EPA 910/9-82-089g
United States
Environmental Protection
Agency
Region 10
1200 Sixth Avenue
Seattle WA 98101
Air 8- Waste Management Division February 1984
xvEPA Residential Wood
Combustion Study
Task 5
Emissions Testing of Wood Stoves
Volumes 1 & 2
-------
RESIDENTIAL WOOD COMBUSTION STUDY
TASK 5
EMISSION TESTING OF WOOD STOVES
. VOLUMES 1 & 2
-------
THIS REPORT CONSISTS OF SEVERAL DIFFERENT PARTS.
THEY ARE LISTED BELOW FOR YOUR CONVENIENCE.
EPA 910/9-82-089a Residential Wood Combustion Study
Task 1 - Ambient Air Quality Impact
Analysis
EPA 910/9-82-089b Task 1 - Appendices
EPA 910/9-82-089c Task 2A - Current & Projected Air Quality
Impacts
EPA 910/9-82-089d Task 2B - Household Information Survey
EPA 910/9-82-089e Task 3 - Wood Fuel Use Projection
EPA 910/9-82-089f Task 4 - Technical Analysis of Wood Stoves
EPA 910/9-82-089g Task 5 - Emissions Testing of Wood Stoves
Volumes 1 & 2
EPA 910/9-82-089h Task 5 - Emissions Testing of Wood Stoves
Volume-s 3 & 4 (Appendices)
EPA 910/9-82-089i Task 6 - Control Strategy Analysis
EPA 910/9-82-089J Task 7 - Indoor Air Quality
-------
DISCLAIMER
*
This report has been reviewed by Region 10, U. S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policies of the U. S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
RESIDENTIAL WOOD COMBUSTION STUDY
TASK 5
EMISSIONS TESTING OF WOOD STOVES
Volume 1 of 4
-------
RESIDENTIAL WOOD COMBUSTION STUDY
TASK 5
EMISSIONS TESTING OF WOOD STOVES
FINAL REPORT
PREPARED BY:
DEL SREEN ASSOCIATES, INC.
Environmental Technology Division
1535 N. Pacific Highway
Woodburn, Oregon 97071
PREPARED FOR:
U.S. ENVIRONMENTAL PROTECTION AGENCY
Region X
1200 Sixth Avenue
Seattle, Washington 98101
TASK MANAGER
Wayne Grotheer
November, 1982
-------
ACKNOWLEDGEMENT
Del Green Associates, Inc. wishes to acknowledge the participation
of OMNI Environmental Services in the emissions testing conducted during
this study. Appreciation is expressed to George Hofer and Wayne Grotheer,
EPA Region X; Barbara Tombleson and John Kowalczyk, Oregon Department of
Environmental Quality; and Norman Edmisten, Engineering Science, for their
aid and participation in development of the study, as well as their review
and input into this report.
-------
THIS REPORT CONSISTS OF SEVERAL DIFFERENT PARTS.
THEY ARE LISTED BELOW FOR YOUR CONVENIENCE.
EPA" 910/9-82-089a Residential Wood Combustion Study
Task 1 - Ambient Air Quality Impact
Analysis
EPA 910/9-82-089b Task 1 - Appendices
EPA 910/9-82-089c Task 2A - Current & Projected Air Quality
Impacts
EPA 910/9-82-089d Task 2B - Household Information Survey
EPA 910/9-82-089e Task 3 - Wood Fuel Use Projection
EPA 910/9-82-089f Task A - Technical Analysis of Wood Stoves
EPA 910/9-82-089g Task 5 - Emissions Testing of Wood Stoves
Volumes 1 & 2
EPA 910/9-82-089h Task 5 - Emissions Testing of Wood Stoves
Volumes 3 & 4 (Appendices)
EPA 910/9-82-089i Task 6 - Control Strategy Analysis
EPA 910/9-S2-089J Task 7 - Indoor Air Quality
-------
DISCLAIMER
This report has been reviewed by Region 10, U. S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policies of the U. S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
TABLE-OF CONTENTS
page
VOLUME 1
Introduction 1
Summary and Conclusions 4
Results 5
Findings and Conclusions 8
Discussion of Major Findings 12
Recommendations for Further Study 20
Experimental Approach 22
Stove Operation 22
Wood Type 25
Test Protocol 25
Stove Selection 28
Sampling and Analytical Methods 31
Test Site and Sample Location 41
Stove Description . 44
Airtight Box 44
Catalytic Box 46
Catalytic Modified Combustion 49
Ceramic 52
Catalytic Retrofit 54
Non-Catalytic Retrofit 56
Results 58
General 58
Airtight Box Stove (Runs 1-9) 60
Catalytic Retrofit (Runs 10-11) 99
Non-Catalytic Retrofit (Runs 12-13) 114
Catalytic Box Stove (Runs 14-15) 127
Catalytic Modified Combustion Stove (Runs 16-17) 142
Ceramic Stove (Runs 18-19) 155
IV
-------
TABLE OF CONTENTS (Cont.)
page
VOLUME 2
Discussion of Results 166
General 166
Comparison of Particulate Results Among Appliances 167
Stove Efficiency 177
Particulate Emissions - First Phase Tests 179
Effect of Fuel Moisture Content 184
Simplified Test Procedures 229
Reasonable Emission Standard 241
Quality Assurance 264
Quality Assurance Records 264
Wood Moisture Determination 265
Particulate Sampling 268
Gaseous Measurements 269
Transmissometer 271
Data Reduction 273
References/Bibliography 274
VOLUME 3
Appendix A - Nomenclature and Sample Calculations
Appendix B - Laboratory Data
Appendix C - Quality Assurance Data
Appendix D - Cost Estimates for Simplified Test Procedures
VOLUME 4
Appendix E - Field Data Runs 1-19
-------
LIST OF TABLES
Table 1
Table 2
Table
Table
Table
Table
Table 7
Table 8
Table 9
Table 10
Table 11
Table 12
Table 13
Table 14
Table 15
Table 16
Table 17
Table 18
Table 19
Table 20
Table 21
Table 22
Table 23
Table 24
Table 25
page
Emission Summary 6
Operating Parameters Summary 7
Results -- Airtight Box Stove 62
Results — Airtight Box Stove (Cold Start Tests) 63
Results -- Airtight Box Stove (By Test Phase) 64
Bacharach Smoke Spot Results — 65
Airtight Box Stove
Visible Emissions Observation Log — 67
Airtight Box Stove
Operation Test Log -- Airtight Box Stove 87
Results « Catalytic Retrofit 100
Results — Catalytic Retrofit (By Test Phase) 101
Bacharach Smoke Spot Results — 102
Catalytic Retrofit
Visible Emissions Observation Log -- 103
Catalytic Retrofit
Catalytic Temperatures -- Catalytic Retrofit 104
Operation Test Log — Catalytic Retrofit 110
Results — Non-Catalytic Retrofit 115
Results ~ Non-Catalytic Retrofit (By Test Phase) 115
Bacharach Smoke Spot Results — 117
Non-Catalytic Retrofit
Visible Emissions Observation Log — 118
Non-Catalytic Retrofit
Operation Test Log — Non-Catalytic Retrofit 123
Results —• Catalytic Box Stove 128
Results — Catalytic Box Stove (By Test Phase) 129
Bacharach Smoke Spot Results — 130
Catalytic Box Stove
Visible Emissions Observation Log — 131
Catalytic Box Stove
Catalytic Temperatures — Catalytic Box Stove 132
Operation Test Log — Catalytic Box Stove 139
-------
LIST OF TABLES (Cont.)
page
Table 26 Results — Catalytic Modified Combustion Stove 143
Table 27 Results ~ Catalytic Modified Combustion Stove 144
Table 28 Bacharach Smoke Spot Results ~ 145
Catalytic Modified Combustion Stove
Table 29 Visible Emissions Observation Log -- 146
Catalytic Modified Combustion Stove
Table 30 Catalytic Temperature -- 147
Catalytic Modified Combustion Stove
Table 31 Operation Test Log -- Catalytic Modified 153
Combustion Stove
Table 32 Results — Ceramic Stove 156
Table 33 Results — Ceramic Stove (By Test Phase) 157
Table 34 Bacharach Smoke Spot Results — Ceramic Stove 158
Table 35 Visible Emissions Observation Log — Ceramic 159
Table 36 Operation Test Log — Ceramic Stove 164
Table 37 Comparison of Operating Parameters and Emissions 170
Table 38 Fuel Load and Combustion Rate Summary 176
Table 39 Results: Stove Efficiency 178
Table 40 Summary of Results: Phase 1 Tests 181
Table 41 -Results: Cold Start vs Hot Start 183
Table 42 Summary of Results: Fuel Moisture Tests 189
Table 43 Normalized Particulate Results': 191
Fuel Moisture Tests
Table 44 Creosote Results 201
Table 45 Creosote Deposition: Theoretical Percent of 211
Total Emissions
Table 46 Summary of Stack Gas Opacity 225
(Transmissometer and Visual Observer)
Table 47 Summary of Average Emission Factors for 230
Carbon monoxide, Gaseous Hydrocarbons, and
Creosote (Literature Review)
Table 48 Simplified Test Procedures Summarized 234
vii
-------
LIST OF TABLES (Cont.)
page
Table 49 Emissions Standard Range 242
Table 50 Summary of Average Emission Rates 246
(Literature Review) .
