United States
Environmental Protection
Agency
EPA-600/7-83-042
August 1983	
Research  and
Development
EVALUATION OF
COMBUSTION MODIFICATION EFFECTS
ON EMISSIONS AND EFFICIENCY OF
WOOD-FIRED INDUSTRIAL BOILERS
Prepared for
Office of Air Quality Planning and Standards
Prepared by
Industrial Environmental Research
Laboratory
Research Triangle Park NIC 27711

-------
                 RESEARCH REPORTING SERIES


Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination  of traditional  grouping  was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

    1. Environmental Health Effects Research

    2. Environmental Protection Technology

    3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    6. Scientific and Technical Assessment Reports (STAR)

    7. Interagency Energy-Environment Research and Development

    8. "Special" Reports

    9. Miscellaneous Reports

This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND  DEVELOPMENT series. Reports in this series result from the
effort funded  under  the 17-agency Federal  Energy/Environment  Research  and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations  include analy-
ses of the transport  of energy-related pollutants and their health and ecological
effects;  assessments of, and development of, control technologies  for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
                        EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for  publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------

-------
                                   ABSTRACT

        Results of full scale tests  to  evaluate  combustion  modifications  for
emission control and efficiency  enhancement  on two  wood-fired industrial
boilers are reported.  These modifications consisted of  lower excess  air  and
variations in the overfire air system operation.
        The boiler at Location 3 is  fueled with  a combination of wood bark and
coal.  The implementation of lower excess air reduced NOV emissions by
                                                         A
37.2 percent and improved thermal efficiency by  1.2 percent.   Variations  in
the overfire air system reduced  NO   by  20.7  percent and  improved efficiency by
1 .6 percent.  The combination of lower  excess air and overfire air system
modification reduced NO  by  18.5 percent and improved efficiency by
0.9 percent.  A 51 percent load  reduction produced  only  a 3.7 percent NO
reduction and a 4.0 percent  loss in  efficiency.
        The boiler at Location 5 uses hogged wood as the primary fuel and oil
as the supplemental fuel.  The effectiveness of  lower excess  air in reducing
NO., was 12.5 percent with a  slight improvement in efficiency  (0.6 percent).
  X
Adjustment of the auxiliary  air  dampers produced a  17.2  percent NOX reduction
and a 1.7 percent improvement in efficiency.  Polycyclic organic matter (POM)
sampling was performed at both baseline and  optimum low-NO  conditions.  On a
yg/m  basis the POM for low-NO   conditions exceeded the  baseline results  by a
factor of two to three.  The results obtained are compared  to previous
sampling on industrial steam boilers.
                                       ii                       KVB72-806015-1308

-------
                                    CONTENTS


Section                                                                    Page

        ABSTRACT                                                           ii

1.0     INTRODUCTION AND  SUMMARY                                           1-1

2.0     TEST EQUIPMENT DESCRIPTION                                         2-1

        2.1   Emissions Sampling  Equipment                                2-1

3.0     WOOD BARK/COAL BOILER,  LOCATION 3                                  3-1

        3.1   Boiler Description                                           3-1

        3.2   Emissions Sampling                                           3-4

        3.3   Baseline Tests                                               3-4

        3.4   Combustion  Modifications                                     3-11

4.0     HOGGED  FUEL BOILER, LOCATION 5                                     4-1

        4.1   Boiler Description                                           4-1

        4.2   Fuel Description                                             4-3

        4.3   Modifications and Tests Performed                           4-5

        4.4   Polycyclic  Organic  Matter (POM)  Sampling                    4-22

5.0     REFERENCES                                                         5-1

6.0     CONVERSION FACTORS                                                 6-1
APPENDIX:
        A.    GASEOUS  AND  PARTICULATE EMISSIONS TEST METHODS
              AND  INSTRUMENTATION
                                       ill
                                                                KVB72-806015-1308

-------
                                    FIGURES
Figure                                                                     Page
3-1     Two-drum Stirling boiler for suspension  and  thin-bed  burning      3-2
        of wood
3-2     Air-swept distributor spout for spreader stoker                    3-3
3-3     NO emissions as a function of G>2 before  overfire  nozzle           3-9
        modification
3-4     Aerodynamic particle diameter - baseline conditions                3-10
3-5     NO emissions as a function of load  for a wood  bark boiler         3-12
3-6     NO emissions as a function of stack 02 for a wood bark            3-13
        boiler
3-7     Aerodynamic particle diameter - low-NO   conditions                3-15
                                              Jt
4-1     Flow diagram of combustion air and  hogged fuel induction
        systems                                                            4-2
4-2     Boiler total particulate emissions  as a  function  of NO            4-13
        emissions for various test conditions on a hogged fuel boiler
4-3     Aerodynamic particle diameter as a  function  of cumulative         4-15
        proportion of impactor catch at low-O2 conditions in  a
        hogged fuel boiler
4-4     Aerodynamic particle diameter as a  function  of cumulative         4-16
        proportion of impactor catch at low-NO   conditions in a
        hogged fuel boiler
4-5     Aerodynamic particle diameter as a  function  of cumulative         4-17
        proportion of impactor catch at low-NO   conditions in a
        hogged fuel boiler
4-6     Total particulate as a function of  wood  ash  content for a         4-19
        hogged fuel boiler
4-7     NO emission as a function of wood moisture content for a          4-20
        hogged fuel boiler
4-8     NO emissions as a function of stack oxygen for three               4-22
        loads in a hogged fuel boiler
 4-9     NO emissions as a function  of  stack  oxygen in the                 4-24
        baseline and low-NOv  configurations
                           A.
 4-10    Mark III adsorbent sampling system                                 4-26
                                       iv                      KVB72-806015-1308

-------
                                    TABLES
                                                                          Page
        Summary of Combustion Modifications at Location 3                 1-2
        Summary of Combustion Modifications at Location 5                 1-3
        Summary of Emissions from Location 3 Wood Bark Boiler             3-5
        Location 3 Fuel Analyses                                          3-8
        Location 5 Fuel Analyses                                          4-4
        Location 5 Ash Analyses                                           4-6
        Summary of Gaseous and Particulate Emissions, Location 5 -        4-9
        Hogged Fuel Boiler
4-4     Additional Emissions Data, Location 5                             4-11
4-5     Emissions of NO at Baseline and Optimum Low-NOx Conditions        4-12
4-6     Summary of POM Analyses for Location 5 - Wood-Fired Spreader      4-28
        Stoker
4-7     POM Emission from Oil-, Coal-, and Wood-Fired Boilers:            4-30
        Comparison with Present Data
                                                              KVB72-806015-1308

-------
                               ACKNOWLEDGMENTS

        The authors wish to acknowledge the assistance of     Robert E. Hall,
the EPA Project Officer, whose direction and evaluation were an important
contribution to the program.
        Acknowledgment is also made of the contributions of the environmental,
engineering, operating and maintenance personnel of the plant at which  the
boilers were located.  Without their willing cooperation,  the program could
not have been conducted.
                 Limitations  on Application of Data Reported

        The pollutant emission data cited in this report pertain  to  two
specific wood boilers.  These data should not be used  to estimate mean
emissions from other types of wood boilers  or to predict emission reduction
potentials of combustion modifications until the modifications have  actually
been carried out on a greater number of such combustion types.
                                      VI
                                                             KVB72-80601 5-1 308

-------
                                  SECTION 1.0
                           INTRODUCTION AND  SUMMARY

        This report is one of a series which describe tests of combustion
modifications to control NO  emissions from  industrial process equipment.
                           X
This work was performed under EPA Contract No.  68-02-2645 over the  time  period
from August 1977 to June 1981.
        The activities reported herein include  tests performed on a wood bark/
coal-fired boiler  (Location 3) and a hogged  wood  fuel boiler  (Location 5).
Oil was the supplemental fuel at Location 5.
        Variations in  load, excess air and overfire air were  the combustion
modifications common to both boilers.  In addition, lowering  combustion  air
preheat and positioning of the supplemental  fuel  oil air damper were  performed
at Location 5.  Polycyclic organic matter (POM) sampling was  also conducted  at
Location 5 at both baseline and optimum  low-NO  conditions.
                                              A
        Table 1-1 summarizes the reductions  in NO and changes in efficiency
measured at Location 3 for each of the combustion modifications.  The overfire
air system modification consisted of increasing each of the overfire  air ports
from 1 inch (2.54 cm)  to 1-1/2 inches  (3.81  cm) in diameter.  As shown in
Table 1-1, the  lowest  NO level obtained  resulted  from implementing  lower
excess air before the  modification of  the overfire air ports.  This arrange-
ment also produced an  increase in boiler efficiency of 1.2 percent.
        Table 1-2 summarizes the NO reductions  achieved at Location 5 and  the
change in efficiency for all modifications except reduced combustion  air
preheat.  This  modification could not  be fully  implemented since the  combus-
tion air temperature could only be reduced by 16  - 22 K.  Also noted  in  this
table is the NO mass emission factor measured after each modification had  been
implemented.
        Of some interest for this particular boiler is the effect of  load
changes on NO emissions.  As noted, increasing  load  (18 percent) actually


                                       1-1                     KVB72-806015-1308

-------
        TABLE 1-1.    SUMMARY OF COMBUSTION MODIFICATIONS AT LOCATION 3
Control
Lower Excess Airt (146)tt
Lower Excess Air§ (184)
Overfire Air Dampers§ (174)
Load Reduction (51%)§,# (140)
NO Reduction,
%
37.2
18.5
20.7
7.9
Efficiency
Change, %
+1.2
+0.9
+1 .6
-4.0
NO After Control
ng/J*
92
150
138
129
 *NO as N02.




 tBefore overfire air system modification.




 §After overfire air system modification.




 ttLoad reduction referenced to nominal operation at 80% of rating.




ttValue in parenthesis is baseline NO (ng/J) before combustion modification.
                                      1_2                    KVB72-806015-1308

-------
        TABLE 1-2.   SUMMARY OF COMBUSTION MODIFICATIONS AT LOCATION 5
                             NO Reduction,   Efficiency  NO After Modification
      Modification                 %         Change, %           ng/J*
Lower Excess Air (40 )§
Increase Overfire Air (46)
Auxiliary Air Damper (36)
Load Changet
+18% (40)
-30% (40)
12.5
21 .7
17.2

27.5
30.0
+0.6
-1 .3
+1.7

+0.9
+1 .8
35
36
30

29
28
*NO as NO2.

tLoad change referenced to nominal operation at 76.5% of rating.

§Value in parenthesis is baseline NO  (ng/J) before combustion modification.
                                                             KVB72-806015-1308

-------
reduced the NO concentration and mass emission factor.  This  characteristic of
peak NO occurring in the mid-load range is somewhat unusual.
        Polycyclic organic matter (POM) samples "were  collected  and analyzed at
both baseline and low-NOx (auxiliary air damper adjustment) conditions.   The
significant finding was that the total POM at the  low-NO,.  condition could be
                                                        X
two to three times higher than that measured under baseline conditions.   This
large difference could be due more to fuel property variations  than to com-
bustion modification although the trend of higher  POM with Tower NO ' has been
                                     •                               A
observed previously on a coal-fired spreader stoker  (Reference  1).
                                       1-4                    KVB72-806015-1308

-------
                                  SECTION 2.0
                           TEST  EQUIPMENT  DESCRIPTION

2.1      EMISSIONS  SAMPLING EQUIPMENT
         Gaseous  emission  measurements  were  made  using analytical instruments
and equipment  contained in a  government-furnished  mobile  instrumentation
laboratory.  A plan  view  of the trailer is  shown in Appendix A.   Total partic-
ulate measurements were made  using  an  EPA Method 5 sampling  train produced by
Joy Manufacturing  Company.   Particulate size  distribution was measured using a
Brink Cascade  Impactor.   Total  oxides  of  sulfur  were  measured by wet chemistry
methods  using  the  sampling train and analytical  procedure of the Goksoyr-Ross
method.   Smoke density was measured using an  automated Bacharach smoke spot
pump.  Stack opacity readings were  made during particulate tests according to
EPA Method  9.
2.1.1    Gaseous  Emissions Sampling  System
         The  laboratory is equipped  with analytical instruments to continuously
measure  concentrations of NO, NO2,  CO, C02, 02,  SO2,  and  hydrocarbons.   The
sample gas  is  delivered to the  analyzers  at the  proper condition and flow  rate
through  the  sampling and  conditioning  system  described below.  Appendix A
describes the  analytical  instrumentation  and  the details  of  the  sampling
system.
         A flow schematic  and  description  of the  flue  gas  sampling and analyz-
ing system  is  presented in detail in Appendix A.   Briefly, the sampling system
used pumps  to  continuously draw flue gas  from the  heater  into the labora-
tory.  A high-capacity heated positive displacement diaphragm pump was  used to
draw a high volume of flue gas  into the analyzers  to  assure  quick response to
source variations.
         Special  precautions were  required to  obtain a representative sample
for the  analysis of  N02,  SO2/ and hydrocarbons.  These precautions consisted
of insuring that the sample was  kept above  its dew point,  to prevent loss  of


                                       2~1                     KVB72-806015-1308

-------
sample components in condensed water.  For this reason, an electrically  heated
9.5 mm (3/8 in.) Teflon sample line was used to bring  the sample  into  the
laboratory for analysis.  The hot pump provided heated sample directly to  the
hydrocarbon, SCU/ and NO  analyzers.
        A portion of the sample pump discharge was also sent through a refrig-
erated condenser  [to reduce the dew point to 275 K  (35°E)],' through, a  large
rotameter with flow control valve, and then to the O_, NO, GO,  and CO- instru-
mentation.  Flow to each individual analyzer was measured and controlled with
smaller rotameters and flow control valves.  Excess  sample was  vented  outside
the laboratory.
2.1.2   Particulate Emissions
        Particulate samples were taken at the  same  sample  location as the gas
sample using a portable effluent sampler produced by  Joy Manufacturing
Company.  These tests were made at baseline and  optimum low-NO  operating
                                                               X
conditions.  This system, which meets EPA design specifications for Test
Method 5, Determination of Particulate Emissions from Stationary Sources
(Federal Register, Volume 36, No. 27, page 24888, December 23,  1971) was used
to perform both the initial velocity traverse  and the particulate sample
collection.  Dry particulates were collected in  a heated case that contained a
110 mm glass-fiber filter for retention of particles  down  to 0.3 microm-
eters. • Condensible materials were collected in  a train of four impingers in a
chilled water bath.
        Particle size was measured at baseline and  optimum low-NO  conditions
                                                                  X
at the same sample location as the Method 5 tests.  A Brink Cascade Impactor
was used for all of the sizing tests because of  its high grain loading
capability.  This impactor was capable of fractionating particles in-situ into
six aerodynamic size ranges  (five collection stages and one backup filter).
The size range capability of this impactor was approximately 0.4 ym to 10 pm.
        The Method 5 sampling train and procedures, and the impactor opera-
tional procedures, are discussed in Appendix A.
                                       2-2                     KVB72-806015-1308

-------
2.1.3   Wet Chemical SO  Measurement
        The Goksoyr-Ross technique was used to sample  the  stack  gas  for  S03.
This method uses controlled condensation of the  stack  gases  in a coil  main-
tained at 333-344 K  (140-160°F) by a water bath.  This  temperature  is  below
the sulfuric acid (H-SO.) dewpoint so  that the SO3  and  the water vapor in the
flue gas condense upon the coil walls  to form H2SO4 droplets.  The  SO2 in the
flue gas passes  through  the coil and is collected in impingers containing a
weak hydrogen peroxide solution.  Following the  impingers, the flue  gas  flows
through a dry gas meter  and is then discharged into the atmosphere.
        The coil rinse and impinger liquid are each analyzed by  means  of an
acid-base titration  with a sodium hydroxide solution.   Both  the  SO3  and  SO2
concentrations may be determined from  this procedure.
2.1.4   Smoke Spot and Opacity Measurement
        On combustion equipment where  smoke numbers normally are taken,  such
as  oil-fired boilers, the smoke number is determined using test  procedures
according to ASTM Designation:  D 2156-65.  The  smoke  number is  determined at
each combustion  modification  setting of the unit.   Examples  are  baseline,
minimum excess air,  low  load, etc., and whenever a  particulate concentration
is  measured.
        Smoke spots  are  obtained by pulling a fixed volume of flue  gas through
a fixed area of  a standard filter paper.  The color (or shade) of the  spots
that are produced is visually matched  with a standard  scale.  The result is a
"Smoke Number" which is  used  to characterize the density of  smoke in the flue
gas.
        Opacity  readings were taken by a field crew member who is a certified
graduate of a U.S. Environmental Protection Agency  approved  "Smoke  School."
Observations are made at the  same time that particulate measurements are made
and as often in  addition as deemed necessary to  gather the maximum amount of
information.  The procedures  set forth in EPA Method  9, "Visual  Determinations
of  the Opacity of Emissions for Stationary Sources," were followed.
                                       2-3                     KVB72-806015-1308

