
0
Outer Secondary Air Passage
i r
O DRB 80 MKB
~ IINR 80 MKB
A XCL 80 MKB
1
2 3 4
Ca/S
Figure 5-3. Summary of S02 capture with near-burner injection of Vicron 45-3.
-------
significant difference in capture for the outer secondary air injection
location. The capture was highest for the DRB with a coal impeller, the only
burner without the air separation plate and the only configuration with an
impeller. The DRB yielded about 40 percent capture at Ca/S of 2, compared to
only about 30 percent for the other two burners.
A summary of the hydrated lime injection results with all three burners
is presented in Figure 5-4. There was no significant difference in capture
for either above burner location for the three burner designs. The 8 ft
level yielded 36 percent SO2 capture Ca/S = 2, while 38 percent reduction was
achieved at the 19 ft level. These furnace elevations corresponded to
temperatures of about 2500 and 2150°F for the 8-ft and 19-ft levels,
respectively. A comparison of hydrated lime and limestone injection at the 8
ft level is made in Figure 5-5 for all three burners. Again, there is no
difference among the burners. The hydrated lime achieves only slightly
higher SO2 capture, about 36 percent at Ca/S = 2, than the limestone, about
33 percent at Ca/S = 2, for the similar operating conditions.
This brief series of sorbent injection tests indicates that for upper
furnace injection, burner design does not affect SO2 capture. This is as
would be expected. While the burner design may dominate near-field gas flows
and temperatures, these factors are mitigated downstream of the burners such
as in the upper furnace near the furnace arch. Burner design only appeared
to affect SO2 capture when sorbent was injected through the nozzles in the
burner exit. Highest capture using that injection location was achieved with
the burner producing the shortest, widest flame, and the highest N0X--
emissions--the DRB equipped with the 75° impeller. This DRB configuration
differed from the HNR and XCL burners by not having an Air Separation Plate
and by using an impeller to disperse the coal instead of a diffuser or Flame
Stabilizing Ring. While it might be expected that the slower mixing, lower
N0X flames from the HNR and SCL burner would yield higher SO2 capture, the
high-velocity sorbent jets may have a better chance to bypass the main flame
zone of the shorter DRB flame.
5-7!
-------
8 ft. above burner Q. (T«2500°F)
19 ft. above burner Q. (T«2150°F)
1 2 3
Ca/S Molar Ratio
O ORB
~ HNR
'A XCL
"? 30
Ca/S Molar Ratio
Figure 5-4. Summary of SO2 capture with Col ton hydrated lime injecti
on.
-------
Colton Hydrated Lime Vicron 45-3
• ;
; O DRB
-('~! HNR
I Aj XCL
CM
O
00
1
4
0
2
3
O-
I
CVI
o
10-
0
1
2
3
4
Ca/S Molar Ratio j Ca/S Molar Ratio
Figure 5-5. Effect of sorbent type on S02 capture with\injection at 2500°F.
-------
6.0 CONCLUSIONS
Three B&W Second Generation Low-NOx burner designs were evaluated for
applicability to the EPA LIMB demonstration site, Edgewater Unit 4. This
evaluation consisted of testing three full-size, 78 x 10® Btu/hr capacity
burners in the EPA LWS to optimize their performance and meet the Edgewater
boiler requirements. The three basic designs were the Dual Register burner,
Babcock-Hitachi N0X Reducing burner, and the XCL burner. Twenty different
configurations of these three basic designs were tested firing Pittsburgh #8
coal, the coal to be used during the LIMB demonstration. The different
configurations represent trials of various burner hardware components to
optimize the burner geometry for low-emission, high-efficiency, and
acceptable flame length. From the screening of the 20 configurations, the
optimum burner, an XCL configuration, was characterized with 3 distinctly
different coals to broaden the range of application. In addition, a brief
series of sorbent injection tests were conduced for a selected configuration
of each basic burner design to determine the effect of burner design on SOg
capture potential. ¦.