Table 51 Particulate Emission Data Summary 248
(Literature Review)
Table 52 Gaseous Calibration Gases 270
VI 1 1
-------
LIST OF FIGURES
Figure 1 Particulate Emissions as a Function of
Fuel Moisture
Figure 2 Particulate Sampling Apparatus
•Figure 3 Sample Location
Figure 4 Airtight Box Stove
Figure 5 Catalytic Box Stove
Figure 6 Catalytic Modified Combustion Stove
Figure 7 Ceramic Stove
Figure 8 Catalytic Retrofit Device
Figure 9 Non-Catalytic Retrofit Device
Figure 10 Gaseous Component (CO, C02, HC) vs Time (Run 1)
Figure 11 Temperature (Combustion, Stack Gas) vs Time (Run
Figure 12 Gaseous Component vs Time (Run 2)
Figure 13 Temperature vs Time (Run 2)
Figure 14 Gaseous Component vs Time (Run 3)
Figure 15 Temperature vs Time (Run 3)
Figure 16 Gaseous Component vs Time (Run 4)
Figure 17 Temperature vs Time (Run 4)
Figure 18 Gaseous Component vs Time (Run 5)
Figure 19 Temperature vs Time (Run 5)
Figure 20 Gaseous Component vs Time (Run 6)
Figure 21 Temperature vs Time (Run 6)
Figure 22 Gaseous Component vs Time (Run 7)
Figure 23 Temperature vs Time (Run 7)
Figure 24 Gaseous Component vs Time (Run 8)
Figure 25 Temperature vs Time (Run 8)
Figure 26 Gaseous Component vs Time (Run 9)
Figure 27 Temperature vs Time (Run 9)
Figure 28 Opacity and Smoke Spot Density vs Time (Run 1)
Figure 29 Opacity and Smoke Spot Density vs Time (Run 2)
page
13
33
42
45
48
51
53
55
57
69
1) 69
70
70
71
71
72
72
73
73
74
74
75
75
76
76
77
77
78
79
IX
-------
LIST OF FIGURES (Cont.)
Figure 30 Opacity and Smoke Spot Density vs Time
Figure 31 Opacity and Smoke Spot Density vs Time
Figure 32 Opacity and Smoke Spot Density vs Time
Figure 33 Opacity and Smoke Spot Density vs Time
Figure 34 Opacity and Smoke Spot Density vs Time
Figure 35 Opacity and Smoke Spot Density vs Time
Figure 36 Opacity and Smoke Spot Density vs Time
Figure 37 Gaseous Components vs Time (Run 10)
Figure 38 Temperature vs Time (Run 10)
Figure 39 Gaseous Components vs Time (Run 11)
Figure 40 Temperature vs Time (Run 11)
Figure 41 Opacity and Smoke Spot Density vs Time
Figure 42 Opacity and Smoke Spot Density vs Time
Figure 43 Gaseous Components vs Time (Run 12)
Figure 44 Temperature vs Time (Run 12)
Figure 45 Gaseous Components vs Time (Run 13)
Figure 46 Temperature vs Time (Run 13)
Figure 47 Opacity and Smoke Spot Density vs Time
Figure 48 Opacity and Smoke Spot Density vs Time
Figure 49 Gaseous Component vs Time (Run 14)
Figure 50 Temperature vs Time (Run 14)
Figure 51 Gaseous Component vs Time (Run 15)
Figure 52 Temperature vs Time (Run 15)
Figure 53 Opacity and Smoke Spot Density vs Time
Figure 54 Opacity and Smoke Spot Density vs Time
Figure 55 Gaseous Component vs Time (Run 16)
Figure 56 Temperature vs Time (Run 16)
Figure 57 Gaseous Component vs Time (Run 17)
Figure 58 Temperature vs Time (Run 17)
(Run 3)
(Run 4)
(Run 5)
(Run 6)
(Run 7)
(Run 8)
(Run 9)
(Run 10)
(Run 11)
(Run 12)
(Run 13)
(Run 14)
(Run 15)
page
80
81
82
83
84
85
86
106
106
107
107
108
109
119.
119
120
120
121
122
135
133
136
136
137
138
149
149
150
150
-------
LIST OF FIGURES (Cont.)
Figure 59 Opacity and Smoke Spot Density vs Time (Run 16)
Figure 60 Opacity and Smoke Spot Density vs Time (Run 17)
Figure 61 Gaseous Components vs Time (Run 18)
Figure 62 Temperature vs Time (Run 18)
Figure 63 Gaseous Components vs Time (Run 19)
Figure 64 Temperature vs Time (Run 19)
Figure 65 Opacity and Smoke Spot Density vs Time (Run 18)
Figure 66 Opacity and Smoke Spot Density vs Time (Run 19)
Figure 67 Summary of Particulate Emissions Results
Figure 68 Total Particulate Emissions vs Burn Rate
Figure 69 Total Particulate Emissions vs
Fuel Load-Combustion Rate Ratio
Figure 70 Filterable Particulate Emissions vs
Fuel Load-Combustion Rate Ratio
Figure 71 Particulate Emissions Results: First Phase Tests
Figure 72 Particulate Emissions Results:
Fuel Moisture Tests
Figure 73 Particulate Emissions as a Function of
Fuel Moisture
Figure 74 Particulate Emissions (Normalized for Burn Rate)
as a Function of Fuel Moisture
Figure 75 The Dependency of Appliance Efficiencies on
Fuel Moisture Content (From Sheltonlc»b)
Figure 76 Creosote Accumulation as a Function of Moisture
Content Using Pinon as Fuel (From She! ton8)
Figure 77 Creosote Accumulation as a Function *of Moisture
Content Using Oak as Fuel (From Shelton8)
Figure 78 Creosote: Transmissometer vs Sample Location
(By Test)
Figure 79 Creosote: Transmissometer vs Sample Location
(By Test Phase)
Figure 80 Creosote (Transmissometer Location) vs
Total Particulate Emissions
page
151
152
160
160
161
161
162
153
168
172
I73
180
185
I88
190
195
196
203
204
206
-------
LIST OF FIGURE (Cont.)
page
Figure 81 Creosote (Sample Location) vs Total 207
Particulate Emissions
Figure 82 Creosote (Average) vs Total Particulate Emissions 208
Figure 83 Creosote (Transmissometer Location) vs 209
Filterable Particulate
Figure 84 Particulate Concentration vs Carbon monoxide 213
Concentration
Figure 85 Particulate Emissions vs Adjusted Carbon 214
monoxide Concentration
Figure 86 Particulate Emissios vs Carbon monoxide 216
Concentrations
Figure 87 Particulate Concentration vs Gaseous 218
Hydrocarbon Concentration
Figure 88 Particulate Emissions vs Adjusted Gaseous 219
Hydrocarbon Concentration
Figure 89 Particulate Emissions vs Gaseous 220
Hydrocarbon Emissions
Figure 90 Particulate Concentration vs Opacity 223
Figure 91 Particulate Emissions vs Opacity 224
Figure 92 Particulate Concentration vs Smoke Spot Density 227
Figure 93 Particulate Emissions vs Smoke Spot Density 228
Figure 94 Wood Moisture Measurement 266 '
xii
-------
INTRODUCTION
Emission testing studies of wood stoves have yielded results
indicating a wide range of emission levels. The particulate emissions
measured range from approximately 2 grams per kilogram of wood burned
for advanced state-of-the-art devices to 75 grams per kilogram of wood
burned for the common airtight box stove. However, variations in the
measured emission rates can be attributed to more than stove design.
Previous studies have indicated that fuel moisture and size, as well as
operating procedures (e.g., burn rate) may have a significant impact on
emissions. Another extremely important factor affecting the measured
emission rates is the test procedure used in determining the particulate
emissions; variations in test procedures can result in significant
differences in measured emission rates.
This particular study was undertaken with several objectives in
mind. The three major objectives of this study were to further identify
the effect of wood moisture on stove emissions, to evaluate several
inexpensive (simplified) test procedures for assessing particulate emissions,
and to define a level of particulate emissions which can be expected from
state-of-the-art improved combustion stoves. In order to achieve these
objectives, this study included conducting a series of emissions tests,
as well as evaluating previous test data reported in the literature.
The emissions tests were conducted in conjunction with OMNI Environ-
mental Services, Portland, Oregon. All tests were conducted at the OMNI
facilities under laboratory conditions. These tests were conducted during
the period June to October 1981. The test program involved 19 emissions
tests with four stoves and two retrofit devices.
1
-------
In order to try to eliminate the variation in emissions due to operating
procedure, a single standard operating procedure was decided upon and used
*
throughout the test program. The objective of this chosen operating procedure
was to maintain a constant heat output rate, as monitored by combustion
chamber temperature and stove surface temperature. A heat output rate
corresponding to a relatively moderate to low burn rate (less than 2.5 kg
wood/hour) was desired since it is believed that is typical of wood stove use
in the Pacific Northwest where moderate winter temperatures (40 to 50 F)
prevail. A single wood type (Douglas fir) was used throughout the study;
wood size was maintained at a consistent level.
The test program began with a series of six emission tests conducted
on an airtight box stove to evaluate the effect of wood moisture on stove
emissions. (Table 2 of the Summary and Conclusions section summarizes the
operating parameters for all tests.) During this test series burns were con-
ducted with wood at three different moisture levels (15,25, and greater than
40 percent moisture on a dry basis).
In addition to the moisture test series on the airtight box stoves,
three improved technology stoves and two add-on devices were chosen for
evaluation during this study in order to determine the effectiveness of
state-of-the-art technology in reducing stove emissions. A primary
objective of this phase of the study was to determine a reasonable and
achievable emission standard. Prior to the testing, the literature was
reviewed in order to identify stoves to be evaluated. Three areas of
technology were identified and one stove was chosen from each general
design area. These areas are catalytic stoves, modified combustion (e.g.,
secondary, downdraft), and combined technology. The three stoves
studied were chosen based on their expected potential for reducing emissions.
2
-------
In addition, two retrofit devices to be used in conjunction with the
airtight box stove were chosen for testing to determine their potential
for reducing emissions from existing stoves. The stoves chosen for
evaluation included a catalytic stove, a catalytic-modified air path
stove, and a ceramic stove; both a catalytic and non-catalytic retrofit
device were tested.