-------
                                  SECTION 3.0
                       WOOD BARK/COAL BOILER, LOCATION 3
3.1
BOILER DESCRIPTION
        This boiler uses by-product wood bark  from  the  plant's  pulping opera-
tion and coal, or coal only.  The unit was  built  in 1966  by  the Wickes Boiler
Company and is rated at  100,000  Ib/hr  (45.4 Mg/h) steam flow firing a combina-
tion of coal and bark or coal only.  The boiler is  equipped  with a  travelling
grate spreader stoker.   The wood bark  is injected pneumatically above the
three coal feeders through three ports located above  the  front  overfire air
ports.  The smaller wood particles burn in  suspension while  the rest of the
bark burns on the grate.  Preheated combustion air  is introduced under the
grate after passing through a tubular  air preheater.  The unit  is equipped
with a multiclone dust collector.  Figure 3-1  is  a  cross-section of a typical
boiler used for suspension and thin-bed burning of  wood.   Figure 3-2 is a bark
distributor which is similar to  the three air-swept distributors used in the
Location 3 unit.  Other  boiler characteristics are:
        Maximum Continuous  High
        Pressure Steam Output:
        Steam Conditions  at
        Superheater Outlet:
        Heating Surface:
                            Firing coal and bark or coal
                            only - 100,000 Ib/hr (45.4 Mg/h)
                            Temperature - 800 °F (700 K)
                            Pressure    - 600 psig  (4.14 MPa)
                                          11,247 ft2  (1,045 m2;
                                           6,060 ft2  (563 m2)
                                           1,768 ft2  (164 m2)
                            Furnace Volume 6,150 ft3  (174 m3)
Boiler
Air Heater
Water Walls
        Bark flow rate to  the boiler was  controlled  as  well as  possible.   Flow
fluctuations and interruptions  occurred from  time  to time  and are  considered
normal operation.  The percentage of bark heat  input was estimated from the
steam chart when bark flow was  interrupted.
        The test unit was  limited in the  amount of wood bark being burned due
to high superheat metal temperature.  Whenever  superheat metal  temperature
                                       3-1
                                                    KVB  72-806015-1308

-------
                           Stack
           Induced Draft
             Fan
Figure 3-1.   Two-drum Stirling boiler  for  suspension
              and  thin-bed burning of wood.
                          3-2
KVB72-806015-1308

-------
                                           Bark Feed
                                                Distributor
                                                Spout Air
                                              Rotating Damper
                                            for Pulsating Air Flow
Figure 3-2.    Air-swept  distributor spout for  spreader stoker.
                                   3-3
KVB72-806015-1308

-------
reached  810°F  (706 K),  the  bark  flow  was  reduced until the tube temperature
dropped  below  800°F  (700 K).   The high  superheat metal temperature is caused
by  the large amount  of  small  bark particles  which are injected above the front
overfire air ports,  burning in suspension above  the  grate near the superheater
tubes.   While  burning coal  only, the  unit encountered no problem in maintain-
ing the  design superheat metal temperature of  750°F  (672 K).

3.2     EMISSIONS SAMPLING
         Gaseous  and  particulate  emissions measurements were made at a single
port in  the duct work downstream of the multiclone dust collector and induced
draft fan.  A  heated sample line was  used to sample  all gaseous emissions.  No
access was available for measurements upstream of the dust collector.  During
the test series  problems were encountered with the heated sample pump and the
problem  could  not be fixed  at the test  site.   For test numbers 3-20 through
3-32 only cold line  data could be taken.   Therefore,  neither N02/ SO2 nor
hydrocarbons could be sampled or measured during these tests.
         At the dust  collector outlet  a  Lear  Siegler  Optical Transmissometer
was  installed  by the boiler owner.  The readings of  this instrument were
recorded in the  control room  data sheets.
         Appendix A describes  the instrumentation employed.

3.3     BASELINE TESTS
        Baseline emission measurements  were  made with the boiler in the "as-
found"* condition firing about 20 percent wood bark  and 80 percent coal.  All
test  results are summarized in Table  3-1.  The results of the wet chemical
        measurements  made during test 3-25 were:
        S02—891 ppm  (corrected  to 3% 02), 794 ng/J
        SO3—4 ppm (corrected  to 3% O2),  4 ng/J
*As-found relates to the unit operation at  the  time  of  the  KVB  test crew
 arrival.  Baseline refers to the unit operating  at  its  nominal conditions.
 As-found and baseline conditions are usually the same,  however,  they  can
 differ.
                                      3-4                   KVB  72-806015-1308

-------
TABLE 3-1.   SUMMARY OF EMISSIONS FROM LOCATION 3 WOOD BARK/COAL BOILER
Nominal Steam
Load
Test
No.
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
3-10
3-11
3-12
3-13
3-14
3-15
3-16
3-17
3-18
3-19
3-20
3-21
3-22
3-23
3-24
3*25
3-26
3-27
3-28
3-29
3-30
3-31
3-32
Date
1979
4/11
4/11
4/11
4/11
4/12
4/13
4/16
4/16
4/16
4/16
4/17
4/17
4/17
4/17
4/18
4/18
4/18
4/18
4/19
4/19
4/19
4/19
4/19
4/20
4/20
4/23
4/24
4/24
4/25
4/25
4/25
4/25
Mg/h
37.2
37.7
37.2
37.2
15.9
36.3
25.0
24.0
25.0
25.9
35.4
35.4
35.4
35.4
35.4
35.4
36.3
35.4
35.4
35.4
35.4
35.4
34.9
34.9
29.5
37.2
37.7.
38.1
18.2
27.2
36.8
46.7
103
Ib/hr
82
83
82
82
35
80
55
53
55
57
78
78
78
78
78
78
80
78
78
78
78
78
77
77
65
82
83
84
4°,
60
81
103
Heat
Input
Rate
MW
29.4
29.8
29.4
29.4
12.6
28.7
19.8
19.0
19.8
20.5
28.0
28.0
28.0
28.0
28.0
28.0
28.7
28.0
28.0
28.0
28.0
28.0
28.7
28.7
23.3
29.4
29.8
30.2
14.4
21.5
29.1
37.0
°2
9.3
10.8
8.2
7.8
11.9
8.7
10.1
8.7
11.0
9.8
9.6
10.4
10.6
9.8
9.7
10.4
9.2
8.5
9.8
10.0
9.6
10.6
9.9
9.9
8.8
8.2
9.6
9.4
11.9
10.8
9.3
9.1
co2
10.3
8.6
12.4
12.8
8.3
9.9
9.8
10.8
8.4
9.4
8.7
8.9
8.1
8.3
8.9
7.7
7.6
10.1
9.0
7.7
7.7
6.8
9.5
8.9
10.2
10.2
9.8
10.2
7.5
8.6
9.5
9.9
NOx
Ppm'
Converter
3ut of
X
i
205
256
202
170
311
204
230
306
438
321
301
310
231
251
275
i
i
i
ti
o
I
01
in
0
$
0
1
o.
o
X


i
ng/J
ervice
i i i
i i i
~
130
162
128
107
196
129
145
193
217
203
190
196
146
159
174
— —
—
—
__
—
	
—
—
--
NO
ppm'
231
394
184
145
205
247
198
170
306
201
216
297
415
312
292
306
227
238
275
221
229
243
218
222
168
192
212
178
205
214
221
268
ng/J
146
249
116
92
130
156
125
107
193
127
136
188
262
197
184
193
143
150
174
140
145
154
138
140
106
121
134
112
130
135
140
169
HC
ppm
0
0
0
0
0
108
206
0
0
924
1895
1995
1341
723
2506
6708
4653
1728
849
i
i
i
0)
0
-rj
e
o
4J
3
O
i
&,
u
o
X



ng/J
0
0
0
0
0
23
44
0
0
196
402
423
284
153
532
1423
987
367
180
— —
—
—
— -
—
	
—
--
—
CO
ppm'
362
574
387
498
613
292
1032
505
207
587
268
304
87
129
288
450
275
238
493
319
308
243
154
292
465
192
391
472
828
281
230
364
ng/J
139
220
149
191
235
112
396
194
80
225
103
117
33
50
111
173
106
91
189
123
118
93
59
112
179
74
150
181
318
108
88
140
so2
ppm'
677
724
647
584
268
1062
396
527
1728
225
1011
1596
2769
2507
1180
645
610
864
1489
i
i
s
I
11
10
o
4J
O
1
O.



ng/J
596
637
569
514
236
934
348
464
1520
198
889
1404
2436
2205
1038
567
537
760
1310
— —
—
—
__
—
__
—
—
—
                                   3-5
                                                         KVB72-806015-1308

-------
                            TABLE  3-1.
(CONTINUED)
Test
No.
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
3-10
3-11
3-12
3-13
3-14
3-15
3-16
3-17
3-18
3-19
3-20
3-21
3-22
3-23
3-24
3-25
3-26
3-27
3-28
3-29
3-30
3-31
3-32
Total
Particulate
* Q lb/106
Bark Btu ng/J
31
31
31
31
50
28 0.244 105
23
23
0
24
15
15
0
0
16 0.504 217
14
6
6
15
15
15
15
15
21
26
16 0.320 138
24 0.400 172
29
10
10
10
10
Solid
Particulate Stack
lb/10S Temperature
Btu ng/J °F K
420
430
430
440
420
0.204 88 428
431
417
408
435
430
428
417
410
0.428 184 428
425
431
435
431
432
435
431
429
430
440
0.274 118 437
0.361 155 435
440
420
419
420
430
489
493
494
500
489
493
495
487
482
497
494
493
487
483
493
491
495
497
495
495
497
495
494
494
500
498
497
500
489
488
489
494
Eff.
%
81.56
79.77
82.77
82.55
79.01
80.99
80.31
81.78
79.16
79.88
79.35
79.70
78.76
79.46
79.68
79.06
77.22
80.39
79.19
77.72
77.61
76.86
80.48
79.34
80.47
79.31
79.56
80.02
76.42
78.65
79.61
77.82
Comments

As found boiler test - high barkflow
0 variation - high
0 variation
0 variation - low
Low load - approx 50% barkflow




Baseline particulate - high barkflow
Medium load - approx 23% barkflow

Medium load, low 0 - approx 23% bark
Medium load - coal only
Medium load - high barkflow^
Baseline - low barkflow
High air - low barkflow
High air - coal only
Normal air - coal only
Baseline particulate - approx 16%
0 variation
0 variation
O variation
Baseline - overfire air variation
Overfire air variation
Overfire air variation
Overfire air variation






bark







Overfire air variation - low excess air
Low NOx - Cascade impactor
SOx test
Low NOx particulate
Baseline particulate - approx 24%
Baseline - Cascade impactor
Load variation
Load variation
Load variation
Load variation



bark





*ppm corrected to 3%
                                       3-6
                   KVB72-806015-1308

-------
Also shown in Table  3-1  is  the  bark  contribution to the total boiler heat
input  (with coal  supplying  the  balance).
        Fuel and  ash analyses are  presented in Table 3-2.   Of note is the high
carbon content  of  the bottom and fly ash.   The boiler efficiencies noted in
Table  3-1 include  the loss  associated with  unburned carbon.   Tests 3-1  through
3-6 were conducted with  the original one  inch (2.54 cm) overfire air
nozzles.  Tests 3-7  through 3-32 were run after the 12 upper and five lower
overfire air ports in the back  of  the boiler were changed  to 1-1/2" diameter
(3.81  cm) to enlarge overfire air  capacity  and increase turbulence.
        Boiler  load  during  the  test  series  was approximately 80  percent of
rated  load except for the load  variation  tests.   NO emissions at the baseline
condition were  146 ng/J  or  231  ppm at 3 percent Oj.  NO emissions as a
function of O~  before the overfire air nozzle modifications  are  shown in
Figure 3-3.  The  smoke limit was found at 7.8 percent O2 at  a load of 82,000
Ib  (37.2 Mg/h)  of  steam  per hour."
        A baseline test  (Test 3-11)  immediately after the  overfire air  port
modification showed  136  ng/J of NO or 216 ppm at 3 percent O2-  NO and  S02
emissions increased  as coal flow increased.  Highest NO emissions occurred
when only coal  was burned,  i.e., 196-277  ng/J (Tests 3-9,  3-13,  and 3-14).
        Solid particulate emissions  during  baseline conditions were measured
with two different amounts  of bark flow.  The highest particulate emissions
(184 ng/J or 0.428 lb/10 Btu,  solid) occurred when approximately 16 percent
of the fuel was made up  by  bark.   At 25 percent bark flow  solid  particulate
emissions decreased  by 16 percent  to 155  ng/J or 0.361 lb/10  Btu.  Particu-
late size distribution was  also measured  using a Brink cascade impactor.  Data
on the size distribution are shown in Figure 3-4 for test  number 3-28,  where
particle diameter  as  a function of cumulative proportion of  impactor catch is
plotted.  Approximately  46  percent of the particulate is below 3 ym
aerodynamic diameter.
                                       3-7                    KVB 72-806015-1308

-------
                                             TABLE 3-2.
LOCATION  3  FUEL  ANALYSES
CD
Test No.
Date
Fuel Type
Ultimate Analysis
Moisture, % weight
Carbon, %
Hydrogen, %
Nitrogen, %
Sulfur, %
Ash, %
Oxygen (diff.), %
Proximate Analysis
Moisture, % weight
Ash, %
Volatile Matter, %
Fixed Carbon, %
Heat of Combustion
Gross Btu/lb
Net Btu/lb
Bottom Ash
Carbon, *
Gross Btu/lb
Fly Ash
Carbon, *
Gross Btu/lb
3-6
4-13-79
Coal

6.63
64.89
4.25
1.48
2.36
8.94
11.45

6.63
8.94
38.78
45.65

12,050
11,660




33.21
4,370
3-15
4-16-79
Coal

6.06
66.12
4.43
1.53
1.96
8.46
11.44

6.06
8.46
37.30
48.18

12,160
11,750


would



3-26
4-23-79
Coal

6.62
61.83
4.33
1.45
3.01
12.20
10.56

6.62
12.20
37.52
43.66

11,580
11,180

1.1.56
not ignite

25.74
3,590
3-27
4-24-79
Coal

6.77
62.60
4.33
1.50
2.30
11.07
11.43

6.77
11.07
37.49
44.67

11,470
11,070






3-6
4-13-79
Wood

51.51
25.58
2.84
0.16
0.030
1.51
18.37

51.51
1.51
37.21
9.77

8,950
NR*






3-15
4-18-79
Wood

53.87
23.93
2.77
0.17
0.029
1.42
17.81

53.87
1.42
37.23
7.48

8,890
NR






3-26
4-23-79
Wood

35.78
33.59
3.77
0.17
0.025
1.61
25.05

35.78
1.51
51.53
11.08

9,060
NR






3-27
4-24-79
Wood

41.23
31.01
3.55
0.20
0.026
1.56
22.42

41.23
1.56
46.51
10.70

8,890
NR






                    *NR = not reported by testing laboratory.

