Urtjted States Off ce of Air Quality
£nviroomentall Protect ori P1anrfn_g and Standards
Agency ^ Research T* aftgle ParJc ^C
Air . 'f r '' .
Municipal Was
Combustors-
Background
hif prmation lor
Proposed Standards
Control of
NOx Emissions
EPA 45CT3 8» 27d
August
_A* %_
TO-,
-------
Page Intentionally Blank
-------
MUNICIPAL WASTE COMBUSTORS --
BACKGROUND INFORMATION FOR
PROPOSED STANDARDS: CONTROL OF NOx EMISSIONS
FINAL REPORT
Prepared for:
Michael G. Johnston
U.S. Environmental Protection Agency
Industrial Studies Branch (MD-13)
Research Triangle Park, North Carolina 27711
Prepared by:
Radian Corporation
3200 E. Chapel Hill Rd./Nelson Hwy.
Post Office Box 13000
August 14, 1989
I -k
*"=""~"'~~"«i I
MAR - 7 1990
-------
DISCLAIMER
This report has been reviewed by the Emission Standards Division
of the Office of Air Quality Planning and Standards/ EPA, and
approved for publication. Mention of trade names or commercial
products is not intended to constitute endorsement or
recommendation for use. Copies of this report are available
through the Library Services Office (MD-35), U.S. Environmental
Protection Agency, Research Triangle Park NC 27711, or from
National Technical Information Services, 5285 Port Royal Road,
Springfield VA 22161.
11
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION .... 1-1
2.0 NOX EMISSIONS ............ 2-1
2.1 NOV EMISSIONS FROM MWC'S WITHOUT ADD-ON
NOX CONTROLS . . 2-1
2.2 FACTORS AFFECTING NOY EMISSIONS ..... 2-7
«
2.3 RELATIONSHIP BETWEEN NOX AND OTHER FLUE GAS EMISSIONS . . 2-10
2.4 REFERENCES . 2-17
3.0 NO EMISSION CONTROLS . 3-1
A
3.1 COMBUSTION CONTROLS ..... 3-1
3.1.1 Low Excess Air and Staged Combustion ....... 3-1
3.1.2 Flue Gas Recirculation 3-2
3.2 GAS REBURNING . . 3-4
3.3 SELECTIVE CATALYTIC REDUCTION . . 3-6
3.4 SELECTIVE NON-CATALYTIC REDUCTION . . 3-9
3.5 SUMMARY OF NOY EMISSION CONTROLS 3-23
" ' '
3.6 REFERENCES. . 3-24
4.0 COST PROCEDURES 4-1
4.1 CAPITAL COST PROCEDURE 4-1
4.1.1 Direct Capital Cost 4-3
4.1.2 Indirect Costs 4-10
4.1.3 Other Costs 4-12
4.2 OPERATING COST PROCEDURE 4-12
4.2.1 Labor and Maintenance 4-12
4.2.2 Electricity. . . . 4-15
4.2.3 Ammonia Consumption 4-18
4.2.4 Other Costs. . . 4-20
4.3 REFERENCES. 4-21
iii
-------
TABLE OF CONTENTS (CONTINUED)
Section
Else
5.0 MODEL PLANT COSTS FOR THERMAL DENOX .....'......,/ 5-1
5.1 MASS BURN/WATERWALL . ... ."/.'., ...... . . , . 5-1
5.2 MASS BURN/REFRACTORY 5-6
5.3 MASS BURN/ROTARY COMBUSTOR. ............... 5-10
5.4 REFUSE-DERIVED FUEL .................... 5.10
5.5 MODULAR COMBUSTORS 5-i§
5.6 FLUIDIZED-BED COMBUSTION. .......... . . . . . . 5.18
5.7 SUMMARY OF m. EMISSION REDUCTION, COST EFFECTIVENESS
AND ELECTRICAL REQUIREMENTS . .... . . . . . . , . . 5.18
5.8 COST SENSITIVITY ANALYSIS . . . . ... . . ....... 5-22
5.8.1 Ammonia Price Variation. ........... 5-22
5.8.2 Electricity Price Variation. ......
5.8.3 NOX Reduction Variation. . ... ...... [\ ', §-24
5.9 REFERENCES. ............ 5_28
IV
-------
LIST OF TABLES i
c
Table Page
2-1 Average N0y Emissions from MWC'S 2-2
2-2 Summary of NOX Emissions Data from MWC'S ......... 2-5
2-3 NOX Variations with Combustor Load 2-9
2-4 Multivariate Analysis of NO Emissions as a Function
of Load, Excess Air, and OOerfire Air Distribution . ... 2-11
3-1 Marion County Emissions Versus Air Supply . .". ..... 3-3
3-2 Results of Testing of SCR System at
Iwatsuki, Japan 3-7
3-3 Results of Testing of SCR System at
Tokyo-Hikarigaoka, Japan. ....... 3-8
3-4 Existing Thermal DeNOx Facilities in United States . . . . 3-11
3-5 Summary of NOV Reduction at Commerce Optimization
Test. ...?....... ... .... 3-13
3^6 Outlet Mercury Emissions Measured from Spray Dryer/
Fabric Filter Systems . . . ... . . ... . . . . . . . 3-17
4-1 Procedures for Estimating Capital Costs for
Thermal DeNOY Applied to New MWC Plants . . . . .. . ... 4-2
i" ,
4-2 Capital Costs Data for Thermal DeNOv Applied
to Mass Burn/Waterwall MWC'S. ............. . 4-4
4-3 Direct and Indirect Capital Costs Data for
Thermal DeNO . ... 4-5
A
4-4 Capital Costs for Thermal DeNO for Two Combustors
at 250 tpd Each (December 1987 dollars) . .... . . . . 4-6
4-5 Cost Analysis Results Using Detailed Costs from
Exxon and Ogden Martin. . . . . . . . . . . ....... 4-9
4-6 Comparison of Actual and Predicted License Fees. . .... 4-13
4-7 Procedures for Estimating Annual Operating Costs
for Thermal DeNOY Applied to New MWC Plants ....... 4-14
-------
LIST OF TABLES (CONTINUED)
4-8 Electrical Power and Ammonia Consumed by Thermal «'
DeNO₯ for Selected MWC Plants . 4-16
x . ,
4-9 Comparison of Actual and Predicted Electrical '
Power Consumed by Thermal DeNOv for Selected
MWC Plants * 4-17
5-1 Model Plant Selection for lll(b) . 5-2 <
<
5-2 Model Plant Specifications and Flue Gas Composition Data . 5-3 -
5-3 Capital Costs for the Mass Burn/Waterwal1 Model Plants- <
Nos. 1 to 3 ($l,000's in December 1987) 5-5 <
5-4 Annualized Costs, Economic and Environmental Impacts for
the Mass Burn/Waterwall Model Plants - Nos. 1 to 3
($l,000's in December 1987) 5-7
5-5 Capital Costs for the Mass Burn/Refractory Model Plant -
No. 4 ($l,000's in December 1987) 5-8
5-6 Annualized Costs, Economic and Environmental Impacts for ?
the Mass Burn/Refractory Model Plant - No. 4 ($l,000's }
in December 1987) 5-9
5-7 Capital Costs for the Mass Burn/Rotary Combustor Model
Plant - No. 5 ($l,000's in December 1987) .... 5-11 *
5-8 Annualized Costs, Economic and Environmental Impacts .\
for the Mass Burn/Rotary Combustor Model Plant - No. 5 ]
($l,000's in December 1987) . 5-12
5-9 Capital Costs for the Refuse-Derived Fuel Fired Model
Plants - Nos. 6 and 7 ($l,000's in December 1987) ;. . . 5-13 ;
'
-------
LIST OF TABLES (CONTINUED)
J_able Page
5-12 Annualized Costs, Economic and Environmental Impacts for
the Modular Model Plants - Nos. 8 to 10 ($l,000's in
December 1987) 5-17
5-13 Capital Costs for the Fluidized Bed Combustion Model
Plants - Nos. 11 and 12 ($l,000's in December 1987) . . . 5-19
5-14 Annualized Costs, Economic and Environmental Impacts for
the Fluidized Bed Combustion Model Plants - Nos. 11 and
12 ($.l,000's in December 1987) . 5-20
5-15 Summary of Costs, Cost Effectiveness, and Electrical
Requirements for New MWC Model Plants Using
Thermal DeNOx 5-21
5-16 Impacts of Varying Ammonia Price ($/Ton) on Thermal DeNO
Annual ized Cost Effectiveness 5~23
5-17 Impacts of Varying Electricity Price ($/kWh) on Thermal
DeNOx Annualized Cost and Cost Effectiveness. 5-25
5-18 Thermal DeNO Annualized Costs and Cost Effectiveness
at 45 and 60 Percent NOX Reduction. ....".. 5-26
Vll
-------
LIST OF FIGURES
Figure . Page
' '.-'.-'. ' ' ''
2-1 Seasonal variations in NOY emissions ... . ... ... . 2-8
A ' "
2-2 NO versus CO emissions for a mass burn
combustor ................ ........ 2-12
2-3 Long-term NO versus CO emissions from the
Millbury MWC. . . . . . . . . . . ........ . . . . 2-14
'.-.. '
2-4 Long-term NO versus 0« emissions from the
Mill bury MWC. . . . . . . . . . ..... ........ 2-15
' : '
2-5 NOV emissions versus COD/CDF emissions ........... 2-16
x . . . ..
3-1 Gas reburning for NO control at a mass burn
combustor .... ..... ............... 3-5
' ''
3-2 Effect of inlet PM on mercury removal efficiency
for MWC' s with spray dryer/fabric filter systems. . . . . 3-19
3-3 Effect of inlet PM on mercury emissions for MWC' s
with spray dryer/fabric filter systems. . . . ...... 3-20
, . .. -
3-4 Effect of inlet CDD/CDF on mercury removal efficiency
for MWC' s with spray dryer/fabric filter systems. . ... 3-21
3-5 Effect of inlet CDD/CDF on mercury emissions for MWC's
with spray dryer/fabric filter systems. . . ...... ... . 3-22
4-1 Comparison of the direct capital cost
equation with the cost data . ....... ....... 4-11
4-2 Comparison of the NH, injection equation
with the contractor/vendor data . . .......... . 4-19
viii
-------
1,0 INTRODUCTION
Emissions of nitrogen oxides (NOY) from municipal waste'combustors.
,A. ' ,
(HMC's) are generally not controlled before being released to the atmosphere.
Methods of control, both through combustion modifications and add-on controls,
are available but have been infrequently applied to MWC's. This report
characterizes NOV emissions from MWC's and assesses the performance and costs
X '
associated with controlling NO emissions.
"
In Section 2.0 of this report, available data on NO emissions from
- ' *»
MWC's without add-on controls are summarized. Some of the NO emissions data
3\
may reflect combustion modifications normally used during MVIC operation. The
various control technologies for reducing NO emissions are reviewed in
'» . - -
Section 3.0. The available performance data and operational experience for
the different NO controls for MWC's are also presented.
A " . "
In Section 4.0, cost algorithms are developed for Thermal DeNOv, one of
, . /\ ,
the add-on control technologies that has been applied to several new MWC's.
A cursory cost analysis for selective catalytic reduction (SCR) is also
presented. In Section 5.0, the cost algorithms for Thermal DeNO are used
to estimate annualized NOV control costs and cost-effectiveness values for
j\ .
12 model plants representative of new MWC's. The sensitivity of Thermal
DeNO annualized costs and cost effectiveness to variations in ammonia and
electricity costs is also investigated.
1-1
-------
-------
2.0 NOY EMISSIONS
A . - -
Nitrogen oxides are formed during combustion through: (1) oxidation of
fuel-bound nitrogen and (2) fixation of atmospheric nitrogen. Conversion of
fuel-bound nitrogen occurs at relatively low temperatures (<2,000°F), while
fixation of atmospheric nitrogen generally occurs at higher temperatures.
Host (75 to 80 percent) of the NOV formed during normal operation of MWC's is
2
associated with fuel-bound nitrogen.
2,1 NOY EMISSIONS FROM MWC'S WITHOUT ADD-ON NOY CONTROLS
A - . ' " A ..-'". ' ' '
The available data on NOY emissions from MWC's without add-on NO
A . J\
controls are listed in Table 2-1 by combustor type (NO emissions following
A
add-on controls are presented in Section 3.0). The data are from test
reports and responses to an EPA survey of MWC facilities. The data cover
52 MWC units (8 mass burn/refractory, 26 mass burn/waterwall, 5 refuse-
derived fuel [RDF], 8 excess-air modular, and 5 starved-air modular) located
at 35 different plants. Each data point represents the average of the NOY
A .'
test runs at the stated unit. Most of these tests were conducted during MWC
compliance testing while the combustor was at full load and at normal
operating conditions. Each test usually lasted from 1 to 3 hours and both
manual (EPA Method 7A) and continuous emission monitoring (CEM) (EPA
Method 7E) methods were used to measure NOV emissions. Table 2-2 summarizes
« - - '
these data. Although none of these units were using add-on NOV controls at
- ' A . '
the time they were tested, several of them used combustion controls to reduce
NOV formation in the combustor.
A -- . ... '.'"'
With one exception, NOX emissions from these facilities ranged from 59
to 375 ppm at 7 percent Og. The remaining unit had emissions of 611 ppm.
The average NOY concentration for all 52 data sets is 211 ppm. On a pound
A - - -'''"-
per million Btu (Ib/MMBtu) basis, this concentration is slightly less than
0.4 Ib/MMBtu. For mass burn/refractory units, the average NO concentration
is 155 ppm and ranges from 59 to 239 ppm. The NO concentration from mass
: ' . A
burn/waterwall units averages 242 ppm and ranges from 68 to 372 ppm. The
68 ppm value was obtained at Long Beach, which uses flue gas recirculation to
reduce NO emissions, and was not included in the average. The remaining
.' - A - ' ' -'.''-'
data were above 154 ppm. For RDF combustors, the average NO concentration
A . - . .
2-1
-------
TABLE 2-1. AVERAGE NO EMISSIONS FROM MWC's
Unit Size
Site3 (tons/day)
Mass Burn/Refractory
McKay Bay 2
Dayton 2
McKay Bay 3
Gal ax
Philadelphia NW 1
Philadelphia NW 2
McKay Bay 4
Dayton 1
Mass Burn/Rotary .Waterwall
Gallatin
Kure
Mass Burn/Waterwal 1
Long Beach (DeNOx off)c
Commerce (DeNOx off)
Baltimore 3
Baltimore 2
Alexandria
Claremont 2
Peekskill
Hampton 1
Nashville Thermal
Baltimore 1
Millbury 2
Mi 11 bury 1
".'
250
300
250
56
375
375
250
300
100
165
460
300
750
750
325
100
750
100
360
750
750
750
Test
Date
09/85
NRb
09/85
NR
02/87
02/87
09/85
NR
02/83
11/80
11/88
06/87
01/85
01/85
12/87
05/87
04/85
06/88
NR
01/85
02/88
02/88
°2
W.
11.8
14.3
.11.6
13.9
13.9
14.8
13.3
14.8
9.1
12.0
10.2
10.0
11.1
12.1
9.4
11.4
NR
11.0
10.6
12.0
10.5
10.3
NO
(ppft)
39.0
33.9
100.4
81.1
86.0
84,3
106; 5
104.8
*-..
124.2
_ -
105.6
52.4
121.0
136.3
122.3
171.3
144.9
NR
156.3
164.0
141.8
169.3
177.5
NOX
(ppm at
7% 02)
59.4
71.4
152.1
160.9
171.1
192.0
216.4
238; 8
146.1
164.9
68.2
154.3
193.7
193.9
207 .8
210.2
218.3
219.2
221.4
222.0
225.7
233.7
Ref.
3
4
3
5
6
6
3
4
7
8
9
10
11
11
12
13
14
15
16
11
17
17
(continued)
2-2
-------
TABLE 2-1 (CONTINUED). AVERAGE NOX EMISSIONS FROM MWC's
Site3
Mass Burn/Waterwall
Peeks kill
Hampton 2
Marion County 2
Claremont 1
Wurzburg
Marion County 2
Pine! las County
Stanislaus 1
(DeNOx off)
Stani si aus 2
(DeNOx off)
Quebec City
Tulsa 1
Tulsa 2
RDF
Mid-Connecticut 11
Biddeford
Niagara Falls
Albany
Lawrence
Modular, Excess-Air
Pigeon Point 2C
North Aroostook
Pigeon Point 3C
Pigeon Point 4C
Pigeon Point lc
Unit Size
(tons/day)
(cont.)
750
100
275
100
330
275
1,000
400
400
250
375
375
675
350
1,000
300
1,000
120
50
120
120
120
Test
Date
11/85
06/88
06/87
05/87
12/85
09/86
02/87
12/88
12/88
03/85
06/86
06/86
07/88
12/87
05/85
06/84
09/87
01/88
NR
01/88
01/88
01/88
(?)
11.7
9.5
9.6
12.2
NR
10.6
9.2
NR
NR
11.6
9.2
8.6
9.9
8.3
NR
NR
12.0
11.7
9.9
11.3
11.2
11.2
NO
156.7
194.7
196.9
161.0
NR
211.8
240.0
NR
NR
205.4
308.5
328.2
153.4
206.5
NR
NR
221.2
69.8
89.7
78.5
81.3
87.7
NOX
(ppm at
7% 02)
236.3
238.6
244.3
258.8
260.7
284.9
285.7
297. Od
304. Od
314.0
367.7
372.2
194.6
228.0
267.9
293.0
345.3
104.8
111.9
114.0
116.9
125.5
Ref.
18
15
19
13
20
21
22
23
23
24
25
25
26
27
28
29
30
31
32
31
31
31
(continued)
2-3
-------
TABLE 2-1 (CONCLUDED). AVERAGE N0₯ EMISSIONS FROM MWC's
Unit Size
Site3 (tons/day)
Modular, Excess -Air (cont.
Pittsfieldc
Pittsfield0
Pope/Dougl as
Modular, Starved-Air
One i da
Tuscaloosa
Red Wing
Prince Edward Island
Cattaraugus
)
120
120
100
50
75
90
36
38
test
Date
10/85
06/86
07/87
08/85
05/85
09/86
11/84
09/84
k
8.9
8.9
13.4
NR
11.3
12.3
11.9
NR
-------
TABLE 2-2. SUMMARY OF NOV EMISSIONS DATA FROM MWC's
-. ... -«
NO Emissions*
(ppm at 7 percent 0?)
Combustor Type
Mass Burn/Refractory
Mass Burn/Waterwal 1
RDF
Modular, Excess-Air
Modular, Starved-Air
All Types
Number of Units
8
26b,c
5
8
5
52
Average
155
240
270
140
215d
210e
Ran?
59 -
154 -
195 -
105 -
86 -
59 -
Je
240
370
345
280
280
370
Averages rounded to nearest 5 ppm.
Includes data from two mass burn/rotary waterwall combustors with N0x
emissions of 146 and 165 ppm. Without these points, the average N0x
concentration still rounds to 240 ppm.
cExcludes data from one unit with flue gas recirculation with NO emissions
of 68 ppm. With this point, the average NOX concentration stilT rounds to
240 ppm.
Excludes one atypical data point of 611 ppm. With this point included,
the average NO concentration is 295 ppm.
A
eExcludes one atypical data point of 611 ppm for a modular starved-air
facility. With this point included, the average is 220 ppm.