Throughout the entire test program measurements were made for
particulate, carbon monoxide, carbon dioxide, oxygen, and hydrocarbon
content of the emissions; the gaseous constituents were monitored con-
tinuously throughout the test burn. In addition, measurements were made
for creosote deposition, opacity, and smoke spot density. Particulate
emissions were measured using an EPA Method 5 Sampling Train modified to
collect condensible organics by including an unheated back-half filter
after the third impinger of the impinger train. Consequently, par-
ticulate emissions were determined and are reported as both total
particulate (front-half filter with condensible fraction) and non-volatile
particulate (front-half catch only).
-------
SUMMARY AND CONCLUSIONS
*
Pertinent results regarding the impact of fuel moisture content on
emissions was obtained; test results indicate "medium" moisture wood is
likely to yield the lowest emissions. Similarly, useful information
regarding the applicability of simplified test procedures was obtained
although more data should be collected before coming to any firm conclusions;
continuous carbon monoxide and hydrocarbon measurements show the greatest
potential as useful simplified procedures. Finally, although a great deal
of information was obtained directly from the data on the testing of the
innovative technology stoves, performance of these appliances was disappoint-
ing; generally results similar to typical well operated units were obtained.
Consequently, a "reasonable emission standard" based on these test data
would reflect an emission level currently achievable by typical well operated
units. Lower emission results (in the range obtained by the ceramic stove
during this study) are expected to be achievable, although emission data is
not available yet to indicate an appliance can consistently meet these low
emission levels at low burn rates (heat output).
Further discussion summarizing the results obtained for each of the
project objectives is presented in this section of the report. A summary
of recommendations for further work also is provided.
-------
RESULTS
Table 1 summarizes the emissions results from the 19 test runs
conducted during this test program. Similarly, Table 2 summarizes the
important operating parameters for the appliances during the emission
testing.
The tests were conducted at an average burn rate of 2.5 kg/hr, dry
basis, (5.5 Ib/hr). During the tests the stove surface temperature
averaged in the 350 - 400° F range, the combustion chamber temperature
averaged 500 - 600° F, and the stack effluent temperature averaged
250 - 300° F.
The average total particulate emission rate for these tests (excluding
fuel moisture and cold start emissions tests) was 19 g/kg. Front-half
(filterable) particulate emissions averaged 21% of the total with an average
filterable emission rate of 3.5 g/kg. The highest emission rate measured
was 62 g/kg (11 g/kg, filterable) at a burn rate of 2.7 kg/hr in the air-
tight box stove fueled on dry (14%) wood. The lowest particulate emissions
obtained, 2 g/kg (1 g/kg filterable) were from the Ceramic stove. However,
this stove operated at a significantly higher burn rate (3.9 kg/hr) than
did the other appliances and burn rate is known to affect emission levels.
Carbon monoxide and hydrocarbon emissions averaged 115 and 8.7 g/kg,
2
respectively. Creosote deposition averaged 330 mg/m kg. Average opacities
measured by the visible emissions observer and transmissometer were 21 and
18 percent, respectively.
The test results are discussed in more detail in the Discussion of
Results section of this report.
-------
TABLE 1
Emissions Summary
Test Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
STOVE TYPE
Box
Box
Box
q
Box*
Box
Q
Box3
Box
q
Box*
Box
Box/Catalytic Add-on
Box/Catalytic Add-on
Box/Non-Catalytic Add-on
Box/Non-Catalytic Add-on
Catalytic Box
Catalytic Box
Catalytic/Secondary Air
Catalytic/Secondary Air
Ceramic
Ceramic
Burn
Rate
kg/hr
(Ib/hr)
1.9 (4.1)
2.4 (5.3)
2.1 (4.7)
8.1 (17.8]
2.7 (6.0)
5.8 (12.8
2.9 (6.4)
2.0 (4.5)
1.7 (3.7)
2.1 (4.6)
2.6 (5.8)
2.4 (5.4)
2.2 (4.9)
1.7 (3.7)
2.2 (4.9)
3.0 (6.7)
2.1 (4.7)
6.4 (14.2
3.9 (8.7)
Partlculate Emissions
Total2
g/dscm
1.3
3.4
1.5
3.6
5.0
3.8
1.5
0.9
.2
.3
.3
.8
.8
2.2
1.5
0.6
1.2
0.2
0.1
Ib/hr
0.09
0.28
0.16
0.72
0.37
0.54
0.12
0.11
0.08
0.10
0.09
0.21
0.17
0.14
0.11
0.09
0.14
0.02
0.01
gAg4
22
54
34
40
62
42
19
24
22
22
17
38
35
38
23
14
30
1
2
Front
Half3
gAg' (x)5
6.3 (29)
14 (21)
11 (33)
9.6 (24)
11 (17)
8.3 (20)
4.4 (23)
6.1 (25)
4.8 (22)
3-7 (17)
3.3 (20)
7.4 (20)
6.1 (17)
5.9 (16)
4.8 (21)
3.5 (25)
5.7 (19)
0.63 (63)
0.97 (49)
Creosote
">g/"2-kg
969
917
218
216
592
291
568
109
240
223
190
273
337
379
273
88
317
56
27
Carbon
Monoxide
gAg4
190
189
160
210
220
170
160
190
110
110
90
200
160
120
50
80
150
20
50
Gaseous
HC
w-7
13.8
16.9
10.5
11.9
12.1
9.2
8.8
11.1
6.6
8.0
6.2
11.8
9.3
10.7
7.3
3.9
5.8
0.4
0.6
I0ry Basis 5Percent of Total Emissions *Cold Start
Opacity8
I
36
30
20
40
37
46
28
20
11
24
16
22
23
10
10
<5
13
-0-
-0-
Stack Gas
Flow
SCUM
30 (1060
38 (1350
50 (1750
91 (3230
34 (1200
66 (2320
36 (1270
55 (1950
32 (1130
37 (1320
32 (1130
54 (1890
42 (H90
30 (1060
34 (1200
71 (2500
56 (1960
52 (IB50
49 (1720
Excess
Air
X
230
160
180
140
180
140
170
410
320
380
240
120
280
140
172
347
498
63
97
Oregon DEQ Method 7 Average of two measurements locations
3EPA Method 5 7As Hexane
Mass Emissions per Mass Dry Fuel Consumed
8
Visual Observer
-------
TABLE 2
Operating Parameters Summary
Test
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Stove Type
Box
Box
Box
Box2
Box
Box2
Box
Box2
Box
Box/Catalytic Add-on
Box/Catalytic Add-on
Box/Non-Catalytic Add-on
Box/Non-Catalytic Add-on
Catalytic Box
Catalytic Box
Catalytic/Secondary Air
Catalytic/Secondary Air
Ceramic
Ceramic
Burn
Rate
kg/hr
(Ib/hr)
1.9 (4.1]
2.4 (5.3]
2.1 (4.7]
8.1 (17.6!
2.7 (6.0
5.8 (12.8
2.9 (6.4]
2.0 (4.5
1.7 (3.7
2.1 (4.6
2.6 (5.8
2.4 (5.4
2.2 (4.9
1.7 (3.7
2.2 (4.9
3.0 (6.7
2.1 (4.7
6.4 (14.2
3.9 (8.7
Fuel Charge
Wet
kg
11.9
12.5
15.7
13.5
11.8
12.4
12.9
17.3
12.5
14.0
12.6
14.0
13.3
7.6
7.5
8.0
8.5
8.7
8.9
Dry
kg
9.4
11.0
11.4
11.9
10.4
10.0
10.3
7.7
5.5
11.4
10.2
11.3
10.8
6.3
6.0
6.5
6.8
7.3
7.1
Fuel
Moisture1
I
26
• 14
126
14
14
24
25
126
126
22
23
24
23
21
25
24
25
21
25
Fuel .
Consumes
^
8.5
10.0
6.5
4.7
9.5
4.5
9.5
5.3
5.2
10.5
9.4
10.5
10.0
5.9
5.6
6.0
6.3
6.5
6.5
Burn
Duration
Minutes
275
250
181
35
208
46
198
155
189
299
223
255
270
209
150
120
175
61
99
Temperature
Combustion
Chamber
•F
513
543
654
675
569
598
610
470
600
602
599
574
553
609
815
627
491
1519
1195
Statk
Gas
•F
265
316
342
544
299
401
372
295
361
290
353
265
251
244
323
375
230
779
604
Dry Basis
Cold Start
1 b moisture . .nn
1b wet wood - Ib moisture ' * luu
-------
FINDINGS AND CONCLUSIONS
The major findings and conclusions resulting from this study are*
as follows:
• The results of these appliance tests indicate that average stove
emissions expected in normal use would be 20 g/kg total particulate
and 4 g/kg filterable particulate (condensible excluded). These
tests were conducted at an average combustion rate of 2.5 kg/hr
which is considered a low to moderate rate. Barnett^-3 reports
that research he has conducted in Upstate New York indicates a
typical burn rate of only 1.5 kg/hr. Higher emissions generally
are produced at lower burn rates; consequently, the 20 g/kg emissions
estimate may actually be lower than emissions from installed stoves
operating at burn rates of less than 2.5 kg/hr.
• Particulate emission levels are a function of combustion rate
with emissions increasing as burn rate decreases. The results
obtained during this study compare qualitatively to results
previously obtained^'1*' •
• Fuel moisture content, does affect emissions. During this study,
emissions (g/104 Btu) increased significantly for both dry wood
(200%) and for wet wood (100%) as compared to the medium moisture
fuel. Optimum fuel moisture is in the 25-35%, dry basis (20 -
30% wet basis) range.