-------
   400
            i     r
i—i—r
   300
          Fuel: s 70% Coal

                  30% Wood Bark

                ( )  Test Number

          Boiler Load: 37.2-37.7 Mg/h
•a

o
<*>

g
a
o
2;
   200
   100
                                    I
                                          I
                                                                 (1)
                                                        (4)
                                                        Smoke  limit
                                                        @  7.8% 0,
      5     6

      >, %, dry
                                                                  10
                                                                        11
          Figure 3-3.   NO emissions as a function of O  before
                        overfire air nozzle modification.
                                    3-9
                                                           KVB72-806015-1308

-------
             10
H1
O
to

00
o
en
o
M
Ul
i
H
CO
O
CD
           o
          w
          EH
          H
          Q
          O
          EH o 9
          p, u.y
         U
         p

         8
         a
            0.3
0-1
  LI   I  I   I   I   I    I	1—I	1   I   I   I   I   I    I
0.01
                                                                               i—rr
                      Location  3 Wood  Bark Boiler

                      Test No.  3-28

                      29% Wood  Bark  and 71%  Coal

                      Load 38.1 Mg/h

                      O^ = 9.4%

                      Brink  Impactor

                      Downstream of  Multiclone
    I  I  I  i   I   i   i    1     I     I    I   I   I   1   I    I     I     I    II
                            5   10   20           60     80   90  95   98  99


                   Cumulative proportion of  impactor catch,  %  by mass
                            Figure 3-4.   Aerodynamic Particle Diameter - Baseline Conditions

-------
3.4     COMBUSTION MODIFICATIONS
        Combustion modification testing included  load  variation,  variation of
excess air and overfire air adjustments with  the  modified  nozzles  described
previously.  NO emissions as a function of  load are presented  in  Figure  3-5.
        A test series  (Tests 3-15  through  3-18) was conducted  to  evaluate  the
effect of stack oxygen on NO emissions.  The  overfire  air  was  left constant -
100 percent open which is normal for  these  tests.  Stack ©2  was controlled by
adjusting the damper on the forced draft fan  which supplies  the undergrate
air.  The baseline condition for the  02 variation  (Test  3-15)  was  9.7  percent
02 with 184 ng/J of NO emissions.   Stack O2 during these tests varied  from a
high value of 10.4 percent  to a  low value  of  8.5  percent.  The effect  of stack
©2 on NO emissions is  shown in Figure 3-6  which includes all the  test  data
measured.  Lowering the stack 02 resulted  in  a decrease  in NO  from 193 ng/J at
10.4 percent 02 to 150 ng/J at 8.5 percent 02.  During the duration of the low
02 test no clinkering  was observed and combustion conditions appeared  to be
good.
        After the O2 variation test series  the boiler  was  returned to  the
baseline condition to  provide a  baseline check point for the overfire  air
tests.  The dampers for the two  overfire air  headers in  the  back  of the  boiler
were set in their normal position  which is  100 percent open.  The  fly  ash
reinjection headers were also in their fully  open position.   The  front over-
fire air dampers were  30 percent open for  the upper header and 90  percent  open
for the lower header.  At baseline conditions (Test 3-19)  NO emissions were
174 ng/J with 189 ng/J CO emissions..   The  dampers  for  the  lower row of over-
fire air jets and for  both  fly ash reinjection headers were  reduced to
50 percent open.  NO emissions at  this test point (Test  3-23)  dropped  to
138 ng/J with 59 ng/J  of CO.
        During the test series,  it was observed that the SO2 measurements  were
varying.  This was due to fluctuations in  wood bark flow.  At  a stable load
condition whenever bark flow decreases more coal  is automatically fed  onto the
grate to keep the load steady.  The higher  coal flow then  increases the  SO2
emissions.
                                                              KVB 72-806015-1308

-------
    300
    200
                                                        (31)
                                          (30)
                              (29)
o
<#>
0-
o
z;
   100
             Fuel: s  80% Coal
                   =  20% Wood Bark
                  (  )  Test Number
                              J	L
                        1
                        1
                        1
            10
20
30
40
     50    60    70    80

Load, 103 Ib steam/hr
90   100
110
            Figure 3-5.
        NO  emissions as a function of load for a
        wood  bark  boiler.
                                       3-12
                                                              KVB72-806015-1308

-------



400

n
T3 300
(N
O
<*>
fl
4J
(0
a
Qj
200
i



100

c
1 1 1 1 1 1 1 1 1 1
^b Baseline before modification
m O Variation before modification
/ \ Overfire air (13)\^
~~ C"} Baseline after modification (2) ^V
1 J O variation after modification
\/ Coal only
(14)Oa(1€
"~ (15)O
O(19)
(6) i ^
d8)CT riS^^
- (26) A (ID ^727) _
(3)" O(28)
• (^
Fuel: = 80% Coal
= 20% Wood Bark
Boiler Load 34.9-38.1 Mg/h
( ) Test Number
1 1 1 1 1 1 1 1 1 1
)123456 789 10 11
                          Stack  Oxygen,  %,  Dry
Figure 3-6.  Location 3 — NO emissions as a function of stack
                               3-13
                                                      KVB72-806015-1308

-------
        Total and solid particulate emissions were measured at  the  low NOX
condition described earlier.  Solid particulate concentration was  118  ng/J
(0.274 lb/106 Btu) with the unit operating at 8.2 percent  stack O2.  NO
emissions were 121 ng/J at this condition.
        The low-NO  cascade impactor test (Test 3-24) is shown  in
                  X
Figure 3-7-  Particulate diameter as a function of cumulative proportion of
impactor catch is plotted.  About 27 percent of the particles are  below 3 um
aerodynamic diameter.  The geometric mean particle size and geometric  disper-
sion coefficient are 6 ym and 1.099 \m, respectively.  A comparison  of the
baseline (Figure 3-4) and low-NOx results (Figure 3-7) indicates that  the
geometric mean particle size for baseline operation is approximately
50 percent of that measured during low-NOx operation  (3.2  ym vs. 6 ym) .
Closing the dampers for the overfire air and fly ash  reinjection (low-NOx
configuration) resulted in the production of larger particulates,  but  at a
reduced mass rate  (118 ng/J vs. 155 ng/J).
        A wet chemistry SOX test (Test 3-25) resulted in 784 ng/J  of S02 and
4 ng/J of SO-,.
        The conclusions from these tests are that closing  the dampers  on both
the overfire air and fly ash reinjection resulted in  lower particulate emis-
sions.  Little change in NO emissions was found when  compared with data
obtained before the modification.  Overfire air adjustment reduced CO  emis-
sions (i.e., more complete combustion) even at low excess  air conditions.
Reduced excess air firing reduced NO emissions and helped  this  unit's  problem
with high superheater metal temperatures.
3.4.1    Efficiency
        Efficiency of the wood bark boiler was calculated  using the  heat loss
method described in the ASME Power Test Code.  The appropriate  fuel  analysis
in Table 3-2 was used in the calculations.  Stack gas losses were  calculated
from the flue gas analyses and radiation loss was estimated from the ABMA
standard Radiation Loss Chart.  The efficiency data for each test  condition is
                                      3-14                    KVB  72-806015-1308

-------
             10
U)
I
          g
          o
          LO
         a:
         u
         EH
a

w
ij
o
M
H
«
-^
DJ

u
               1

             0.9
             0.5
                                          i    r
                                        O
                                                                   1    I
                                            i   r
                                                 T
                                                                          Location 3 Wood Bark  Boiler

                                                                          Test No. 3-24

                                                                          21% Wood Bark,  79%  Coal

                                                                          Load: 34.9 Mg/h

                                                                          O  =9.9%
                                                                           2

                                                                          Brink Impactor

                                                                          Downstream of Multiclone
         Q
         §
         a
to
 I
CO
o
CTl
o
u>
o
oo
            0.1
                   I   I   I
                        I
I
I
J	I    I   I   I   I	I
I	I
I   I
                 0.01
                        2        5   10     20         50                  95   98  "


                            Cumulative  proportion of impactor catch, % by mass
                             Figure  3-7.    Aerodynamic Particle Diameter - Low-NO  Conditions
                                                                                  x

-------
presented in Table 3-1.  Efficiency varied from 82.77 percent at  8.2 percent
02 and a load of 82,000 Ib/hr (37.2 Mg/h) steam flow to 76.86 percent  at
10.6 percent 02 and 78,000 Ib/hr (35.4 Mg/h) steam flow.  Low excess air
firing resulted in improved efficiency without adversely affecting other
operating conditions.
                                      3-16                    KVB  72-806015-1308

-------
                                  SECTION 4.0

                         HOGGED  FUEL  BOILER,  LOCATION  5


4.1     BOILER DESCRIPTION

        Full-scale  combustion modification  tests  were carried out on a hogged

fuel power boiler located at a  large pulp and  paper mill.   The boiler is a

spreader stoker  rated  at 58.6 MW  thermal input (200,000 Ib/hr steam flow —

25.3 kg/s) when  firing wood.  Wood is  the primary fuel  and oil is used as a

supplementary fuel  only  when the  demand  exceeds  58.6  MW (200,00 Ib/hr — 25.3

kg/s).  The unit was placed into  operation  in  September,  1977-  The design

specifications of the  Combustion  Engineering boiler are as follows:

        1.  CAPACITY - 200,000  Ib/hr (58.6  MW)-wood only;  250,000 Ib/hr
                        (73.3 MW)-wood  and oil

        2.  DESIGN  PRESSURE - 700 psig (4,923  kPa)

        3.  OPERATING  PRESSURE  -  390 psig  (2,785  kPa)

        4.  STEAM OUTLET TEMP - 600  deg. F  (316  K)

        5.  STOKER  - traveling  grate (heating  surface = 352 sq. ft. or
                     32.7 m2)

        6.  OIL  BURNERS  -  (4) at  one elevation,  oriented tangentially

        7.  IGNITORS - High energy arc (1)  per burner

        8.  BOILER  HEATING SURFACE 15,242 sq.  ft. (1,416 m2)

        9.  WATER WALL HEATING  SURFACE 5,482 sq.  ft.  (509 m2)

        A schematic of the combustion  air,  flue  gas,  and wood handling systems

is given in Figure  4-1.  The hogged  fuel was fed  from the storage bin via

conveyor belt to the screw feeders  (large pieces  of wood are  recycled through

the hogger where they  are chopped and  sent  back  through the system). The wood

flow rate was controlled by the screw  feeder RPM. The  wood flow was deter-

mined volumetrically by  integrating  over the screw feeder revolutions.  A

strip chart in the  control room displayed the  mass flow rate  of the wood fuel


                                       4-1                     KVB72-806015-1308

-------
 I
 to
-J
to
 I
CO
o
en
o
i->
LTl
I
h-1
CO
O
CD
        TO

      SCRUBBER

        AND

       STACK

      (Sample

     Ports C)
                             NO.  10 POWER BOILER
                                       STEAM COIL

                                       AIR HEATER
                         I.D.  FAN
                                            HOG FUEL

                                            STORAGE BIN
                                          BELT CONVEYOR

                                          FROM STORAGE BIN

                                          TO FLIGHT  CONVEYOR

                                          AND SCREW  FEEDERS
                                               F.  D.  FAN
                                                                                         SPREADERS
UNDERGRATE

   AIR
                 Figure 4-1.   Flow diagram of  combustion air and hogged  fuel  induction systems.

-------
which was determined using  an  average  density  for  the wood.   The wood mass
flow measurement was not  very  accurate because of  varying densities.   This
fact was verified after the tests  by calculating wood flows  based on steam
flow, actual  fuel heating value  (as determined by  laboratory analysis),  and
efficiency data.
        The balanced draft  combustion  air  system consisted of an air heater,
undergrate air  system  (four zones  front to back),  overfire air ports  (four
elevations each with four ports  located in the corners and oriented tangen-
tially), and  an air system  for the four oil guns (controlled by five  dampers
located near  each gun).
        The flue gas,  after leaving the boiler proper, passed through the air
heater making a right-angle bend at the air heater hoppers.   From this point
the  flue gas  passed through the  multiclone and on  through a venturi-type
scrubber and  out the stack. A bypass  stack was used when the scrubber was
off-line.
        Most  of the emissions  sampling was done at the boiler outlet above the
entrance to the air heaters (Ports A).  Simultaneous particulate measurement
was  made at the multiclone  outlet during one test  (Ports B)  and at the stack
(Ports C) during another  test  to determine the particulate removal efficiency
of different  sections  of  the flue gas  system.

4.2     FUEL  DESCRIPTION
        The hogged fuel consists of sawmill wastes purchased from neighboring
mills.  All hogged fuel arrives  by barge.   Bark and wood waste from fir and
hemlock logs  constitutes  approximately 90  percent  of the hogged fuel.  Mois-
ture content  of this fuel can  vary from 44 percent to 58 percent depending on
source and season.  Salt  content of the fuel varies from 0.7 percent to
1.6 percent.  The salt content of  the  hogged fuel  is the result of storing or
transporting  logs in salt-containing waterways. Most of the hogged fuel is
unloaded directly into an inside storage bin.
        Three wood fuel analyses and two No.  6 oil analyses are presented in
Table 4-1.  Analyses of ash samples obtained during Test 5/2-1d are shown in
                                       4-3                    KVB72-806015-1308

-------
                    TABLE 4-1.    LOCATION 5 FUEL ANALYSES
Sample Identification:
    1.  Wood fuel sample Loc. 5
    2.  Wood fuel sample Loc. 5
    3.  Wood fuel sample Loc. 5
    4.  No. 6 Fuel Oil Loc. 5, 9-13-79, Test  5/1-1
    5.  No. 6 Fuel Oil Loc. 5, 10-23-79, Test  5/7-2C
9-20-79, Test 5/2-lb
10-3-79, Test 5/2-4a
10-17-79, Test 5/7-2
Wood Fuel Samples:

Proximate Analysis:
Moisture, %
Volatile Matter, %
Ash, %
Fixed Carbon, %
Ultimate Analysis  (Dry Basis):
Carbon, %
Hydrogen, %
Nitrogen, %
Sulfur, %
Ash, %
Oxygen, %  (by difference)

Heat of Combustion  (Dry Basis)
Gross Btu/lb  (kJ/kg)

No. 6 Fuel Oils

Ultimate Analysis:

Carbon, %
Hydrogen, %
Nitrogen, %
Sulfur, %
Ash, %
Oxygen, % (by difference)

Heat of Combustion:

Gross Btu/lb (kJ/kg)
Net Btu/lb (kJ/kg)
1
51.76
37.59
2.55
8.10
i):





• S):













1
47.13
5.66
0.16
0.11
5.29
41.65
8207
(19090)
4
89.
8.
0.
1.
0.
0.
2
44.64
42.27
3.64
9.45
2
48.53
5.50
0.21
0.088
6.57
39.10
8438
(19626)

21
27
45
52
019
53

50
38
2
9
3
49.81
5.69
0.27
0.091
4.04
40.10
8674
(20175)
5
89.69
7.65
0.39
1.71
3
.05
.29
.02
.64
Avg.
48.49
5.62
0.21
0.10
5.30
40.28
8440
(19630)





0.030
0.53

    17,550 (40703)
    16,800 (39075)
17,220 (40052)
16,520 (38424)
                                       Avg.
                                     Ash-Free
                                       51.20
                                         5.93
                                         0.22
                                         0.11

                                       42.53
                                         8912
                                       (20728)
                                        Avg.
                                        89.45
                                        7.96
                                        0.42
                                        1.62
                                        0.02
                                        0.53
17,385 (40435)
16,660 (38749)
                                      4-4
                            KVB72-806015-1308

-------
Table 4-2.  The carbon content of both  ash  streams  was  included in the boiler
efficiency calculations.

4.3     MODIFICATIONS AND TESTS  PERFORMED
        The combustion modifications  performed  at Location  5  were the
following:
        1.  Excess air variation
        2.  Load variation
        3.  Overfire air variation
        4.  Auxiliary air damper adjustments
        5.  Combustion air  preheat  variation
        Excess air variation was accomplished by adjusting  the  induced- and
forced-draft  fan dampers to increase  or decrease combustion air flow while
maintaining as constant a fuel flow as  possible.  Since the wood fuel composi-
tion was  quite variable  (it was  not possible to control the fuel composition
during the tests), some difficulties  were encountered in achieving steady
conditions.   At times the fuel flow had to  be adjusted  to maintain the desired
steam flow, thus, oxygen levels  tended  to fluctuate by  plus or  minus approxi-
mately one percentage point.  The average oxygen was varied from 5.0 percent
to 9.3 percent during these tests.
        In the load variation series  of tests,  steam flow was varied from
baseline  (approximately 70-80 percent of full capacity) to  90 percent of
capacity  and  then to approximately  50 percent of capacity.  Two different
oxygen conditions were set  up at both the high  and  the  low  steam flow rates.
In this test  series and in  other tests  at Location  5,  some  load fluctuations
occurred.  This is partly because of  variability of the wood  fuel,  but also
because the boiler was designed  as  a  load-following unit.  The  other boiler in
the powerhouse at Location  5 is  a black liquor  recovery boiler.  In operation,
the recovery  boiler is baseloaded and the hogged fuel boiler  takes load
swings.  Although the plant was  quite cooperative in maintaining constant load
on the hogged fuel boiler during emissions  tests, some  unsteadiness was
                                       4-5                     KVB72-806015-1308

-------
               TABLE 4-2.   LOCATION  5  ASH ANALYSES
                            Test 5/2-1d
Bottom Ash
Moisture, %
Ash, %
Sulfur,  %
Carbon,  %
As Received
    0.88
   93.94
    0.16
    3.85
Dry Basis
    0
  94.77
   0.16
   3.88
Heat of Combustion  (gross and net]
Btu/lb
kJ/kg
    331
    770
   334
   777
Fly Ash
                                        As Received
                Dry Basis
Moisture, %
Ash, %
Sulfur, %
Carbon, %
    4.71
   44.88
    0.27
   47.37
    0
 47.10
   0.28
 49.71
Heat of Combustion  (gross and net]
Btu/lb
kJ/kg
   7,047
   16,390
  7,395
  17,200
                                                        KVB72-806015-1308