6.1 Burner Performance and N0X Emissions
The three basic burner designs, the DRB, HNR, and XCL burners, represent
an evolution of development. Each incorporates the common design features of
a central, cylindrical coal nozzle surrounded by two concentric annular
secondary combustion air passages. The optimization tests screened available
burner adjustments as well as the various burner component configurations for
each design. The three basic components of each burner, the coal injector,
inner secondary air zone, and outer secondary air zone, were evaluated in
these screening tests. The results from these tests can be easily
generalized for all three low-N0x burners with respect to importance to
performance. In each case, the coal injector was the dominant factor that
determined the key performance characteristics of N0X, flame length, and
carbon burnout. Both the design of the coal injector and the available
adjustments, such as Impeller position, could produce up to about 67 percent
reduction in N0X. The outer secondary air zone, the degree of swirl
f 6-1
-------
generated, and the air flow rate through the outer passage, was second in
importance to burner performance. The inner air zone factors of swirl and
air flow rate generally had the least affect on burner performance.
Consistent and recurring throughout the screening tests of all three
burners was the close correlation of NQX emissions with flame length--low-NOx
emissions were achieved with long flames while short flames were, associated
with high N0X. Given the physical constraints of a practical application,
such as the Edgewater boiler, the minimum level of N0X emissions achievable
will be limited by the furnace depth (about 22 ft at Edgewater Unit 4).
Specific results for the three burner designs are summarized below.
Dual Register Burrier. The Low Velocity ORB designed to fit within the
same exit as the other two candidate burners, the HNR and XCL, produced
excessively long flames for three of the four configurations. Only a non-
optimized 75° impeller design produced a flame less than the 22 ft furnace
firing depth. N0X emissions for the three configurations which produced
flames over 22 ft long were low, ranging from 264 ppm to 386 ppm. The 75°
impeller-equipped configuration produced 708 ppm with an 18 ft long flame.
Available data from a B&W sponsored test program suggest that the performance
of this Low Velocity DRB is not representative of the current standard Phase
V DRB. Flames less than 22 ft long could be achieved by the Phase V DRB with
its slightly higher velocities, albeit with slightly higher N0X emissions
(292-372 ppm). To achieve that performance, however, required tightly closed
burner settings which produced burner pressure drops over 6 in. W.G.
Babcock-Hitachi NR Burner. The Babcock-Hitachi NR burner relies on
biasing the secondary combustion air to the outer zone coupled with a very
high degree of swirl for flame shaping and N0X control. Minimum W0X
emissions were 222 ppm with a flame over 22 ft long using burner settings
typical of Babcock-Hitachi practice. The two other configurations evaluated
produced higher N0X emissions, 289-348 ppm, but with correspondingly shorter
flames, 18-20 ft long. In each case, however, burner pressure drop was about
7 in. W.G.
6-2
-------
XCL Burner. The XCL burner, which represents the latest development in
the B&W Dual Register Burner evaluation, was tested in 13 configurations
during 2 series of tests. This burner design demonstrated the most potential
to meet the LIMB demonstration because of its inherent flexibi1ity. N0X
emissions ranged from 194-700 ppm with flames from 12 to over 22 ft long.
Only 5 configurations yielded flames less than 22 ft long, with N0X emissions
from 288 to 546 ppm. The unique B&W DeN0x stabilizer achieved the lowest
emissions but required burner settings producing a burner pressure drop of
8.20 in. W.G. The other configurations were based on either a 20° or 30°
coal impeller design. The impeller equipped XCL configurations could achieve
a wide range of N0X and flame length by the adjustment of the impeller
position, all with burner pressure drop less than 5 in. W.G. At optimum
conditions, the 20° impeller in an expanded coal nozzle gave 338 ppm N0X
while the 30° impeller in the standard coal nozzle gave 374 ppm.
From these numerous burner configurations, two stand out as suitable for
application for the LIMB demonstration. All configurations tested met the
requirements of a firing capacity of 78 x 10® Btu/hr per burner, a throat
diameter.no greater than 35 inches, and mechanical reliability meeting
commercial standards. The Edgewater boiler also imposed the constraint on
flame length, 22 ft, and on maximum tolerable burner pressure drop, about 5
in. W.G. In addition, the burners had to produce a stable flame with low
emissions but high combustion efficiency. The two configurations meeting all
those conditions were:
• XCL burner with 30° impeller in the standard coal nozzle with
appropriate outer vane design.
• XCL burner with 20° impeller in an expanded coal nozzle.