• Only one of the improved technology stoves tested resulted in any
significant reduction in emissions. The emissions from the
Ceramic stove were measured at 1-2 g/kg; a reduction of 90%
from the average 20 g/kg. However, this emission reduction must
be qualified, since the burn rate for this appliance was sig-
nificantly higher (5.4 vs 2.5 kg/hr) than the other appliances.
Nonetheless, the net efficiency measured for this appliance was
similar to the other appliances (65%) even though high stack
losses were measured. Further testing (heat output rate and
efficiency) of this appliance is warranted to monitor and
determine the actual heat release rate characteristics of this
stove. If the manufacturer's claim that this ceramic unit
"stores" the heat generated during the rapid combustion for slow
release during the non-burn periods, the operation of this unit in
a high burn batch charge manner would achieve the homeowner's goal
of steady heat output, as well as a clean burn with few emissions.
This appliance would then provide a basis for defining a reasonable
emission level for well designed, clean burning units.
-------
The catalytic devices did not reduce emissions significantly under
the operating conditions of these tests. Catalytic operation was
considered marginal for these test operating conditions. A
literature review indicates mixed results have been obtained with 7
catalytic devices. Two other investigators (Oregon DEQld and Barnett )
also have reported no significant reduction in measured emissions
from box catalytic stoves. However, Barnett7 recently has reported
significant reduction (85%) in emissions from a prototype catalytic
stove and Shelton2,3 has reported significant reductions (up to 50%)
in creosote from the same catalytic add-on device tested in this study.
(Discussed in more detail later in this section). The conclusion from
this study is that simple addition of a catalytic device to a standard
appliance is not guaranteed to reduce emissions. More testing in this
area may be warranted in order to characterize the range of operating
parameters which will result in reduced emissions from catalytic
appliances. This testing could be conducted by a simplified test
procedure (e.g., CO) to determine under what conditions adequate
catalytic combustion is achieved.
The Catalytic Modified Combustion appliance did not result in
significantly lower emissions than the norm, although this appliance
when operating properly did have an emission rate at the lower end
of the normal range (14 g/kg).
The non-catalytic add-on device did not reduce emission levels under
the test conditions of this study. However, two other investigations
(Oregon DEQ and She!ton ) found emission reductions of 50 and 211,
respectively for this device. (Shelton's work was for creosote
deposition).
Test results obtained confirmed the importance of operating conditions
in obtaining meaningful results. As previously noted, a correlation
between burn rate and emissions was obtained. Furthermore, for two
of the stoves tested (catalytic box and catalytic modified combustion),
a significant difference in emission rates (50% reduction) was noted
for the two runs in each of the paired samples and it is believed
these significant differences were directly related to operating
changes (i.e., burn rate and combustion chamber temperature), and
not simply to measurement error.
Due to the generally poor performance of the improved technology
appliances, these test data do not provide a good basis for
establishing a technology forcing Emission Standard. Strictly
based on these test data, a standard of 15-20 g/kg (5-10 g/kg
filterable) would be justifiable. Nonetheless, the results of the
ceramic stove tests (at an elevated burn rate), as well as other
-------
data from the literature (Barnett's catalytic protype, Oregon DEQ
results on a furnace) suggest that an emission rate of 5 g/kg (1 g/kg
filterable) at a typically used burn rate (1.0 - 2.5 kg/hr) may
ultimately be achievable in a well designed appliance. Construction
of smaller stoves to be operated at low burn rates more typical of
normal home use may help to achieve this goal.
The correlation between total particulate matter and several
"simplified" procedures ~ creosote, carbon monoxide, hydrocarbons,
opacity, and smoke spot density — were all examined. Results of
these correlations were:
• A reasonable correlation (correlation coefficient 0.8) with par-
ticulate matter was obtained for carbon monoxide and hydrocarbons
(adjusted for excess air). Either of these parameters could be
used to monitor and characterize particulate emissions levels.
Additional data would definitely need to be collected prior to
using such a correlation for actually certifying appliance
emissions rates, should a certification program be initiated.
• A correlation between particulate matter and creosote was established.
A correlation coefficient of 0.80 was obtained (two data points
deleted). However, multiple creosote samples were collected during
this study and relatively poor precision among samples was obtained
(correlation coefficient of 0.6 between samples at two different
locations during same test). Another investigator (Barnett ) has
reported "excellent" correlation between coreosote and particulate.
However this correlation was based on average values, and as in
this study, all tests were conducted under similar stove ogesating
conditions. The results of a recent (February 1982) study * by
She!ton cast serious doubt on the utility of creosote deposition
as a general simplified method for estimating emission rates. In
his study Shelton showed that creosote deposition was decreased by
75% by use of a stack barametric damper; basically, the reduction
was due to the effect of dilution air and increased flue gas flow, and
no corresponding reduction in actual emissions to the atmosphere
would be expected.
• A correlation was obtained between particulate matter and opacity
with a correlation coefficient of 0.8. However, this correlation
is for particulate matter concentration (g/dscm). This has limited
benefit as a simplified test method since concentration will be
effected by excess air. Any increase in excess air will result
in a decreased opacity level. Consequently, opacity may not be
an accurate indication of total emissions. Nonetheless, opacity
could be used as a screening device to indicate emission levels
known to exceed a particular level. In other words, if opacity
exceeded a given level, one could be assured the emissions exceeded
a corresponding level (q/kq); however, if the opacity was less than a
given level, one could not be certain the emissions did not exceed a
10
-------
corresponding level (g/kg) because of the potential effect of
dilution (excess) air on the particulate concentration. (High
type II error ~ false acceptance.)
• No correlation was obtained between particulate emissions and
smoke spot density; the method was not sensitive enough to measure
corresponding changes in emissions.
Filterable (Method 5 Front-half) particulate averaged 21% of the
total particulate catch. This percentage was fairly constant except
for the ceramic stove which had significantly less total emissions;
in this case the filterable fraction was a higher percentage of the
total. This simply indicates that improved combustion has a greater
impact on the condensible fraction than on the filterable fraction.
Nonetheless, filterable particulate could be used as a test method for
establishing emission levels; use of this method would simply narrow
the emission range (total particulate range of 1 to 62 g/kg versus
filterable range of 0.6 to 14 g/kg). Since a great deal of the
emissions reduction occurs in the condensible fraction, use of a
filterable only method would result in a slight penalty to the low
emitting units when comparing emissions. For example, based on
total particulate, the ceramic stove indicates a reduction of 90%
from the average emission value (2 vs 20 g/kg), but only a 75%
emission reduction for filterable particulate (1 vs 4 g/kg)
No trend was apparent from these data (3 tests only) to indicate the
effect on emissions of a cold start as compared to a start from a
hot bed of coals.
11
-------
DISCUSSION OF MAJOR FINDINGS
The major findings relative to the three major objectives of this study —
effect of fuel moisture content on emissions, evaluation of simplified test
procedures, and evaluation of innovative appliances to define Reasonable
Emission Levels are discussed in more detail below. The author intentionally
has avoided providing a great deal of data and data analysis in this section.
The reader is referred to the Discussion of Results section of this report
for evaluation of the actual data and information obtained from the litera-
ture review.
Effect of Fuel Moisture on Emissions
The results of this study indicate particulate emissions are significantly
affected by wood fuel moisture. Figure 1 graphically presents the results
obtained. Based on these tests the optimum fuel moisture range is 25-35
percent, dry basis. It is important to note that the results presented are
results obtained with the appliances operating at nearly the same combustion
rate. This is important in interpreting the results since emissions have
been shown to be dependent upon combustion rate. Consequently, the high
emissions obtained with the dry fuel are expected to be largely due to the
fact that the air inlet was restricted during these tests in order to main-
tain a heat output rate consistent with the medium moisture tests. This
operation is considered typical of how a homeowner would operate an appliance,
since the ultimate objective of a user generally is a constant heat output.
The results obtained during this study mostly are consistent with
fi 8
results obtained by Shelton ' during two studies (comparison of results
discussed in more detail in the discussion of Results section). In one
12
-------
70
60
CO
50
o>
c
o
Ol
IO
Q.
40
30
FIGURE 1
Particulate Emissions As A Function
of
Fuel Moisture
• g/kg
• g/kg, average
a g/104 Btu
• g/104 Btu, average
20-
10
(Wet)
(Dry)
10
11
20 30 40
25 43 67
Fuel Moisture, %
50
100
60
150
-------
study Shelton measured efficiency as a function of fuel moisture (emissions
were not actually measured). Based on this study by Shelton (assuming
emissions to be inversely proportional to combustion efficiency) one would
expect minimum emissions to occur in the 25-35% moisture range with emissions
Q
increasing for both higher and lower moisture levels. In She!tons second study ,
creosote accumulation was measured as a function of fuel moisture; the results
of the study indicate decreasing creosote emissions with increasing fuel
moisture. She!ton's results confirm that higher emissions are to be expected
for dry fuel, but contradict the results obtained during this study for high
moisture fuel. However, burn rate data were not provided for Shelton's
study. Shelton's general approach was to maintain constant heat output.
Consequently, it is assumed that a slightly higher burn rate was used for
the wet fuel tests in order to maintain the same heat output rate as the
medium fuel tests (an increased burn rate being required to compensate for a
latent heat loss due .to moisture). This assumption is unconfirmed; nonetheless,
an increased burn rate would account for some of the apparent reduction in
emissions for the wet fuel. Furthermore, creosote may not be a true indicator
of total emissions emitted to the atmosphere, especially when operating
conditions are significantly changed (fuel moisture content, fuel type,
excess air).