-------
unavoidable.  In general,  load  could  be  held to within about ±5 percent of
some average value over  the  course  of the  day's tests,  however,  over several
days the load variations were on  the  order of +15 percent and more for a few
tests.  In some cases, No. 6 oil  had  to  be burned in order to keep the fire
steady and to maintain load.
        Three values of  overfire  air  (OFA) were evaluated:   a baseline value
of 5.7 percent of total  combustion  air,  a  high of 9.7 percent,  followed by
zero OFA.  As shown in Figure 4-1,  there were four elevations of overfire air,
each with four tangentially-oriented  ports.  At baseline conditions only the
fourth elevation  (highest  above grate) was used.  At the high overfire air
condition the top two elevations  were used.
        The overfire air elevations were either on or off;  no throttling of
the flow was possible.   The  lowest  two elevations of overfire air are not used
by the plant because it  was  determined soon after the boiler installation that
stack opacity increased  when they were used and efficiency was  decreased.  In
addition, when more than two elevations  of overfire air were used, grate
temperatures became dangerously high  due to the reduced undergrate air flow.
        The auxiliary air  dampers are part of the oil air system shown in
Figure 4-1.  They are located near  each  of the four oil guns which are mounted
in the four corners above  the overfire air ports.  There are five air regis-
ters for each of the oil guns.   In  the baseline condition when no oil was
fired there was no air flow  through these  registers.  In the modified condi-
tion the lowest of the five  registers (auxiliary air damper "AA") was opened
10 percent and air was thus  injected  at  approximately the level of the oil
guns without any oil flow.   Thus, the adjustment provided an alternative to
overfire air as a means  of staged combustion air.  The other oil air registers
were also adjusted but did not  appear to give NOX emissions as  low as the
auxiliary air damper "AA"  adjustment.
        The combustion air temperature was varied from baseline level to a low
inlet air temperature by bypassing  the steam coil portion of the air heater
for that particular test.  The  combustion  air temperature dropped from 517°F
                                       4-7                    KVB72-806015-1308

-------
(543 K) to 477°F (521 K) at high load and from  495°F  (531  K)  to  467°F (515 K)
at baseline load.  Thus, only modest changes in combustion air temperature
were achievable.
        A complete set of emissions measurements including gaseous,  total and
solid particulate and particulate size, wet chemical  SOx,  and polycyclic
organic matter  (POM) was made for the baseline  condition.   The same  series of
tests  (except for wet chemical SO  measurement) was run  at the optimum low NOx
conditions which was a combination of low excess air  and auxiliary air damper
adjustment.  Nearly complete sets of gaseous O2, CO,  C02,  NO, NOX, S02,  and HC
emissions were obtained on all tests.  Appendix A provides a  description of
the measurement equipment.
4.3.1   Results
        The data from Location 5 is summarized in Tables  4-3  and  4-4.   The
particulate measurements in Table 4-3 were obtained at  the  boiler outlet
(upstream of multiclone) while those in Table 4-4 were  obtained downstream of
multiclone.  The data yields two important observations:
        1.  NOX emissions were low with only a single reported concen-
            tration over 100 ppm.                           >
        2.  Particulate emissions prior to any dust collecting device
            were high and variable.  The range of total particulate
            concentrations measured at the boiler outlet  was  1270 -  3780
            ng/J (2.96 - 8.79 lb/106 Btu).  The overall fly ash removal
            efficiency based on one set of simultaneous particulate
            measurements at the boiler outlet and at the  stack was
            84 percent.
The average NO emissions for the baseline and optimum low NO  configurations
are shown in Table 4-5.  The average reduction in NO concentration based on
those values is 17.2 percent.
        Boiler efficiency increased at the low NO  conditions to  72.3  percent
                                                 X
(average for Tests 5/5-2, 5/7-2b, and 5/7-2c) from the  baseline efficiency of
71.1 percent (average for Tests 5/5-1, 5/7-1b, and 5/7-1c)  for a  gain  of
1.7 percent.  The auxiliary air damper "AA" adjustment  thus gave  reduced NO
emissions and the maximum efficiency condition at baseline  load.
                                      4-8                     KVB72-806015-1308

-------
TABLE 4-3.   SUMMARY OF  GASEOUS  AND PARTICULATE EMISSIONS, LOCATION 5 -
                           HOGGED FUEL BOILER
Steam Flow
Test No.
5/1-1
5/2-la
5/2-lb
5/2-lc
5/2-ld
5/2-le
5/2-lf
5/2-2
5/2-3
5/2-4
5/2-4a
5/3-1
5/3-2
5/3-2a
5/3-3
S/3-3a
5/4-1
5/4-2
5/4-3
5/4-4
5/4-2a
5/5-1
5/5-2
5/5-3
5/6-1
5/6-2
5/6-la
5/6-2a
5/7-1
5/7-2
5/7-la
5/7-2a
5/7-lb
5/7-2b
5/7-lc
5/7-2C
Date
1979
9-17
9-21
9-20
9-24
9-25
9-26
10-1
9-21
9-21
9-21
10-3
10-4
10-4
10-4
10-4
10-4
10-1
10-1
10-1
10-1
10-2
10-16
10-16
10-16
10-15
10-15
10-15
10-15
10-17
10-17
10-18
10-18
10-22
10-22
10-23
10-23
Kg/s
28.0
19.3
16.6
14.4
20.0
17.6
21.0
19.3
19.3
19.3
20.2
17.5
22.7
22.7
13.5
13.5
19.5
19.5
19.5
19.5
19.2
18.1
18.1
18.1
26.2
23.6
21.4
20.4
18.8
18.8
18.9
18.9
17.6
17.6
18.0
18.0
Flow
103
lb/hr
222
153
132
114
159
140
167
153
153
153
160
139
180
180
107
107
155
155
155
155
152
144
144
144
208
187
170
162
149
149
150
150
140
140
143
143
Heat Input
Rate
MW
78.0
53.9
46.3
40.2
56.0
49.2
58.6
53.9
53.9
53.9
56.3
48.9
63.3
63.3
37.5
37.5
54.5
54.5
54.5
54.5
53.3
50.7
50.7
50.7
73.3
65.6
65.6
56.9
52.5
52.5
52.8
52.8
49.2
49.2
50.4
50.4
1C6
Btu/hr
266
184
158
137
191
168
200
184
184
184
192
167
216
216
128
128
186
186
186
186
182
173
173
173
250
224
204
194
179
179
180
180
168
168
172
172
°2
*
8.3
7.2
8.7
9.6
6.9
7.1
5.4
6.5
9.3
5.7
5.0
7.4
4.8
7.7
10.3
7.0
7.3
8.1
7.4
8.1
7.7
8.3
5.8
7.4
6.5
7.1
7.4
8.0
6.9
5.1
8.1
5.1
8.7
5.6
7.6
6.2
%
12.7
12.0
10.0
9.0
12.4
13.6
14.5
13.6
10.2
12.9
15.0
12.6
15.9
12.4
8.9
12.9
12.8
10.3
11.7
11.0
12.3
11.8
14.8
12.8
10.9
9.9
12.1
12.9
13.5
15.8
10.6
16.2
10.6
14.4
12.3
13.9
N
ppm'
103
80
81
86
50
45
59
106
88
69
69
61
49
53
72
53
85
67
57
58
67
89
62
81
86
99
89
85
82
61
76
55
61
54
66
65
°x
ng/Jf
56
43
44
47
27
24
32
57
48
37
37
33
27
29
39
29
46
36
31
31
36
48
34
44
47
54
48
46
44
33
41
30
33
29
36
35
NO
ppm*
97
74
72
72
48
43
55
101
82
65
68
61
49
53
72
51
85
67
53
56
65
85
61
78
83
96
84
83
78
56
73
51
59
54
62
59
ng/J+
53
40
39
39
26
23
30
55
44
35
37
33
27
29
39
28
46
36
29
30
35
46
33
42
45
52
45
45
42
30
40
28
32
29
34
32
HC
ppm»
81
47
213
557
91
44
139
35
204
96
130
40
112
141
315
72

223
214
89
39
60
48
68
79
50
36
45
57
50




103
92
ng/J
15
9
40
105
17
a
26
7
38
18
24
8
21
27
59
14

42
40
17
17
11
9
13
15
9
7
8
11
9




19
17
CO
ppm*
929
1522
>2656
>3144
>2585
>2258
>2312
894
>3077
847
2172
2310
>2222
>2707
>3364
>2571
1130
865
1853
1298
1780
1312
>2153
2038
1460
700
1055
670




>1463
>1165
1988
2422
ng/J
306
502
>875
>1036
>852
>744
>762
295
>1014
279
716
761
>732
>892
>1109
>847
372
285
611
428
587
432
>710
672
481
231
348
221




>4S2
>384
655
798
so..
ppm*
44
13
30
20
42
97
3
0
0
3
39
8
0
0
0
0
0
0
0
0
47















ng/J
33
10
23
15
32
73
2
0
0
2
29
6
0
0
0
0
0
0
0
0
35















                                    4-9
                                                          KVB72-806015-1308

-------
                         TABLE 4-3.
(CONTINUED)
Test No.
5/1-1
5/2-la
5/2-lb
5/2-lc
5/2-ld
5/2-le
5/2-lf
5/2-2
5/2-3
5/2-4
5/2-4a
5/3-1
5/3-2
5/3-2a
5/3-3
5/3- 3a
5/4-1
5/4-2
5/4-3
5/4-4
5/4-2a
5/5-1
5/5-2
5/5-3
5/6-1
5/6-2
5/6- la
5/6-2a
5/7-1
5/7-2
5/7-la
5/7-2a
5/7-Lb
5/7-2b
5/7-lc
5/7-2C
Total Solid
(Blr.Out.) (Blr.Out.)
Fuel Mix
Wood/Oil lb/106 lb/106
% of heat in Btu ng/J atu ng/J
75/25 8. 50 3660 9.37 3660
100/0
100/0 5.12 2200 5.07 2180
80/20 7.41 3190 7.27 3120
100/0
100/0 2.96 1270 2.87 1240
100/0
100/0
100/0
100/0
100/0 8.79 3780 8.77 3770
100/0
100/0
100/0
100/0
100/0
100/0
100/0
100/0
100/0
100/0 3.82 1640 3.81 1640
100/0
100/0
100/0
85/15
85/15
100/0
100/0
100/0
100/0 5.64 2430 5.62 2420
100/0
100/0 3.50 1505 3.48 1500
100/0
100/0
91/9
91/9 5.83 2510 5.82 2500
Air Heater
Gas Out
Temperature
°F
463
465
455
441
471
460
453
465
470
470
462
465
480
460
420
440
465
440
455
445
454
428
433
430
454
405
394
440
450
440
440
440
430
453
440
443
K
513
514
508
500
517
511
507
514
516
516
512
514
522
511
489
500
514
500
508
503
508
493
496
494
508
480
474
500
505
500
500
500
494
507
500
501
Boiler
Eff.
»
79.2
68.6
67.5
75.6
68.8
69.8
70.5
69.6
67.1
69.0
70.4
69.1
70.3
69.2
67.7
69.8
69.1
68.2
68.8
68.6
69.2
69.8
71.2
70.4
75.6
73.0
67.6
70.0




68.9
70.4
74.7
75.4
Comments
As Found
Baseline
Baseline-Brink
Baseline-Multiclone efficiency-bad wood
Baseline-Brink and SO -bad wood
Ba^.Gline-POM-bad wood
Baseline-Brink
Low O -unsteady conditions
High 0,
Minimum O
Minimum 0 -Brink
Baseline
High Load-Low 0
High Load-High 0
Low Load-High O
Low Load-Low O.,
2
Baseline
High OFA
Zero OFA
Repeat Baseline
High OFA
Baseline
Aux.Air "AA" dpr. adj.
Oil Air "C" adj.
High Load
Low C.A. Temp. -High load
Low C.A. Temp.
Baseline
Baseline
Low NO -Brink
Baseline
Lew NO -Brink-POM
Basel ine
Low NO -Brink
Baseline
Low MO -Total part ic. removal efficiency
* dry, corrected to 3* 0 , dry
t as NO.
                                       4-10
                       KVB72-806015-1308

-------
                                  TABLE 4-4.    ADDITIONAL EMISSIONS  DATA, LOCATION 5
03
O
O1
O

Steam Flow
Test No. Date,
(Location) 1979 kg/a 103 Ib/hr
5/2 - Ic 9-24 14.4 114
(Multiclone
Outlet)
5/7 - 2c 10-23 18.0 143
(Stack)
5/2 - 1d 9-25 20.0 159
(Boiler
Outlet)
Wet Total Solid Particulate
Chemical SO, Particulate Particulate Oa „ ,
£ KG mO va J.
02 Efficiency
%, dry ppm* ng/J lb/106 Btu ng/J lb/106 Btu ng/J Percent Comments
9.6 — — 3.02 1300 2.85 1230 59.2 Multiclone efficiency only-
multiclone partially plugged

9.0 — -- 0.92 394 0.88 377 84^2 Total particulate collection
efficiency
7.0 18 13.5 — — — — -- Zero detectable SO3 -
Goksoyr-Ross method

        *Dry, corrected to 3%
U)
O
oo

-------
    TABLE 4-5.  EMISSIONS OF NO  AT BASELINE AND OPTIMUM LOW-NO  CONDITIONS


Condition
Baseline
Low NO
Mean, Standard Deviation
No. of
ppm* ng/Jt Data Pts .
67.9, 13.5 36.8, 7.3 14
56.2, 4.0 30.4, 2.0 5
   (aux.  air dpr.  "AA"
    adjustment)
 *dry,  at  3  percent
 tNO  as NO0
        The particulate  emissions  at  the  optimum low NO  condition ranged from
 1500  -  2420 ng/J  (3.50 - 5.83  lb/106  Btu).   It  was  assessed that this emission
 level was not  different  from that  found at  baseline,  which was 1270 - 3190
 ng/J  (2.96 - 7.41  lb/106 Btu).
        However, when all of the total particulate  data from all of the tests
 conducted upstream of the multiclone  are  plotted as a function of NO emis-
 sions,  as is done  in Figure 4-2, a trend  of lower particulate emissions with
 lower NO emissions is noted.  The  reason  for  this behavior is unclear at
 present.  Correlations were made of particulate emission with excess air level
 and with the pressure drop across  the grate,  however,  no meaningful result was
 obtained with  either variable.  The wood  fuel composition and the size of the
wood fuel may  be the most important variables affecting the particulate
emissions from the hogged fuel boiler, and  may  influence NOV emissions as
                                                            X
well.

        The Brink  cascade impactor tests  conducted  at the boiler outlet were
analyzed to determine if  the solid particulate  size distribution was related
                                      4-12
                                                               KVB72-806015-1 308

-------
-p
ffl
O
rH
\
£1
EH
O
E-i
    3870
    (9.0)
    3440
    (8.0)
    3010
    (7.0)
    2580
    (6.0)
    2150
    (5.0)
en
2
O
H
in
en
§   1720
„   (4.0)
u
£   1290
<   (3.0)
     860
    (2.0)
     430
    ri.o)
                            50       60        70
                                NO, ppm, Dry at 3% 0,
                                                         80
90
100
    Figure 4-2.
                 Boiler total particulate  emissions as a function of NO
                 emissions for various  test conditions on a hogged fuel
                 boiler.
                                     4-13
                                                           KVB72-806015-1308

-------
to boiler operation.  (Three of these distributions are  shown  in  Figures 4-3
through 4-5.)  These distributions, characterized by  their  mass median
particle size, are summarized below:
Test.
No.
5/2-4a
5/7-2b
5/2-1f
5/7-2
5/2-1b
5/7-2a
uperamng
Condition
Minimum ©2
Low NOv
X
Baseline
Low NO
X
Baseline
Low NO
iieuj-dii LJ
ym
200
250
600
38
250
22
lb/106 Btu
8.77
6.92
5.79
5.62
5.07
3.48
ng/J
3770
2980
2490
2420
2180
1500
        These data indicate  that  there was no  clear  relationship between the
mass median diameter and either the particulate  emission  factor or the boiler
operating condition.   (As will be discussed  shortly,  there  is  a relationship
between the particulate temperature factor and wood  ash content.)
        Although the size of the  fuel was not  characterized,  it is anticipated
that for fuels of equal ash  content, smaller pieces  of  fuel would be more
easily entrained in the flue gas  after combustion  and would lead to larger
particulate concentrations having smaller aerodynamic particle diameters.
Larger pieces of fuel would  tend  to burn on  the  grate and fall into the ash
pit after combustion and less fine material  would  be carried out of the boiler
in the flue gas.  The boiler did  not have the  instrumentation necessary to
separately measure the amount, if any, of sander dust or  wood shavings used
during a test.
        While the total impactor  catch weight  was  considerable, much of the
particulate material was large and, therefore, was caught in the precutter
cyclone before reaching the  stages.  Figures 4-3 through  4-5 show that only
20 - 40 percent of the total catch actually  impacted on the stages.  During
some of the tests the impactor jets became plugged and  sampling had to be
discontinued prematurely.  Thus,  the sample  weight on the stages was sometimes
less than ideal.
                                       4-14                     KVB72-806015-1308

-------



















1
H
l_n






-J
NJ
1
co
o
0
I-J
Ln
1
f— '
U)
o
CD
-LU
8

6


4
6
3 3
o
Q
K 2
W
W
H
Q
a i
o
BO. 8
rf!

rl 0.6
o
H
1 0.4
Q
O
§0.3
0.2



0.1
J 1 1 1 1 1 1 1 1 1 1 1 | | | 1 I 1 1 | |l '
~ LOCATION #5 - HOGGED FUEL BOILER ~~
TEST #5/2-4A
FUEL - Hogged Wood
— LOAD - 160 x 10 Ibs/hr. _
O 0 - 5.0%
~~ 1 BRINK IMPACTOR ~~
/ BOILER OUTLET
— / TOTAL CATCH WEIGHT - 0.1560 g —
r~\ (includes cyclone catch)
1^
— / —
(i
W
/
/
_ / _
	 JL 	
_ O _
	 r _
/
— i —
1 -
o


— —
— —



1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 II II l
































.01.05.1 .2.512 5 10 20 30 40 50 60 70 80 90 95 98 99 99.8 99.9 99.99

CUMULATIVE PROPORTION OF IMP ACTOR CATCH, PERCENT BY MASS



Figure 4-3. Aerodynamic particle diameter as a function of cumulative proportion
of impactor catch at low-O  conditions in a hogged fuel boiler.