2-5
-------
is 266 ppm with a range of 195 to 345 ppm. For excess-air modular units, the «
NOX emissions average 138 ppm and range from 105 to 282 ppm. The data for *
excess-air modular units are heavily weighted by the data from Pigeon Point
and Pittsfield, which have Vicon units that employ flue gas recirculation 4
(FGR) (approximately 35 percent of the total air supply). This technology *
accounts for 70 percent of the total design throughput capacity of modular ,
excess-air units. The North Aroostook and Pope/Douglas combustors do not i
employ FGR. For modular starved-air facilities (including the 611 ppm ]
emission rate from Cattaraugus), the average NOX concentration is 294 ppm. . I
Excluding Cattaraugus, the average is 215 ppm with a high concentration of *
279 ppm. s
An analysis of variance of the NOX emissions data was performed to *
determine if there are any significant differences between the emissions from *
the different MWC combustor types. This 'analysis*, the Duncan Range Test, j
compares the means and ranges of the data from each combustor type and <
determines, to a 95-percent confidence level, whether the data from different '
combustor types are distinct. The analysis shows that NOV emissions from (
. ...'. 'X - ^
mass burn/waterwall, starved-air modular, and RDF combustors are similar, and ,
that NOV emissions from mass burn/refractory and excess-air modular combustors {
' ' \
are similar. However, NO emissions from mass burn/waterwall and mass (
burn/refractory combustors are also statistically similar, leaving no distinct '
differences between the two similar groups of combustors. Thus, although the J
average NO emissions for the different combustors show some variation, the <
variations are not large enough to support a conclusion that different MWC (
combustor types have different NO emission values. |
The observed variations in NOV emissions could be due to normal daily \
X ; ,
variations as well as seasonal factors. For example, continuous N0v
' ' ' ' . n ., ' \
measurements were collected between July and September 1988 as part of a test <
program at the MWC facility in Mi 11 bury, Massachusetts. Although combustor *
operation during the testing was maintained as close to normal as possible, <
these data range from less than 50 ppm to nearly 500 ppm at 7 percent 02. ;
Similarly, at the MWC in Marion County, Oregon, variations in NOY emissions \
' ' " . ' ..'. ''.'' '.-.'- ...-.;. -.. : :- ' {
of 120 ppm during a single day under normal operating conditions were ,
AO
observed. "
2-6
-------
2.2 FACTORS AFFECTING NOV EMISSIONS
A
In Figure 2-1, NO emissions are shown by month for each combustor type
A . -. " -
to show seasonal variations. For mass burn/waterwal1 combustors, NO
1 '/\
emissions are generally higher in the summer months than in the winter
months. (The 140 ppm value recorded in June was from Commerce, which burns
primarily commercial refuse). However, NO emissions between 210 and 290 ppm
A ---..
were observed for all the months with data. Insufficient data are available
for the other combustor types to determine similar trends. The observed
higher NO emissions from mass burn/waterwall units during the summer months
/t
may be due to higher nitrogen content of the fuel because the raw refuse
contains more yard wastes, which have a high nitrogen content.
Previous investigations of NOV emissions from coal-, oil-, and gas-fired
43
utility boilers have found that combustor load can affect NO emissions.
dd dR dfi
At MWC facilities in Marion County, * Peekskill, and Quebec City^° NO
«
emissions were measured during short-term tests at different combustor loads.
In addition, at Marion County and Quebec City, NO emissions were measured at
" ' - '
different excess air rates and overfire air distributions. These data are
summarized in Table 2-3.
During the Marion County tests, the NOX emissions at low load and normal
air supply (76 percent of full load, Run 6a) averaged 257 ppm at 7 percent 02
while the five tests at normal load and normal air supply (Runs 1, 2, 10,
lla, lib) averaged 286 ppm at 7 percent 02, a difference of about 10 percent.
However, the low load NO measurement is within the range of the normal load
« .
measurements (255 to 309 ppm). Comparison of low load versus normal load at
Peekskill (Runs 1.1-13 versus Runs 2-7) and Quebec City (Runs 2, 10, and 11
versus Runs 5, 6, and 12) are inconclusive, due to simultaneous changes in
load and excess air. Comparisons of the effects of high load versus normal
load at Peekskill (Runs 8-10 versus Runs 2-7) and Quebec City (Runs 7 and 9
versus Runs 5, 6, and 12) on NOX emissions failed to find any clear impact of
load on NO emissions. Based on these data, changes in load within the range
tested (70-115 percent of design) do not appear to have any significant
impact ©n NOX emissions.
2-7
-------
X
XI
< +'
o- o o o e
88 JT 8 ' * « '*' * *
.
o ' o e
-1
"*
1
I
T-hf
I
x
i
(b %L © Wdd) 8UO!88|IU3
2-8
-------
TABLE 2-3. NO VARIATIONS WITH COMBUSTOR LOAD
Site
Marion County
Marion County
Han on County
Marion County
Marion County
Marion County
Marion County
Marion County
Marion County
Marion County
Marion County
Marion County
Marion County
Marion County
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Peekskill
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Quebec City
Run
1
2
10
lla
lib
3a
3b
4
5
6a
6b
7
8
9
2
3
5
6
7
8
9
10
11
12
13
2
10
11
5
6
12
7
9
3
4
14
15
y.'.
Load (%
(of full)
100
100
100
100
100
95
95
98
103
76
71
77
74
78
100
100
100
100
100
113
112
113
87
87
87
71
71
71
100
100
100
114
114
100
99
101
101
NO
(ppm)
264
262
228
218
240
218
230
190
240
220
142
184
150
219
191
193
179
181
174
160
164
190
147
155
133
155
127
128
158
155
149
155
185
168
164
127
137
NOX
(ppm, n 02)
308
309
269
255
288
203
317
220
276
257
232
195
188
282
239
279
242
249
242
232
230
256
240
251
220
272
224
200
184
181
190
198
236
262
256
199
193
°2
.(*).
9.0
9.1
9.1
9.0
9.3
6.0
10.8
8.9
8.8
9.0
12.4
7.8
9.8
10.1
9.8
11.3
10.6
10.8
10.9
11.3
11.0
10.6
12.4
12.3
12.5
13
13
12
9
9
10
10
10
12
12
12
11
Comments3
LEA
HEA
LOA
HOA
HEA
LEA
LOA
HOA
HEA
HEA
LOA
LOA
Tests where air supply was purposely varied are noted.
HEA = high excess air; LEA = low excess air; HOA = high overfife air;
LOA = low overfire air. Other tests may have shown similar variation
(i.e., similar 0« levels), but these tests were not designed around air
supply changes.
2-9
-------
Tests to evaluate the impact of high excess air (HEA) during normal load (
operation at Marion County (Run 3b) and Quebec City (Runs 3 and 4) suggest ]
that HEA increases NOX emissions. However, during low load tests at Marion
County, NOX emissions were lower with HEA (Run 6b) than with normal air J
supply (Run 6a). Emissions of NOY during tests at Marion County with low *
: . X '..' ' ' ,
excess air (LEA) at normal load (Run 3a) and low load (Run 7) were both lower j
than tests at normal air supply and corresponding loads. Tests at Marion \
County (Runs 4, 5, 8, and 9) and Quebec City (Runs 14 and 15) during which I
the distribution of air above and under the grate was varied suggests that \
low overfire air (LOA) reduces NO emissions. The impact of high overfire }
n ' ". ' ' .,"''. i
air (HOA) on NO emissions, however, appears small. Further discussion of {
-' f
the use of LEA and overfire air distribution as NOX control techniques is j>
presented in Section 3.1. ;,
A multivariate analysis of the effects of load, excess air, and overfire (
air distribution on NO emissions was performed with the data from Marion *,
- " " ..-'.' t
County and Quebec City. The results are summarized in Table 2-4. No single ' ' ]
variable yields a significant correlation. Stronger correlations result as J
each additional variable is included in the analysis, suggesting that NO «
. ' - ; '' . A '-
emissions are dependent on all three variables. However, the final corre-
lation coefficients are not high, suggesting that other parameters such as .
fuel composition or heating value also affect NOYemissions. ]
x ...-. {
2.3 RELATIONSHIP BETWEEN NOX AND OTHER FLUE, GAS EMISSIONS ' i
It is generally thought that NO emissions increase as combustion i
efficiency increases. This implies that an inverse relationship between NOX f
and CO emissions should exist. The available NOX and CO emission data from |
two facilities were used to investigate this relationship. The relationships <
between NO and 09 emissions and between N0y and COD/CDF emissions were also ;
J\ tm .« ' |
investigated. 1
Figure 2-2 presents NO and CO emissions data measured at the Olmsted
-------
TABLE 2-4. MULTIVARIATE ANALYSIS OF NOV EMISSIONS AS A FUNCTION OF LOAD,
EXCESS AIR, AND OVERFIRE AIR DISTRIBUTION
Test
Correlation Coefficient
Marion County
Quebec City
NO vs. load
#*
0.2631
0.0666
NOV vs. excess air
«
0.0328
0.4259
NOX vs. overfire air distribution
0.2295
0.0846
NO vs. load, excess air
0.4579
0.5209
NOX vs, load, excess air,
overfire air distribution
0.6157
0.7296
2-11
-------
300
225
c o
o *-
°W -D
.2 £ 150
O
75
30
r
60
90
I
120
150
CO Emissions, ppm
corrected to 7% 02
!
Figure 2-2. NOX versus CO for a Mass Burn Combustor.
2-12 : - \ . . .. ''
-------
conditions (zero excess air). Taken as a whole, these data support the
existence of an inverse relationship between NO and CO, with NOV emissions
/\ . XV
increasing with decreasing CO emissions. At CO levels below 60 ppm, however,
there is no apparent trend in the NO measurements.
/\. '
Figure 2-3 presents 1,330 1-hour average CEM measurements of NO and CO
collected at the Millbury MWC between July 15 and September 15, 1988. The
average NOX value is 223 ppm at 7 percent Op. Eighty-five percent of the
measurements are between 175 ppm and 275 ppm. Ninety-nine percent of the
measurements are less than 360 ppm. Within the measured range of CO
emissions (25-60 ppm), no trend in NO emissions occurs. These results are
" '.."
consistent with the data from Olmsted County in Figure 2-2. The NOX and 0,,
data from Hi 11 bury are plotted in Figure 2-4. Most of the 02 values are
between 8 and 13 percent. As with the NO and CO measurements, there is no
A
apparent relationship between NO and 09.
-. A Cm
Two of the facilities with above average NO concentrations (Pinellas
County and Marion County, both of which have Martin combustors) have reported
very low CDD/CDF concentrations. This suggests that the combustion
conditions associated with CDD/CDF destruction may contribute to NOY
«
formation. A plot of NO emissions versus CDD/CDF emissions for eight
different MWC plants is shown in Figure 2-5. Examining all of the data as a
set as well as the data from each individual plant, NOV emissions do not vary
'. .. - ' f^
significantly as the CDD/CDF concentration changes. For CDD/CDF
concentrations of 30 to 1,200 ng/dscm, NOY emissions are consistently between
A
200 and 330 ppm.
2-13
-------
D
D
D
D
a
D
a
n
a
a
n
cP
_ o
_ .o
CM
O
&'
£
Q.
a.
O
O
CM
O
O
O
CO
O
CD
m
o
^t
n
o
CN
o
O
o
00
CM
O
UD
CJ
o
CM
O
CM
CM
O
O
CM
O
co
O
co
O O
CM O
2-14
-------
en
CM
O
E
a.
a
X
O
z
440
420
400 -
380 -
360
340..-
320 -
300
280 -
260
240 -
220 '-
200
180 -
160 -
1 40,
120 H
100
D
D
a
D
n n
a
a
D
man
D
a
n
a
D
a
n n
DP n
r-j Q
D
D- ^ %
D on D g D
D
a a
14
16
. !nLet..02-
Figure 2-4. Long-term MOXversus O2emisslons for the MiHbury MWC
-------
400 -
350 -
^s 300 -
O*
. *
O 4>
' .- 0- "' ' ' ' '' ;
r i i i I i i 11 i i i i
) 200 400 600 800 1000 1200
Inlet CDD/COF Concentration (ng/dscm @7% Oa)
Bidkfeford (RDF| A ^anon
(RDF) °
X PinoiSas County (MB/WWJ
Prince Edward istend
Figure 2-5, NOX emissions versus
emissions
-------
I 2.4 REFERENCES ' .
1. California Air Resources Board. Air Pollution Control at Resource
Recovery Facilities. Sacramento, CA. May 24, 1984. p. 79.
> 2. Reference 1, p. 75.
"V ' . '
s 3. Clean Air Engineering, Inc. Report on Compliance Testing for Waste
5 Management, Inc. at the McKay Bay Refuse-to-Energy Project located in
Tampa, FL. October 1985.
-! ,
.4. Letter and enclosures from John W. Norton, County of Montgomery, OH, to
c Jack R. Farmer, U. S. Environmental Protection Agency, Research; Triangle
^ Park, NC. May 31, 1988.
s
; 5. Letter anil enclosures from W. Harold Snead, City of Gal ax, VA, to
'' Jack R. Farmer, U. S. Environmental Protection Agency, Research Triangle
Park, NC. July 14, 1988.
7 - . '. - -
6. Neulicht, R. (Midwest Research Institute). Emissions Test Report: City
' of Philadelphia Northwest and East Central Municipal Incinerators.
Prepared for the U. S. Environmental Protection Agency, Philadelphia, PA.
* October 31, 1985.
s - - -
:s 7. Cooper Engineers, Inc. Air Emissions Tests of Solid Waste Combustion in
^ a Rotary Combustion/Boiler System at Gal Tatin, TN. Prepared for West
t County Agency of Contra Costa County, CA. July 1984.
( 8. Cooper and Clark Consulting Engineers. Air Emissions Tests of Solid
i- Waste Combustion in a Rotary Combustion/Boiler System at Kure, Japan.
? Prepared for West County Agency of Contra Costa County, CA. June 1981.
k ' '
[" 9. Ethier, D. D., L. N. Hottenstein, and E. A. Pearson (TRC Environmental
J Consultants). Air Emission Test Results at the Southeast Resource
( Recovery Facility Unit 1 (Long Beach), October to December 1988.
t Prepared for Dravo Corporation. Long Beach, CA. February 28, 1989.
C pp. 13, 14.
t 1-0. McDannel, M. D., L, A. Green, and B. L. McDonald (Energy Systems
Associates) Air Emissions Tests at Commerce Refuse-to-Energy Facility.
May 26 - June 5, 1987. Prepared for County Sanitation District of Los
Angeles County. Whittier, CA. July 1987.
11. Entropy Environmentalists, Inc. Baltimore RESCO Company, L. P.,
Southwest Resource Recovery Facility. Particulate, Sulfur Dioxide,
Nitrogen Oxides, Chlorides, Fluorides, and Carbon Monoxide Compliance
^ Testing, Units 1, 2, and 3. Prepared for RUST International, Inc.
,t January 1985.
2-17
-------
12. Zurlinden, R. A., et. al., (Ogden Projects, Inc.). Environmental Test i
Report, Alexandria/Arlington Resource Recovery Facility, Units 1, 2, «!
and 3. Prepared for Ogden Martin Systems of Alexandria/Arlington, Inc. i
Alexandria, VA. Report No. 144 A (Revised). January 1988. {
13. Almega Corporation. SES Claremont, Claremont, NH, NH/VT Solid Waste \
Facility, Unit 1 and Unit 2: EPA Stack Emission Compliance Tests, I
May 26, 27, and 29, 1987. (Prepared for Clark Kenith, Inc.). Atlanta, J
GA. July 1987. }
14. Fossa, A. J., et. al. Phase I Resource Recovery Facility Emission 1
Characterization Study, Overview Report. New York State Department of j
Environmental Conservation. Albany, NY. May 1987. I
15. McDonald, B» L., M. D. McDannel and L. A. Green (Energy Systems 5
Associates), Air Emissions Tests at the Hampton Refuse-Fired Steam
Generating Facility, April 18-24, 1988. Prepared for Clark-Kenith,
Incorporated. Bethesda, MD. June 1988.
16. Letter and enclosures from J. T. Nestle, Jr., Nashville Thermal Transfer
Corporation, to Jack R- Farmer, U.S. Environmental Protection Agency,
Research Triangle Park, NC. March 31, 1988.
17. Entropy Environmentalists. Emissions Testing at Wheelabrator Millbury,
Inc. Resource Recovery Facility, Millbury, MA. Prepared for Rust
International Corporation. February 8-12, 1988.
18. Radian Corporation. Results from the Analysis of MSW Incinerator
Testing at Peekskill, NY. Prepared for the New York State Energy
Research and Development Authority. Albany, NY. August 1988.
19. C. L. Anderson, et. al. (Radian Corporation). Characterization Test
Report, Marion County Solid Waste-to-Energy Facility, .Inc., Ogden Martin
Systems of Marion, Brooks, Oregon. Prepared for U. S. Environmental \
Protection Agency. Research Triangle Park, NC. EPA Contract J;
No. 68-02-4338. EMB Report No. 86-MIN-04. September 1988. |T
20. Hahn, J. L. (Cooper Engineers, Inc.). Air Emission Testing at the
Martin GmbH Waste-to-Energy Facility in Wurzburg, West Germany.
Prepared for Ogden Martin Systems, Inc. Paramus, NJ. January 1986.
21. Vancil, M. A. and C. L. Anderson (Radian Corporation). Summary Report,
CDD/CDF, Metals, HC1, SO-, NO , CO and Particulate Testing, Marion
County Solid Waste-to-Energy Facility, Inc., Ogden Martin Systems of
Marion, Brooks, Oregon. Prepared for U. S. Environmental Protection
Agency. Research Triangle Park, NC. EPA Contract No, 68-02-4338. EMB
Report No. 86-MIN-03A. September 1988.
2-18
-------
22, Entropy Environmentalists, Inc. Stationary Source Sampling Report,
Signal RESCO, Pinellas County Resource Recovery Facility,
St. Petersburg, FL, CARB/DER Emission Testing, Unit 3 Precipitator
Inlets and Stack. February and March 1987.
23. Hahn, J. L. and D. S. Sofaer (Ogden Projects, Inc.). Air Emissions Test
Results from the Stanislaus County, California Resource Recovery
Facility. Presented at the International Conference on Municipal Waste
Combustion. Hollywood, FL. April 11-14, 1989. pp. 4A-1 to 4A-14.
24. Lavalin, Inc. National Incinerator Testing and Evaluation Program: The
Combustion Characterization of Mass Burning Incinerator Technology;
Quebec City (DRAFT). Prepared for Environmental Protection Service,
Environment Canada. Ottawa, Canada. September 1987.
25. Seelinger, R.,'et. al. (Ogden Projects, Inc.) Environmental Test
Report, Walter B. Hall Resource Recovery Facility, Unit 1 and 2.
Prepared for Ogden Martin Systems of Tulsa, Inc. Tulsa, OK.
September 1986.
26. U. S. Environmental Protection Agency. Municipal Waste Combustion
Multipollutant Study: Refuse Derived Fuel, Suimary Report, Mid-
Connecticut Resource Recovery Facility, Hartford, CT. EMB Report
No. 88-MIN-09A. Research Triangle Park, NC. January 1989.
27. Klamm, S., G. Scheil, M. Whitacre, and J. Surnam (Midwest Research
Institute) and W. Kelly (Radian Corporation). Emissions Testing at an
RDF Municipal Waste Combustor (DRAFT). Prepared for the
U. S. Environmental Protection Agency, Research Triangle Park, NC.
May 1988.
28. New York State Department of Environmental Conservation. Emission
Source Test Report--Preliminary Report on Occidental Chemical
Corporation EFW. January 16, 1986.
29. Reference 14.
30. Entropy Environmentalists, Inc. Stationary Source Sampling Report,
Ogden Martin Systems of Haverhill, Inc., Lawrence, Massachusetts,
Thermal Conversion Facility. Particulate, Dioxins/Furans and Nitrogen
Oxides Emission Compliance Testing. September 1987.