In summary, the optimum fuel moisture range appears to be 25-35% dry
basis. Dryer fuel is expected to result in increased emissions. The
emissions expected for very wet fuel are less clear; this study indicates
slightly greater emissions, whereas one of Shelton's studies indicates
less emissions. However, the fact that more fuel must be consumed when
14
-------
wet fuel is used (due to the heat loss from vaporization of water) warrants
using properly seasoned wood, even if no emission reduction can be confirmed.
Evaluation of Simplified Test Procedures
Of the'five simplified test procedures evaluated during this study,
carbon monoxide (CO) and/or total hydrocarbon (THC) measurements offer
the greatest advantages and flexibility. For both CO and THC a correlation
coefficient of 0.8 was obtained between concentration (adjusted for excess
air) and total particulate matter. One of the advantages of this procedure
is that if concentration of the gaseous pollutant correlates with measured
total particulate (g/kg) than stack gas volumetric flow rate does not need
to be measured. Another advantage of measurement of CO concentration is
that cost and level of effort can be adjusted depending upon the level of
accuracy required. Different techniques (varying from orsat to laboratory
*
quality continuous monitoring instrumentation) can be used for the measure-
ments. For example, the cost of using this method to evaluate a single
burn cycle is estimated at between $300 and $600 depending upon whether
an orsat method or instrumental technique is used.jnth a corresponding
capital cost range of $600 to $13,000. Comparable estimated costs for
conducting a single Method 5 test are $2,100 (level of effort) and
$20,000 (initial capital expenditure). Further study should be conducted
and additional data obtained to verify the correlation of CO and THC to
total particulate and to examine use of the method over a wider range
of conditions.
Creosote deposition measurements are extremely inexpensive, requiring
almost no capital expenditure and very little level of effort.
15
-------
The test can be conducted for as long or as short a period of time as
desired with little additional cost or level of effort over that required
for operation of the appliance. However, the results obtained during this
study, as well as information obtained from the literature review indicate
that problems may exist with precision of the technique and with the range
of conditions for which it may apply. These problems are discussed in more
detail in the Discussion of Results section.
The usefulness of opacity measurements is limited to that of a
screening technique, due to the techniques dependence upon excess air
dilution. As mentioned earlier in this summary, evaluation of the smoke
spot density data did not result in any apparent correlation.
One simplified technique, high volume sampling, not evaluated in
this study due to budgetary constraints, warrants investigation. There
is no real cost benefit to using this method. The advantage of this
method as a simplified technique is that the sampling period is extremely
short which permits discreet samples to be taken during various periods
of the burn cycle. This is useful for evaluating changes in stove per-
formance as operating conditions are changed. A comparative test series
of this method with EPA Method 5 would provide the data necessary to
determine how the filterable particulate measured by the High Volume
method compares to the total particulate measured by EPA Method 5.
Table 48 in the Discussion of Results section summarizes the advantages,
disadvantages, and costs of the simplified methods evaluated during this
study.
16
-------
Evaluation of Innovative Technology Appliances to Define Reasonable
Emission Levels
Of the three innovative technology appliances and two retrofit
devices tested during this study, only one appliance — the ceramic
stove -- resulted in significant emissions reduction (90%) from the
baseline level established from a typical box (step-type) stove operating at
similar conditions. However, due to the high burn rate of this appliance,
the meaningful ness of the low emission rate obtained (1-2 g/kg) is
somewhat obscured. Further testing of this appliance's heat output
rate characteristics is warranted to determine if the unit will provide
a steady heat output (due to storage capabilities of the ceramic) even
when operated in a high burn -- batch type firing mode.
A literature review was conducted. Average emissions reported by
various researchers for tests, conducted at burn rates of less than 3.5 kg/hr
are as follows (these data are presented and discussed in more detail
in the Discussion of Results section'):
Investigator Total Particulate Filterable Particulate
g/kga
DGA/OMNI • 27
i .
Oregon DEQld ' 32C '
Sanbornlh 27
Barnett13'7
Butcher10 (1979)
Butcherlk (1980)
6
,
15°
18
7b
9b
4b
a Dry basis
High volume
c Includes results from tests with burn rate up to 4.5 g/kg
17
-------
Total participate is defined as the combined filterable (front-half) and
condensible fraction of the EPA Method 5 sampling system (with or without
modification to include second back-up filter after impinger train).
Filterable particulate is the front-half filter catch of the Method 5
sampling system, or the particulate matter collected by an unheated filter
of a high volume sampling system (with or without dilution air added prior
to contact with the filter). These data indicate an average emission rate
of greater than 25 g/kg.
Results from this study (e.g., results of catalytic-modified com-
bustion unit when operated at proper combustion temperature) indicate an
emission level of from 15-20 g/kg total particulate (5-10 g/kg filterable)
is reasonable to expect from well operated commercial units currently
available. Limited data from several tests indicate that emissions levels
as low as 5 g/kg (1 g/kg filterable) from improved technology appliances
are achievable but not necessarily at the low burn rates typically used in
small home appliances; such a low emission level should be considered
"technology forcing" at this time. The limited data are:
Appliance
Ceramic Stove
Furnace
Catalytic.,
Appliance
# Tests
2
2
3
Burn Rate
kg/hr
5
14
- 1.6
Total Part.
g/kg
2
4
-
Filterable Part.
g/kg
1
1
1
Although no significant emissions reductions were noted for the two
retrofit devices (catalytic and non-catalytic) tested during this study,
2 3
She!ton's ' recent results indicate a reduction in creosote deposition of
18
-------
up to 20% for the same non-catalytic device and up to 50% for the same
catalytic device. Oregon DEQ noted an emissions reduction of 50% for
the non-catalytic device during their study (1 test run only). The
disparity in results for the catalytic device indicates that a reduction
in emissions from catalytic action can not be automatically assumed; instead,
like emissions reduction achieved from other combustion techniques reduction of
emissions from catalytic action are critically dependent upon numerous
operating conditions. Factors which affect catalytic action include flue
gas temperatures, oxygen available, residence time at the catalyst, and
the proper mixing of effluent gases and combustion air. To what degree the
conditions necessary for emissions reduction by catalytic action are
dependent upon appliance design, appliance operating conditions, and/or
operator actions is unclear at this point. Proper catalytic action is
no doubt a function of all these, but the limiting factors have yet to
»
be clearly defined.
In any event, She!ton's results indicate a reduction of up to 50% .
(based on creosote deposition) may be achievable from retrofit devices
(note-that the non-catalytic device costs about $150 and the catalytic
unit about $300). This level of emission reduction would likely result
in average emissions in the 15-20 g/kg range for existing units.
The effect of test methods and appliance operation -- especially
appliance operation — on measured emissions is significant. Consequently,
it must be understood that in defining any emissions level certain para-
meters regarding these two variables also must be defined and subsequently
go hand-in-hand with the defined emission level. For this discussion of
emission levels, the author has basically assumed the appliance is operated
19
-------
at a combustion rate of less than 3 kg/hr, and that emissions are
measured by EPA Method 5 with condensible analysis. For any emissions
measurement program to determine or compare emissions from various
appliances (whether for certification or research) stove operating pro-
cedures and test methods must be carefully defined if the results are
to be meaningful. Development of such a standard test protocol is well
outside the scope of this project.
RECOMMENDATIONS FOR FURTHER STUDY
The results obtained during this study indicate the need for further
study of various items related to wood stove emissions and the measurement
of wood stove emissions. Recommendations for further study are summarized
below:
• A standard operating/test protocol by which to compare stove emissions
is required. Considerable effort will be required to develop such a
protocol, but considering the impact operating conditions have on
emission rates, the effort is necessary. The protocol should include
requirements for testing at two or three (low, medium and possibly
high) heat output rates and should establish a specific criterion
for fuel load/stove volume. Other details such as fuel moisture,
fuel type, fuel size, and stove operating cycle (when to start and
stop emission test) must also be addressed. Obviously, specific
emission test methods criteria also must be established.
• Evaluation of "typical" burn rates actually used in the Pacific
Northwest is suggested. Duetto the significant dependence of emissions
on burn rate, it would prove profitable to conduct a study to determine
the actual combustion rates (heat release rate) normally used by home-
owners in the Pacific Northwest. This information can be used to better
estimate emission factors from existing installations, to determine the
rates at which stoves should be operated during emission testing, and to
determine proper stove sizing for the "typical" Northwest installation.
• Techniques for measuring low velocity flow rates during stove emission
testing should be evaluated with the goal of establishing an acceptable
standard procedure. Development of a low flow measurement technique
will not only provide an independent technique for determining mass
flow which can be used to validate stoichiometric mass balance, but
also will provide the basis required for conducting proportional sampling
(particulate sampling rate proportional to volumetric flow.rate).
20
-------
Conduct additional testing to verify the correlation of CO and THC
to total particulate and to examine the use of this simplified method
over a wider range of operating/emission conditions.
Conduct additional testing to further evaluate the potential of creosote
measurements as a simplified test procedure. A two-fold evaluation is
recommended. First, evaluate the precision of the method by conducting
a series of tests using simultaneous multiple creosote measurements,
and refine the method as necessary. If adequate precision is obtained,
then evaluate the creosote technique in relation to the standard par-
ticulate measurement technique (EPA Method 5 with back-half analysis).
Conduct a controlled comparative study of the emissions measured by
EPA Method 5 and a high volume sampling technique. This evaluation
would give a basis of comparison for the emissions data already
collected by the different methods, and would also help to define
the potential of the high volume method as a "simplified" emissions
measurement technique.
Conduct additional testing of catalytic stoves and develop a "screening"
protocol for catalytic operation. Conflicting results have been obtained
regarding the ability of catalytic devices to reduce emissions from
wood stoves; however, the potential for emissions reduction warrants
further study of catalytic systems. Combustion conditions (therefore
stove operating conditions) are extremely important to proper catalyst
operation. Consequently, an important aspect of any study -of catalytic
appliances is the establishment of proper operating procedures and
conditions. A screening protocol should be developed (e.g., use of
CO/HC monitoring) to use in identifying catalytic stove operating
boundaries prior to emission testing.