-------
                 10
 i
 h->
 (Tv
to
 I
30
o
rr>
o
h-1
01
i
H
U)
O
00
                o
                LT>
               Q

                v  2
               K
               Q
^0.8
               o
               H
               P
  1.4



 JO.3





 0.2
 0.1
                    .JillI\T
                             1    T
                     1    I   TT~I   I    \     I    I     TT
                                           Tl    L.
                                                                                  HOGGED FUEL BOILER
                                                     LOCATION #5 -

                                                     TEST #5/7-2

                                                     FUEL - Hogged Wood

                                                     LOAD - 149 x 10  Ibs/hr.

                                                     0  - 5.1%

                                                     BRINK IMPACTOR

                                                     BOILER OUTLET

                                                     TOTAL CATCH WEIGHT - 0.1177 g

                                                     (includes cyclone)
                                                    O
          111   III
                 I
I    I   I   I   I   I    I
I	I
I   1
I   I
I
      .01.05.1.2 .5
    12     5   10   20  30 40 50 60 70  80   90   95    98 99

     CUMULATIVE PROPORTION OF IMPACTOR CATCH,  % BY MASS
                                         99.8  99.9 99.99
 Figure 4-4.
Aerodynamic particle  diameter as a function of cumulative proportion

of impactor catch  at  low NO  conditions in a hogged fuel boiler.

-------
i
{—>
-~j
tv>
I
oo
o
(T>
O
M
Ul
U)
o
00
              10


               8


               6
             o
             LO
            Q
             .  2
            K
            £
            H
            Q

            W  1


            So.8
u
H
0.6




0.4


0.3




0.2
  0.1
         TT~I—TT~T	\	1	1—I   I   I   I   I—I     I    I     TT
                                                                                TT
                                                  LOCATION  #5  -  HOGGED FUEL BOILER

                                                  TEST  #5/7-2A

                                                  FUEL-Hogged  Wood

                                                  Load  -  150 x 10   Ibs/hr.


                                                  BRINK IMPACTOR

                                                  BOILER  OUTLET

                                                  TOTAL CATCH  WEIGHT  - 0.094  g

                                                  (includes cyclone)
         1  1         1   1

                                  1
1    1   1   1      1    1
1    1
11
     .01.05.1  .2  .5
                    12    5   10   20  30 40 50 60   70  80    90   95    98

                     CUMULATIVE PROPORTION OF IMPACTOR  CATCH, %  BY MASS
                                     99 99.8 99.9 99.99
   Figure  4-5.
                             Aerodynamic particle diameter as a function of cumulative proportion
                             of impactor catch at low NO  conditions in a hogged fuel boiler.

-------
        The fuel composition data obtained show that  there  is  some  correlation
of total particulate emission with the ash content of  the wood fuel.   This
correlation is shown in Figure 4-6.  Since a large portion  of  the fuel ash
(probably over 70 percent) is fly ash rather than bottom ash one would expect
such a correlation.  Because of the spread in the data it was  not possible to
separate the effects of different operating conditions on emissions.   Thus,
the curve shown is an overall curve not specific to only one set of
conditions.  The curve in Figure 4-6 was fit by li-near least square regression
analysis.  A correlation coefficient approaching unity indicates a  good fit
and a coefficient approaching zero indicates a poor fit.  As noted,  the corre-
lation in terms of wood ash content can explain 68.9  percent of the variation
in total particulate.
        To determine whether or not there was a relationship between  the
moisture content of the fuel and total particulate emission, the data were
plotted and a linear regression analysis performed.   The result was a correla-
tion coefficient of only 10.1 percent indicating a weak relationship  between
particulate emission and wood moisture content.  The  data indicates that,
under different boiler operating conditions, total particulate can  vary by as
much as a factor of three for nearly the same fuel moisture content.   Although
it is reasonable to suppose that the amount of moisture in  the fuel affects
the burnout of the wood pieces, other factors appear  to play a more signifi-
cant role in the combustion and carryover of the wood.  More data would be
required at the various operating conditions to determine their true  effects
on particualte emission.
        Figure 4-7 was generated by linear regression  of the NO emission
versus fuel moisture data.  A somewhat better correlation was  found (r2 =
26.3 percent), however, there is still a large amount of data  scatter.  Based
on previous tests of steam and water injection in boilers,  it  might be
expected that increased moisture content of the fuel  would  result in  lowered
flame temperatures and reduced NO emission, however,  this does not  appear to
be the case for the hogged fuel boiler at Location 5.
        In a similar manner, NO emission was evaluated as a function  of fuel
nitrogen in an attempt to determine whether or not any correlation  existed.

                                      4-18                     KVB72-806015-1308

-------
  3
  4-1
  m
   (U
  4J
   rO
  r-l
   3
   0
  •H
  -P
  r-l
  03
  4J
  O
  EH
       10.0
     (4300)
       8.0
      (3440)
        6.0
      (2580)
   4.0
(1720)
        2.0
       (860)
                                     5/2-4a
                             5/2-lc Q
QMin.  02

   High  OFA
   Baseline

/\ Optimum  Low NO

LOAD: 18.0  Kg
       steam/s
   (143,000  Ib
   steam/hr)
           y=l.983x+1.138  (Curve Fit-Linear Regression)
           Correlation coefficient = r2 = 0.689
                               I
                                            I
         J_
            0123456
                          Wood Ash Content (Wet Basis),  %

Figure 4-6.    Total particulate  as a function of wood ash content for
              a hogged  fuel  boiler.
                                    4-19
                                                     KVB72-806015-1308

-------
        80
        70
        60
      CN
     t*>
     n
     -P
     fl3
        50
     I  40
     ft
     c
     0
     -H
     m
     to
     •H

     w
30
        20
       10
            I         I


          High OFA


          Baseline


       /\ Optimum Low NOX


              . 0
                          5/2-4a
                         5/4-2a


LOAD:  18.0 Kg steam/s

(143,000 Ib stearn/hr)    5/7-2c
                                                   o
                   10        20       30        40       50

                   Wood Moisture Content (Wet Basis), %
                                                         60
Figure 4-7.    NO  emission  as  a function of wood moisture content

              for a  hogged fuel boiler.
                             4-20
                                            KVB72-806015-1308

-------
The data were too scattered  to  develop  a  meaningful curve.   However,  the data

indicate that there was no significant  increase  in NO emission as fuel nitro-

gen increased.   (The weight  percent  of  nitrogen  in the fuel varied over a
narrow range from 0.08 percent  to  0.14  percent on a wet basis.)

        Very low SO2 emissions  were  measured either by wet chemical technique

or continuous gaseous analyzer.  The SCU  emission data were presented in

Table 4-2.  There are two reasons  for the low SG>2 emission:

        1. Low fuel sulfur content (0.09  - 0.11  percent by weight on a
           dry basis).

        2. Ash sulfur retention.   Analysis of the fly ash and bottom ash
           yielded sulfur values  (dry basis) of  0.28 percent and
           0.16  percent, respectively.  These values exceeded that for
           the wood fuel indicating  some  degree  of sulfur concentration
           in these ash streams.

Changing fuel sulfur content and,  no doubt,  sulfur content of the fly ash and

bottom ash (sulfur in the fly ash  and bottom ash was measured for only one
test) caused some variation  in  SO2 emission, however, the absolute level of

S02 was generally less than  50  ppm,  dry at 3 percent 0,,.  The error in the

continuous analyzer S0? data at this low  level for these tests approached 20 -

30 percent due to problems with the  analyzer.

        Gaseous  SO2 measurement was  made  for the first twenty-one tests with

the DuPont 400 Photometric Analyzer.  A wet chemical SO  test (Goksoyr-Ross
                                                        A
method) yielded  18 ppm  (13.5 ng/J) of SO2 and no detectable SO-j.  By compari-

son, the DuPont  analyzer indicated 42 ppm of SO2 during the same test.  The
DuPont analyzer  was down for repair  during the remainder of the test program

at Location 5.

        The results of the excess  air variation  test series at three different

boiler loads are shown in Figure 4-8.  The maximum NO reduction obtained by

reducing the excess air from baseline (7.2 percent oxygen)  to 5.7 percent

oxygen was 12 percent.  The  efficiency  at the reduced excess air condition was

slightly higher  than that at baseline condition.  CO emissions were also

reduced in two of the three  lowered  excess air tests.
                                       4-21                     KVB72-806015-1308

-------
*
m
a
a
o
2
    140
    120
    100
    80
    60
    40
     20
                    1    I     T
                -L--L
   1     I    I     I    \     I
                                             STEAM FLOW


                                          kg/S (10  Ib/hr)
                  19'3


                  22.7 (180)



                  13.5 (107)




     FUEL:   Hogged Wood (100%)
 O

5/2-2
                                                         5/3-3
                                             5/3-2A
                           5/3-2
                                      5/3-3A

56    7    8   9


STACK OXYGEN, % DRY
                 10
                                                         11  12   13  14
   Figure 4-8.  NO emissions as a function of stack oxygen for three

                loads in a hogged fuel boiler.
                                 4-22
                     KVB72-806015-1308

-------
        Similar behavior of NO  emission  versus  excess  oxygen was found to
                              X


occur at full capacity and half capacity.   NO  emissions at both high and low



loads were less than emissions at baseline  by  approximately 30 percent.  CO



emissions, however, were high for both  full and  half capacity tests.  Unburned



hydrocarbons  (as methane) are reported  in Table  4-3; however, the data appear



to be quite variable.





        NO  emissions were reduced  by increasing the overfire air in another
          A.


series of modification tests.  The  reduction in  emissions  from baseline was



21 percent (from 85 ppm to 67 ppm)  and  CO emissions were reduced at the same



time from a baseline level of 1130  ppm  to 865  ppm.   However,  the efficiency



was reduced by  1.3 percent at the high  overfire  air condition.  These calcula-



tions are made by comparing the baseline  value from test 5/4-1 to the high



overfire air  test 5/4-2.   (Test 5/4-2a  was  also  conducted  at high overfire air



conditions, but was done on a different day.)





        Shutting off the overfire air produced no significant change in NO



emissions as  is seen by comparing the emissions  from test  5/4-3 to those of



5/4-4, however, CO emission increased considerably at zero overfire air.  No



definite relationship of particulate emissions with overfire air variation can



be determined from the data.





        Lowering the combustion air temperature  by only a  modest amount had



little impact on NO  emissions, but increased  CO emissions.  The effect of the
                   A.


optimum low NO  register adjustment (10 percent  open auxiliary damper "AA",
              X


others closed) on NO  emissions is  shown  in Table 4-4.   The effect of this
                    X


modification  on efficiency has already  been discussed.   Other oil air regis-



ters were adjusted but failed to achieve  the same NO  reduction.
                                                     X




        The effect of the auxiliary air damper "AA" adjustment coupled with



lowered excess air on NO emissions  is shown in Figure 4-9.








4.4     POLYCYCLIC ORGANIC MATTER  (POM) SAMPLING
        POM sampling was  conducted  at Location 5 for baseline conditions and



for the optimum  low NOx condition with the  boiler operating at 70 - 75 percei



of capacity.  One POM  test,  at  the  boiler outlet,  was run at each condition.
                                       4-23                     KVB72-806015-1308

-------
  160
  140
  120
cTioo
0\°
on
-P
rd
   80
a
   60
   40
   20
                                               = BASELINE          _

                                           ["I  = LOW NO

                                           FUEL:  Hogged Wood (100%)
                                                 except for Tests
                                                 5/7-1C and 5/7-2C ~"
                                                 where 9% of heat
                                                 input came from
                                                 No.  6 Oil.

                                           STEAM FLOW: 18.3 Kg/s
                                                     (145,000 Ib/hr)
                                                          5/5-1
                              5/5-2
              5/7-2B
                    5/7-2
                        5/7-2A
                                                                      10
                              STACK OXYGEN, % DRY
Figure 4-9.
                NO emission as a function of stack oxygen in the
                baseline and low- NO  configurations.
                                  4-24
                                                      KVB72-806015-1308

-------
The sampling system is a modified Method  5  sampling  train  developed by
Battelle Columbus Laboratories.  A  combination  of  conventional  filtration with
collection of organic vapors by means  of  a  high surface  area  polymeric
adsorbent (XAD-2) proved highly efficient for collection of all but the  more
volatile organic species.  The modified sampling system  consists of the
standard EPA train with the adsorbent  sampler  (Figure  4-10) located between
the filter and the impingers.  With this  system filterable particulate can be
determined from the filter catch and the  probe  wash  according to Method  5,
whereas the organic materials present  can be determined  from  the analysis of
the filterable particulate and the  adsorbent sampler catch.   The impingers  are
only used to protect the dry-gas meter, and their  contents are  discarded.
4.4.1   POM Emissions
        Sample time was  extended  to  two  hours  to  provide  a  large  enough  sample
for analysis.  Following the  sampling  period,  the organic resin module was
sealed and returned to the  laboratory  for  analysis.   The  sampling probe  and
glassware were washed with  a  50-50 mixture of  methylene chloride  and  methanol
per laboratory instructions.   The filter and wash were  also sent  to the
laboratory following weighing.
        These samples were  analyzed  by capillary-El,  GC-MS  utilizing  a  30M
SE-52 column with hydrogen  as  a carrier  gas.  All data  were collected by
single ion monitoring  (SIM) to improve selectivity and  sensitivity.
        The extractions  from  the  samples received from  KVB,  reagent blanks,
calibration standards and extracts from  spiked filters  and  XAD-2  cartridge
blanks were analyzed using  GC/MS  procedures.
        After running all the  samples, the areas  of the peaks  were determined
by use of a computer.  For  the calibration solutions,  the peak area of  the
standard compound is ratioed  to the  nearest internal standard  peak.  A  least
square analysis is performed which provides a  slope and intercept used  to
quantitate the sample solutions.  The  intercept is taken  to represent the
quantitation limit.  A correlation coefficient is also  calculated which  gives
an indication of the linearity of the  calibration.  An  acceptable curve  has a
correlation coefficient  of  at  least  0.990.
                                       4-25                     KVB72-806015-1308

-------
 GLASS WATER
 JACKET
 8-MM GLASS
 COOLING COIL
    ADSORBENT
  GLASS WOOL PLUG
  RETAINING SPRING
                                           28/12 BALL JOINT
                                            FLOW DIRECTION
                                        GLASS FRITTED
                                        DISC
FRITTED STAINLESS STEEL DISC

15-WM SOLV-SEAL JOINT
Figure 4-10.   Mark III adsorbent  sampling  system.

                          4-26                      KVB72-806015-1308

-------
        The correlations obtained were as follows:
        Phenanthrene               0.997
        Anthracene                 0.997
        Fluoroanthrene             0.998
        Phyrene                    0.998
        Benzo(a)anthracene         0.996
        Chrysene                   0.997
        Benzo(e)pyrene             0.997
        Benzo(a)pyrene             0.997
        Perylene                   0.997
        Indeno-pyrene              0.994
        Benzo(g,h,i)perylene       0.994
        Coronene                   0.999

        Using the computer determined areas  for  the  sample  peaks,  these  areas
were ratioed  to that of the nearest  internal standard.   These  ratios  were  then
inserted into the appropriate quantitation equation  to  determine  the  total
amount of the particular POM  (expressed in yg) in the sample.   Finally,  the
values for the reagent blank were subtracted from the amount calculated  to
yield the POM actually in the sample.  The results are  shown in Table 4-6.
        •The results of the analyses  are presented in yg per total  sample.   The
quantitative detection limit was 0.5 yg,  thus  samples with  POM's  present at
levels lower  than this are reported  as <0.5  yg (the  standard deviation at
lower levels was prohibitively high  for accurate quantitation).   Samples
reporting POM values of ND (none detected) are at a  level of less  than 0.1  yg
(the approximate qualitative detection limit).   The  standard deviation on
points around 0.5 yg averaged around ±20  percent,  at levels around 5  yg  it
averaged around ±15 percent, and at  levels above 12  yg  the  standard deviation
averaged around ±10 percent.
        The POM analyses for the low NO   condition are  presented  in the  first
                                       X
six columns of Table 4-6 for the XAD-2 module, the filter/probe wash/cyclone
and total.  The corresponding data for the unmodified or baseline condition
are shown in  the last six columns.   The POM  emissions in the  low  NO,,  condition
                                                                    A.