Performance of these two configurations is summarized in Table 6-1. In
addition to,meeting all Edgewater boiler requirements the two impeller-
equipped XCL burner configurations offer a very effective handle to optimize
performance to suit the application. This control is. the adjustable position
6-3
-------
TABLE 6-1. OPTIMUM BURNER CONFIGURATIONS FOR EDGEWATER
UNIT 4 {78 x 106 Btu/hr SRT = 1.20)
1
1 i
t
Burner
NOx
@ OS O2 '
(ppm)
Flame
Length
(ft)
Fly Ash
Carbon
{wt %)
Burner 4P
(in. W.G.)
XCL w/30°Impeller,
Standard Coal
Nozzle
XCL w/20° Impeller,
Expanded Coal
Nozzle
374 ¦
338
20-22
21
4.42
4.92
3.30
4.90
'"6-4;
-------
of the coal impeller. For both designs, flame length and N0X emissions can
be varied simply by moving the impeller a matter of inches. The impeller
adjustment can thus be used to tune the burner for maximum N0X reduction
within the constraints of available firing depth.
In application to the Edgewater Unit #4 boiler, N0X emissions are
expected to be somewhat higher than those recorded in the LWS furnace. This
is because of the multiple burner configuration and the resulting more
confined, hotter thermal environment of the boiler unit. A parameter often
used to describe firing density, and to correlate N0X emissions in different
boilers is the burner zone heat liberation, which relates thermal input at
the burners to the area of cooled wall surface in the burner zone. This
parameter takes value of 175 x 10^ Btu/hr.ft^ for the LWS furnace, and 245 x
10^ Btu/hr.ft^ for the Edgewater boiler. This difference is expected to
yield an increase of approximately 60 ppm in N0X emissions from the boiler
for comparable burner operating conditions in the LWS. Optimum burner
configurations referenced in Table 6-1 are therefore expected to give
approximately 435 ppm and 400 ppm N0X, respectively, in the Edgewater boiler.
6.2 S02 Reduction Potential
A brief series of sorbent injection tests was performed for a selected
configuration of each burner design, DRB, HNR, and XCL burners. Two near
burner locations and two upper furnace locations were evaluated at nominal
full load conditions. Vicron 45-3 limestone was injected through the three
locations closest the burner and Col ton hydrated lime was injected through
the two upper furnace locations. At the lower level of ports, with a gas
temperature of about 2500°F, limestone achieved 33 percent capture at Ca/S
molar ratio of 2 while the hydrated lime achieved 36 percent capture. At the
uppermost level » associated with a gas temperature of 2150°F, the hydrated
lime achieved 38 percent capture at Ca/S molar ratio of 2.
The two near burner locations considered were with the coal and through
four high-velocity nozzles in the outer secondary air passage. Limestone
injected with the coal achieved only 28-32 percent capture at Ca/S molar
6-5
-------
ratio of 2 for all three burners. The injection of limestone through the
outer secondary air passage yielded higher SOg capture for the DRB equipped
with the 75° impeller (40 percent at Ca/S molar ratio of 2) than for the HNR
and SCL burners (30 percent). The di fference in results is not fully
understood, but appears to be associated with the near burner flow field as
suggested by flame shape.
6-6
-------
I.
7.0 REFERENCES
1. Brackett, C. E. and Barsin, J. A. The Dual Register Pulverized Coal
Burner--A N0X Control Device. In: Proceedings of the N0X Control
Seminar. Electric Power Research Tnstitute, EPRI SR-39, 1976.
2. Barsin, J. A. Pulverized Coal Firing N0X Control. In: Proceedings of
the EPRI N0X Control Symposium, 1978.
3. Campobenedetto, E. J. The Dual Register Burner--Field Test Results.
Presented at Engineering Foundation Conference on Clean Combustion of
Coal, Franklin Pierce College, Rindge, New Hampshire, August 1977.
4. Barsin, J. A. Fossil Steam Generator N0X Control Update. In:
Proceeding of the 1982 Joint Symposium on Stationary Combustion 1UX
Control, Vol. I, EPA-600/9-85-022a {NTIS PB85-235604), July 1985.
5. Pohl, J. H., Chen, S. L., Heap, M. P., and Pershing, D. W. Correlation
of N0X Emissions with Basic Physical and Chemical Characteristics of
Coal. In: Proceedings of the 1982 Joint Symposium on Stationary
Combustion N0X Control, Vol. II, EPA-600/9-85-022b (NTIS PB85-235612),
July 1985.
------- |