Additional testing to determine the actual instantaneous heat release
characteristics of the ceramic stove should be considered. Any additional
testing should include a detailed evaluation of the stoves heat output
rate and efficiency while operated in a batch type operation (i.e..,
initial fuel load with high burn rate followed by waiting period prior
to recharging).
21
-------
EXPERIMENTAL APPROACH
STOVE OPERATION
Of major consideration during any wood stove test program is
determination of the stove operating procedures to be followed during
the emission testing. Two basic approaches are available. In the
first case, stove damper and inlet air settings are established at the
beginning of a test burn, and no adjustments are made throughout the
burn. For this case, one might suspect that run to run variations
would be reduced. In the second case, stove operating parameters
(e.g., combustion chamber temperature, excess air, etc.) are monitored
during the burn and adjustments are made to the air inlets and/or draft
in order to maintain the monitored parameter within a certain range.
In this study, the latter approach was followed. The parameters to
be monitored were combustion chamber temperature and stove skin tem-
perature.
Of primary consideration in establishing the operating parameters
for this test series was how the typical Pacific Northwest homeowner
would normally operate a stove. Due to the moderate climate in the
Pacific Northwest, a low to medium burn rate was 'desired - certainly
not a high burn rate. A typical homeowner would make adjustments to
stove operation based on heat output. That is, a homeowner would not
adjust the stove damper based on stack temperature, emissions, or even
burn rate, but instead would make adjustments depending upon whether
more or less heat is desired from the stove. Therefore, the test pro-
22
-------
tocol established a desire to maintain a constant heat output. Unfor-
tunately, this parameter is not easily monitored. Consequently, the
objective of our operating technique was to obtain and maintain a near
constant combustion chamber temperature (BOOT) and stove surface
temperature (350-400*F). Obviously, the stove surface temperature
reacts slowly to changes in combustion rate. Therefore, although both
parameters were continuously monitored, the combustion chamber tempera-
ture was generally the parameter used to determine whether or not air
inlet and damper adjustments were required.
Prior to conducting any emission tests, experimental burns were
conducted with the airtight stove to determine the basic operating
procedure to be followed and to establish draft settings and burn rate.
After several tests an operating procedure was determined which
resulted in a combustion chamber temperature of 500" F, a stove surface
temperature of 350 - 400 *F, and a burn rate of 2.5 kg/hr. These
ranges for combustion chamber and stove surface temperatures then
became the guidelines for the rest of the test program. However, it
must be noted that this combustion chamber temperature is marginal in
regards to the required temperatures for operation of the catalytic
stoves. During testing of the catalytic stoves, care was taken to
assure that an adequate "light-off" temperature (normally 1000 • F) was
reached and an attempt was made to maintain the combustion chamber
temperature (e.g., the temperature of the gas stream feeding the
catalyst) at a level above the manufacturer's suggested minimum tempera-
ture for maintaining catalytic reaction.
23
-------
All emission tests were conducted on a single charge of wood fed
to a hot bed of coals. The normal procedure followed was to first
charge the stove with a known mass of wood to bring the stove surface
temperature up to the desired level and to establish the hot coal bed.
Once this initial charge was reduced to less than one-tenth its initial
weight (dry basis), the stove was ready for the test burn charge.
After taring the stove scale, a weighed amount of kindling and wood was
loaded to the stove and the emission test was begun. The doors
remained open for 30 seconds to 3 minutes until it could be determined
that the fire was well established. The doors were closed and the air
control dampers were opened (usually full open at the very beginning of
the burn). Once the combustion temperature reached the appropriate
level, the dampers were regulated to a partially closed position;
monitoring of the combustion chamber temperature continued throughout
the test run. The test burn and emission testing was considered com-
plete when less than 10% of the wood mass charged (dry basis) remained.
During the burn, the stove generally was opened once near the end of
the burn, primarily to redistribute the wood into a single pile in the
chamber.
24
-------
WOOD TYPE
All tests were conducted using split Douglas fir as the fuel.
Where bark was present, the bark was not removed. The size of the
pieces generally was in the 12 - 16 inch (girth) range. With the excep-
tion of the high moisture test runs, the wood was seasoned; the high
moisture test runs were conducted with freshly cut, green Douglas fir.
Prior to each test run, moisture measurements were made on each
piece of wood to be used for that burn. These moisture values were
recorded and the average moisture value determined. The actual pro-
cedure used for wood moisture determination is discussed later in the
Sampling and Analytical Procedures Section.
TEST PROTOCOL
Moisture Tests
The purpose of this test series was to determine the impact of
wood moisture content on emissions. The series of tests conducted
consisted of six test burns using the airtight box stove. Two burns
were conducted at each of three moisture levels3; low (15-20 percent),
medium (25-30 percent), and high (greater than 40 percent). The test
protocol for this test series called for conducting all tests under the
same stove operating conditions, as discussed in a previous section of
this report. Consequently, instead of operating the stove with the
same damper/air settings for each moisture content; the air settings
were adjusted in order to achieve similar stove skin temperatures,
combustion chamber temperatures, and consequently, burn rate.
dry basis
25
-------
During this series of tests three of the emission samples (one
for each wood moisture level) were "split" samples. That is, Method 5
particulate sampling trains were changed during the run so that emis-
sion rates from different parts (phases) of the burn could be indepen-
dently determined. Furthermore, three "cold start" -tests also were con-
ducted as part of this test series. The cold start tests were short runs
conducted on the initial charge of wood to the cold stoves. Emissions
measured during start-up a cold stove were to be compared to emissions
measured when a fresh fuel load is added to a hot coal bed. Consequently,
this phase of the test program actually included 12 emission test for 9 burns.
Reasonable Emission Level
This series of emission tests were conducted in order to determine if
improved technology stoves would have a significant impact on reducing wood
stove emissions. If a significant reduction in emissions could be measured,
then a "reasonable" emission level to be expected from wood stoves could be
developed from these data. The test protocol called for sampling three
different improved technology stoves, as well as two retrofit devices to
be used on existing airtight box stoves. Two test runs on each stove were
conducted under the same stove operating conditions. For all test runs,
medium (25-30 percent) moisture wood was used as the fuel. Split emission
samples were collected during one run for each of the stoves tested.
Consequently, a total of 15 emission tests were conducted for 10 test burns.
For the catalytic stoves, the stove operating procedures included the
additional criterion of operating the stove at a high enough burn rate to
assure a gas temperature sufficient for catalytic oxidation. However, the
26
-------
starting point for determining the burn rate was the same operating
conditions (combustion chamber temperature 500°F, and stove surface
temperature 350 - 400°F) as with the initial airtight stove tests.
Thus, even during the test4ng for a reasonable emission levels, the
stoves were intentionally operated at a moderate to low burn rate.
The burn rate for one stove in this test series (Ceramic) could
not be controlled since the manufacturer provides no inlet air adjust-
ments. A damper is provided in the stack, and this damper was placed
in the fully closed position during testing.
Simplified Test Procedure
One of the objectives of this project was to evaluate "simplified
procedures" for measuring the emissions from wood stoves. The protocol
for this phase of the project called for simultaneous sampling using
several methods so that the results could be directly compared. This
multiple sampling was conducted throughout the test program during the
moisture and reasonable emission levels test series. No additional
test burns were conducted specifically for this purpose. Test pro-
cedures utilized for the simplified sampling included continuous
measurement of carbon monoxide, carbon dioxide, oxygen and hydro-
carbons; measurement of stack gas opacity;' measurement of creosote
accummulation; and, a smoke spot density test.
27
-------
STOVE SELECTION AND DESCRIPTION
A stove typical of those currently in use was desired for testing
the impact of wood moisture on emissions. Consequently, a simple air-
tight box stove was chosen for these tests. Since this stove type
is considered representative of stoves currently in use, this stove '
also was used in conjunction with testing of the retrofit devices.
The airtight box stove tested did not include any design features one
would expect to reduce emissions (e.g., baffles, secondary combustion,
thermostatic damper). Consequently, this stove could be used to
establish an emissions baseline for wood stoves operated at the test
conditions.
As already mentioned the test protocol called for the testing of
"improved" technology stoves to determine if and to what extent emis-
sion reductions could be obtained. A literature search of technical
papers and manufacturer's information was conducted to identify the
stoves to be tested. Initially, several stove design categories were
identified and considered likely to result in reduced emissions. These
general categories were: downdraft, secondary combustion, catalytic,
and design combinations such as catalytic/downdraft.
The final decision on which stoves to test was based on several
factors. The factors considered in the selection process were per-
ceived potential for reduced emissions, stove availability in the
Pacific Northwest, and results of emission data, where available.
Review of the market place did not reveal any true downdraft stoves
that were actively being marketed in the Pacific Northwest. Conse-
28
-------
quently, this stove category was dropped from consideration. Two sec-
ondary combustion air units and one ceramic stove were identified as
potential candidates. Due to limited resources, only one stove could
be tested. Based on discussion with the manufacturers or manufacturer's
representatives the decision was made to include the ceramic stove in
the test program, since this stove incorporates a unique technology —
combustion at high temperature with (theoretically) storage of the
heat in the ceramic for later slow release. Two stoves utilizing cat-
alysts to reduce emissions and increase thermal efficiency were iden-
tified and chosen for testing. One stove was typical of the majority
of the prototype catalytic stoves just entering the marketplace at the
time of this test program. That is, the stove is similar in configura-
tion to the typical airtight box stove but with a ceramic honeycomb
structure coated with a catalyst placed just prior to the stove outlet.