                                      4-27                     KVB72-806015-1308

-------
                                                                  TABLE  4-6.

                                  SUMMARY OF  POM  ANALYSES FOR LOCATION 5  -  WOOD-FIRED SPREADER STOKER
POM
Phenanthrene
Anthracene
Methyl Anthracenes/
Phenanthrene a
Fluoranthene
Pyrene
Methyl Pyrene/Fluoranthene
Benzo (c)phenanthrene
Benz (a)anthracene
Chrysene
Methyl Chrysenes
Dimethylbenz anthracenes
Benzofluoranthenes
Benzo(a)pyrene
Benzo (e)pyrene
Perylene
Methylcholanthrenes
Indeno (1 ,2 ,d-cd)pyrene
Benzo (g,h,i)perylene
Dibenz anthrancenes
Dibenzpyrenes
Co rone ne
TOTAL **
Sample volume, m
Low NO Test
X
XAD-2
Pg
0.7
ND*
0.7
<0.5
ND
<0.5
ND
ND
ND
ND
ND
ND
ND
ND
ND
<1.2
ND
ND
ND
ND
ND
pg/m3
.70
ND
.70
<.5
ND
<.5
ND
ND
ND
ND
ND
ND
ND
ND
ND
<1.2
ND
ND
ND
ND
ND
1.4O.6
Probe Rinse,
Cyclone & Filter
Pg
2.5
<0.4
2.7
o.e
1.0
0.7
ND
ND
<0.5
ND
ND
ND
<0.6
0.6
ND
<1.2
ND
<1.2
ND
0.1
2.2
pg/m
2.5
<.4
2.7
0.8
1.0
.7
ND
ND
<.5
ND
ND
ND
<.6
.6
ND
<1.2
ND
<1.2
ND
.1
2.2
10.6<14.5
Total
pg
3.2
<.4
3.4
0.8<1.3
1.0
.7<1.2
ND
ND
<.5
ND
ND
ND
<.6
.6
ND
<2.4
ND
<1.2
ND
.1
2.2
pg/m
3.2
<.4
3.4
.8<1.3
1.0
.7<1.2
ND
ND
<.5
ND
ND
ND
<.6
.6
ND
<2.4
ND
<1.2
ND
.1
2.2
12.0<18.1


Baseline Test
XAD-2
P9
0.9
ND
1.0
ND
ND
<0.5
ND
ND
ND
ND
<0.5
<0.5
ND
ND
ND
<1.2
ND
ND
ND
ND
0.1
pg/m3
.7
ND
.a
ND
ND
<.4
ND
ND
ND
ND
<-4
<.4
ND
ND
ND
<-9
ND
ND
ND
ND
.1
1.6<3.7
Probe Rinse.
Cyclone & Filter
pg
1.6
ND
1.1
<0.5
<0.5
<0.5
ND
ND
ND
ND
ND
ND
ND
ND
ND
<1.2
ND
ND
ND
ND
ND
pg/m
1.0
ND
.8
<.4
<.4
<.4
ND
ND
ND
ND
ND
ND
ND
HD
ND
<.9
ND
ND
ND
ND
ND
1.8<3.9
Total
pg
2.5
ND
2.1
<.5
<.5
<1.0
ND
ND
ND
ND
<.5
<.5
ND
ND
ND
<2.4
ND
ND
ND
ND
0.1
pg/»3
1.7
ND
1.6
<.4
<.4
<.8
ND
ND
ND
ND
<.4
<.4
ND
ND
ND
<1.8
ND
ND
ND
ND
.1
3.4<7.6


 *ND
      Not Detected,  less than 0.1 yg
**Two  totals are shown, e.g., 1.4<3.6  where 1.4 is total of all quantified amounts and
  3.6  is total of quantified amounts plus all values indicated as <,  which indicates
  that a compound was  observed but cannot be quantified at a value below the amount
  shown.

-------
were two to three times higher  than  the  emissions at baseline.  This large

difference could be due more  to the  fuel change  than to the combustion modifi-

cation although the trend of  higher  POM  with lower NO  has been observed

before (Ref. 1 ).

        A comparison of previous  POM measurements of coal- and oil-fired

boilers with the present data is  provided in Table 4-7.  The POM emissions

from the hogged fuel boiler in  the  low NO  mode  were higher than those from
                                          X
the two coal-fired spreaker-stokers  and  the residual-oil fired unit tested on
an EPA thirty-day field test  program (Refs. 1,2,3).  The baseline POM

emissions, however, were approximately equal to  those of one coal-fired

spreader-stoker (Ref.  1) and  were slightly less  than the oil-fired unit.  In

both baseline  and low  NO  conditions,  the POM emissions from the hogged fuel
                        X
boiler at Location 5 were well  below those of the pulverized coal unit tested

on the Thirty-Day program  (Ref. 4).

        The following  conclusions were made based on the present study:

        1 . POM was higher in  the  low NO   mode than under baseline
           conditions.

        2. POM emission for this  hogged  fuel boiler was greater in the
           low NO  configuration  than two coal-fired spreader stokers
           and one residual-oil-fired boiler tested previously by KVB;
           at  baseline, the POM emission was on  a par with that of one
           coal-fired  spreaker-stoker and slightly below that of the
           oil-fired unit.

        3. At  baseline and in the low NO  mode,  POM emission was well
           below that  of a pulverized coal boiler previously tested.

        4. The POM emissions  are  likely  to be dependent on the type of
           wood burned.  The  wood fuel composition could not be con-
           trolled because of the nature of the  wood sources.
           Therefore,  further study  is required  to determine the exact
           extent of its influence  on POM emissions.
                                       4-29                    KVB72-806015-1308

-------
                                  TABLE 4-7.    POM EMISSION FROM OIL-,  COAL-,  AND WOOD-FIRED

                                            BOILERS:   COMPARISON WITH  PRESENT DATA
i
U)
O

30-Day
(Ref .
30-Day
(Ref.
30-Day
(Ref.
30-Day
(Ref.
Hogged
Location
Tests, Site 1
1)
Tests, Site 2
2)
Tests, Site 3
4)
Tests, Site 4
3)
Fuel Boiler
Boiler Capacity Control
and Type Fuel Technology
Spreader-Stoker Coal LEA
12.6Kg/s (100,000 Ib/h)
Residual-Oil-Fired No. 6 Oil SCA (BOOS)
lO.OKg/s (79,000 Ib/h)
Pulverized Coal-Fired Coal SCA + LEA
32.8Kg/s (260,000 Ib/h)
Spreader-Stoker Coal LEA
20.2Kg/s (160,000 Ib/h)
3
Total POM, yg/m
Low NO
Baseline x
1.6K10.8 4.17<10.9
7.59<11.4 4.57<9.68
66.9<70.8 64.57<68.21
1.24<2.52 1.035<2.76
Spreader-Stoker Wood LEA + Air Register 3.6<7.9 12.0<18.1
      Location 5  (Present Data)  25.2Kg/s  (200,000  Ib/h)
Adjustment

-------
                                  SECTION  5.0

                                  REFERENCES
1.       Carter,  W. A.,  and Buening, H. J., "Thirty-Day Field Tests of
        Industrial Boilers, Final Report, Site 1 - Coal-Fired Spreader
        Stoker," EPA-600/7-80-085a, April 1980.

2.       Carter,  W. A.,  and Tidona, R. J., "Thirty-Day Field Tests of
        Industrial Boilers, Final Report, Site 2 - Residual-Oil-Fired Boiler,"
        EPA-600/7-80-085b, April 1980.

3.       Carter,  W. A.,  and Hart, J. R., "Thirty-Day Field Tests of Industrial
        Boilers, Final Report, Site 4 - Coal-Fired Spreader Stoker," EPA-
        600/7-80-085d,  April 1980.

4.       Carter,  W. A.,  and Buening, H. J., "Thirty-Day Field Tests of
        Industrial Boilers, Final Report, Site 3 - Pulverized Coal-Fired
        Boiler," EPA-600/7-80-085c, April 1980.
                                      5-1                    KVB72-806015-1308

-------







cr>
1
h- '






§
-j
to
I
-806015-:
To Obtain
g/Mcal
106 Btu
Btu
lb/106 Btu
ft
in.
ft2
ft3
Ib
Fahrenheit
Fahrenheit
psig
psia
iwg (39.2 °F)
106 Btu/hr
GJ/hr
*These conversions
The values given
From
ng/J
GJ
gm cal
ng/J
m
cm
2
m
3
m
kg
Celsius
Kelvin
Pa
Pa
Pa
MW
MW
depend
are for
                                                    SECTION 6.0
                                               CONVERSION FACTORS
                                       SI Units to Metric or English Units
                                        Multiply By
                                        0.004186
                                        0.948
                                        3.9685x10
                                        0.00233
                                        3.281
                                        0.3937
                                        10.764
                                        35.314
                                        2.205
-3
                                        tp =  9/5 (tc)  + 32
                                        tp =  1.8tK -  460
                                        P   .   =  (P  )(1.450xlO~
                                        psig     pa
                                        P   .   =  (P  ) I
                                        psia     pa
                                        P.    = (P   ) (4
                                        iwg     pa
                                        3.413
                                        3.60
                                                                              To Obtain ppm
                                                    Multiply*
                                                  Concentration
at 3% O0 of
Wood
CO
HC
NO or NO
x
SO. or SO
2 x
Oil Fuel
CO
HC
NO or NO
x
SO,, or SO
2 x
Coal Fuel
CO
HC
NO or NO
x
SO or SO
in ng/J by
2.31
4.05
1.41
1.01
2.93
5.13
1.78
1.28
2.69
4.69
1.64
1.18
o
03

-------
                                          English and Metric Units to SI Units
                                                                                                           Multiply*
 1
 M
 --J
 NJ
 I
 CO
 O
 cn
 o
 M
 Ln
 I
 M
 OJ
 O
00

To Obtain
ng/J
ng/J
GJ
m

cm
2
m
3
m
kg
Celsius
Kelvin

Pa

Pa

Pa
MW
MW



*mv.^<-'^ ^~~,.~ 	

From
lb/106 Btu
g/Mcal
106 Btu
ft

in.
2
ft
3
ft
Ib
Fahrenheit
Fahrenheit

psig

psia

iwg (39.2 °F)
106 Btu/hr
GJ/hr





Multiply By
430
239
1.055
0.3048

2.54

0.0929

0.02832
0.4536
tc = 5/9 (tp - 32)
t = 5/9 (t - 32) + 273
K F
P = (P . + 14.7) (6.895x10 )
pa psig
P = (P . ) (6.895x10 )
pa psia
P = (P. ) (249.1)
pa iwg
0.293
0.278



i
To Obtain C
ng/J of in
Wood
CO
HC
NO or NO (as NO )
x 2
SO or SO
2 x


Oil Fuel
CO
HC
NO or NO (as NOJ
x 2
SO,, or SO
2 x


Coal Fuel
CO
HC
NO or NO (as NO0)
x 2
SO^ or SO
2 x
Toncentrat
ppm @ 3% i

0.432
0.247
0.710

0.988




0.341
0.195
0.561

0.780




0.372
0.213
0.611

0.850

The values given  are  for typical fuels.

-------
                                APPENDIX A-1.0
                GASEOUS AND PARTICULATE EMISSIONS  TEST  METHODS
                              AND INSTRUMENTATION
         All emission measurement instrumentation  was  carried  in  a  9.8  m x
2.4 m (8 x 42 ft) mobile laboratory trailer.  A plan view  of the  trailer is
shown in Figure A-1.  The gaseous species  measurements were made  with
analyzers located in the trailer.
         The emission measurement instrumentation  used was the following:
               TABLE A-1.  EMISSION MEASUREMENT INSTRUMENTATION
  Species
             Manufacturer
Measurement Method
Model
 No.
Hydrocarbon
Carbon Monoxide
Oxygen
Carbon Dioxide
Nitrogen Oxides
Particulates
Sulfur Dioxide
Particle Sizing
Smoke Spot
Opacity
Sulfur Oxides
  (sox)
         Beckman Instruments
         Beckman Instruments
         Teledyne
         Beckman Instruments
         Thermo Electron Co.
         Joy Manufacturing C
         DuPont Instruments
         Brink
         Bacharach
Flame lonization
IR Spectrometer
Polarographic
IR Spectrometer
Chemiluminescent
EPA Method 5 Train
UV Spectrometer
Cascade Impactor
ASTM 2156-65
EPA Method 9
Goksoyr-Ross
402
865
326A
864
10A
EPA
400
BMS1 1
RCC
A-1 .1
GAS SAMPLING AND CONDITIONING SYSTEM
         A flow schematic of the flue  gas  sampling  and  analyzing system is
shown in Figure A-2.  The sampling system  uses  three  positive-displacement
diaphragm pumps to continuously draw flue  gas  from  the  stack  into the
                                      A-1
                                                     KVB72-806015-1308

-------
Calibration  Gas
  Bottles
                                               Door and Stairs
                                                                                                    Spare Calibration

                                                                                                     Gas Bottles
       Air

   Conditioning/
     Heater
oooo
f~\ Sample Handling/
^ Conditioning
O Room
	 -— TT-

. . Counter Top/
|Sirik~[ Cabinets
Instrument
Console
Air Conditioning/Heating Duct and Ven
J J 11 J

Counter Top/Cabinets
ts
I
•Over
Lv ^» •
••» ••> «^ •

Fume
Hood
OOOO
. ErfiMfi. S-ta^Aifi. 	
/Storage
Room
                                                                          2.4m x 12.8m Double Axle

                                                                          Semi Trailer
to
 i
oo
o
CTi
o
H
Ul
             Figure  A-l.  Instrumentation trailer floor  plan.
o
CO

-------
 I
oo
o
cr>
o
M
ui
 i
                                                                                                   Hot

                                                                                                 Sample        Dry Sample Lines

                                                                                                  Line    (Typical Set-Up Six Lines):
                              o
                                 Hot Pump

                                 Pressure
             Hot Pump
             Vacuum
               HOTBOX
                                   Hi-Span
                                   Ini-opan

                                   (Dry  Sample)
                                                                                                                    Filters (6)

                                                                                                                     (7 micrometers)
                                                                                             Condenser
                                                                                               16
                                                                                              Hot/Cold
                                                                                              Switch
                                                                    | Manifold   )


                                                                          1
                                                             Refrigeration Condenser


                                                         Q-Osample Pressure
                                                                                                                         span
OJ
o
oo
Figure  A-2.   Flue gas sampling and analyzing  system.

-------
laboratory.  The sample pumps pull from six unheated  sample lines.   Selector
valves allow composites of up to six points to  be  sampled at one time.  The
probes are connected to the sample pumps with 0.95 cm (3/8") or 0.64 cm (1/4")
nylon line.  The positive displacement diaphragm sample  pumps provide unheated
sample gas to the refrigerated condenser  (to reduce the  dew point to 35°F), a
rotameter with flow control valve, and to  the G>2,  NO,  CO,  and C02
instrumentation.  Flow to the individual analyzers is measured and controlled
with rotameters and flow control valves.   Excess sample  is vented to the
atmosphere.
         To obtain a representative sample for  the analysis of NO2,  S02 and
hydrocarbons, the sample must be kept above its dew point, since heavy hydro-
carbons may be condensible, and S02 and NO2 are quite soluble in water.  For
this reason, a separate electrically-heated sample line  is used to bring the
sample into the mobile laboratory for analysis.  The  sample line is  0.95 cm
(3/8-inch) Teflon line, electrically traced and thermally insulated  to main-
tain a sample temperature of up to 478 K  (400°F).   A  heated diaphragm pump
provides hot sample gas to the hydrocarbon, SO2 and NOX  analyzers.