31. Letter and enclosures from Philip Gehring, Plant Manager, Pigeon Point
Energy Generating Facility, to Jack R. Farmer, Director, Emission
Standards Division, OAQPS, U. S. Environmental Protection Agency.
June 30, 1988.
2-19
-------
32. York Services Corporation. Final Report for a Test Program on the
Municipal Incinerator Located at Northern Arobstook Regional Airport,
Frenchville, ME. Prepared for Northern Aroostook Regional Incinerator.
Frenchville, ME. January 26, 1987.
33. Midwest Research Institute. Analysis of Data from Phase I Testing at
the Vicon Incinerator Facility in Pittsfield,Massachusetts. Prepared
for New York State Energy Research and Development Authority. Albany,
NY. February 10, 1986. pp. 19, 22.
34. Midwest Research Institute. Results of the Combustion and Emission
Research Project at the Vicon Incinerator Facility in Pittsfield,
Massachusetts. Prepared for New York State Energy Research and
Development Authority. Albany, NY. June 1987. pp. 4-5 and 4-9.
35. Interpoll Laboratories. Results of the July 1987 Emission Performance
Tests of the Pope/Douglas Waste-to-Energy Facility MSW Incinerators in
Alexandria, MN. Prepared for HDR Techserv, Inc. Minneapolis, MN.
October 1987.
36. Reference 14.
37j PEI Associates, Inc. Method Development and Testing for Chromium,
Municipal Refuse Incinerator, Tuscaloosa Energy Recovery, Tuscaloosa,
AL. Prepared for U. S. Environmental Protection Agency. Research
Triangle Park, NC. EMB Report 85-CHM-9. EPA Contract No. 68-02-3849.
January 1986.
38. Cal Recovery Systems, Inc. Final Report, Evaluation of Municipal Solid
Waste Incineration. (Red Wing, Minnesota facility) Submitted to
Minnesota Pollution Control Agency. Report No. 1130-87-1. January 1987,
39. Environment Canada. The National Incinerator Testing and Evaluation
Program: Two Stage Combustion (Prince Edward Island). Report
EPS 3/UP/l. September 1985.
40. Reference 14.
41. Entropy Environmentalists. Emission Test Report, Municipal Waste
Combustion, Continuous Emission Monitoring Program, Wheelabrator
Resource Recovery Facility, Millbury, Massachusetts. Prepared for the
U. S. Environmental Protection Agency, Research Triangle Park, NC. EPA
Contract No. 68-02-4336. October 1988.
42. Ogden Martin Systems of Pennsauken, Inc. Pennsauken Resource Recovery
Project, BACT Assessment for Control of N0y Emissions, Top-Down
Technology Consideration. Fairfield, NJ. ^December 15, 1988. pp. 11,
13.
2-20
-------
43. Ling, K. J., R. J. Milligan, H. I. Lips, C. Castaldini, R. S. Merrill,
and H. B. Mason (Acurex Corporation). Technology Assessment Report for
Industrial Boiler Applications: NO Combustion Modification. Prepared
for U.'S. Environmental Protection Agency. Research Triangle Park, NC.
EPA-600/7-79-178f. December 1979. pp. 2-15, 2-48, 2-49, 2-72.
44. Reference 19.
.45. Reference 18.
46. Reference 24.
47. Linz, David 6. (Gas Research Institute). Emissions Reduction from Waste
Combustion Using Natural Gas. Presentationto U. S. Environmental
Protection Agency, Research Triangle Park, NC. November 30, 1988.
2-21
~rr
-------
-------
3.0 NOV EMISSION CONTROLS
^ - . .
There are two basic approaches to controlling NO emissions:
A.
(1) combustion modifications and (2) add-on controls. Combustion
modifications include staged combustion, low excess air (LEA), and flue gas
recirculation (FGR). Add-on controls include natural gas reburning,
selective non-catalytic reduction (SNCR), selective catalytic reduction
(SCR), and wet flue gas denitrification. Of these techniques, only
combustion modifications, reburning with natural gas, SNCR, and SCR have been
successfully demonstrated with MWC's or show significant potential for
effective and economical NO control. Thus, detailed descriptions of NO
A . ' ' . . /\
controls will be limited to these technologies. With each description,
measured NO emission reductions and possible problems with implementation on
n
MWC's are also provided.
3.1 COMBUSTION CONTROLS
Combustion modifications can achieve moderate NO emission reductions
i "
from MIC's by limiting the amount of NO formed in the combustion process.
Low excess a;ir, staged combustion, and FGR are combustion controls for NO
1 . .'...".
described in this section.
3.1.1 Low Excess Air and Staged Combustion
Low excess air and staged combustion can be used separately or together.
With LEA, less air is supplied to the combustor than normal, lowering the
supply of oxygen available in the flame zone to react with nitrogen in the
combustion air. With staged combustion, the amount of underfire (primary)
air is reduced, generating a starved-air region. By creating a starved-air
zone, part of the fuel-bound nitrogen is converted to ammonia (NH3).
Secondary air to complete combustion is added as overfire (secondary) air.
If the addition of overfire air is properly controlled, NH3, NOX, and 02
react to form N~ and water.
A Japanese mass burn/refractory combustor using automatic controls to
obtain LEA/staged combustion conditions demonstrated up to 35 percent
..-'- , ; - . . '!'. 1 ' ': -
reduction in NOX emissions over using manual controls. At Marion County,
3-1
-------
the effects of low excess air and low and high overfire air were evaluated. \
The NO data from these tests are presented in Table 3-1. Compared to normal '
f'
operating conditions at Marion County (75 percent excess air), LEA j
(40 percent excess air) conditions reduced NOX emissions from 286 ppm to \
203 ppm, a decrease of 29 percent. Under low load conditions, LEA reduced }
'I
NOX emissions from 257 ppm (at 70 percent excess air) to 195 ppm (at i
58 percent excess air), a decrease of 24 percent. During tests of the )
combustor with only underfire air (low overfire air), but at normal excess {
air conditions, NOX emissions decreased by 27 percent at low load (188 ppm |
versus 257 ppm) and 23 percent at normal load (220 ppm versus 286 ppm). «
During parametric combustor tests at Quebec City, use of low overfire air }.
reduced NOX emissions by 25 percent compared to tests conducted^ at similar f
load and excess air levels. The reason low overfire air generates less NO |
" i
is not certain, but it may be at least partially caused by high excess air at 1
the grate reducing the peak flame temperature, which in turn decreases |
thermal NOX formation. NOX measurements taken at Marion County during j
testing with high overfire air anl normal load (276 ppm) and low load \
(252 ppm) were roughly equal to tests conducted at similar load and normal '
\
air distribution (286 ppm and 257, ppm, respectively). These data suggest i
that use of high overfire air may be ineffective in reducing NO emissions I
'" \ 'X I
from mass burn waterwall combustors. <
3.1.2 Flue Gas Recirculation 1
In FGR, cooled flue gas is mixed with combustion air, thereby reducing {
the oxygen content of the combustion air supply. The flame temperature is ;
lowered and less oxygen is present in the flame zone, reducing thermal N0x I
generation. At the Long Beach, CA, mass burn combustor, where FGR is used to |
supply 10 percent of the underfire air, reductions in NOX emissions have been \
observed, although no quantitative results are available. At the Kita {
facility in Tokyo, Japan, a Volund mass burn/refractory combustor, where FGR {
is used to supply 20 percent of the combustion air, NO reductions of 10 to f
'' ' . o , X i
25 percent have been reported. At higher FGR rates, little increase in NQX J
reduction was observed. The modular excess-air combustors at Pigeon Point ^
and Pittsfield are Vicon units that have FGR built into the system. In Vicon |
-------
TABLE 3-1. MARION COUNTY EMISSIONS VERSUS AIR SUPPLY
Air
Supply
Normal
LEA
LOA
HOA
Normal
LEA
LOA
HOA
Runs I
.1, 2, 10,
lla and lib
3a
4
5
6a
7
8
9
Load
[% of Full)
100
95
98
103
76
77
74
78
NO Emissions
.(ppm, 7% 02)
286C
203
220
276
257
195
188
282
Excess
Air (%)
75
40
74
73
70
58
88
94
% NOY .
Reduction0
--
29
23
4
' --
24
27
(io)d
Tests where air supply was purposely varied are noted. LEA = low excess
air; HOA - high overfire air; LOA * low overfire air.
"'Compared to NOX emissions at normal air supply and similar load.
"Average NO emissions for the 5 runs.
Percent increase in NOX emissions.
3-3
-------
combustors, flue gas from ducting at the boiler exit (prior to flue gas I
cleaning) is injected into the primary combustion chamber. Recirculated flue 1
gas supplies approximately 35 percent of the combustor air. Emissions of NO ]
measured at Pigeon Point and Pittsfield range from 100-140 ppm at 7 percent i
0... There are no data available comparing NOV emissions with and without FGR ,'
i_ ... .... ''.' X - ; i
for a Vicon combustor. \
' ' ' ( ' ' C
Combustion modifications for NOX control may not increase emissions of )
other pollutants. However, if the modifications are not properly applied, \
higher emissions of CO, HC, and other products of incomplete combustion |
(PIC's) may result. For example, if the excess air is decreased too much, \
visible emissions and higher CO concentrations may result,5 If too much flue J
gas is recirculated, the flame zone can become unstable, causing poor (
combustion and higher CO emissions. ' Also, corrosion and slagging in the )
boiler may occur. I
' . ' *
3.2 GAS REBURNING \
Gas reburning is a NOV'control technique that overlaps combustion J
. - - ' . , - .. * . ' : . '; ':' .' -; ......' . './.'. /-.-.;. .-.. - |
modification techniques. A schematic of the natural gas burning method \
applied to a mass burn combustor is shown in Figure 3-1. Low excess air is (
provided at the combustor grate, with recirculated flue gas introduced above |
the grate. Natural gas is added to this LEA zone to generate a fuel-rich |
zone. Air is supplied above the fuel-rich zone to complete combustion. This \
process is designed to reduce NO formation without increasing CO emissions. |
Natural gas reburning at MWC's is a new technology being evaluated by |
the Gas Research Institute. The goal of gas reburning is to achieve up to (
75 percent NO reduction. To date, most of the data on reburning are for I
8 ' ' ' ' ' }
pulverized coal-fired (PC) boilers. Testing for MWC's is currently underway \
in a 6 tpd pilot-scale combustor. In the pilot-scale unit, NOY emissions f
- . " " ' . _ . A - I*
without gas reburning ranged from 190 to 260 ppm at 7 percent Og. With gas
reburning, the NO emissions were 110 to 125 ppm at 7 percent 02, an average
reduction of 50 percent. The maximum NOY reduction measured was 60 to
A
70 percent. During these tests, 15 percent (heat input basis) natural gas,
15 percent flue gas recirculation (for mixing the natural gas), and 30 to
40 percent excess air were used. Neither CO nor hydrocarbon emissions
increased with gas reburning.
-------
Air
Natural Gas
Municipal
Solid Waste
Figure 3-1. Gas Returning for NC^ Control
at a Mass Burn MWC.9
3-5
«0
IO
M
-------
3.3 SELECTIVE CATALYTIC REDUCTION \
Selective catalytic reduction is an add-on control technology for NO ;
removal. Ammonia (NH3) is injected into the gas flue downstream of the i
boiler where it is mixed with the NOX contained in the flue gas and passed j
through a catalyst bed. In the catalyst bed, NOX is reduced to N2 by I
reaction with NHV The overall reactions between NOV and NH~ are: }
I
(1) 4 NO + 4 NH3 + 02 ----> 4 N2 + 6 H20 j
(2) 2 N09 + 4 NH, + 0,' -----> 3 N9 + 6 H90 )
2 3 2 2 2 j|
The reactions between NOV and NH, occur at temperatures of 375-750°F, I
A . 0 . ' . £'
depending on the specific catalyst. ^
Selective catalytic reduction has been tested at coal, oil, and natural c
,,.-'" ' --.-. i"
gas-fired facilities in the U. S. Reductions of NOX emissions of 60 to 1
85 percent have been measured at these facilities with NH,:NOV molar ratios \
rt 1 1 '
of 0.6 to 0.9 and temperatures between 570 and 750 F. Currently there are j
no applications of SCR to MWC's in the U. S. NO emission reductions of 26 C
. . ' " " .'". . J,
to 86 percent have been measured at two Japanese mass burn MWC sites using |
special low temperature catalysts (V20g - tiOj,, temperatures of 375 to I
535°F).12 The SCR system at the 65 ton/day MWC in Iwatsuki, Japan, |
demonstrated an average NOY reduction of 77 percent (versus design of 80 f
^ . '.. ' f
percent) during two performance tests conducted approximately 1 and 2 months |
after plant startup. This SCR unit, located downstream of a spray 1
dryer/fabric filter system, operated at an average temperature of 395°F and a i
NH,:NOV molar ratio of 0.7. Data from these tests are reported in Table 3-2. 1
3 X . . - f
At the Tokyo-Hikarigaoka 150 ton/day MWC, the SCR system demonstrated an |
average NOV reduction of 44 percent at a temperature of 475°F and a NH-:NO j
X - - V' X . ' \
molar ratio of 0.57. These tests were conducted approximately 3 months after {
startup; the data are presented in Table 3-3. This SCR unit was retrofit
between an ESP and a wet scrubber. Because of space constraints, the SCR
unit was sized for 51 percent NOX removal.
There are several operating considerations with SCR. First, the SCR
operating temperature at both Iwatsuki and Tokyo-Hikarigaoka exceed the
3-6
-------
TABLE 3-2. RESULTS OF TESTING OF SCR SYSTEM AT IWATSUKI, JAPAH
CO
I
Test
Date Unit No.
2-19-87 IB
3-6-87 IB
2-19-87 2B
3-6-87 2B
Average IB
Average 2B
Average Overall
Run No.
1
2
3
4
5
6
1
2
3
4
5
6
SCR Inlet
Temperature ( F)
385
399
396
396
392
392
388
388
401
401
405
397
393
397
395
Molar Ratio
-------
TABLE 3-3. RESULTS OF TESTING OF SCR SYSTEM AT TOKYO-HIKARIGAOKA, JAPAH
co
i
00
NO Concentration
Test SCR Inlet Molar Ratio Outlet NHj fppm at 7X O.) NO Removal
Date Run Ho. Temperature (°F) (NH :NO ) Concentration (ppm) Inlet Outlet Efficiency (X)
3-17-87 1 478 0.44 2.6 150 98 34
2 478 0.80 15 148 73 51
3-18-87 3 475 0.43 0.5 166 123 26
4 471 0.69 6.1 162 94 42
3-19-87 5 475 0.56 13 153 66 57
6 475 0.50 14 158 73 54
Average 475 0.57 8.5 156 83 44
-------
fabric filter outlet temperature needed to achieve maximum control of
CDO/CDF, HC1, and SO-. As a result, either flue gas reheat will be needed or
reduced control of CDD/CDF, HC1, and S02 will occur. Second, performance of
SCR can be detrimentally affected by catalyst poisoning by either metals or
acid gases. Also, entrained particulate can blind or deactivate the
catalyst. Third, because ammonia is injected into the flue gas, ammonia
emissions can result. In a properly operated system, ammonia emissions are
;. . -. - 1.3 - -"'.' '
typically less than 10 ppm. At the Tokyo-Hikarigaoka MWC, outlet ammonia
emissions averaged 8.5 ppm and ranged from 0.5 to 14 ppm. Fourth, depending
on the location of the catalyst bed (i.e., after the economizer or after
particulate/acid gas removal), flue gas reheat may be necessary to reach the
desired catalyst operating temperature. Flue gas reheat can be a significant
14 ' ' '
expense.
3.4 SELECTIVE NON-CATALYTIC REDUCTION
Selective non-catalytic reduction (SNCR) refers to add-on NOY control
.,'-..- - . '.-'"" " " "" :
techniques which reduce NOX to N« without the use of catalysts. These
techniques include Exxon's Thermal DeNO , which uses injection of ammonia;
"< A' . " " .
the Electric Power Research Institute's NO OUT process, which injects urea
and chemical additives; and EMCOTEK's two-stage urea/methanol injection
process. To date, only Thermal DeNOv has been demonstrated on MWC's in the
- ' -. . " ' : '',"
U.S., although, the other techniques have been tested in Europe and Japan.
Because of this, discussion of SNCR techniques focuses on Thermal DeNOx.
'.With Thermal DeNO , ammonia is injected into the upper furnace area of
! . . A . ' ' |1 .".'"
the combustpr. Ammonia and NO react according to the foilowilng competing
reactions:
(1) 4 NO + 4 NH3 + 02 ----> 4 N2 + 6 H20
4NH3 + 502 -> 4 NO + 6 H20
At 1,600 to 1,800°F, the first reaction dominates and NOY is reduced to N?.
-» .. . . . A .-..''-.' t
Above 2,000 F, the second reaction dominates and NH3 is oxidized to NO.
Below 1,600 F, both reactions proceed slowly and NH, remains unreacted.
': . v ' '15
Reductions as high as 65 percent are projected for MWC's by Exxon.
3-9
-------
Because of the viability in combustion characteristics of MSW, furnace J
temperatures in the upper furnace can vary rapidly. This necessitates i
installation of ammonia injectors at several furnace elevations to assure ]
injection at proper temperatures. The sensitivity of ammonia-based SNCR ^
reactions to temperature is one of the primary reasons behind development of ]
the urea-based NOXOUT and EMCOTEK processes. ^
Thermal DeNO has been applied at several MWC's in Japan and at three JJ
" ' ' \i
state-of-the-art mass burn/waterwal1 combustors in California (Commerce, I
Stanislaus County, and Long Beach). Each of the operating MWC's in the U. S. 1
using Thermal DeNO is summarized in Table 3-4. I
1 ''
-------
TABLE 3-4. EXISTING THERMAL DENO FACILITIES IN UNITED STATES
CO
facility Location
Commerce
Long Beach
Stanislaus Co.
Unit 1
Unit 2
Startup Combustor Number Combustor t£CD. NO Emissions,
Date Type of Combustors Size, tpd Type ppm g -7Z 0-
2/87 MB/WW 1 300 SD/FF 119°
7/88 MB/HW 3 460 SD/FF 56**
8/88 MB/WW 2 400 SD/FF
93
1136
Estimated
NO Reduction,
ji
Percent
44
50
69
63e
*MB/HW = mass burn/waterwall.
SD/FF = spray dryer/fabric filter.
Average of 10 short-term optimization tests at NH injection rate of 2.4 Ib/ton (see Table 3-5).
d
Average of three compliance tests.
At 12 percent CO .
-------
slurry is injected through arotary atomizer, with the rate of slurry
addition controlled by an S02 monitor/controller at the stack. The amount of
dilution water in the lime slurry is controlled to maintain temperature at
the outlet of the SD. Flue gas exiting the SD flows through the reverse-air
FF. Design flue gas flow to each FF is 118,000 acfm at 285°F. Each FF has
10 compartments of teflon-coated fiberglass bags and a net air-to-cloth ratio
2 ' ''"''.
of 1.8 acfm/ft . Ducting is provided to route flue gas from one FF to
another if one unit goes down. Flue gas is exhausted through a common stack.