Some secondary air also is added just prior to the catalyst face in
order to assure adequate oxygen is present for combustion. The second
catalytic stove tested can be categorized as a catalytic/modified com-
bustion stove because this stove includes features in'addition to the
catalyst (e.g., tertiary air and baffling) to further enhance the com-
bustion process. This stove was chosen as the combination technology
stove.
Two retrofit devices with the potential for reducing emissions
(e.g., retrofit devices intended for other than simply increasing heat
transfer efficiency) were chosen for testing. Both the devices were
designed to reduce emissions after primary combustion as opposed to
29
-------
changing basic stove operating parameters for emissions reduction
(e.g., thermostatically controlled barametric dampers). The only
catalytic retrofit known to exist on the market was chosen for testing.
This device essentially retrofits the typical ceramic honeycomb
catalyst structure directly to the stove exit flue. The second device
tested retrofits in exactly the same manner, but is non-catalytic; the
device consists of a heavy stainless steel pad designed to capture
particulate matter as it leaves the stove and retain the material for
removal or combustion at a later time.
Although many other stove designs are available and many of those
investigated offered potential emission reductions, budget constraints
limited the stoves and retrofit devices which could be tested. The
design features of the devices chosen for testing are discussed in
further detail in the following sections.
30
-------
SAMPLING AND ANALYTICAL METHODS
General Test Procedure
As already mentioned each test burn was conducted with a single
charge of wood of known mass and moisture content. Emission -testing
was begun as soon as the test charge was ignited and testing was ter-
minated when one-tenth the initial charge mass (dry basis) remained
in the chamber. Testing was continuous throughout the run except in
several cases where "split" samples were taken. Data have indicated
that more emissions may be generated during the initial phase (first
hour) of a burn. Since one of the purposes of this test program was
to correlate data from several test procedures (e.g., opacity, smoke
spot) with particulate mass data, it was desired to take as many dis-
tinct particulate samples as possible and preferably to take these
measurements at different mass emission rates. Consequently, for
several of the test runs, the burn was divided into two phases and
separate ("split") emissions samples were taken for each phase. In
this case a second sample train was prepared and ready for testing
prior to beginning the test burn. At the appropriate time, one sample
train was removed and immediately replaced by the second sampler. This
entire process took less than five minutes. For these two-phase test
burns, the criteria used to determine the beginning of the second phase
was when the remaining wood mass equaled less than 75% of the initial
charge (dry basis). This generally occurred at about three quarters to
one-hour into the burn. Two phase samples were taken for Runs 1, 2, 3,
10, 13, 14, 17, and 19. The following sections briefly describe the
31
-------
test and analytical procedures followed during this test program.
Particulate Measurements
Particulate measurements were conducted according to the pro-
cedures of Oregon Source Test Method 7. This method is essentially -
EPA Test Method 5 except that the sampling train is modified to include
an unheated filter between the third and silica gel impingers. As per
Method 5, the front filter is maintained at a temperature of 250" F,
and the impinger train is maintained in an ice bath. Figure 2 is a
schematic of the sampling apparatus. Analysis includes measurement of
the condensible fraction collected in the impinger train. As per
Oregon Method 7, during this study the following sample fractions were
collected and analyzed:
1. Filterable (front half) Particulate
The filterable particulate is determined by combining two
sample fractions.
a. Acetone rinse of nozzle, probe, and front-half of front
filter holder - the wash is evaporated to dryness and_
the particulate gravimetrically determined.
b. Front filter - the glass fiber filter is desiccated and
brought to constant weight to gravimetrically determine
particulate.
2. Condensible (back-half) Particulate
The back-half particulate is determined by analyzing four
separate sample fractions.
32
-------
FIGURE 2
Particulate Sampling Apparatus
CNECX
VALVE
OHinCt MAROUETEH
33
-------
a. Impinger contents, organic fraction - the contents of
the impingers are extracted with dichloromethane
(CHLCLp, methylene chloride). This extracted organic
fraction then is evaporated to dryness and the remaining
residue is gravimetrically determined.
b. Impinger contents, water fraction - after extraction
of the impinger contents with organic solvent, the
remaining water fraction is brought to dryness and the
residue gravimetrically determined.
c. Impinger acetone wash - after removing the impinger con-
tents for analysis and completing the distilled water
impinger wash (which is added to the impinger contents),
the impinger train is rinsed with acetone. This acetone
wash is evaporated to dryness and the residue gravi-
metrically determined.
d. Filter
The back-half filter is desiccated, and the particulate
mass is gravimetrically determined.
3. Total Particulate
Total particulate is determined by combining the results
of the front-half and back-half analyses.
During this test program, sample recovery was completed at the
test site location immediately following completion of the test run.
All samples then were sent to the laboratory for analysis.
34
-------
The only major deviation from the established test method is
that single point sampling was conducted throughout the test program.
The center of the flue duct was chosen as the sample point. Secondly,
although probe nozzle sizes and sampling rates were chosen to approxi-
mate isolinetic sampling, isokinetic sampling was not of major concern;
a criterion of ±10% isokinetic was not used to determine whether or
not a test run was valid.
Stack Volumetric Flow
A calibrated S-type pi tot tube was used during the testing to
measure the stack gas velocity head. A micromanometer was used in
conjunction with this pi tot tube due to the extremely low flow rates
and, consequently, low pressure differentials. Although the pitot tube
was calibrated prior to use, measurements obtained during testing are
considered inaccurate. This conclusion was reached based on mass
(carbon) balance calculations. This is not surprising since the
apparent flow rates measured were very low (2-5 fps) with velocity head
pressure differentials in the .001 to .004 inches water range. Even
with a micromanometer, measurement of pressure differentials in this
range is marginal. Secondly, the small size of the stack (8-inch diameter)
can potentially cause an error in velocity measurements due to blockage
effects of the probe (the stack area is small enough that the cross-
sectional area of the probe is significant and can result in erroneous
pressure differential readings); the expected result would be pressure
radings biased high.
Consequently, volumetric flow rates were determined using
stoichiometric calculations and measured carbon monoxide, carbon dioxide,
35
-------
and gaseous hydrocarbons (as Hexane) concentrations to calculate the
total volumetric flow for the entire test period. The average concentra-
tions calculated from measurements (5-minute intervals) by the continuous
monitors were used for these calculations. Fuel mass burned was deter-
mined from direct measurements of the platform balance and of fuel
moisture content.
The assumptions and calculations followed are presented in detail
in Appendix A.
Carbon Monoxide. Carbon Dioxide, and Oxygen
Measurement of these gases were conducted by two separate tech-
niques during the sampling. An Orsat was used to measure CO, C0p» and
Oo, periodically during the testing (at half-hour to one-hour intervals).
In addition, a non-dispersive infrared (.NDIR) analyzer was utilized to
continuously monitor the CO and CO- content of the effluent; an oxygen
analyzer was utilized for continuous oxygen measurement (Runs 10 - 19
only). All instruments were calibrated before and after each test period
(e.g., before and after each run or daily when more than one run was
conducted in a day).
The sample for all gaseous measurements were taken from the same
point in the stack via a single probe and line; the sample line then
was split three ways to serve the NDIR, oxygen analyzer, and orsat.
Hydrocarbons
Gaseous hydrocarbons were measured with a NDIR calibrated on
hexane. A gaseous sample was continuously taken from the effluent
stream and fed to the analyzer. The sample was taken via the same
sampling probe and line as the CO, C0o> and 0~ samples. The analyzer
36
-------
was calibrated on a daily basis using hexane; no attempt was made to
determine the analyzer's response to other organic gases.
Opacity
Stack gas opacity measurements were made by two separate pro-
cedures throughout the testing program. Visual observations by a
certified observer were periodically conducted. Observations were
generally taken for ten-minute periods at intervals of one-half to
one-hour during the test run. Observations were made from the roof of
the building, and in most cases, trees were used as the observation
background.
An across stack transmissometer also was used during the test
program to continuously monitor the stack gas opacity. The transmis-
someter consists of an incandescent light source on one side of the
stack transmitting a light beam to a photovoltaic cell on the other
side of the stack. The photocell is calibrated to measure any attenua-
tion of the light by the stack effluent. A strip chart recorder is
used to continuously record the transmissometer output. The transmis-
ometer was designed and built according to the specifications of the
smoke generator required for Reference Test Method 9. Briefly these
specifications require the following:
1. A photopic spectral response of the photocell
2. A photocell angle of view less than or equal to 15 degrees
3. A light source angle of projection less than or
equal to 15 degrees
4. A calibration error less than or equal to ±3 percent
opacity (at three opacity levels)
37
-------
5. A response time of less than 5 minutes
6. A zero/span drift of less than ±1 percent opacity
in 30 minutes
A calibration was conducted prior to initiating the test program
and then repeated again in the middle of the program and at the end
of the program. Cleanliness of the photocell and light source were
routinely checked and no problems were encountered in this regard.
Creosote
Creosote deposition was measured by suspending a small steel plate
in the flue during the test burn; the plate was suspended from the top
of the stack by a fine wire. The plate was tared before use and the
subsequent weight gain due to creosote deposition during the test burn
was gravimetrically determined. The plates measured 3x5 inches
* 2-22
yielding a calculated surface area of 0.21 ft (1.94 x 10 m ) for
two sides.
The creosote collection plates were placed at two different loca-
tions during the emission testing. One plate was lowered into the
stack to a level just above the particulate sampling location. A
second plate was lowered into the stack to a level near the transmis-
someter. Consequently, two different samples were simultaneously
collected for each run.