A-1.2    INSTRUMENTAL CONTINUOUS MEASUREMENTS
         The laboratory trailer is equipped with analytical instruments to
continuously measure concentrations of NO, NO_, CO, CO_,  0~, S0_, and hydro-
carbons.  All of the continuous monitoring instruments and sample handling
system are mounted in the self-contained mobile laboratory.  The entire system
requires only connection to on-site water, power,  and sampling lines to become
fully operational.  The instruments themselves  are shock mounted on  a metal
console panel.  The sample flow control measurement,  and selection,  together
with instrument calibration are all performed from the console face.  Three-
pen recorders provide a continuous permanent record of the data taken.  The
sample gas is delivered to the analyzers at the proper condition and flow rate
through the sampling and conditioning system described in the previous
section.   The sections below describe the  analytical  instrumentation.
                                      A-4                    KVB72-806015-1308

-------
A-1.2.1  Nitric Oxide  (NO)  and  Total  Nitrogen Oxides (NO )
                                                         X
         Both the total  nitrogen  oxides  (NO ) and nitric oxide (NO) concen-
                                            A.
trations are measured  from  a  sample gas  obtained using a heated sample line at
394 K  (250°F).  In addition,  the  nitric  oxide concentrations are measured
sequentially from samples obtained using the unheated sample line that is
connected to the same  analyzer  in the laboratory trailer.  In the latter case,
water is first removed from the sample gas  by a refrigeration unit.  The
analytical instrument  that  is used for all  of these measurements is the Thermo
Electron Model 1OA chemiluminescent gas  analyzer.
         For NO analyses, the sample  gas is passed directly into the reaction
chamber where a surplus  of  ozone  is maintained.  The reaction between the NO
and the ozone produces light  energy proportional to the NO  concentration which
is detected with a photomultiplier and converted to an electrical signal.  Air
for the ozonator is drawn from  ambient through an air dryer and a 10-micro-
meter filter element.
         The chemiluminescent reaction with ozone is specific to NO.  To
detect NO2, a thermal  converter has been designed to dissociate the NO2 to NO
by the bi-molecular reaction: 2 NO2 ->• 2  NO  + O-.  A model 700 thermal
converter is used in conjunction  with the chemiluminescent  gas analyzer as
shown in Figure A-3.   The converter is a coil of resistance-heated stainless
steel tubing whose purpose  is to  drive the  N02/N0 ratio to  its chemical
equilibrium value at the converter temperature and pressure.  The unit is
designed to operate at a temperature  of  923 K (1200 °F) and pressure of .3 kPa
(10 torr).  For these  conditions  and  typical stack gas 02 concentrations, the
equilibrium N02 concentration is  0.2  percent of the total NOx concentration.
Therefore, when a gas  sample  containing  any N02 is passed through the con-
verter, essentially all  the NO? would be converted to NO.  The resulting total
NO is then measured using the chemiluminescent analyzer and the difference
between the actual NO  and the "total  NO" would be the sample N02
concentration.  The "total  NO"  is interpreted as NO .
                                                    .A
                                       A_5                    KVB72-806015-1308

-------
                    Capillary
                    0.020cm x 3.8cm
     Optical
    .Filter

   Photo-
   multiplier

Oxygen
Regulator
                  Analyzer
                             -J
                    Reference        I



                           Sample V.     )
                           Regulator"-]—
                      Reaction
                      Chamber
                                             Sample
                                             Gage
                                        I
                                              Flow
                                              Meter

                                          Capillary
                                          0.013cm
                                           x
                                          2.9cm
                          Inlet
                                                Outlet
                                        Capil-
                                        -lary
                                        3.051cm
                                         x
                                        3.8cm
      Oxygen
      or
      Dry Air
unit

NO
1
Anplifier

r-
I Me
-. -«._ J_ i
i
HV
5uppl\


— o
Low p
>ter p
i 1—
r


— '
1
                               fciS
                               Plugged
                                     si—6—6—1
                                                   Room
                                                   Air
                                                  Water
                                                   Trap
                              —i      Trap

                              ^T~
                                            To
                                            Bypass
                                            Pump
                               0.013cm 0.051cm i
                                Capillaries   i


                                  Model  700   !
                                             Heated
                                             Sample
                                             Line
Figure A-3.
Schematic of NO /NO  chemiluminescent analysis system.
               X
                                A-6
                                          KVB72-806015-1308

-------
         N0~ may react upon contact with  H^O  (liquid phase)  to form HNO,
(nitric acid).  Under field test  conditions,  the  exhaust gas may contain
significant H20 (depending upon the process and  the  ambient  meteorological
conditions), and it is necessary  to convert the  N02  to NO before the H20 is
allowed to condense in the sampling system.   By  using the heated sample line
and the Thermo Electron Model  700 heated  NOX  module,  NO2 concentrations will
effectively be measured.  In reference  to Figure  A-3, the sample is maintained
above the H20 dew point up to  and through the 127 ym (0.005  in.) capillary in
the heated module.  Downstream of this  capillary,  the flow network is main-
tained at 1.3 kPa (10 torr), where the  partial pressure of the H_0 in the
sample is sufficiently low to  prevent any condensation at ambient temperature.
         When using the heated system,  NO, NO0,  and  NO  are  measured on a wet
                                              £         X
basis.  When not using the heated system,  a condenser is placed upstream of
the analyzer and NO is measured on a dry  basis.
         Specifications
         Accuracy:  1% of full scale
         Span stability:  ± 1% of full  scale  in  24 hours
         Zero stability:  ± 1  ppm in 24 hours
         Power requirements:   115 ± 10V,  60 Hz,  1000 watts
         Response:  90% of full scale in  1 sec  (NOX  mode);
                    0.7 sec (NO mode)
         Output:  4-20 ma
         Sensitivity:  0.5 ppm
         Linearity:  ± 1% of full scale
         Vacuum detector operation
         Range:  2.5, 10, 25,  100, 250,  1000, 2500,  10,000 ppm full scale
                                       A_7                    KVB72-806015-1308

-------
A-1.2.2  Carbon Monoxide  and  Carbon  Dioxide (CO and CO-)

         Carbon monoxide  and  carbon  dioxide concentrations are measured using
Beckman Model  864 and  865 short-path-length nondispersive infrared analyzers
(see Figure A-4).  These  instruments measure the differential in infrared
energy absorbed from energy beams  passed  through a reference cell  (containing
a gas selected to have minimal  absorption of infrared energy in the wavelength
absorbed by the gas component of  interest)  and a sample cell through which  the
sample gas flows continuously.  The  differential absorption appears as a
reading on a scale of  0%  to 100%  and is  then related to the concentration of
the species of interest by calibration curves  supplied with the instrument.  A
linearizer is  supplied with each  analyzer to provide a linear output over the
range of interest.  The operating  ranges  for the CO analyzer are 0-100 and  0-
2000 ppm, while the ranges for  the C02 analyzer are 0-5% and 0-20%.
         Specifications
         Span  stability:   ± 1%  of  full scale in 24 hours
         Zero  stability:   ± 1 ppm  in 24 hours
         Ambient temperature  range:   273  to 322 K (32°F to 120°F)
         Line  voltage:  115 ± 15V  rms
         Response:  90% of full scale in  0.5 or 2.5 sec
         Linearity:  Linearizer board installed for one range
         Precision:  ± 1% of  full  scale
         Output:  4-20 ma
A-1.2.3  Oxygen (O2)

         A Teledyne Model 326A  oxygen analyzer is used to automatically and
continuously measure the  oxygen content of  the flue gas sample.  The analyzer
utilizes a micro-fuel cell which is  specific for oxygen, has an absolute zero,
and produces a linear output  from  zero through 25 percent oxygen.  The micro-
fuel cell is a sealed electrochemical transducer with no electrolyte to change
or electrodes to clean.   Oxygen in the flue gas diffuses through a Teflon
membrane and is reduced on the  surface of the  cathode.   A corresponding

                                      A-8                   KVB72-806015-1308

-------
    Reference
       Cell
    Detector
                          Infrared  Source
  - Sample in from source

Sample
Cell
mm
  *• Sample out

 Diaphragm
 Distended
                            Absorbs infrared
                            energy in region of
                            interest
                            Other molecules
                 Control
                  Unit
Figure A-4.   Schematic of NDIR analyzer.
                    A-9
                                            KVB72-806015-1308

-------
oxidation occurs at the anode internally and an electric  current is  produced
that is proportional to the concentration of oxygen.   This  current is measured
and conditioned by the instrument's electronic circuitry  to give an  output in
percent CU by volume for operating ranges of 0% to  5%,  0% to 10%,  and 0%
to 25%.
         Specifications
         Precision:  ± 1% of full scale
         Response:  90% in less than 40 sec
         Sensitivity:  1% of low range
         Linearity:  ± 1% of full scale
         Ambient temperature range:  273 K to 325 K (32°F to 125°F)
         Fuel cell life expectancy:  40,000% +-hrs
         Power requirement:  115 VAC, 50-60 Hz, 100 watts
         Output:  4-20 ma
A-1.2.4  Total Hydrocarbons (HC)
         Hydrocarbon emissions are measured using a Beckman Model 402 high-
temperature hydrocarbon analyzer.  The analyzer utilizes  the flame ionization
method -of detection which is a proven technique for a  wide  range of  concen-
trations (0.1 to 120,000 ppm).  A flow schematic of the analyzer is  presented
in Figure A-5.  The sensor is a burner where a regulated  flow of sample gas
passes through a flame sustained by regulated flows of air  and a premixed
hydrogen/nitrogen fuel gas.  Within the flame the hydrocarbon components of
the sample stream undergo a complex ionization that produces electrons and
positive ions.  Polarized electrodes collect these  ions,  causing current to
flow through electronic measuring circuitry.  Current  flow is proportional to
the rate at which carbon atoms enter the burner.
         The analysis occurs in a temperature-controlled  oven.  The  sample is
extracted from the stack with a stainless steel probe  which has been thermally
treated and purged to eliminate any hydrocarbons existing in the probe
itself.  An insulated heat-traced teflon line is used  to  transfer the sample

                                      A-10                   KVB72-806015-1308

-------

Figure A-5.   Flow schematic of hydrocarbon analyzer  (FID).
                           A-ll
                                                    KV372-806015-1308

-------
to the analyzer.  The entire heated network is maintained at  a  temperature
sufficient to prevent condensation of heavier hydrocarbons.
         The flame ionization detector is calibrated with methane,  and  the
total hydrocarbon concentration is reported as the methane  equivalent.   FID's
do not respond equally to all hydrocarbons but generally provide  a  measure of
the carbon-hydrogen bonds present in the molecule.  The FID does  not detect
pure carbon or hydrogen.
         Specifications
         Full-scale sensitivity:  adjustable from 5 ppm CH^ to  10%  CH^
         Ranges:  Range multiplier switch has 8 positions:  X1,  X5,  X10,  X50,
                  X100, X500, X1000, and X5000.  In addition, span  control
                  provides continuously variable adjustment within  a dynamic
                  range of 10:1
         Response time:  90% full scale in 0.5 sec
         Precision:  ± 1% of full scale
         Electronic stability:  ± 1% of full scale per 24 hours with ambient
                                temperature change of less  than 5.6 K (10°F)
         Reproducibility:  ± 1% of full scale for successive  identical  samples
         Analysis temperature:  ambient
         Ambient temperature:  273 K to 317 K (32°F to 100°F)
         Output:  4-20 ma
         Air requirements:  250 to 400 cc/min of clean, hydrocarbon-free
                            air, supplied at 2.07 x 105 to  1.38 x 106
                            n/m2 (30 to 200 psig)
         Fuel gas requirements:  75 to 80 cc/min of fuel consisting of
                                 100 percent hydrogen supplied  at 2.07  x 1O5
                                 to 1.38 x 106 n/m2 (30 to  200  psig)
         Electric power requirements:  120 V, 60 Hz
         Automatic flame indication and fuel shut-off valve


                                      A-12                  KVB72-806015-1308

-------
A-1.2.5  Sulfur Dioxide
         A Dupont Model 400 photometric  analyzer is used for measuring SO-.
This analyzer measures the difference  in absorption of two distinct
wavelengths  (ultraviolet) by  the  sample.   The  radiation from a selected light
source passes through the sample  and  then into the photometer unit where the
radiation is split by a semi-transparent mirror into two beams.  One beam is
directed to a phototube through a filter which removes all wavelengths except
the "measuring" wavelength, which is  strongly  absorbed by the constituent in
the sample.  A second beam falls  on a  reference phototube, after passing
through an optical filter which transmits only the "reference" wavelength.
The latter is absorbed only weakly, or not at  all, by the constituent in the
sample cell.  The phototubes  translate these intensities to proportional
electric currents in the amplifier.   In  the amplifier, full correction is made
for the logarithmic relationships between the  ratio of the intensities and
concentration or thickness  (in accordance with Beer's Law).  The output is
therefore linearly proportional,  at all  times, to the concentration and thick-
ness of the  sample.  The instrument has  a lower detection limit of 2 ppm and
full scale ranges of 0-200 and 0-2000  ppm.
         Specifications
         Noise:  Less than 1/4%
         Drift:  Less than 1% full scale in 24 hours
         Accuracy:  (± 1% of  analyzer  reading) + (+ 1/4% of full scale range)
         Sample cell:  304 stainless  steel,  quartz windows
         Flow rate:  0.05 dm3/s  (6 cfh)
         Light source:  Either mercury vapor,  tungsten, or "Osram"
                        discharge type lamps
         Power rating:  500 watts maximum, 115V,  60 Hz
                                       A-13                   KVB72-806015-1308

-------
         Reproducibility:  1/4% of scale
         Electronic response:  90% in 1 sec
         Sample temperature:  378 K  (220°F)
         Output:  4-20 ma d.c.

A-1.3    PARTICULATE MATTER TOTAL MASS CONCENTRATION
         Particulate matter is collected by  filtration  and  wet impingement in
accordance with US-EPA Method No. 5.  Nomograph  techniques  are utilized to
select the proper nozzle size and to set the isokinetic flow rates.
         Gas samples for particulate sampling can be  taken  from the  same
sample port as those for gas analysis and passed through the Joy Manufacturing
Company Portable Effluent Sampler.  This system, which  meets the EPA design
specifications for Test Method 5, Determination  of  Particulate Emissions from
Stationary Sources (Federal Register, Volume 36, No.  27,  page 24888,
December 24, 1971, and revisions thereof) is used to  perform both the initial
velocity traverse and the particulate sample collection.
         Dry particulates are collected in the heated case  that may  contain a
cyclone to separate particles larger than 5 ym and  a  110-mm glass-fiber filter
to retain particles as small as 0.3 )am.  Condensible  particulates are col-
lected in four Greenburg-Smith impingers immersed in  a  chilled water bath.
         The sampling probe is positioned through an  exhaust port and attached
to the sampling box.  The probe consists of  a sampling  nozzle, heated probe,
gaseous probe,  thermocouple, and pitot tube.  The ball  joint from the heated
probe connects to the cyclone and glass filter holder assembly.  These assemb-
lies are positioned in the heated sampling box which  is maintained at 394 K
(250°F), in order to eliminate condensation.  The sample then passes from the
heated section to four Greenburg-Smith impingers immersed in an ice  bath.
Only the second impinger has the original tip, the  other three have  had the
tip removed to decrease the pressure drop through them.  The first and second
impingers are filled with 150 milliliters of distilled/deionized water.  The
third impinger is left dry.  The fourth impinger is filled  with approximately
200 grams of indicating silica gel to remove entrained  water.  The use of

                                      A-14                   KVB72-806015-1308

-------
silica gel assures  that  a  dry  sample  is  delivered to the meter box.  After
sampling, the spent silica gel is  discarded and not used for any further
analysis.
         An umbilical  cord connects  the  last impinger,  the pitot tube, and the
heating elements to the  meter  box  which  is located in a convenient place
within 15 m of the  sampling ports.   The  meter box contains a vacuum pump,
regulating valves,  instantaneous and  integrating flow meters, pitot tube
manometers, vacuum  gauge,  and  electrical controls.
         Particulate matter (solids  and  condensibles) is collected in three
discrete portions by the sampling  train: the probe and  glassware upstream of
the filter; the filter;  and the wet  impingers.   The probe and glassware are
brushed and rinsed  with  acetone; the  matter is  captured for gravimetric analy-
sis.  The probe and glassware  are  then rinsed with distilled water and the
rinsings transferred to  a  second container for  analysis.  The filter is desic-
cated and analyzed  gravimetrically.   The combined impinger liquid is heated to
drive off uncombined water and the residue retained for analysis.  The parti-
culate matter analysis is  illustrated schematically in  Figure A-6.
         US EPA Method 5 considers the particulate matter captured in con-
tainers (1) and (3); the filter, probe brushing,  and probe acetone rinse.  As
EPA source standards are based on  solid  particulates only, care is taken to
differentiate between  solid and the  total (including condensible) particu-
lates.  The water wash is  performed  because KVB's test  experience has shown
that a significant  amount  of water-soluble material may sometimes be captured
by the probe.
         The dry sample  volume is  determined with a dry test meter at a mea-
sured temperature and pressure and then  converted to standard conditions.  The
volume of condensed water  in the impingers is measured  in miHiliters and the
corresponding volume of  water  vapor  is then computed at standard conditions.
The dry sample volume and  water vapor volume are then summed to give the total
sample volume.  The dry  sample volume is used in the data reduction
procedures.
                                       A_15                  KVB72-806015-1308