The Stainslaus Waste-to-Energy Facility in Crows Landing, California
consists of two identical Martin GmbH waterwall combustors, each capable of j
combusting 400 ton/day MSW. Each combustor is equipped with Exxon's Thermal <
DeNO (ammonia injection) for NO control. Emissions are controlled \
** * i
downstream of the boiler with a Flakt spray dryer/fabric filter system. In \
the SD, slaked lime slurry is injected through two-fluid nozzles, with the ]
amount of slurry controlled according to the stack S0« concentration and the r!
dilution water flow controlled according to the SD outlet temperature. A !
residence time in the SD of 15 seconds is used to dry the slurry and obtain a (
flue gas temperature of 285°F at the SD outlet. Flue gas exiting the SD }
flows through the pulse-jet FF at 94,000 acfm and 285°F. The FF has six }
compartments of teflon-coated fiberglass bags (1,596 bags total) land a net {
air-to-cloth ratio of 3.2 afm/ft2. ;
'-'''". *
Because of the limited operating time of these units, long-term (
performance and reliability data are limited. Available performance data are *
based mainly on short-term compliance testing using continuous emission (
showed variations in performance with ammonia injection location and NH3:NO i
* rt ' '.'" ', ' - . . *
molar ratio. These data are summarized in Table 3-5. The objective of {
3-12
-------
TABLE 3-5. SUMMARY OF NO REDUCTIONS AT COMMERCE OPTIMIZATION TEST
Controlled NO Emissions
Injection
Location'
Top Row
Bottom Row
NH
Ib/hr J
15
30
45
15
30
45
Rate
lb/ton"
i:2-
2.4
3.6
1.2
2.4
3.6
Average NH
Molar Ratio
0.85
1.53
2.36
0.89
1.88
2.36
Number
of Tests
5
10
11
2
2
2
JDRB 8 7? 0 )
Average
117
92
93
126
120
94
High *
145
107
148
136
125
107
Low
105
63
58
116
114
80
HO Reduction «) >C
Average
22.3
43.9
44.0
11.8
19.8
42.1
High
27.0
57.5
62.0
12.5
24.4
48.4
Low
14.5
21.9
18.0
11.0
15.3
35.7
Pounds per ton refuse fed. Calculated based on 300 tpd capacity of combustor.
to
I
u>
Based on NO emissions with Thermal DeNO turned off for each test.
x x
Percent NO reductions do not correspond directly to those for NO emissions.
-------
these tests was to determine the optimum ammonia injection elevation. During \
testing, the ammonia injection location was varied between a top row and a
bottom row of injection nozzles. The ammonia injection rate was also varied, <
ranging from 0 to 3.6 Ib NHg per ton refuse at each injection location. f;
Injection through the top row of nozzles generally resulted in lower NO «!
1 : .'.., .. >
best system performance is achieved with ammonia injection throughjone or i
more of the upper three injector rows depending on real-time monitQnng of !
combustor conditions and NO,, levels. Maximum 1-hour NOV emissions from }
x x '.. . . . j
February through May 1989 were less than 150 ppm at 7 percent 02 on all but 6 {
days (out of 110 days total). All of the 24-hour averages were less than ,f
19 !
120 ppm at 7 percent O-. \
Emissions of NOV measured during three short-duration tests on Unit 1 at ,|
" - ' .',''.
the Long Beach facility averaged 56 ppm at 7 percent 02 with the Thermal
DeNO system operating normally. Three runs performed 1 month later without
Thermal DeNOv measured average NOV emissions of 68 ppm at 7 percent CL,
X . . . X ,. . " .'.-.. ' ^
suggesting a NO reduction of roughly 20 percent due to Thermal DeNQx- NQX
-------
measurements during bbth test periods are based on grab sampling and wet
chemistry analysis using South Coast Air Quality Management District (SCAQMD)
Method 7.1. These uncontrolled NOV levels are significantly lower than
20
typically measured by the plant CEM system. When neither the FGR or
Thermal DeNOv systems are in operation, NO emissions measured by the plant
A . . ^
CEMS are typically 190-230 ppm at 7 percent Op. With FGR only, NOX emissions
based on the plant CEMS are typically 160-190 ppm. When both FGR and Thermal
DeNO are operated, NO emissions are reported to be consistently less than
A - ' rt '
120 ppm, and frequently less than 50 ppm. These data indicate that the
Thermal DeNO system reduces NO emissions at Long Beach by 30-70 percent.
A A
At the Stanislaus County MWC, three tests were performed on each of the
21
facility's two units. Without ammonia injection, the NO emissions from
. ' J\
Unit 1 averaged 297 ppm at 12 percent CO-. With ammonia injection of
29 Ib/hr (1-7 Ib NH3 per ton MSW), the NOX emissions averaged 93 ppm at
12 percent CO-, a reduction of 69 percent. Similar results were obtained for
Unit 2, where NO emissions averaged 304 ppm at 12 percent C09 without
rt . : £f
ammonia injection and 113 ppm at 12 percent C02 with an ammonia injection
rate of 25 Ib/hr (1.5 Ib NH3 per ton MSW), a reduction of 63 percent.
As with SCR, there are potential problems associated with Thermal DeNO..
A
Ammonia or ammonium chloride emissions may result when the NhL is injected
outside the desired temperature window, at a higher than normal rate, or when
residual HC1 levels in the stack exceed roughly 5 ppm. At the Long Beach
MWC, a detached ammonium chloride plume has been observed downwind of the
stack when the Thermal DeNO is used. At the Stanislaus County MWC, an
A
ammonium chloride plume was observed at an NH, injection rate 50 percent
higher than the normal feed rate of 1.5-1.7 lb/ton.22 At the Commerce MWC,
ammonia emissions following the unit's spray dryer/fabric filter have not
been measured above 2 ppm at 7 percent 0«. However, an ammonium chloride
plume is frequently present.
Corrosion of the boiler tubes by corrosive ammonia salts which are
formed from unreacted ammonia and sulfur dioxide or hydrogen chloride has
been hypothesized to be a potential problem with Thermal DeNOx. However, no
3-15
-------
boiler corrosion problems attributable to ammonia salts have been observed *
with the U. S. systems during the limited amount of operating time.24'25 In '
Japanese MWC's ammonia is generally injected into refractory sections, not in I
boiler tubes where corrosion can occur, (
^
Increased CO emissions with ammonia injection has also been suggested as |
" 5*fi '"''" '""'" ^
a potential problem with Thermal DeNQ,. At Commerce, measured CO emissions <
' ' « * ' ' \
while the DeNOx was operating normally (15 ppm at 7 percent Q2) were essen- (
tially the same as the CO emissions without the DeNO (14 ppm at 7 percent 1
?7 x * rr r t
01 (
V- .. . . . - .- ..... ' . \
A recently identified concern with Thermal DeNO,, is that the {
' '" : .-.'-.- ,;- .!- :\ . .-.: A. ' . . . X'- '' ;'. r
ammonia injected into the flue gas may reduce control of mercury emissions by
a spray dryer/fabric filter. Outlet mercury emissions from MMC's with spray
dryer/fabric filter systems are presented in Table 3-i. Compliance tests at
?7 9ft
Commerce (June 1987), Long Beach (November 1988)," and Stinisiaus County
(December 1988) showed relatively high mercury emissions (180 to
900 ug/dscm at 7 percent 02) compared to facilities without SNCR |Biddeford,
Quebec City, and Mid-Connecticut). At Commerce, mercury concentrations prior
to and following the spray dryer/fabric filter were simultaneously measured
during a single run and indicated little or no removal of mercury. During
the tests at Commerce, portions of the probe rinse from the spray
dryer/fabric filter inlet and outlet samples were inadvertently discarded, (
As a result, the calculated concentrations and removal efficiencies are |
estimates. However, because mercury is generally volatile, relatively little i
/ . !»
mercury was probably present in the discarded samples. Thus the calculated I
values are believed to be representative. Uncontrolled mercury J
concentrations were not measured at Stanislaus County and Long Beach, but the $
- 'i
measured outlet emissions suggest little removal of mercury. Because these |
three facilities have spray dryer/fabric filter systems as well as ammonia V
injection for NOV control, it has been suggested that the poor mercury I
A- ' I"
removals may be due to the ammonia in the flue gas. >
A possible explanation for the impact of Thermal DeNOx on mercury f
control is that mercury is normally in a combined ionic form (principally J
) that can absorb or condense onto particulate matter at the low (
3-16//- . :.." . ;. ; f
1 . ' ' ' . . . . ' ' #
-------
TABLE 3-6. OUTLET MERCURY EMISSIONS MEASURED FROM SPRAY DRYER/FABRIC FILTER SYSTEMS
Facility
Location
Commerce (1)
Commerce (2)
Commerce (2)
Long Beach
Stanislaus
County
Unit 1
Unit 2
Mid-Connect icut
Marion County
Biddeford
Quebec City
Inlet Inlet PM Inlet CDD/CDF Outlet
Temperature to Emissions, Emissions, Mercury Emissions,
APCD Type" Fabric Filter, °F gr/dscf i 12X CO, ng/dscm 9 7X Cl ug/dscm 8 7X 0,
= 2 .2 2
SD/FF/SNCR 270 1.8 28.1 200 - 940
SD/FF/SHCR NAb 2.01 NA 39.4
SD/FF/SNCR NA 1.23 619 67.9
SD/FF/SHCR 300 1.58 NM° 180
SD/FF/SNCR 287 NM NM 499
SD/FF/SHCR 292 KM NM 462
SD/FF 277 2.41 1,019 50
SD/FF 280 0.88 43.0 239
SD/FF 278 3.2 ' 903 0
SD/FF 283 2.92 1,960 14.7
SD/FF = spray dryer/fabric filter.
SNCR = Selective non-catalytic reduction.
NA » not available.
NM = not measured.
-------
operating temperatures of the fabric filter {less than 3Q0°F}.^ By \
injecting ammonia into the flue gas, however, pockets of reducing atmosphere <
may form which reduce mercury to an elemental form, which is more volatile i
and difficult to collect, \
However, data collected more recently at Commerce (May 1988} (
demonstrated mercury removals of 91 percent while firing a mixture of }
60 percent commercial refuse and 40 percent residential refuse and 74 percent
while firing a mixture of 95 percent commercial refuse and 5 percent
31 ..,.-: . . ..- '.
residential refuse. During both of these tests the ammonia injection (
system was operating. These test results indicate that ammonia injection may $
not be the reason for the observed low mercury removals. (
Another theory gaining acceptance regarding the removal of mercury is J
that carbon in the flue gas enhances adsorption of mercury and that Thermal |
DeNOv has no effect. This theory suggests that the poor removals of )
' ' : ' ' -V
mercury at the MWC's with Thermal DeNO are a result of good combustion f
leaving little carbon in the fly ash onto which the mercury could adsorb. In |
Figure 3-2, mercury removal efficiency from spray dryer/fabric filter systems 1
operating at 300°F or less is plotted as a function of the PM concentration. \
at the combustor exit. The data suggest increased mercury removal with f
... " . . . ./ ''..'. +
increasing inlet PM concentration. Mercury emissions as a function of inlet f
PM are shown in Figure 3-3. The trends are similar to those in Figure 3-2, J
The data from the 1987 test at Commerce represent maximum estimated emissions \
and are separated by run because the results varied widely. 1
Little direct data are available on the carbon content of the fly ash f
from the facilities in Table 3-6. However, it is expected that CDQ/CDF \
- -.'-'' - I
concentrations at the combustor exit are indicative of good combustion, and \
. '- .. . . -.. -. ' . -.. .... , ' ' '33' . f
thus provide a surrogate measure for the carbon content of the fly ash. |
Data on mercury removal efficiency and mercury outlet concentration versus . )
CDD/CDF at the combustor exit are shown in Figures 3-4 and 3-5, respectively. .'\
Both of these figures support the theory that reduced carbon content in the ;
fly ash increases mercury emissions. \
Because of the limited amount of mercury emissions data from MWC's with ;
Thermal DeNOv and the apparent strong relationship between fly ash $
X '. -,.- f
3-18
-------
Co
iuu -
90 -
* 80 -
o
c
"o
£ 70 -
ui
"a
o
1 60 -
DC
h.
O
| 50 -
40 -
0.
, .A,
.
A
"
Mid-Connecticut
+ Qu«b»c City
O Blddcford
A Comm»rc« (1988)
X Comm»rc«(1887)
X
I II I II I I I I I II I
8 1.2 1.6 2 2.4 2.8 3.2 3.6
Inlet PM Concentration (gr/dscf » 12% CQ)
Figure 3-2. Effect of inlet PM on mercury removal efficiency
for MWCs with SD/FF systems.
-------
*-^
O*
r*.
©
o
1
3
0
8!
e
8
OB
0
1UUU
900 -
800 -
700 -
600 -
500 -
400 -
300 -
200 -
100 -
X Mid-Connecticut
+ Quebec Crty
o Biddeford
A Commerce (1988)
x Commerce (1987)
a Long Beach
v Marion County
-
' *
n x
A
A +-'
u ~| r. j , , , , , . ,.---- .,- |- j- j- f j ~-j
Q-8" 1.2 1.6 2 2.4 2.8 3.2 3.6
InUtA PM ConoentraMon (gr/dscf @ 12% CQ)
! Mivklua! runs during mam tssJ p@dod ^ Commefce.
3-3., of irtel py on mercurf
-------
100
_O
£
"3.
' O
~ I
80
70 -
60 -
Mid-Connecticut
+ Quebec City
o Blddofford
A Commerce (1988)
80
400
BOO
1800
I
1SOO
2000
2400
2800
Inlet CDD/COF Concentration (ng/dscm » 7% Q,)
Figure 3-4. Effect of inlet CDD/COF on mercury removal efficiency
for MWCs with^^ spray dryer/tebric finer systems
-------
cf
#
N
on
CO
I
ro
fNJ
o
o
o
X
*«
33
O
uuu
900 -
800 .?
700 -
600 -
500 -
400 -
300 -
200 -
1 00 -
0 -
x + Quebec
o Biddefor
A Commer
X Commer
v Marlon C
Mid-Con
. . " .'-.-
X . ' ' . ' .
A
. - A .; . . .-
I T ' . ' 1 1 - . I I 1 i
City
d
ce(1988)
ce (1987)
Jounty
nectlcut
i i -r 1
0
400 800 1200 1600
Inlet CDD/CDF Concentration (ng/dscm P 7%
2000
2400
2800
Figure 3-5. Effect 'of inlet CDD/CDF ©n mercury emissions for
MWCs with spray dryer/fabric fiiters systems
-------
concentration and carbon content versus mercury control, the hypothesized
detrimental effect of Thermal DeNOx on mercury control by a spray dryer fabric
filter cannot be proved with certainty.
3.5 SUMMARY OF NOY EMISSION CONTROLS
,'-..." A "
There are advantages and disadvantages to the control of NO emissions
from MWC's with both combustion modifications and add-on NO controls.
A
Combustor modifications, such as low excess air and staged combustion, can be
implemented relatively easily without substantial additional cost. However,
consistent and quantifiable NO emission reductions have not been demonstrated
A
with these technologies. The highest potential NO emission reduction
appears to be about 30 percent. Higher NO reductions would result in
increased CO, HC, or other PIC emissions.
Natural gas reburning offers the potential to achieve 60 to 70 percent
NO reductions without increasing CO emissions. The technology has only been
A | .. ' . " "
tested on a pilot-scale MWC, however, and further testing needs to be done
before applying reburning to full-scale MWC's.
Selective catalytic reduction appears able to yield high NOV reductions.
A1
Reductions of NOV at a full-scale MWC in Japan averaged nearly 80 percent,
A
with a low of 62.5 percent measured for one run. However, catalyst poisoning
and deaetivation may substantially decrease performance with time.
Thermal DeNOx has been used on three MWC's in the U. S. Reductions of
NO emissions during short-term tests may be as high as 65 percent, but can
A
vary widely during normal operation. Controlled NOY emissions of 150 ppm at
A
7 percent Og or less are consistently achievable with SNCR for long- and
short-term tests. Because of the significant variability in Thermal DeNOx
performance over time and the lack of CEM data, it is not currently possible
to relate measured NO emission reductions during short-term compliance tests
to long-term performance levels. Visible plume formation may occur as
combustor operating conditions vary. Uncertainty also exists regarding the
possible relationship between Thermal DeNO and mercury emissions.
y .
&
>' 3-23
"IT
-------
3.6 REFERENCES 1
'''"' ' ' ' - '.'.'.'- . . :.'.' ' 1
1. Californii Air Resources Board. Air Pollution Control at Resource i
Recovery Facilities. May 24, 1984. pp. 85 - 87. I
f
2. Telecon. Charlie Tripp, SERRF Operations Officer, with Michael A. j
Vaneil, Radian Corporation. January 26, 1989. Operation history of j
Long Beach MWC. c
' % ' ' 1
3. Reference 1, pp. 87 - 92. J
4. Reference 1, p. 90. I
5. Fossil Fuel Fired Industrial Boilers - Background Information, Volume 1: \
Chapters 1 - 9. U. S. Environmental Protection Agency, Research t
Triangle Park, N.C. EPA-450/3-82-006a. March 1982. p. 4-105. '?
fc
6. Reference 5, p. 4-114. J
7. Reference 2. \
8. Kokkinos, A., Reburning for Cyclone Boiler NO Control. Journal of Air \
and Waste Management Association (JAPCA). Vof. 39, No. 1, January 1989. \
pp. 107-108. «
9. Abbasi, H. A., and M. J. Khinkis (Institute of Gas Technology), }
D. C. Itse and C. A. Penterson (Riley Stoker Corporation), Y. Wakamura
(Takuma Company) and D. G. Linz (Gas Research Institute). Pilot-scale
Assessment of Natural Gas Reburning Technology for NO Reduction from
MSW Combustion Systems. Proceedings of the International Conference on
Municipal Waste Combustion. Hollywood, Ft. April 11-14, 1989, p. 13-5.
10. Reference 9. pp. 13-1 to 13-29. I
11. Sedman, C. B. and T. G. Brna. Municipal Waste Combustion Study: Flue |
Gas Cleaning Technology. U. S. Environmental Protection Agency, f
Research Triangle Park, N.C. EPA/53Q-SW-87-021d. June 1987. p. 3-3. !
12. Mitsubishi SCR System for Municipal Refuse Incinerator, Measuring t.
Results at Tokyo-Hi karigaoka and Iwatsukiy Mitsubishi Heavy Industries, f
Ltd. July 1987. .\
#'
13, Reference 11, pp. 3-3, 3-4. v
."''..-'.:'. .'.' . ,: ' ..= '.'' . . .:. '. I
14. Ogden Martin Systems of Pennsauken, Inc. Pennsauken Resource Recovery }
Project, BACT Assessment for ^^ Control of -'.-MO Emissions , Top -Down |
Technology Consideration. Fairfield, N.J. December 15, 1988. *
pp. 26, 27.^ ' : ,
- - -
3-24. . '. . . ' ' . ' ' '."''. f
-------
. . . .
> 15. Letter and enclosures from Krider, D. E., Exxon Research and Engineering
> Company, to Martinez, J. A., Radian Corporation. January 11, 1989.
,. 3 p. Costs for Thermal DeNOx.
16. McDannel, M. D., L. A. Green, and B. L. McDonald (Energy Systems
s Associates). Air Emissions tests at Commerce Refuse-to-Energy Facility,
j* May 26 - June 5, 1987. Prepared for County Sanitation Districts of Los
) Angeles County. Whittier, CA. July 1987.
v .
f :
) 17. Telecon. Ed Wheless, Commerce Refuse-to-Energy Facility, with David M.
i White, Radian Corporation.
^
| 18. Reference 14, pp. 47-57.
J! 19. Plant Visit. Johnston, M. 6., EPA/OAQPS and D. M. White, Radian, with
i E. Wheless, Commerce Refuse-to-Energy Facility. Review of Commerce
V Operating Experience.