-------
                     PARTICUUTE MATTER MASS DETERMINATION
sampling
train
component
particulate
matter
transfer
procedure
container
processing
analysis
result
Probe Cyclone
Y y y
Brushing

Acetone
Rinse

Distilled
Water
Rinse
\_/ i


Filter
i
t
                                               Impingers
                                                Distilled

                                                Water
                                                Rinse
Bake at 215"F  to drive off uncombined  1^0 and  Acetone
K
                                                        V
            Gravimetric to 0.1 milligrams
                           I
                                            Dig
                       Samples  stored for Compositional Analysis
      Figure A-6.    Processing and analyzing particulate matter.
                                     A-16
                                                               KVB72-806015-1308

-------
         A point of interest  is  the  method  chosen to calculate particulate
emissions in ng/J or  lb/10^ Btu  from the  experimental data.  The particulate
sampling train, properly operated, yields particulate mass per unit flue gas
volume.  Having measured g/m  , it  is necessary to establish the flue gas
volume per unit heat  input if  emissions  in  ng/J are desired.  The original
Method 5 involved determining  a  velocity  traverse of the stack, the cross-
sectional area, the flue flow  rate,  and  fuel heating value.  A revised and
more accurate method  has been  promulgated by the Environmental Protection
Agency that utilizes  a fuel analysis (carbon content, hydrogen content, high
heating value, etc.)  and the  measured 02  in the exhaust to calculate the gas
volume generated in liberating 1.055 GJ  (a  million Btu's).  The velocity
traverse approach generally results  in a  20 to 30 percent higher value and is
believed to be less accurate.
                                                             KVB72-806015-1308
                                      \~ -L /

-------
 A-1.4    SMOKE SPOT
         On  combustion equipment where smoke numbers normally  are  taken,
 such as  oil-fired  boilers, KVB, Inc. determines the smoke number  using
 test procedures  according  to ASTM Designation:  D 2156-65.  The smoke
 number is determined at each combustion modification setting  of the
 unit.  Examples  are baseline, minimum excess air, low load, etc.,  and
 whenever a  particulate concentration is measured.
         Smoke  spots are obtained by pulling a fixed volume of flue gas
 through  a fixed  area of a standard filter paper.  The color (or shade) of
 the  spots that are produced  is  visually matched with a standard scale.
 The  result  is  a  "Smoke Number" which is used to characterize the density
 of smoke in the  flue gas.
         The sampling device is a hand pump similar to the one shown
 in Figure A-7  .   It is a commercially available item that can pass 36,900
 +_ 1650 cu cm of  gas at 289K and 1 atmosphere pressure through an enclosed
 filter paper for each 6.5  sq cm effective surface area of the filter
 paper.
                            Sampling Tube
T-' ..
T '*—-.- . 	 	 	 (



                                                       Handle'
              Figure A-7.   Field-service-type smoke tester.

        The smoke spot sampler is provided with a motor-driven
actuator to ensure a constant sampling rate independent of variations
in stroke rate that can occur when the sampler is operated manually.
        The smoke scale required consists of a series of ten spots  numbered
consecutively from 0 to 9, and ranging in equal photometric steps from white
through neutral shades of gray to black.  The spots are imprinted or other-
wise processed on white paper or plastic stock having an absolute surface
                                      A-18                    KVB72-806015-1308

-------
reflectance  of between  82.5 and  87.5%, determined photometrically.   The  smoke
scale spot number  is defined as  the reduction  (due to  smoke)  in  the  amount of
light reflected by a soiled spot on the filter divided by  10.
        Thus the first  spot, which is the color of the unimprinted scale,  is
No. 0.  In this case there is no reduction in reflected incident light directed
on the spot.  The  last  spot, however, is very dark, reflecting only  10%  of the
incident light directed thereon.  The reduction in reflected  incident light
is 90%, and this spot is identified as No. 9.  Intermediate spot numbers are
similarly established.  Limits of permissible reflectance  variation  of any
smoke scale spot will not exceed +_ 3% relative reflectance.
        The test filter paper is made from white filter paper stock  having
absolute surface reflectance of  82.5 to 87.5%, as determined by  photometric
measurement.  When making this reflectance measurement, the filter paper is
backed by a white  surface having absolute surface reflectance of not less
than 75%.
        When clean air  at standard conditions is drawn through clean filter
paper at a flow rate of 47.6 cu  cm per sec per sq cm effective surface area
of the filter paper, the pressure drop across the filter paper falls between
the limits of 1.7  and 8.5 kPa (1.3 and 6.4 cm of mercury).
        The sampling procedure is exactly that specified in D 2156.  A clean,
dry, sampling pump is used.  It  is warmed to room temperature to prevent
condensation on the filter paper.  When taking smoke measurements in the
flue pipe, the intake end of the sampling probe is placed  at the center  line
of the flue.  When drawing the sample, the pressure in the flue  gas  stream
and the sampler is allowed to equalize after each stroke.
        The smoke density is reported on the Mobile Lab Data Sheet as the  Smoke
Spot Number on the standard scale most closely corresponding to test spot.
Differences between two standard Smoke Spot Numbers are interpolated to
the nearest half number.  Smoke  Spot Numbers higher than 9  are reported
as "Greater than No.  9."
        This procedure  is deemed to be reproducible to within +_ 1/2  of a
Smoke Spot Number  under normal conditions where no oily stain is deposited
on the disk.

                                 A~19                       KVB72-806015-1308

-------
        KVB's field experience with industrial boilers has been that the
human factor involved in the interpretation of the smoke spot by an experi-
enced observer does not cause a significant lack of precision.

A-1.5   OPACITY
        Opacity readings are taken by a field crew member who is a certificated
graduate of a U.S. Environmental Protection Agency approved "Smoke School".
Observations are made at the same time that particulate measurements are
made and as often in addition as deemed necessary to gather the maximum
amount of information.  The procedures set forth in EPA Method 9, "Visual
Determinations of the Opacity of Emissions for Stationary Sources" are
followed.
        Observations are Imade and recorded at 15 second intervals while
particulate concentration is being measured and after the unit has stabilized
at other times.  Before beginning observations, the observer determines that
the feedstock or fuel is the same as that from which the sample was taken
for the fuel analysis.
        Before beginning opacity observations, the observer makes arrangements
with the combustion unit operator to obtain the necessary process data for the
standard KVB Control Room Data Sheet.  The control room data are recorded for
the entire period of observations, as is customarily done by KVB during an
emissions test.  The process unit data that are obtained include:

        a.  Production rates
            1.  maximum rated capacity
            2.  actual operating rate during test
        b.  Control device data
            1.  recent maintenance history
            2.  cleaning mechanism and cycle information
        The observer requests the appropriate plant personnel to
briefly review and comment on the opacity measurements and process
data and the observer comments on:
                                     A-20                    KVB72-806015-1308

-------
        a.   the basis  for  choosing the  observation periods  used.
        b.   why it  is  believed  the periods  chosen constitute periods
             of greatest opacity.
        c.   why the observations span a time period  sufficient to
             characterize the opacity.
        Consideration  is given  to postponing the EPA Method 5  particulate
tests during periods of cloudy  or rainy weather because  of  the inability
of the observer to  monitor the  smoke.
A-l.5.1  Sulfur Oxides (SO )
                          x
        Goksoyr-Ross Method—Wet Chemical  Method
        The Goksoyr-Ross Controlled Condensate  (G/R)  method is used for the
wet chemical SO /SO  determination.   It  is a  desirable method because of its
               fc   «J
simplicity and clean separation of particulate  matter, S0_ and ¥. SO  (SO ) .
This procedure is based on the separation  of  H_SO.  (SO )  from S0_ by cooling
the gas stream below the dew point of H_SO but above the HO dew point.
                                       £,  f±                £
Figure A-8 illustrates schematically  the G/R  test system.
         Particulate matter is first removed from exhaust gas stream by means
of a quartz glass filter placed in the heated glass filter holder.  Tissue-
quartz filters are recommended because of their proven inertness to H2S04.
The filter system is heated by a heating tape so that the gas out temperature
of 533 K (500°F) is maintained.  This temperature is imperative to ensure that
none of the H^SO, will condense in the filter holder or on the filter.
         The condensation coil where the H^SO. is collected is cooled by water
which is maintained at 333 K (140°F) by a heater/recirculator.  This tempera-
ture is adequate to reduce the exhaust gas to below the dew point of H_SO..
                                     A-21                     KVB72-806015-1308

-------
        Adapter for Connecting Hose
Asbestos Cloth
 Insulation

Glass-Cloth Heating
   Mantle   ^~^-

          Stack
         Gas Flow
        Rubber
        Vacuum
         Hose
Recirculator

  'hermometer
                               Styrofoam Ice Chest
Vacuum
Gauge

Dry Test
                                                                       -way
                                                                      Valve
                                                                      Drierite
 Figure A-8.   Schematic of Goksoyr-Ross controlled condensation system  (CCS).
                                     A-22
                                                              KVB72-806015-1308

-------
         Three impingers are shown in Figure A-8.  The  first impinger is
filled with 3 percent H202 to absorb SO2.  The second impinger  is  to remove
carryover moisture and the third contains Drierite and  a  thermometer to mea-
sure the exhaust gas temperature to the dry gas meter and pump.   The sampling
rate is 0.04 dm3/s (0.08 cfm).
A-1.5.2  Analysis Procedure
         For both SO2 and H2S04 determination, the analytical procedure is
identical.  The H2SO, sample is washed from the back part of the  filter holder
and the coil using a 5 percent isopropyl alcohol solution.   The sample from
the first impinger which is assumed to be absorbed and  reacted  S02 in the form
of H2SO^ is recovered with distilled water washing.  The  amount of H2S04 in
the condensate from the coil and from the H202 impinger is  measured by H+
titration.  Bromphenol Blue is used with NaOH as the titrant.

A-1.6    PARTICLE SIZE DISTRIBUTION
         Particle size distribution was determined using  a  Brink Model BMS-11
cascade impactor (Figure A-9).
A-1.6.1  Design
         The Brink sampling probe is a modular cascade  impactor apparatus
suitable for sampling dust in a wide range of flue gas  conditions.   The pri-
mary components are a cyclone separator, a cascade impactor,  an absolute
filter, and a critical orifice.  The cascade impactor comes  with  sampling tips
of various sizes and is constructed of 316 stainless steel.   The  impactor
consists of a number of stages arranged in series.  From  one to five stages
can be used.  Each impactor stage has an orifice and a  collection  cup.   These
are shown in Figure A-9.  The orifice diameter and the  distance between the
orifice and the cup determine the particle collection characteristics of the
stage.  These dimensions are listed in the table in Figure  A-9.  An absolute
filter follows the final impactor stage.  This back-up  filter was  a Gelman
Type A Glass Fiber Filter.  Special collection substrates (e.g.,  glass  fiber,
aluminum foil, etc.) placed on the collection plates can  be  used.   The stack
                                      A-23                     KVB72-806015-1308

-------
                                          DIMENSIONS OF  CASCADE  IMPACTOR JETS

                                                             Dimensions,  Cm
                                                                     Spacing of
                                                                    Jet  Opening*
                                                                       0.747
                                                                       0.533
                                                                       0.419
                                                                       0.282
                                                                       0.220
                                          *From collection cup surface
                               COLLECTION
                                   CUP
                                SPRING


                               •JET SPINDLE

                                GASKET
-3  SLOTS
The in-line impactor has five stages. Particles
in the range of 0.3 to 3.0 microns are collected
by  successive impingement.
                                                        Collection cups are positioned so that
                                                        the distance from the jet decreases
                                                        as the jet diameter becomes smaller.
                                                        Annular slots around cup minimize
                                                        turbulence.
    Figure A-9.    Design  of a single stage  from a Brink-type cascade impactor.
                                           A-24
                                                                    KVB72-806015-1308

-------
sampler is designed to operate at  2.8  l/min or  less.   An  ideal flow rate is
around 2
A-1.6.2  Operation
         The impactor will be carefully loaded with the stage  cups  and the
preweighed stage substrates.  The Brink should be tightened with  wrenches to
make certain the high temperature No. 116 asbestos gaskets are sealed.   The
appropriate nozzle for isokinetic sampling is now added.  The  flue  gas
temperature should be above the dew point but less than 450°F. After mounting
the impactor on the preheated sample probe, it will be inserted into the duct
to be preheated for at least 30 minutes.  The inlet nozzle will be  pointed
downstream of the flow during the heating phase.  A predetermined flow rate is
established immediately and is maintained constant, since any  attempt to
modulate flow to compensate for changes in the duct flow rate  to  provide
isokinetic sampling will destroy the utility of the data by changing the cut
points of the individual stages.  Establishment of the correct flow rate
quickly is especially important for the short sampling times typical of high
dust loaded streams.
         Sample times will normally vary from 5 to 45 minutes  depending on the
dust loading.  A total sample weight of <8 mg per stage should be collected.
The stages of the Brink impactor will yield cuts of 0.25, 0.50, 1.0,  1.5, and
2.5 ym.  After the sampling cycle has been completed, the impactor  is cooled
and disassembled.  Proper disassembly is critical to make sure the  collected
material stays where it originally impacted.  After the desiccating of the
collection media, weighing is performed to determine the net particle accumu-
lations .
A-1.6.3  Data Presentation
         To determine the concentration of particulates  for  any  size range,
first determine the percentage of total particulates  for  each  stage.  Then the
cumulative percentage is determined beginning with  the  last  stage  of the
impactor.  By plotting the effective cutoff diameter  and  the cumulative
percent on logarithmic probability graph paper,  the particle concentration by
weight for any specific size can be determined.
                                                               KVB72-806015-1308

-------
                                TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing/
1. REPORT NO.
 EPA-600/7-83-042
                                                       3. RECIPIENT'S ACCESSION"
4. TITLE AND SUBTITLE
 Evaluation of Combustion Modification Effects on
  Emissions and Efficiency of Wood-fired Industrial
  Boilers
                                    5. REPORT DATE
                                    August 1983
                                    6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R. J. Tidona, W. A.  Carter, H. J. Buening,  and
  S. S.  Cherry
                                                       8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                       10. PROGRAM ELEMENT NO.
 KVB,  Inc.
 18006  Skypark Boulevard
 Irvine, California  92714
                                    11. CONTRACT/GRANT NO.
                                    68-02-2645
12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
                                    13 TYPE OF REPORT AND PERIOD COVERED
                                    Final; 4/79 - 10/79	
                                    14. SPONSORING AGENCY CODE
                                     EPA/600/13
is. SUPPLEMENTARY NOTES  iERL-RTP project officer is Robert E. Hall, Mail Drop 65, 919 /
541-2477.
is. ABSTRACT
               report gives results of full-scale tests to evaluate combustion modifi-
 cations (lower excess air and variations in the overfire air system operation) for
 emission control and efficiency enhancement on two wood-fired industrial boilers.
 One boiler,  rated at 100,000 Ib steam/hr, is fueled with a combination of wood bark
 and coal. Implementation of lower excess air reduced NOx emissions  by 18. 5 per-
 cent and improved thermal efficiency by 0.89 percent.  Variations in the overfire air
 system reduced NOx by 20. 7 percent and improved efficiency by 1.63 percent.  The
 second boiler,  rated at 200,000 Ib steam/hr when fired with hogged wood, can
 achieve 250,000 Ib steam/hr when fired with hogged wood and an oil supplement.
 The effectiveness of lower excess air in reducing NOx was about 14 percent with a
 slight (0.6 percent) improvement in efficiency. Adjusting the auxiliary air dampers
 reduced NOx by 17.2 percent and improved efficiency by 1. 7 percent.  Polycyclic
 organic matter (POM) was sampled at both baseline and optimum low-NOx conditions.
 Under baseline conditions,  POM emissions were similar to those of a coal-fired
 spreader stoker and an oil-fired boiler, but were well below  those of a pulverized-
 coal-fired boiler tested previously. For the wood-fired boiler, POM  emissions in the
 low-NOx mode were higher than those at baseline.
17.
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                           b-IDEN-HFIERS/OPEN ENDED TERMS
                                                                    c. COSATI Field/Group
Pollution
Combustion
Wood
Fuels
Fossil Fuels
Nitrogen Oxides
Boilers
Dust
Industrial Processes
Particle Size
Stoichiometry
Pollution Control
Stationary Sources
Combustion Modification
Particulate
Overfire Air
13 B
2 IB
11L
2 ID

07B
13 A
11G
13H
14G
07D
 3. DISTRIBUTION STATEMEN1

 Release to Public
                        19. SECURITY CLASS (This Report)
                        Unclassified
                         21. NO. OF PAGES
                              88
                        20. SECURITY CLASS (This page)
                        Unclassified
                                                                    22. PRICE
EPA Form 2220-1 (9-73)
                                        A-26

-------