"v ..-'
; 20. Letter from Rigo, H. 6., Rigo & Rigo Associates, Inc., to Johnston, M.,
> U. S. Environmental Protection Agency. March 13, 1989. 2 pp.
'». Compliance Test Report Unit No. 1 -- South East Resource Recovery
} Facility, Long Beach, CA.
it...-.-.'
| 21. Telecon. Gary Reeves, Stanislaus County APCD, with Michael A. Vancil,
? Radian Corporation. January 19, 1989. Operation history of Stanislaus
} County MWC.
'.']' .
/.' 22. Reference 20.
-\ . . -
y 23. Telecon. Ed Mheless, Plant Manager, Commerce Refuse-to-Energy Facility,
^ with Michael A. Vancil, Radian Corporation. Januarys, 1989. Operation
; of Thermal DeNO .
\ -
I 24. Reference 20.
v 25. Reference 2.
'v1 ' '
1 26. Lyon, Richard K. Thermal DeNO . Environmental Science and Technology.
'.. Vol. 21, Number 3, 1987. pp. $31 - 236.
> 27. Reference 16.
> .''-.
> . ' 28. Ethier, D. D., L. N. Hottenstein, and E. A. Pearson (TRC Environmental
Consultants). Air Emission Test Results at the Southeast Resource
: Recovery Facility Unit 1 (Long Beach), October to December 1988.
5 Prepared for Dravo Corporation. Long Beach, CA. February 28, 1989.
pp. 13, 14.
3-25
-------
29. Hahn, J. L. and D. S. Sofaer (Ogden Projects, Inc.). Air Emissions Test ;
Results from the Stanislaus County, California Resource Recovery !
Facility. Presented at the International Conference on Municipal Waste I
Combustion. Hollywood, FL. April 11-14, 1989. pp. 4A-1 to 4A-14.
30. Metzger, M. and H. Braun. In-situ Mercury Speciation in Flue Gas by ^
Liquid and Solid Sorption Systems. Chemosphere. Vol. 16, No. 4. 1987. }
pp. 821-832. I
31. Telecon. Vancil, H. A., Radian Corporation, with Pasek, R., California )
Air Resources Board. April 25, 1989. test results from 1988 testing at ^
Commerce. ]
32. Donnelly, J. R. and B. Brown. Joy/Niro SDA MSW Gas Cleaning I
Systems--New Developments. Presented at the 1989 APCA Conference.
Anaheim, California. June 26-30, 1989. Paper No. 89-109.6. ^
33. Personal communication. Hahn, J., Ogden Projects, Inc. and M. Johnston, 1
EPA/OAQPS. April 1989. }
3-26
-------
4.0 COST PROCEDURES
* Procedures are developed in this section for estimating capital and
> annual operating costs for applying Thermal DeNO to new MWC's. As discussed
!J . X ".--
j in Section 3.0, Thermal DeNOx is a selective noncatalytic reduction,(SCNR)
) technique for controlling NO emissions which is being commercially used by
\ X
C. three full-scale .MWC's in California. To be consistent with other cost
3 analyses performed for this regulatory development, costs for Thermal DeNOv
i 12 x
! are presented in December 1987 dollars. ' Section 4.1 presents the pro-
It cedures for estimating capital costs, and Section 4.2 presents the procedures
I for estimating annual operating costs. The procedures presented in both
') sections will be used to estimate costs of Thermal DeNOY for twelve lll(b)
\ . - . . i X
( model plants in Section 5.0. Each model plant represents a subcategory of
new MWC's. Each subcategory represents a different type and size of MWC
{ ' expected to be built in the future. It should be emphasized that these
procedures provide "study estimates" (i.e., +30 percent accuracy) of Thermal
DeNO costs for an individual application.
A
4.1 CAPITAL COST PROCEDURE
Table 4-1 presents the procedure for estimating capital costs for
Thermal DeNO applied to new MWC plants. The total capital investment
. A
includes direct purchased costs for equipment, indirect and contingency
costs, licensing (royalty) fee, preproduction costs, and NO monitoring
in,
equipment costs. The direct purchase costs include costs for the following
equipment: a low-pressure air compressor, ammonia storage tank, ammonia
vaporizer, injection nozzles, piping, and associated instrumentation.
Indirect costs include field labor overheads, erection fee, and contractors'
engineering and design fees. The contingency cost accounts for:
(a) unforeseen expenses that may occur such as equipment modification,
increases in field labor costs, increases in startup costs, etc. and
(b) risks associated with meeting performance guarantees and the operating
experience level of the technology. A licensing fee charged by the process
vendor (Exxon Research and Engineering Company) is also included in the total
capital costs. Preproduction costs include operator training, equipment
4-1
-------
TABLE 4-1. PROCEDURES FOR ESTIMATING CAPITAL COSTS FOR THERMAL DeNO . 1
APPLIED TO NEW MWC PLANTS*'0 x
Direct Costs. 103 $ = 0.444 (Q * N)0'621 +151 {
'.-'' . ' ' *'
Indirect Costs. 103 $ = 0.33 direct costs + 10 {
= 0.147 (Q * N)°'621 +60 {
3 ' '" ' ' - ' ' '
Contingency. 10 $ = 20% of direct and indirect costs |
License Fee. 103 $ = 3.35 + 7.01 x 10"4 * Q * N {
Preproduction. 10 $ = 2% of the sum of the direct capital, indirect capital, \
and contingency + one month of the direct operating cost at {
full load excluding monitors '*,
NOV Monitor, 103 $ = 24 * N \
''X ' . . . . ^ :
Total Capital Investment - Direct Costs + Indirect Costs + Contingency + |
License Fee + Preproduction + NO Monitor f
' ' ' ':
/
... . -
«.-.' .-'- : ' '
Costs are in December 1987 dollars.
h ' ' " ' ' ' ' " '0
Q = 125 percent of the calculated flue gas flowrate per combustor at 450 F, )
acfm. }
N = number of combustors. J
4-2
-------
- ..-.
checkout, extra maintenance, and inefficient use of chemicals and other
materials during plant startup. The total capital investment includes
separate NOV monitoring equipment per combustor to ensure continuous
** -
emissions compliance.
Table 4-2 presents the capital cost data base used to develop the cost
procedures. The data base contains capital estimates for Thermal DeNO
''."", A
applied to 12 proposed mass burn/waterwal1 MWC facilities, ranging in size
from 150 to 3,000 tpd. Most of these cost estimates were provided by Exxon
and none of them contained any itemization of equipment or other costs. Only
one data source reported actual flue gas flowrate as shown in Table 4-2. The
flue gas flowrates of the other plants were estimated assuming an excess air
level of 80 percent. These flue gas flowrates represent typical conditions
'' -
associated with new mass burn/waterwal1 facilities (see Table 5-1 in this
report). It is assumed that the costs for Thermal DeNOv applied to mass
'' x
burn/waterwal 1 combustors are similar to those for the other combustor types,
since NO^ emissions for mass burn/waterwal 1 combustors are within the range
. X ' ' '. -'..-
for all other combustor types as discussed in Section 2.0. Sections 4.1.1,
4.1.2, and 4,1.3 discuss the bases and rationale for the capital cost
. ' ' ' ' '
.procedure. :.".
4.1,1 Direct Capital Cost . ' . '''
- - '.''
Table 4-3 presents the direct capital costs from Table 4-2 for the
' . '--''" :
12 mass burn/waterwall facilities corrected to December 1987 dollars using
the Chejjjgal Engineering Plant Cost Index. As shown by Table 4-3, no
apparent trend can be observed between direct capital costs and plant size
either in tpd or acfm.
To better define direct and indirect costs, itemized capital cost data
' . ''-",.
were obtained from Exxon and Ogden Martin Systems, Inc. (the developer of the
Stanislaus County MWC plant in California that is equipped with Thermal
". -, ' ' '-.
DeNOv) for a 500 tpd plant consisting of two mass burn/waterwall
13":
eombtsstors. These two cost estimates are presented in Table 4-4. For
engineering equipment costs, the Ogden Martin costs are consistently higher.
The ammonia CEM and level of safety equipment included in the Ogden Martin
4-3
-------
TABLE 4-2. CAPITAL COSTS DATA FOR THERMAL DeNO APPLIED TO MASS BURN/WATERWALL MWC'S
Total Plant
Capacity, tpd
150
500
500
500
650
800
960
1 ,000
1 ,200
1,440
1,5006
3,000
Total Plant
Flue Gas
Flowrate
acfm
34,162
113,873
113,873
115,500b
148,035
182,197
218,636
227,746
273,295
332,970b
213,636
683,239
Direct Capital
Costs, $1,000
150
700
654
987
315
300
291
HA
NA
1,609
645 (674)
1,455
Indirect Capital
Costs, $1,000
220
270
351
917
682
450
221
NA
HA
1,136
295 (344)
3,147
Total
Capital Costs
Costs, $l,000a
375
970
1,010
1,900^ /'
997
750
512
960
2,660
2,745
840 (1,018)
4,600
License
Fee, $1,000
45
100
100
96
158
158
195
NA
HA
355
296
729
Design NO
Reduction, X
50
60
40
36
NAd
... "
36
65
NA
36
50
HA
Cost Basis,
Quarter (year)
4th (1987)
4th (1988)
3rd (1988)
3rd (1988)
3rd (1988)
3rd (1986)
3rd (1988)
3rd (1988)
1st (1988)
3rd (1988)
1st (1987)
3rd (1988)
Reference
3
3
4
5
6
6
3
7
8
9
: 3
8
Excludes the licensing fee.
Actual flue gas flowrate reported. Flue gas flowrate for the other plants was estimated assuming an excess air level of 80 percent and a
flue gas tesfserature leaving the combustor of 450 F.
Excludes reported costs of $600,000 for an amnonia slip continuous emission measurement system and ammonia safety equipment.
dHA -not available. .,'. ' ' .. ',. - , ' . ;/ ..., , ;-.-,.... .
Excludes the cost of an air compressor. The costs in parenthesis include the costs for an air compressor estimated using
References 10 to 12.
-------
TABLE 4-3. DIRECT AM3 INDIRECT CAPITAL COSTS DATA FOR THERMAL DeNO
Total Plant
Capacity, tpd
150
500
500
500
650
800
960
1,000
1,200
1,440
1,500
3,000
j^ Costs in December
<* ' bAt 450°F.
Total Plant
Flue Gas Flowrate,
_ o
acfm
34,162
113,873
113,873
115,500
148,035
182,197
218,636
227,746
273,295
332,970
341,619
683,239
1987 dollars.
Design NO
Reduction, *%
50
60
40
36,
NA1
55
36
65
NA
36
50
NA
Direct Capital
Costs >
$1,000
e
671f
631
9528
314
314e
281
NA
NA
1,550
704e
1,520
Indirect Capital
Costs
$1,000°
221
259
231
395h
532
471
213
NA
NA
459
359
3,139
Indirect/
Direct Cost
Ratio
1.46
0.39
0.33
0.42
1.69
1.50
0.76
NA
NA
0.30
0.51
2.17
Excludes contingency costs.
Ratio of indirect to direct capital costs.
Costs used to develop scaling factor for direct capital cost equation in Table 4-1.
Includes the cost of three 50 percent capacity air compressors. The other
plants have one 100 percent air compressor.
K '
Excludes NH. slip CEM and ammonia safety equipment costs presented in Table 4-4.
Excludes general and administrative expenses costs.
NA = not available.
-------
TABLE 4-4. CAPITAL COSTS FOR THERMAL DeNO
TAL COSTS FOR THERMAL DeNOv FOR TWO COMBUSTORS
AT 250 TPD EACH (December 1987 dollars}''5
1. Engineering Equipment Costs:
o Ammonia injection header
and nozzles
o Ammonia circulation heaters
o Air compressors
o Ammonia storage tank
o Ammonia safety equipment
o Ammonia slip CEM
o Electrical equipment
o Instrumentation and
control s
Total Engineering and
Equipment (1)
2. Direct Installation Costs:
o Earthwork and concrete
o Structural steel and
buildings
o Piping including valving
and supports
o Electrical and controls
o Equipment erection and
painting
Total Direct Installation
Costs (2)
Total Direct Costs (3) * (l)+(2)
3. Indirect Costs:
o Construction management,
indirects and fees
o Design engineering
o Exxon engineering
o General and administrative
expenses
Total Indirect Costs (4)
Exxon
11,600
4,050
93,500
21,900
N/AD
N/A
N/A
86,300
217,400
N/A
N/A
124,100
205,500
83,900
414,000
631,400
79,300
70,500
62,700
N/A
21 3 , 000
Ogden
Martin
103,000
7,700
152,400
24,100
289,400
289,400
31,000
151.000
1,048,000
67,000
58,000
173,000
145,000
41,500
484,000
1,532,000
82,000
217,000
96,000
256,000
651,000
Percent
Difference
790
90
63
10
-
-
74
f*
250°
- .
-
39. .
-30
-51
17
97b
3
210
54
. -
206
Continued
4-6
-------
TABLE 4-4 (CONCLUDED). CAPITAL COSTS FOR THERMAL DeNO- FOR TWO COMBUSTORS
"x.
AT 250 TPD EACH (DecemberAl987 dollars)
Exxon Licensing Fee (5)
Contingency (6)
Total Capita] Costs * (3)+
|4)*'(5)+(6)
Exxon
96,000
126,500
1,067,000
Ogden
Martin
92,600
233,400
2,510,000
Percent -
Difference
-4
85
135
Calculated as 100 * (Ogden Martin estimate - Exxon estimate)/Exxon estimate.
N/A = not applicable.
cExcludes ammonia.CEM costs.
4-7
-------
design are based on site-specific requirements and are not -expected to be ;
required in most Thermal DeNO systems. However, if these two items are s
. - - /\ . . .
excluded, the Exxon and Ogden Martin equipment costs are similar except for:
* ' ' - ' \
(a) ammonia injection header and nozzles, (b) air compressors, and j
(c) instrumentation and controls. As shown in Table 4-5, costs for items (a) )
': ' ' ' '.'' . - I
and (b) were compared to costs estimated from literature sources. Costs \
, i
estimated from literature sources for these two areas are comparable with f
Exxon and are lower than those provided by Ogden Martin. For instrumentation «;
and controls [item (c)], Exxon did not include automatic controls designed to \
.meet continuous NO emission limits. As shown in Table 4-4, the direct t
* ' ' f
installation costs as a percent of equipment costs are similar for both Exxon j
and Ogden Martin. . . {
To account for the differences in equipment cost estimates between Exxon j
and Ogden Martin, the following two-step approach was used to derive the ";
direct capital equation in Table 4-1. First, Exxon's direct capital costs /
presented in Table 4-4 were adjusted to include Ogden Martin's costs for t
' ' ' *J
instrumentation and controls, earthwork and concrete, and structural steel /
and buildings (Exxon costs did not include site preparation costs). A cost f
of $30,000 was also added to the direct capital cost for ammonia safety i
equipment consisting of water sprays and ambient ammonia monitoring. J
Instrumentation and control costs ($151,000) were assumed to be fixed; that <
is, they do not vary with combustor size. f
"i
Second, the direct capital costs of Thermal DeNOv excluding >
' A ' :."-. j
instrumentation and control costs were assumed to be related to the total
plant flue gas flowrate by the following equation:
DC = a (T_FLW)b (1) j
where: <
DC - direct capital costs, 1,000$ . j.
T_FLW = total plant flue gas flowrate, acfm \
a = coefficient !
b = scaling factor [
''
4-8
-------
TABLE 4-5. COST ANALYSIS RESULTS USING DETAILED COSTS FROM EXXON AND
06DEN MARTIN
1. Cost Comparison with Literature for Engineering Equipment
Ogden
Literature Exxon Martin
Ammonia injection header and nozzles 20,200a 11,600 103,000
"'" K " "
Air compressors 72,800 93,500 152,400
2. Indirect Cost as Percentage of Direct Costs and Contingency Cost as a
Percentage of the Direct and Indirect Costs
... ' . . .
Ogden
Exxon Martin
Indirect costs 33 42
Contingency 15 17
Extrapolated based on flue gas flowrate for a 500 MW coal-fired boiler
equipped with SCR using 0.6 costing rule. Costs include only the NH3/air
mixer and injection grid of NH3/air/flue gas. Cost data are from
Reference 14.
From Reference 15. Based on three 50 percent capacity industrial service
air compressors (Ingersoll-Rand Type 40 series) rated at 50 psig. [Note:
Exxon provided costs for air compressors based on three 50 percent capacity
compressors. Ogden Martin did not indicate the basis for their air
compressor costs].
cExcludes the costs for ammonia slip CEM, ammonia safety equipment, and
general and administrative expenses.
4-9
-------
Costs were adjusted to other plant sizes based on a scaling factor of 0.621. >,
This scaling factor was estimated from the Exxon direct capital cost !
estimates in Table 4-3 for Thermal DeNOx systems designed for 50 to -1
55 percent NO,, reduction. The other cost data in Table 4-3 were not used ^
x . .. . . *
because of differences in design bases and costing procedures. The <
coefficient, a, in Equation 1 was determined from the adjusted costs from
Step 1 above and the scaling factor. r,
Figure 4-1 presents the plot of the direct capital cost equation in j
Table 4-1 and the cost data from Table 4-3. As shown in the figure, the »
costs estimated by the equation are within the three cost data points for the "
500 tpd (115,000 acfm) plant size. However, the costs estimated by the (
equation are higher than most of the other reported costs. As discussed |
above, the cost equation is based primarily on the itemized direct cost data «:
provided by Exxon and Ogden Martin for a 500 tpd plant. Although no itemized }
cost data were provided for the other plants, it is believed that the lower f
costs for the other plants reflect system designs that did not include all of ]
the needed equipment and installation expenses. \
The cost equation in Table 4-1 is a based on flue gas flowrate instead {'
of waste throughput (tpd). Although tpd of refuse is a rough estimate of I
flue gas flowrate, it does not differentiate between mass burn and RDF t
combustors or differences in design excess air levels. To accommodate ;
short-term variations in feed waste composition and operating conditions, the c
flue gas flowrate used in the equation is based on 125 percent of the design f
flue gas flowrate. I
4.1.2 Indirect Costs !
'..'''. ' . . ... -,.-', 4
Indirect capital costs are typically a function of the direct capital |
costs. The indirect cost factor of 0.33 in Table.4-1 is based on the Exxon ^
data from Table 4-5. This factor corresponds to the Exxon cost data for the |
500 tpd plant. A startup cost of $10,000 for travel and supervision was |
added to the indirect costs since startup was not included in the indirect |
costs provided by Exxon. f
' t
4-10
-------
I-..
CO
CM
ID
O
O
d)
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Direct capital cost equation
D
n
Data used as the
basis for the
cost equation
n u
600
200 400
Flue Gas Flowrate, 103 acfm
Figure 4-1. Comparison^of the direct capital cost equation with the cost data.
-------
' . . ' " . -. - - , . ' i
4.1.3 Other Costs *
'. ' ' . <
A contingency based on 20 percent of the direct and indirect capital
estimated from guidelines developed in Reference 19. Total capital costs for r
NOV monitoring equipment is the incremental costs for NOV of a combined \\
on * V
NOX/S02/02 monitor (in December 1987 dollars). y
4.2 OPERATING COST PROCEDURE |
Table 4-7 presents the procedure for estimating annual operatingciosts t
for Thermal DeNOY. The total annualized operating costs include labor-related I
- - ' ^ ' ' '''' '
costs (operating, supervision, maintenance, and overhead), electricity, <'
ammonia consumption, operation and maintenance of the NOX monitor, and /
additional capital-related charges such as taxes, insurance, administration, j
and capital recovery. Operating costs for Thermal DeNO were obtained for j
'. -X . ,
the 12 mass burn/waterwal1 MWC facilities from data provided by Exxon and .
from other sources. The following four sections discuss the bases and (
t
rationale for the operating cost procedure. . ''''- I
(,
4.2.1 Labor and Maintenance i
Exxon indicated that Thermal DeNOx requires little additional \
maintenance and labor beyond that for thecombustors. For this reason, {
operating and maintenance Tabor costs were estimated using the smallest labor *
'. . - - - - . - 9V .-.... *
requirement (0.5 hour/shift) prescribed by EPA/CEIS. Supervision costs are ;
15 percent of the operating labor costs. These labor estimates are ;
' ' y\ ' "*
consistent with those estimated by others. /
-------
TABLE 4-6. COMPARISON OF ACTUAL AND PREDICTED LICENSE FEES3
Total
Total Plant Plant Flue 6ash Actual License Predicted License Percent
Capacity, tpd Flowrate, acfnT Fee, $1,000 Fee, $1,000C Error
V,
1
V
1"
r
V
V
1 - . ' -
f
i
>. ' '
V ;
£ . ' -
I
t
i
150
500
500
500
650
800
960
1,000
1,200
1,440
1,500
3,000
aln December
bAt 450°F.
34,162
113,873
113,873
115,500
148,035
182,197
218,636
227,746
273,295
332,970
341,619
683,239
1987 dollars.
45
96
96
93
165
166
188
NAe
NA
323
309
762
33
103
103
105
133
163
195
203
243
237
303
602
-27
7
7
11
-19
- 2
4
' '--- .
/
-27
- 2
-21
°Costs estimated from equation presented in Table 4-1.
Percent error = (Predicted-Actual) License Fee ,nn
Actual License Fee x 1UU
eNA = not available.
4-13
-------
TABLE 4-7. PROCEDURES FOR ESTIMATING ANNUAL OPERATING COSTS FOR THERMAL
riown flDPi TFH in NFU uur DI AMTca»D
DeNO APPLIED TO NEW MWC PLANTS
Operating Labor (Basis; 0.5 roan-hour/shif^ wage of $12/hr);
OL .- 0.75 * N * HRS
Supervision: 15% of the operating Tabor costs or 0.15 * OL
Mai ntenance Labor; (Basis: 0.5 man-hour/shift, 10% wage premium over the
operating labor wage)
MAINT - 0.825 * N * HRS or 1.1 * OL
Maintenance Materials: 2 percent of the sum of the direct capital, indirect \
capital, and process contingency 'j
Electricity; ELEC = (0,000391 * FLW + 0.963 * NH3) * N * HRS * ERATE j
Ammonia: AMM = NH3 * HRS * ARATE/2,000 \
NOX Monitoring; NOXM - 19,000 * N j
Overhead; 60% of all labor costs including maintenance materials }
Taxes. Insurance, and Adiini stratiye Charges:. 4% of the total capital cost /
excluding license fee and <'
monitors i
Capital Recovery (Basis: 15 year equipment life and 10% Interest rateV: . . ' <
13.15% of the total capital investment {
-'-'. . ..." ' - <
. - . 1 - .. - .- "--- ' - " ' T _ >,
uAll costs are in December 1987 dollars. <
OL "operating labor, $/yr t
N » number of combustors «
HRS » operating time at full rated capacity, hours/year (
MAINT = maintenance costs, $/yr (
ELEC = electricity costs, $/yr _ j
FLW » flue gas flowrate per combustor at 450 F, aefm (;
NH, « ammonia injection rpte, Ib/hr
3 - (0.015 +"0.0016, * NO₯R) * TPD * N *
." . ' ' - : A '-''.. "* - ""''-.
where: NO R = NO reduction, percent <;
TPD - comfiustor size, tpd ^
HHV= higher heating value for refuse, Btu/lb (defaults: f
4,595 for MSW, 8,552 for RDF, and 5,080 for cofired RDF . . \
with wood) \
NO .- NO emissions without Thermal DeNOx control, ppmv at |
x 7 percent 0? . f
ERATE - electrical power cost, $7kWh (default: $0.046/kWh) f
AMM = ammonia costs, $/yr *«
ARATE * ammonia cost rate, $/ton (default: $200/ton) 4
NO M = NO monitoring operating and maintenance costs, $/yr '«
" " " - - ' \
- '.-.'.'-: ' ' . 4-14 ';- .-- .; - . ;- .'' .-^.--: . . )
-------
Maintenance materials are estimated at 2 percent of the total capital
" ' 24 ' " '
costs excluding both the monitor costs and license fee. The maintenance
cost estimates shown in Table 4-7 do not include any costs for increased
maintenance of the boiler tubes from ammonia salt deposition that may be
caused by Thermal DeNO . It is assumed that based on design and operation
1" "
improvements gained from the initial Thermal DeNO facilities, the potential
. . - . A
of boiler tube fouling caused by ammonia salt deposition will be minimal.
Consequently, cleaning of the boiler tubes can be performed during normally
scheduled downtime periods. To be consistent with previous costing analysis
for this source category, operating and maintenance labor wages are $12/hr
25
and $13.20/hr (10 percent above $12/hr), respectively.
4.2.2 Electricity
The equation for estimating electricity costs (ELEC) is based on power
26
consumption data provided by Exxon and others, as shown in Table 4-8.
Electricity is consumed primarily by the ammonia vaporizer and the air
compressor. The electricity consumed by the ammonia vaporizer is directly
related to ammonia injection rate, and the electricity consumed by the air
compressor is proportional to the size of the combustor (i.e., flue gas
flowrate). The electrical power requirements presented in Table 4-8 were
linearly correlated with ammonia injection rate and flue gas flowrate,
resulting in the following equation:
EPOWER = °-000391 * FLW * N + 0.963-* NH3 * N (2)
where; r
EPOWER = slectri-cal power requirement, kW
FLW = flue gas flowrate per combustor at 450°F, acfm
NH3 = ammonia injection rate per combustor, Ib/hr
(see Equation 4)
N = number of combustors
Table 4-9 shows that, with the exception of the 150 tpd plant, Equation 1 is
within +40 percent of the data. Annual electricity cost (ELEC) is calculated
4-15
-------
TABLE 4-8. ELECTRICAL POWER AND AMMONIA CONSUMED BY THERMAL DeNO FOR SELECTED MWC PLANTS
Total Plant
Capacity, tpd
1,
1,
. 1,
1,
1,
3,
150
500
500
500
500
650
960
000
200
440
440
500
000
Flue Gas
Flowrate, acfm
34
113
113
115
115
148
218
227
273
332
332
683
il62
,873
,873
,500
,500
,035
,636
,746
,295
,970
,970
-
,239
NH, Injection
Rate, Ib/hr
(Ib/ton MSW)
25
55
(4
(2
.0)
6)
NAb (NA)
61
97;
t
.. HA*
71
110
218
214
335
97
354
(2
(4
(H
(1
(2
(4
(3
(5
(1
(2
9)
.7)
.8)
.6)
,4)
.6)
.6)
.6)
.8)
Electrical
Power, kU
38
155
NA
113
118
NA
110
171
353
360
360
54°
NA
Design NO
Reduction,**
50
60
40
36
36
NA
36
65
ISA
36
36
50
NA
References .
3
3 ' '
4
_5 ' .
5
6
' 3 -
7
a ' '.' '
9
9
3
6
8At 450°F.
HA « not available.
Power requirement for ammonia vaporization and heating only.
-------
TABLE 4-9.
COMPARISON OF ACTUAL
CONSUMED BY THERMAL
1
AND PREDICTED ELECTRICAL POWER
DeNO FOR SELECTED MWC PLANTS
A
Total Plant
Capacity, tpd
150
500
500
500
960
.1,000
1,200
1,440
. 1,440
1,500
Estimated using
Percent error =
Actual Electrical
Power, kW
38
155
113
118
110
171
353
360
360
54
Predicted Electrical
Power, kWa
37
98
104
139
154
195
317
336
453
93
Percent
Error
- 2
-37
- 8
17
40
14
-10
- 7
26
73
Equation 2.
(Predicted-ActualV Electrical Power lnn
Actual Electri cal Power
4-17
-------
by multiplying the above power requirement rate equation by the annual
operating hours and electricity price ($/kWh), as shown by the equation in
Table 4-7. The default electrical price (ERATE) used in Table 4-7 is
$0.046/kWh. This price was used in previous costing analyses for MWC's.27
4.2.3 Ammonia Consumption
The ammonia injection rate (NH3) was determined based on operating and
design parameters. The following equation (Equation 2) is derived for
estimating ammonia consumption expressed in terms of Ib NH3/ton HSH using
data reported in the compliance test for the Commerce MWC {presented in
Section 3.4) and data reported by Exxon for NO reductions of 36 to
oo X
65 percent (see Table 4-8):
NH,_T = [0.352 * 0.0385 * (NOyR)] * NOV * HHV
6 X 2li 4,595
where NH3_T = N.H3 injection rate, . Ib/ttin'MSN
NO R = NOV reduction, percent.
A . ' A . . "
NO = NO emissions without Thermal DeNO^ control, ppmv at
X X ' X '"."'..',
7 percent 0-.
HHV - higher heating value,of refuse, Btu/lb (this correction
factor (HHV/4,595) can be used to convert Ib NH3/ton MSW
to a Tb NH3/ton RDF or Ib NH3/ton cofired RDF using the
respective heating values for RDF and cofired ROF.)
Figure 4-2 presents the plot of the above equation and the data obtained by
Exxon and others.
From the data used to develop Equation 3, ammonia consumption ranges
from 1.8 Ib NH3/ton MSW at 36 percent reduction to 2.6 Ib NH3/ton MSr at.
65 percent NO reduction. Assuming an uncontrolled NOY emission level of
A" " - A ' .
213 ppm at 7 percent 02, the NH3-to-NOx stoichiometric ratio ranges from 1.4
to 2.2. Two data points at 50 percent NO reduction reported by Exxon were
A ' p -
excluded in developing Equation 2, because the reported ammonia Injection
rates at this NOX reduction were inconsistent with each other and with the
other data points. The large differences in ammonia consumption provided by
Exxon for both data points at 50 percent reduction were attributed to the
4-18
-------
w
i
A
.2
'p
o
10
I
z
5.5 -
5 -
4.5
4
3.5 -
3 -
2.5 -
2 -
1.5
30
D
D
D
D
Data used as the
basis for the NH3
injection equation
n
injection equation
40 50
Percent NOx Reduction
r-
60
70
Figure 4-2. Comparison of the NH3 injection equation with the contractor/vendor
data.
-------
differences in uncontrolled NO emissions used. By the same token, the
" . - "
ammonia consumption rates provided by Ogden Martin for achieving 36 percent
NOX removal for the 500 and 1,440 tpd plants were not considered, because the
high ammonia injection rates may lead to high NH3 slip. In addition, ammonia
consumption rate did not agree with the ammonia injection rates measured at
Commerce (2.0 Ib NHj/ton MSW at 45 percent N0X reduction and 2.7 lb NH3/ton J
MSW at 60 percent NO reduction). I
''"'' J
Equation 3 is based on normalizing uncontrolled NOW emissions to i
...'.-". ' - x ,-..- - -. . . <
213 ppmv at 7 percent. 09 .. Ammonia injection rate (NH,), expressed in Ib/hr, |
. . > -.. . -.?'... i
is calculated using Equation 4: '
NH3(lb/hr) = (0.015 + 0.0016 * NOXR) * TPO * N *|y^ * 4^595 (4) \
where N = number of combustors f
....-'. i
TPD = cpmbustor size, tpd |
HHV = higher heating value for the refuse* Btu/lb C
-.".- . . i
Annual ammonia costs (AMM), as shown in Table 4-7, are calculated by
multiplying Equation 3 by the annual hours of operation and the ammonia price
in dollars per ton. Based on contacts with ammonia producers and readily
available information, ammonia costs per ton across the country vary between
$90 and $230/ton.29"31
4.2.4 Other Costs
Operating and maintenance costs for the NOx "nltorlng equipment are the
incremental costs for NO of a combined NOSO/O monitor (in December 1987
dollars). Overhead and capital charges such as taxes, insurance, admini-
stration, and capital recovery are estimated using the same procedure used in
33 ' ' " ' ' ' ' '
previous costing analyses. Downtime costs are not included in the annual
operating costs. It is assumed that the operating experience of this tech-
nology gained from now to the time of the NSPS proposal (November 1989) will
result in little or no downtime costs.
4-20
-------
4.3 REFERENCES
1. Radian Corporation, lll(b) Model Plants Description and Cost Report.
Prepared for the U. S. Environmental Protection Agency. Research
Triangle Park, N.C. September 9, 1988. 92 p.
2. Radian Corporation. Municipal Waste Combustion Retrofit Study.
Prepared for the U. S. Environmental Protection Agency. Research
Triangle Park, N.C. September 15, 1988. 486 p.
3. Letter and enclosures from Krider, D. E., Exxon Research and Engineering
Company, to Martinez, J. A., Radian Corporation. January 11, 1989.
3 p. Thermal DeNO costs for MWC's.
. . A
4. Letter and enclosures from Fellows, W. D., Exxon Research and
Engineering Company, to White, D. M., Radian Corporation. February 17,
1989. 3 p. Thermal DeNO costs for a 500 ton per day MSW incineration
plant. x
5. Ogden Martin Systems, Pennsauken Resource Recovery Project: BACT
Assessment for Control of NO Emissions Top-Down Technology
Consideration. Fairfield, N. J. December 15, 1988. 103 p.
6. Letter and enclosures from Schubert, J. E., Commonwealth of Virginia,
Department of Air Pollution Control, to Johnston, M. G., EPA.
December 8, 1988. 17 p. Thermal DeNOv costs for MWC's.
' ' A
7. Letter and enclosures from Krider, D. E., Exxon Research and Engineering
Company, to White, D. M., Radian Corporation. November 21, 1988: 5 p.
Thermal DeNO costs for a 1,000 ton/day mass burn/waterwall plant.
/\
8. Letter and enclosures from Strobridge, D. E., Camp Dresser and McKee,
Inc., to Andrews, B., Florida Department of Environmental Regulation.
March 9^ 1988. 13 p. Pasco County, Florida Air Permit Application.
9. HDR Engineering, Inc., et al. Solid Waste Facility Permit Application
for the Union County Resource Recovery Project, Addendum IV, Volume III,
Top-Down BACT Analysis. Trenton, NJ. December 1988.
10. McCabe, W. L. and J. C. Smith. Unit Operations of Chemical Engineering:
3rd Edition. New York, McGraw-Hill. 1976. p. 201.
11. Hall, R. S., J. Matley, and K. J. McNaughton. Current Costs of Process
Equipment. Chemical Engineering (New York). 89:80-124. April 5, 1982.
12. Guthrie, K. M. Process Plant Estimating Evaluation and Control.
Solana, CA\ Craftsman Book Company of America. 1974. p. 167.
13. References 4 and 5.
4-21
-------
14. Maxwell, J. D., and L. R, Humphries. (Tennessee Valley Authority). . <
Economics of Nitrogen Oxides, Sulfur Oxides, and Ash Control Systems for
Coal-Fired Utility Power Plants. Prepared for the U. S. Environmental 1
Protection Agency. Washington, DC. Publication No. EPA-600/7-85-006. *
February 1985. j
i
15. Richardson Engineering Services, Inc. Process Plant Construction |
Estimating Standards: Volume 4 - Process Equipment. Mesa, AZ. 1988.
Account No. 100-251. pp. r to 3. (
16. Letter from Sedman, C. B., EPA, to Chang, J., Acurex Corporation. «
July 14, 1986. EPA guidelines for costing flue gas cleaning technology I
for MWC. /
17. Reference 6. ,'
18. Radian Corporation. Cost Procedures for Municipal Waste Combustion. t!
Prepared for the U. S. Environmental Protection Agency. Research 5
Triangle Park, N.C. September 30, 1988. (
19. Electric Power Research Institute. TAG - Technical Assessment Guide. I
Volume 1: Electricity Supply - 1986. Palo Alto, CA. Report
-------
31. Chemical Marketing Reporter - Chemical Profiles. October 3, 1988.
32. Reference 20.
33. Reference 18.
4-23
-------
-------
5.0 MODEL PLANT COSTS FOR THERMAL DENOX
This section presents the costs of Thermal DeNO for the 12 lll(b) model
plants. Table 5-1 presents key design information for the lll(b) model
plants. Table 5-2 presents plant specifications and flue gas composition
data for each model plant. Reference 1 describes the rationale in Selecting
these model plants and presents cosnbustor capital and operating costs (without
Thermal DeNOY) for each model plant. Procedures presented in Section 4.0 of
. A ,, ' .
this report were used to estimate the capital and operating costs of Thermal
DeNO for the 12 model plants. As presented in Section 3.4, Thermal DeNOY
A A
has been demonstrated to achieve 45 percent NO reduction. Therefore, Thermal
"
be:NOv costs are based on this NO reduction efficiency. Sections 5.1, 5.2,
X - A
5.3, 5.4, 5.5, and 5.6 present the Thermal DeNO costs for the mass burn/
A "
waterwall, mass burn/refractory, mass burn/rotary, refuse-derived fuel (RDF),
modular combustor, and the fluidized-bed combustor (FBC) model plants,
respectively. Also presented in each section are the annual N0₯ emission
A .
reductions (tons/year and Mg/year), cost effectiveness ($/ton and $/Mg), and
annual electrical consumption (MWh/year) for Thermal DeNOx for each model
plant. Section 5.7 summarizes Thermal DeNO costs, cost effectiveness, and
A :-.'.''
electrical requirements for each model plant.
Section 5.8 presents the results of the cost sensitivity analysis for
/thermal-DeNOY as a function of ammonia and electrical prices across the U.S.
A
This section also estimates the costs of Thermal DeNO for achieving 65
A ' - "
percent NOY emission reduction. The analysis was performed using the 800 tpd
A
mass burn/waterwal1 model plant and the 2,000 tpd RDF model plant.
5.1 MASS BURN/UATERWALL
Table 5-3 presents the capital costs for the 200, 800, and 2,250 tpd
mass burn/waterwal1 model plants. This table shows the combustor capital
costs as well as the itemized costs for Thermal DeNO . Thermal DeNOv capital
A A
costs range from $1,010,000 for the 200 tpd plant to $3,740,000 for the
2,250 tpd plant. The increase in total plant capital costs due to Thermal
DeNO ranges from 3.4 percent for the 2,250 tpd plant to 5.7 percent for the
#%
200 tpd plant.
5-1
-------
TABLE 5-1. MODEL PLANT SELECTION FOR lll(b)
I
rvs
Model
Plant
Number
-1
2
3
4
5
6
7
8
9
10
11
12
Unit size,
Combust or type (tpd)
Mass burn/waterwall
Mass burn/waterwall
Mass bum/waterwall
Mass burn/ refractory
Mass burn/ rotary
combustor (waterwall)
Refuse-derived fuel
Refuse-derived fuel
Modular excess air
Modular /starved air
Modular/starved air
Fluidized-bed
combustion (BFB)
Fluidizsd-bed
combustion (CFB)
100
400
750
250
350
500b
500b
120
25
50
450
450
Number of Total plant Annual operating Heat
combustors capacity, (tpd) hours recovery
2
2
3
2
3
4
4
2
2
2
2
2
200
800
2,250
500
1,050
2,000
2,000
240
50
100
900
900
5,000
8,000
; 8,000
8,000
8,000
8,000
8,000
8,000
5,000
8,000
8,000
8,000
Steam
Electricity
Electricity
Electricity
Electricity
Electricity
Electricity
Electricity
None
Electricity
Electricity
Electricity
Fuel
; 100X MSW
100X MSW
100% MSW
100Z MSW
10 OX MSW
100X RDF
50X RDF/
50* wood
100X MSW
100X MSW
100% MSK
100% RDF
100JS RDF
24 hr/day K 333 days/yr m 8,000 hr/yr
100 he/wk x 50 wk/yr = 5,000 hr/yr
Unit size represents RDF for Model Number 6 and represents combined RDF and wood for Modal Number 7.
SBFB = Bubbling Fluidized-bed; and CFB - Circulating fluidized-bed
-------
TABLE 5-2. MODEL PLAST SPECIFICATIONS AND FLUE GAS COMPOSITION DATA
in
Item
Facility Specification
No. of combustors per model
Total daUly charge rate, tpd
Annual operating hours
Ash content of feed waste, X
Excess combustion air, X of
theoretical
PM emission factor, X of
feed waste ash
Baseline PM emission rate,
gr/dscf :
Stack height, ft
Stack diameter, ft
Number of stacks
Flue Gas Data Per Combustor0
Volume flovrate:
dscfm
scfm
acfm
Outlet temperature F
Small
MB/WW
(No. 1)
2
200
5,000
22.2
80
10
0.08
140
4.0
2
11,500
13,300
22,800
450
Medium
MB/WW
(No. 2)
i
2
800
8,000
22.2
80
10A
0.05
200
6.0
2
46,000
53,100
91,100
450
Large
MB/UW
(No. 3)
3
2,250
8,000
22.2
80
10
0.05
230
7.0
3
86,200
99, 500
171,000
450
M3/REF
(No. 4)
2
500
8,000
22.2
200
10
0.08
150
9.0
2
48,100
52,500
90,200
450
MB/RC
(Ho. 5)
3
1,050
8,000
22.2
50
10
0.05
125
5v.O
3
33,400
39,600
68; 100
450
Model
RDF
(No. 6)
4
2,000
8,000
7.5
50
80
0.05
200
8.0
4
58,700
68,500
118,000
450
Plants*
8DF
(Coflred)
(No. 7)
4
2>000
8,000
4.3
50
80
0.05
200
8.0
4
52,000
62,600
107,000
450
MI/EA
(No. 8)
2
240
8,000
22.2
100
0.50
0.08
70
6.0
1
15, 300
17,500
30,000
450
MI/SA
(No Heat
Rec.) MI/SA FBC
(Ho. 9) (No. 10) (No. 11)
222
50 100 900
5,000 8,000 8,000
22.2 22.2 7.5
100 100 60
0.50 0.50 80
0.1 0.08 0.01
60 60
5.0 5.0
1 1 2
3,200 6,400 56,400
3,600 7,300 65,200
14,100 12,500 99,700
1,600 450 350
FBC
(No; 12)
2
900
8,000
7.5
60
80
0.01
2
56,400
65,200
99,700
350
Continued
-------
TABLE 5-2 (CONCLUDED). MODEL PLANT SPECIFICATIONS AND FLUE GAS COMPOSITION DATA7
Item
Emission Concentrations per
combustor at 7X 0 (dry) :
NO , ppmv
Particular Matter:
mg/dscm
(gr/dscf1
CO; ppmv
CDD/CDF, se/dscm
Acid gas:
HC1, ppnrr
S02> pfttfr
Annual Emissions per
d
combustor at 7Z 0 (dry):
NO , tons/i-r
PM, tons/r:
CO, tons IT"
CDD/CDF -(x 10~2), Ibs/yr
HC1, tons4Tr
SO , tons, yr
. . . _. .
Small
MB/WW
(No. 1)
213
4,600
(2)
50
200
500
200
72
408
5
3,56
69
50
- -^ -^ ; . -.
Medium
MB/WW
(No. 2)
213
4,600
(2)
50
200
500
200
463
2,610
33
22.8
439
320
Large
MB/WW
(No. 3)
213
4,600
(2)
50
200
500
200
1,300
4,890
62
42.8
823
601
1 MB/REF.
(No.. 4)
213
4,600
(2)
100
300
500
200
290
1,630
41
21.4
274
200
MB/RC
(Nov 5)
213
4,600
C2)
100
300
500
200
607
2,280
58
29.8
333
280
Model
RDF
(No. 6)
213
9,200
(4)
100
200
500
300
1,420
8,030
102
35.2
669
666
Plants
RDF
(Cofired)
(No. 7)
213
9,200
(4)
100
200
250
150
1,260
7,130
90
31.1
326
368
MI/EA
(No. 8)
213
4,600
(2)
100
200
. '
500
200
139
783
22
6.84
132
96
MI/SA.
(No Heat
Rec.)
(No. 9)
213
230
(0.1)
50
300
500
200
18
5
2
1.34
17
13
MI/SA
(No. 10)
213
230
(0.1)
50
300
500
200
58
16
4
4.28
55
40
FBC
(No. 11)
200
23
(0.01)
50
20
350
240
301
1.6.4
20.8-
3.16
457
379
FBC
(No; 12)
200
23
(0.01)
100
400
:"-'
350
240
301
16.4
42.0
63 . 3
457
379
MB/HW - aass burn/waterwall, MB/REF - mass bum/ refractory, MB/RC - mass burn/rotary coiabustor, MI/EA - raodular/exeess air, MI/SA - modular /starved air
.RDF - Eefvs* -derived fuel, and PBC - fluidised-bsd combustion.
From Repc-t to Congress, Publication No. EPA/530-SW-87-021e.
.Calculates based on the facility specifications in this table and the feed waste composition data from Table 1-3 in Reference 1.
' Emissions at combustor exit. Annual emissions from the stack,eacept for NO are included in Section 7.0_in Reference 1, At baseline, excluding Model Plant
No. 9, s-»ek emissions of PM are assumed to comply with the 0.05 gr/dsc£ or 0.08 gr/dscf limits as rsquirad by 40 CFR 60, Subparts Db or E. (Model Plant 9 is
smaller isan tha 50 tpd combustor size cutoff in Subpart E.) Baseline controls would not affect emissions of the other pollutants listed, and stack emissions
would be the same as listed above. Annual emissions for NO can be estimated from data in this section.
-------
TABLE 5-3. CAPITAL COSTS FOR THE MASS BURN/WATERWALL MODEL PLANTS -
NO. 1 TO 3 ($l,000's in December 1987)
Total Combustor Capital Cost
Thermal DeNQ.. Capital Cost
.X
Direct Cost
Indirect Cost
Process Contingency Cost
Licensing Fee
Preproduction
NOV Monitoring Equipment
No. 1
200 tpd
Plant
17,860
550
191
148
43
25
48
No. 2
800 tpd
Plant
50,000
1,090
371
293
163
50
No -i,3
2,250 tpd
Plant
110,000
1,940
651
519
452
98
72
Total Thermal DeNO Cost
1,010
2,020
3,740
Total Plant Capital Cost
Percent Cost Increase Attributed
to Thermal DeNO..
18,870 52,020 113,740
5.7
4.0
3.4
5-5
-------
Table 5-4 presents the annualized costs for the 200, 800, and 2;250 tpd <
mass burn/waterwall model plants. Thi? table shows the combustion annualized 1
y
costs as well as the itemized Thermal DeNO annualized costs. at 45 percent ]
*^> - .. ' ^
NO reduction. Annualized cpsts for Thermal DeNO range from $279,000 for \
*^ A j
the 200 tpd plant to $1,140,000 for the 2,250 tpd plant. The increase in ,|
total plant annualized costs attributed to Thermal DeNO ranges from ^
3.7 percent for the 2,250 tpd plant to 5.8 percent for the 200 tpd plant. £
Cost effectiveness compared to uncontrolled range from $2,150/Hg ($I,950/ton) (
for the 2,250 tpd plant to $9,450/Mg ($8,570/ton) for the 200 tpd plant. \
Table 5-4 also presents estimates of annual electrical requirements and £j
NO emission reductions for Thermal DeNO at each model plant. The ,4 {
r j
electrical requirements range from 173 MWh/yr for the 200 tpd plant to |
3,110 MWh/yr for the 2,250 tpd plant. Emission reductions of HQ corres- {
)
ponding to 45 percent NOX reduction range from 30 Mg/yr (33 tons/yr) for the 1
200 tpd plant to 531 Mg/yr (586 tons/yr) for the 2,250 tpd plant. The j
annualized costs, electrical requirements, and NO emission reductions are ^
based on 5*000 hours of operation for the 200 tpd plant and 8,000 hours of I
i
operation for the 800 and 2,250 tpd plants. I
5.2 MASS BURN/REFRACTORY [
Table 5-5 presents the capital costs for the 500 tpd mass <
burn/refractory model plant. This table shows the combustor capital costs as «
well as the itemized costs for Thermal DeNOv. Thermal DeNO capital costs j
.-- X . . « . - . ' j
are $2,010,000 for this plant. The increase in plant capital costs
attributed to Thermal DeNOv is 5.4 percent. 3
x ..... t
Table 5-6 presents the annualized costs for the 500 tpd mass ;
burn/refractory model plant. This table shows the combustor annualized costs *
as well as the itemized Thermal DeNOv annualized costs at 45 percent N0y ]
A - ' . ' "
reduction. Annualized costs for Thermal DeNOv are $549,000. The increase in *
. ' .' X ' .". ' i
total plant annualized costs attributed to Thermal DeNQx is 4.6 percent. ,
Cost effectiveness of removing NOY is $4,640/Hg ($4,210/ton). ;
« ..-.-:' .--.''.' ' C
Table 5-6 also presents estimates of annual electrical requirements and j
NO emission reductions for Thermal DeNOx at this plant. The electrical J
requirement for this plant is 899 MWh/yr. Emission reduction of NOX is {
5-6
-------
TABLE 5-4. ANNUALIZED COSTS, ECONOMIC AND ENVIRONMENTAL IMPACTS FOR THE MASS
BURN/WATERWALL MODEL PLANTS - NO. 1 TO 3
($l,000's in December 1987)
Combu stor Annual i zed Cost
Thermal DeNO., Cost
x. .
Direct Cost:
- Operating Labor
- Supervision
- Maintenance
- Electricity
- Ammonia
- NOX Monitoring Equipment
Total Direct Cost
Indirect Cost:
- Overhead
- Taxes, Insurance, and
Administration
- Capital Recovery
Total Indirect Cost
Total Annual i zed Cost
Total Plant Annual i zed Cost
Percent Cost Increase Attributed
to Thermal DeNO
X
NQH Reduction, tons/yr (Mg/yr)
A '
Cost Effectiveness, $/ton
($/Mg)
No. 1
200 tpd
Plant
4,850
8
1
26
8
9
_38
89
21
37
132
190
279
5,130
5.8
33(30)
8,570
(9,450)
No. 2
800 tpd
Plant
14,370
12
2
48
51
56
_38
207
37
72
265
374
582
14,950
4.1
208(189)
2,790
(3,080)
No. 3
2,250 tpd
PI ant
31,000
18
3
82
143
156
.57
459
62
1 28
491
681
1,140
32,140
3.7
586(531)
1,950
(2,150)
Electricity Use of Thermal
DeNOv, KWh/yr 173 1,110 3,110
5-7
-------
TABLE 5-5. CAPITAL COSTS FOR THE MASS BURN/REFRACTORY MODEL PLANT - NO. 4
($l,000's in December 1987)
500 tpd
Plant
.. . . . ,.. ..
Total Combustor Capital Cost 37 s 550
' .-.. ' ..-: - .: '-.-
Thermal DeNO.. Capital Cost
-
- A
Direct Cost
Indirect Cost
Process Contingency Cost
Licensing Fee
Preproduction
NO,, Monitoring Equipment
x
Total Thermal DeNOx Cost
Total Plant Capital Cost
Percent Cost Increase Attributed
to Thermal DeNO
1,090
369
291
161
47
2,010
39,560
5.4
i
5-8
-------
TABLE 5-6. ANNUALIZED COSTS, ECONOMIC AND ENVIRONMENTAL IMPACTS FOR THE
MASS BURN/REFRACTORY MODEL PLANT - NO. 4
($1,000's in December 1987)
500 tpd
Plant
Combustor Annualized Cost 11,870
Thermal DeNO^, Cost
Direct Cost:
- Operating Labor 12
- Supervision 2
- Maintenance 48
- Electricity 41
- Ammonia 35
- NO Monitoring Equipment 38
Total Direct Cost 176
Indirect Cost:
- Overhead 37
- Taxes, Insurance, and
Administration 72
- Capital Recovery 264
Total Indirect Cost 373
Total Annualized Cost 549
Total Plant AnnuaTized Cost 12,420
Percent Cost Increase Attributed
to Thermal DeNOv 4.6
*>
NOV Reduction, tons/yr (Mg/yr) 130(118)
A
Cost Effectiveness, $/ton 4,210
($/Mg) (4,640)
Electricity Use of Thermal
DeNO. MWh/yr 899
5-9
-------
118 Mg/yr (130 tons/yr). The annualized costs, electrical requirement, and '
NOY emission reduction are based on 8,000 hours of operation.
A . ' ' - : 1
1
5.3 MASS BURN/ROTARY COMBUSTOR '
Table 5-7 presents the capital costs for the 1,050 tpd mass burn/rotary
combustor model plant. This table shows the combustor capital costs as well *
as the itemized costs for Thermal DeNO . Thermal DeNO capital costs are /
^\ ' . A
$2,180,000 for this plant. The increase in plant capital costs attributed to i
Thermal DeNO is 3.2 percent. )
' i
Table 5-8 presents the annualized costs for the 1,050 tpd mass «,
burn/rotary combustor model plant. This table shows the combustor annualized I
costs as well as the itemized Thermal DeNOx annualized costs at 45 percent c,
NOV reduction. Annualized costs for Thermal DeNOv are $680,000 for this I
X ., ' ' , X "i
plant. The increase in total plant annualized costs attributed to Thermal ^
DeNCL is 3.5 percent. Cost effectiveness of removing NO is $2,740/Hg (
($2,490/ton). I
Table 5-8 also presents estimates of annual electrical requirements and <
NO emission reductions for Thermal DeNO at this plant. Electrical require- ->
A ' . . Jn. ' :-
ment for this plant is.1,340 MWh/yr. Emission reduction of NO is 248 Mg/yr I
. - « ''.-.-.." 't
(273 tons/yr). The annualized costs, electrical requirement, and NOX emission )
reduction are based on 8,000 hours of operation. (
' ' i'
5.4 REFUSE-DERIVED FUEL ' i
£
Table 5-9 presents the capital costs for the 2,000 tpd RDF and the j
2,000 tpd cofired RDF/wood model plants. This table shows the combustor ?
capital costs as well as the itemized costs for Thermal DeNO . Thermal '. . ]
s\
DeNOx capital costs are $3,570,000 for the 2,000 tpd RDF plant and $3,380,000 ?
for the 2,000 tpd cofired RDF plant. The capital costs for Thermal DeNO^ ;
increase the total plant capital costs by 2.6 percent for the 2,000 tpd RDF '{
plant and 2.4 percent for the 2,000 tpd cofired RDF plant. .;
Table 5-10 presents the annualized costs for the 2,000 tpd RDF and *
cofired RDF plants. This table shows,the combustor annualized costs,,as well <
as the itemized Thermal DeNOw annualized costs at 45 percent NO reduction. I
x * "
5-10
-------
TABLE 5-7. CAPITAL COSTS FOR THE MASS BURN/ROTARY COMBUSTOR MODEL PLANT
NO. 5 ($1,GOO's in December 1987)
1,050 tpd
Plant
Total Combustor Capital Cost 69,140
Thermal DeNO Capital Cost
A
Direct Cost 1,160
Indirect Cost 394
Process Contingency Cost 311
Licensing Fee 182
Preproduction 56
NO Monitoring Equipment 72
A
Total Thermal DeNO Cost 2,180
A
Total Plant Capital Cost 71,320
Percent Cost Increase
Attributed to Thermal DeNO,. 3.2
5-11
-------
TABLE 5-8. ANNUALIZED COSTS, ECONOMIC AND ENVIRONMENTAL IMPACTS FOR THE
MASS BURN/ROTARY COMBUSTOR MODEL PLANT - NO. 5
($l,000's in December 1987)
1,050 tpd
Combustor Annual! zed Cost 19,520
Thermal DeNO.. Cost
~~ X '- . . '"' '-,,
Direct Cost:
-'' . . '
- Operating Labor 18
- Supervision 3
- Maintenance 57
- Electricity 62
- Ammonia 73
- NO Monitoring Equipment _ 57
Total Direct Cost 270
Indirect Cost:
- Overhead 47
- Taxes, Insurance, and
Administration 77
- Capital Recovery 286
Total Indirect Cost 410
Total Annualized Cost 680
Total Plant Annualized Cost 20,200
Percent Cost Increase Attributed
to Thermal DeNOy 3-5
" .
NOV Reduction, tons/yr (Mg/yr) 273(248)
*v
Cost Effectiveness, $/ton 2,490
($/Mg) (2,740)
Electricity Use of Thermal
DeNOY, MWh/yr 3
/* , . ' '.'-.': - - ; . '''
5-12
-------
TABLE 5-9. CAPITAL COSTS FOR THE REFUSE-DERIVED FUEL FIRED MODEL PLANTS
NO. 6 AND 7 ($l,000's in December 1987)
No. 6
2,000 tpd
Plant
No. 7
2,000 tpd
Cofired Plant
Total Combustor Capital Cost
135,000
143,800
Thermal DeNO Capital Cost
J\
Direct Cost
Indirect Cost
Process Contingency Cost
Licensing Fee
Preproduction
NOY Monitoring Equipment
A.
Total Thermal DeNOx Cost
Total Plant Capital Cost
Percent Cost Increase Attributed
to Thermal DeNO.,
1,850
620
494
415
97
96
3,570
138^570
2.6
1,760
590
469
380
92
96
3,380
147,180
2.4
5-13
-------
TABLE 5-10. ANNUALIZED COSTS, ECONOMIC AND ENVIRONMENTAL IMPACTS FOR THE
REFUSE-DERIVED FUEL FIRED MODEL PLANTS - NO. 6 AND 7
($l,000's in December 1987)
Combustor Annualized Cost
Thermal DeNO., Cost
X
Direct Cost:
- Operating Labor
- Supervision
- Maintenance
- Electricity
- Ammonia
- NO,, Monitoring Equipment
X
Total Direct Cost
Indirect Cost:
- Overhead
- Taxes, Insurance, and
Administration
- Capital Recovery
Total Indirect Cost
Total Annualized Cost
Total Plant Annualized Cost
Percent Cost Increase Attributed
to Thermal DeNOx
NOV Reduction, tons/yr (Mg/yr)
X
Cost Effectiveness, $/ton
i ($/Mg)
Electricity Use of Thermal
DeNO . MWh/yr
x .
.
No. 6
2,000 tpd
Plant
33,200
24
4
85
142
168
76
499
68
122
470
660
1,160
34,360
3.5
541(582)
1,810
(1,990)
3,090
5-14
No. 7 «
2,000 tpd ;
Cofired Plant J
<
35,070 *
\
f
24 .. ' . :
4 : . ^
82
130
154
76. .- .. ;
470 I
. , . j
66-
116
445
627
1,100
36,170
£
3.1
569(516)
1,930
(2,130)
2,820
-------
Annualized costs for Thermal DeNOY are $1,160,000 for the 2,000 tpd RDF plant
. ' ' ' A - . --_'..-.
and $1,100,000 for the 2,000 tpd cofired RDF plant. The respective increases
in total plant annualized costs attributed to Thermal DeNO are 3.5 and
3.1 percent. Cost effectiveness is $l,990/Mg ($l,810/ton) for the 2,000 tpd
RDF plant and $2,130/Mg ($l,930/ton) for the 2,000 tpd cofired RDF plant.
Table 5-10 also presents estimates of annual electrical requirements and
NO. emission reductions for Thermal DeNO at each model plant. The
' ' r A ';. . A '
electrical requirements are 3,090 and 2,820 MWh/yr for the 2,000 tpd» RDF and
25000 tpd cofired RDF plants, respectively. Emission reductions of NO are
582 Mg/yr (641 tons/yr) for the 2,000 tpd RDF plant and 516 Mg/yr
(569 tons/yr) for the 2,000 tpd cofired RDF plant. Uncontrolled NO emissions
' . ' . ' * A
in terms of ppm at 7 percent 02 are about the same for RDF and wood/RDF
firing, since NO emissions from wood firing alone are about the same as MWC
..'' A .
firing. The anhualized cost, electrical requirements, and NO emission
reductions are based on 8,000 hours of operation for both plants.
5.5 MODULAR COMBUSTORS
Table 5-11 presents the capital costs for the 240 tpd modular excess
'".. "- -. - - .
air, the 50 tpd modular starved air, and the 100 tpd modular starved air
model plants. This table shows the combustor capital costs as well as the
itemized costs for Thermal DeNO . Thermal DeNO capital costs range from
$616,000 for the 50 tpd modular starved air plant to $1,140,000 for .the
240 tpd modular excess air plant. The increase in total plant capital costs
due to Thermal DeNO ranges from 8.7 percent for the 240 tpd plant to
" . /\ x
49 percent for the 50 tpd plant.
Table 5-12 presents the annualized costs for the three modular plants.
This table shows the combustor annualized costs as well as the itemized
Thermal DeNO annualized costs at 45 percent NO reduction. Annualized costs
A X
for Thermal DeNOx range from $190,000 for the SO tpd plant to $337,000 for
the 240 tpd plant. The increases in total plant annualized costs attributed
to Thermal DeNOY range from 7.7 percent for the 240 tpd plant to 31 percent
..'"' . .
for the 50 tpd plant. Cost effectiveness range from $5,950/Mg ($5,400/ton)
for the 240 tpd plant to $25,700/Mg ($23,300/ton) for the 50 tpd plant.
5-15
-------
TABLE 5-11. CAPITAL COSTS FOR THE MODULAR MODEL PLANTS - NO. 8 JO 10
($l,000's In December 1987)
Total Combustor Capita] Cost
Thermal DeNO.. Capital Cost
No. 8
240 tpd
Excess Air
No. 9
No. 10
50 tpd 100 tpd
Starved Air Starved Air
13,150
1,270
5,510
Direct Cost
Indirect Cost
Process Contingency Cost
Licensing Fee
Preproduction
NOY Monitoring Equipment
J\
Total Thermal DeNOY Cost
A
Total Plant Capital Cost
Percent Cost Increase Attributed
to Thermal DeNOv
X
624
216
168
56
27
48
1,140
14,290
8.7
330
119
90
14
15
.48.
616
1,890
48.5
426
150
115
25
19
48
' " 783
6,290
14.2
5-16
-------
TABLE 5-12.
ANNUALIZED COSTS, ECONOMIC AND ENVIRONMENTAL IMPACTS FOR THE
MODULAR MODEL PLANTS - NO. 8 TO 10 -
($l,000's in December 1987) ?
Combustor Annual i zed Cost
Thermal DeNO., Cost
X
Direct Cost:
- Operating Labor
- Supervision
- Maintenance
; - Electricity
- Ammonia
- NOX Monitoring Equipment
Total Direct Cost
Indirect Cost:
- Overhead
- Taxes, Insurance, and
Administration
. - Capital Recovery
Total Indirect Cost
Total Annual i zed Cost
Total Plant Annual i zed Cost
Percent Cost Increase Attributed
to Thermal DeNOv
X
NO Reduction, tons/yr (Mg/yr)
Cost Effectiveness, $/ton
($/Mg)
Electricity Use of Thermal
DeNO. MWh/yr
in
No. 8
240 tpd
Excess Air
4,360
12
2
33
16
17
^8
118
28
41
150
219
337
4,700
7.7
63(57)
5,400
(5,950)
348
No. 9
50 tpd
Starved Air
605
8
1
19
2
2
38
70
17
22
81
120
190
795
31.4
8.2(7.4)
23,300
(25,700)
45
- '',' -''
No. 10
100 tpd
Starved Air
1,830
12
2
27
7
-.". 7
*;> 38
92
24
28
103
155
248
2,080
13.6
26(24)
9,530
(10,500)
145
5-17
-------
Table 5-12 also presents estimates of annual electrical requirements *
and NOX emission reductions for Thermal DeNOx at each model plant. The
electrical requirements range from 45 MWh/yr for the 50 tpd plant to ]
348 MWh/yr for the 240 tpd plant. Emission reductions of NO ranged from ' i
7 Mg/yr (8 tons/yr) for the 50 tpd plant to 57 Mg/yr (63 tons/yr) for the \
240 tpd plant. The annualized costs, electrical requirements, and NO ; I
emission reductions are based on 5,000 hours of operation for the 50 tpd }
modular starved air plant and 8,000 hours of operation for the other two *
plants. \"' ' ~\
5.6 FLUIDIZED-BED COMBUSTION 1
Table 5-13 presents the capital costs for the 900 tpd bubbling bed and 1
the 900 tpd circulating bed model plants. This table shows the combustor ^
capital costs as well as the itemized costs for thermal deNO . Thermal *
' ' ' . . . X '. '.
DeNO capital costs are $2,270,000 for each model plant. The increase or f
*\ " - ' . i
total plant capital costs due to thermal deNO is 3.1 percent. *V
A . vt
Table 5-14 presents the annualized costs for both plants. This table f
shows the combustor annualized costs as well as the itemized thermal deNO
annualized costs at 45 percent NOX reduction. Annualized costs for'thermal
deNOv is $658,000 for each plant. The increase in total plant annualized
- , . ..- jr
costs attributed to thermal deNO is 3.4 percent. Cost effectiveness is
$2,670/Mg ($2,430/ton) for each plant.
Table 5-14 also presents estimates of annual electrical requirements anc
NOY emission reductions for thermal deNOv at each plant. The electrical
A X
requirement is 1,380 MWh/yr for each plant. Emission reduction of N0x:is
246 Mg/yr (271 tons/yr). The annualized costs, electrical requirements, and
NOY emission reductions are based on 8,000 hours of operation for both
A . '
plants.
5.7 SUMMARY OF NOX EMISSION REDUCTION, COST EFFECTIVENESS, AND
ELECTRICAL REQUIREMENTS
Table 5-15 summarizes the information on NOV emission reductions,
". .. '..-.'' X x* .
capital costs, annualized costs, cost effectiveness, and electrical require-
ments for the 12 model plants. Also shown are annual tonnages of waste
combusted by each model plgant.
5-18
X \
/
-------
TABLE 5-13. CAPITAL COSTS FOR THE FLUIDIZED BED COMBUSTION MODEL PLANTS
NO. 11 AND 12 ($l,000's in December 1987)
No. 11 No. 12
900 tpd 900 tpd
Bubbling Circulating
Bed Plant Bed Plant
Total Combustor Capital Cost 73,870 73,870
Thermal DeNOx Capital Cost
Direct Cost 1,220 1,220
Indirect Cost 413 413
Process Contingency Cost 327 327
Licensing Fee 199 199
Preproduction 57 57
NOY Monitoring Equipment 48 48
A
Total Thermal DeNO₯ Cost 2,270 2,270
"
Total Plant Capital Cost 76,140 76,140
Percent Cost Increase Attributed
to Thermal DeW) 3.1 3.1
5-19
-------
TABLE 5-14. ANNUALIZED COSTS, ECONOMIC,AND ENVIRONMENTAL IMPACTS FOR THE
FLUIDIZED BED COMBUSTION MODEL PLANTS - NO. 11 AND 12
($l,000's in December 1987)
Combustor Annual i zed Cost
Thermal DeNO., Cost
A
Direct Cost:
- Operating Labor
- Supervision
- Maintenance
- Electricity
- Ammonia
- NO Monitoring Equipment
Total Direct Cost
Indirect Cost:
- Overhead
- Taxes, Insurance, and
Administration
- Capital Recovery
Total Indirect Cost
Total Annual i zed Cost
Total Plant Annual i zed Cost
Percent Cost Increase Attributed
to Thermal DeNOv
X
NO Reduction, tons/yr (Mg/yr)
/\ .
Cost Effectiveness, $/ton
($/Mg)
Electricity Use of Thermal
DeNOY, MWh/yr
A . ' --.'"
No. 11
900 tpd
Bubbling
Bed Plant
19,300
12
2
52
64
71
_3J
239
40
81
298
419
658
19,960
3.4
271(246)
2,430
(2,670)
1,380
No. 12
900 tpd
Circulating
Bed Plant
19,300
12
2
52
64
71
_38
239
40
81
298;
4 19
(58
19,960
3.4
271(246)
2,430
(2,670)
1,380
5-20
-------
TABLE 5-15. SUMMARY OF COSTS, COST EFFECTIVENESS, AND ELECTRICAL REQUIREMENTS FOR NEW MWC MODEL PLANTS USING THERMAL DeNO
tn
I
f\J
No.
1
2
3
It
5
6
7
8
9
10
11
12
Model Plant
Type3
MB/WW
MB/WW
MB/WW
MB/REF
MB/RC
RDF
RDF (Cofired)
MI/EA
MI/SA
MI/SA
FBC
FBC
TPYb
41,700
267,000
750,000
167,000
350,000
667,000
667,000
80,000
10,400
20,800
300,000
300,000
NO
Emiss Ion Keduet ion
tons/yr (Mg/yr)
33
208
586
130
273
641
569
63
8.2
26
271
271
(30)
(189)
(531)
(118)
(248)
(582)
(516)
(57)
(7.4)
(24)
(246)
(246)
Thermal DeNO
Capital
Costs, $1,000
1,010
2,020
3,740
2,010
2,180
3,570
3,380
1,140
616
783
2,270
2,270
Annual
Thermal DeNO Cost Electrical
Ahnualized Effectiveness Requirements,
Costa, $1,000 $/ton ($/Mg) MWh/yr
279
582
1,140
549
680
1,160
1,100
337
190
243
658
658
8,570
2,790
1 ,950
4,210
2,490
1,810
1,930
5,400
23,300
9,530
2,430
2,430
(9,450)
( 3,080)
( 2,150)
( 4,640)
( 2,740)
( 1,990)
( 2,130)
( 5,950)
(25,700)
(10,500)
( 2,670)
( 2,670)
173
1,110
3,110
899
1,340
3,090
2,820
348
45
145
1,380
1,380
aMB/WW = mass burn/waterwall
MB/REF = mass burn/refractory
MB/RC = mass burn/rotary combustor
RDF = refuse-derived fuel
MI/EA = modular incinerator/excess air
MI/SA = modular incinerator/starved air
TPY == tons per year of refuse
-------
5.8 COST SENSITIVITY ANALYSIS
This section presents the variations in costs and cost effectiveness of
Thermal DeNOV with changes in ammonia and electrical power costs. Costs of
A - '...-' . -
anhydrous ammonia ($/ton) and electrical power ($/kWh) can vary widely across
the country. A survey of anhydrous ammonia and electrical power costs across
the country indicates that ammonia costs range between $70 and $230/ton and
electricity costs range between $0.0275 and $0.08/kWh.2"5
The sensitivity of Thermal DeNO costs to regional ammonia and «
electricity prices was estimated for two model plants. The 2,000 tpci RDF
plant was selected, since this plant had the highest annualized costs and
lowest cost effectiveness of the model plants evaluated in Sections 5.1 to
5.7. The other model plant selected was the 800 tpd mass burn/waterwall
plant. This plant was the smallest plant size with a cost effectiveness of
near $3,000/ton or less. The ammonia price was varied from a baseline cost
of $200/ton, which was used to cost the model plants in Sections 5.1 to 5.7,
to $100 and $400/ton. The results of jthrs analysis are presented in
Section 5.8.1. Electricity price was also varied from $0.046/kWh, which was
used in Sections 5.1 to 5.7, to $0.0275 and $0.08/kWh. The ammonia pfice used
when varying the electricity prices was $200/ton. The results of varying the
' "'..-... . "
electricity prices are presented in Section 5.8.2.
In addition, costs and cost effectiveness of Thermal DeNO at 60 percent
1 " A
NO reduction are reported in Section 5.8.3 for both model plants. Ammonia
J\ _. . .
and electrical prices were the same as used previously in Sections 5.1 to 5.7
(i.e., $200/ton for ammonia and $0.046/kWh for electricity), i
5.8.1 Ammonia Price Variation j
Table 5-16 presents the impacts of varying ammonia prices ($100/ton and
$400/ton) on Thermal DeNOv annualized costs and cost effectiveness for the
/\
800 tpd mass burn/waterwall model plant and the 2,000 tpd RDF model plant.
As shown in this table, the cost and cost effectiveness of Thermal DeNO are
insensitive to the ammonia price variations. A 50 percent decrease in the
ammonia price (from $200 to $100/ton) results in a small decrease ins
annualized costs and cost effectiveness (up to 8 percent) for both model
5-22
-------
TABLE 5*16. IMPACTS OF VARYING AMfJQHIA PRICE ($/TOH) OH THERMAL DeHO ANNUALI2ED COST ATO COST EFFECTIVENESS
Combustor
Type*
MB/WW
RDF
Combustor Ammonia
Size, tpd Price, $/ ton
800 200°
400
100
2,000 200°
400
100
Annual Ized
Cost, $1,000
582
638
553
1,160
1,330
1,070
Cost
Effectiveness
$/ton
2,790
3,060
2,660
1,810
2,080
1,680
b
Percent Change
Aianonla
Price
_
100
-50
.
100
-50
Annual Ized
Cost
_
10
-5
.
15
-8
Cost
Ef f ect Ivenes a
_
10
-5
.
15
-8
tn
ro
to
MB/WW m mass burn/waterwall
RDF = refuse-derived fuel
Percent change Is calculated from the cost results at $200/ton for ammonia.
Used to estimate model plant costs In Sections 5.1 to 5.7
-------
plants. The respective annualized costs and cost effectiveness based on j
$100/ton for ammonia are $553,000 and $2,660/ton for the 800 tpd plant and {
1,070,000 and $l,680/ton for the 2,000 tpd RDF plant. Similarly, a ]
100 percent increase in ammonia price (from $200 to ;$400/.ton) results in a 1
small increase in annualized costs and cost effectiveness (up to 15 percent) j
for both plants. The respective annualized costs and cost effectiveness J
based on $400/ton for ammonia are $638,000 and $3,060/ton for the 800 tpd ]
plant and $1,330,000 and $2,080/ton for the 2,000 tpd plant. 1
5.8.2 Electricity Price Variation ' ' \
. ' r ' X
Table 5-17 presents the impacts of varying electricity prices on Thermal j
DeNO annualized costs and cost effectiveness for the 800 tpd mass hum/ )
: /% - - -;.'-. . :
waterwall model plant and the 2,000 tpd RDF model plant. Thermal DeNO 4
,: - .,... ^ J
annualized costs are estimated based on electricity prices of $0.046, j
$0.0275, and $0.080/kWh. As shown in this table, the coslts and cost I
effectiveness of Thermal DeNOv are relatively insensitive to the electricity j
. ' !.''' i
price variation seen across the country. A large change in electricity 1
. ' - ;' '"."- , &
prices (up to 74 percent) results in a small change in annualized costs and }>
cost effectiveness (up to 9 percent) for both model plants. The respective \
' ' ^
annualized costs and cost effectiveness based on $0.0275/kWh are $561,000 j
and $2,690/ton for the 800 tpd plant and $1,100,000 arid $l,72p/ton for the |
2,000 tpd RDF plant. Similarly, the respective annualized costs and cost J
effectiveness based on $0.08/kWh are $620,000 and $2,980/ton for the 800 tpd |
plant and $1,270,000 and $l,980/ton for the 2,000 tpd plant. I
! . ^'!
5.8.3 NO.. Reduction Variation
'
Table 5-18 presents the annualized costs and the cos^t effectiveness for
al DeNO at 60 percent NO reduction for both the 800 tpd mass burn/
X X: . ' .,..---
waterwall model plant and the 2,000 tpd RDF model plant. The cost results at
60 percent
this table.
Thermal DeNO at 60 percent NO reduction for both the 800 tpd mass burn/
X X: . ' .,..---
del plant and the 2
60 percent NOV reduction are compared to those at 45 percent NO reduction in
A . . A .
5-24
-------
TABLE 5-17. IMPACTS OF VARYING ELECTRICITY PRICE ($/KWH) OH THESMAL DeSO ANNUALIZED COST AM) COST EFFECTIVENESS
uv
tn
Combustor
Type*
MB/WW
RDF
Cocnbustor Electricity
Size, tpd Price, $/kWh
800 0.046b
0.0275
0.08
2,000 0.046b
0.0275
0.08
Annual ized
Cost, $1,000
582
561
620
1,160
1,100
1,270
Cost
Effectiveness
$/ton
2,790
2,690
2,980
1,810
1,720
1,980
Percent Change
Electricity
Price
.
-40
74
-40
74
Armualized
Cost
.
-4
7
-5
9
Cost
Effectiveness
..
-4
7
-5
9
9MB/WW - mass burn/waterwall
RDF = refuse-derived fuel
Percent change is calculated from the cost results at $0.046/kWh for electricity.
Used to estimate model plant costs in Sections 5.1 to 5.7
-------
TABLE 5-18. THERMAL DeNO ANNUALIZED COSTS AND COST EFFECTIVENESS AT 45 AND 60 PERCENT NO REDUCTION
Cost Percent Change
Combustor
Type*
MB/WW
RDF
Combustor Percent NO Annual ized
Size, tpd Reduction Cost, $1,000
800 45b 582
60 604°
b
2,000 45 1,160
60 1,230°
Effectiveness Annualized
S/ton Cost
2,790 -
2,180 4
1,810
1,440 6
Cost
Effectiveness
-22
_
-20
SMB/WW = mass burn/waterwall
T
I
._ RDF » refuse-derived fuel
Used to estimate model plant costs in Sections 5.1 to 5.7
°Costs do not include the capital expense of Combustor modifications to improve the gas residence time and
mixing of ammonia with the flue gas for achieving 60 percent NO reduction.
-------
The annualized cost at 60. percent NO reduction is $604,000 for the
800 tpd plant and $1,230,000 for the 2,000 tpd RDF plant. The increase in
annualized costs over those at 45 percent NOX reduction is 4 percent for the
800 tpd plant and 6 percent for the 2,000 tpd RDF plant. The cost increase
at 60 percent NO reduction includes higher costs for ammonia and
^ .
electricity, but does not include the capital expense of combustor
modifications to increase flue gas residence time and mixing needed to
achieve this NO reduction level.
The cost effectiveness at 60 percent NO reduction is $2,180 and
" -
$l,440/ton for the 800 tpd and 2,000 tpd model plants, respectively. Cost
effectiveness decreases by roughly 21 percent from those at 45 percent NO
reduction for both plants.
5-27
-------
5.9 REFERENCES
1. Radian Corporation, 1
Prepared for the U. S
Triangle Park, N.C.
2. Chemical Marketing Reporter. Volume 233. Number 1. January 4, 1988
fah '
of anhydrous ammonia.
. Number 14. October 3, ,988.
5' vo«79y *nfoation ^ministration. Typical t
1987. -n«M.f^* of Energy> pUbT1ca|?0;a;^
5-28
------- |