Bacharach Smoke Spot Density
The Bacharach Smoke Spot tester was utilized to periodically
measure smoke density during the emissions test. This device consists
of a hand operated pump which is used to draw a sample from the stack
38
-------
and through a "spot" of filter paper. The degree of soiling of the
filter is then visually determined by comparison to a numerically
indexed chart. The smoke spot index is from 0 (no soiling) to 9
(maximum soiling). This device is frequently used to check soot levels
in the effluent of boilers. The manufacturer suggests that 10 strokes
of the hand pump be used for each sample (i.e., ten units of gas are
drawn through the filter for each spot test). However, during the
initial tests it was found that, due to the high emission levels, ten
strokes of the pump resulted in an excessive sample. Consequently,
during the remainder of the test program duplicate samples were taken;
one sample with five strokes and one sample with ten strokes. Only the
five stroke sample is reported, since this technique provided the better
sample (soiling ranges of 4 - 9).
For the smoke spot test multiple sampling locations were used to
determine if there might be a variation in results due to location within
the flue. Samples were routinely taken from two locations; the
sampling location for gaseous emissions, and the stack outlet. Results
for both locations are reported.
Wood Moisture Determination
Wood moisture was determined for each piece of firewood fueled.
The moisture content was measured and recorded within one day of its
use. A Delmhorst Moisture Meter operating on the principle of elec-
trical resistance was utilized for these measurements. This type
instrument is typically used for measuring moisture content of lumber
and timber.
39
-------
In addition to these measurements several wood samples were
analyzed for moisture content using ASTM procedure D 2016 "Moisture
Content of Wood". In this procedure a cross section of the log is cut
and weighed prior to and after heating to 211* F (105* C) to determine
the percentage weight loss due to water evaporation. These tests were
conducted strictly as a quality assurance check.
Net Efficiency Calculation
The net efficiency calculations presented in this report are
based on the heat loss method. In this method the useable heat obtained
from the appliance is calculated as the difference between the energy
input and the heat loss in the effluent gas stream. The energy input is
calculated from the gross calorific value and mass of fuel used during
the test. The heat loss is calculated as the sum of the sensible and
*
latent heats of the components of the effluent stream. During these
tests, the volumetric flow of the effluent stream was calculated from a
Stoichiometric Mass Carbon balance. The method used assumes all
carbon is accounted for in the measured quantities of carbon monoxide,
carbon dioxide, particulate (assume 70% carbon), and gaseous hydro-
carbons (measured as Hexane). The calculated volumetric flow is used to
calculate the sensible and latent heat losses. All other energy is
considered usable. The basic equation is as follows:
Net Efficiency, % = Heat Input - Heat Loss x 10Q
Heat Input
The complete procedure used is presented in Appendix A.
40
-------
TEST SITE AND SAMPLE LOCATION
Figure 3 is a schematic diagram of the testing site. The stove
being tested was placed on an electronic platform balance to continuously
monitor the change in wood mass during the test burn. A slip joint section
was used in the flue pipe to allow slight movement of the stove and platform,
as well as to accommodate expansion and contraction of the flue pipe. All
Method 5 particulate samples were taken through a single port located
approximately 8 stack diameters from the stove exit. Gaseous samples were
drawn from a sample port located approximately one foot lower in the duct.
A transmissometer system was permanently located above the Method 5 sampling
location.
This entire sample site set-up was located in an indoor unheated
area. During testing, the rain cap was removed from the stack exit.
41
-------
FIGURE 3
Test Site Configuration
Roof
T
3'
_L
10" diameter
Transmissometer
jL
\
Gaseous
Sample
Point
9.5'
Stove
Floor
T 0.5'
Particulate
Sample Point
11.75'12-00'
8'
• See
Table A
Scale
i
T
Rain cap removed during testing
A - Length of reduction is 0.6' when D = 8" and is 1.3' when D
D - 8" for all runs except Runs 14 and 15 when D = 6".
6",
42
-------
FIGURE 3, TABLE A
Stove Height (Inches)
Box Stove 30
Box Stove/Catalytic Add-on 42
Catalytic Box Stove 33
Ceramic Stove 12
Catalytic/Modified Combustion Stove 33
43
-------
STOVE DESCRIPTIONS
AIRTIGHT BOX STOVE
This stove is typical of the standard airtight stove constructed
of plate metal and welded at the seams. Figure 4 is a schematic of the
stove. The stove and hence the fire box are basically a simple large
box. In this case the fire box was lined with firebricks. Two ports on
the front of the stove are used to adjust combustion air. This stove
did not include any internal baffles, or other combustion modifying
devices, nor did it have any fans or blowers to aid in heat transfer.
The fire box on this stove has a volume of 4.9 cubic feet and an
approximate surface area (six sides) of 22.5 square feet. The manu-
facturer suggests that this stove will heat a space of 1250-1750 square
«
feet; no Btu output ratings are given.
During testing, the thermocouple for monitoring combustion chamber
temperature was placed through the back of the stove at the center of
the horizontal axis, and just above the top of the firebrick.
44
-------
, BO*
Gas
Hot-
-------
CATALYTIC STOVE
This stove's configuration is typical of an airtight box stove,
although the stove has several modifications to accommodate the cata-
lytic combustor. A schematic of the stove is presented in Figure 5.
It is of welded plate steel construction. The firebox is lined with
refractory brick on the bottom and partially up the side walls. The
fire box on this stove is relatively long and narrow (28 X 11 inches)
with a total volume of 2.6 cubic feet. The approximate surface area of
this stove is 25.4 square feet. Charging of the stove is done through
a single door on the narrow end of the stove.
As previously mentioned this stove has several significant mod-
ifications to the typical box design in order to accommodate the cata-
lyst. The catalyst used in this stove is of the precious metal type
coated on a ceramic honeycomb support. The honeycomb support in this
model stove is six inches in diameter. The primary combustion air
enters through two air control dampers on the stove door. Preheated
secondary air is controlled by a separate mechanism and is fed in to
the stove at a point immediately preceeding the catalyst. After passing
through the catalyst the gases flow around a baffle prior to exiting
through the flue. This extended pathlength of the exhaust fumes
permits additional residence time within the stove theoretically enhanc-
ing combustion and heat transfer. As a safety device to prevent back
flashing out the door, a catalyst by-pass damper is automatically
activated whenever the charge door is open. This by-pass is located
just inside and above the door, and consequently permits an immediate
46
-------
increase in draft and influx of air whenever the door is opened. This
stove does not include any fans or additional heat exchangers to aid in
heat transfer.
During the testing a thermocouple was placed into the firebox at
a point midway down (horizontal axis) the back wall and about eight
inches above the top of the firebrick lining. In addition, a second
thermocouple was placed just above the catalytic combustor in order to
monitor the temperature of the gases exiting the combustor.
47
-------
FIGURE 5
Catalytic Stove*
Flue
Gas
Catalyst
Catalys
By-pas:
Damper
TC - Thermocouple
* Not to scale
48
-------
CATALYTIC MODIFIED COMBUSTION STOVE
This stove encompasses several unique features (atypical of the
basic box stove) only one of which is a catalytic combustor. Figure 6
is a schematic of the stove. The firebox of this stove is basically
rectangular in shape with charging conducted through a single door on
the stove's end. The combustion chamber on this stove is not firebrick
lined. The combustion chamber volume is 3.0 cubic feet. Surface area
of the stove is estimated at 35 square feet.
First of all, note that this stove does not use the ceramic honey-
comb configuration for support of the catalyst medium. Although a
precious metal catalyst is utilized, it is coated on one surface of
flat ceramic "plates" of the stove manufacturer's own design. The
plates are located at the top of the combustion chamber with two series
of plates running the length of the chamber; one set of plates on each
side of the chamber. During operation, the effluent gas stream is
swept across the catalyst coated plate surface prior to exiting the
stove.
In addition to the catalyst, the other design feature of this
stove directed at increasing the combustion efficiency is the control
and direction of the combustion air paths. Both primary and secondary
combustion air enter through a single thermostatically controlled air
inlet damper. The primary air is directed up through the grating sup-
porting the burning wood, whereas the secondary air, after preheating,
is directed into the firebox chamber just above the burning wood mass.
The effluent gases pass upward to the top of the combustion chamber
49
-------
where the gases are directed by a set of baffles across the surface of
the catalyst coated plates. After passing over the plates, the gases
again are directed over a series of baffles to induce mixing and
enhance combustion,* prior to exiting through the flue pipe. In this
design, tertiary combustion air also is used. The tertiary air enters
the stove at the level of the wood grating via a manually operated air
damper. The air first travels the length of the stove via a channel
next to the wood grate, travels up a channel in the back of the stove
and then is directed through holes in the catalyst plates to mix with
the effluent combustion gases at the catalyst plate surface.
This stove does not utilize blowers to increase heat transfer.
However, finned heat exchangers are located along the surface of the
stove to increase heat transfer. These exchangers are located between
the firebox and the stove's outer shell.
During the emission testing, thermocouples were placed at three
locations within the stove chamber. These locations included the fire-
box chamber, the "third level air stage" just prior to the catalyst,
and the post catalyst stage. These locations are depicted in Figure 6.
Prior to conducting the emissions testing, discussions with the
manufacturer indicated minimum temperatures of 400° F for the tertiary
air and 650*F for the combustion chamber were required in order to
achieve catalytic reaction. Consequently, these temperatures were
considered as necessary minimum values during the emission testing.
50
-------
FIGURE 6
Catalytic - Modified Combustion Stove*
Flue
Gas
Catalyst Plates
1. Primary/Secondary air inlet
2. Primary combustion air
3. Secondary combustion air
4. Tertiary air inlet
* Not to scale
5. Tertiary air (through channel up
stove back)
6. Tertiary air inlet at catalyst
7. Heat exchange fins
TC - Thermocouple
51
-------
CERAMIC STOVE
The ceramic stove tested is significantly different in design
than the typical steel plate welded box stove. Figure 7 is a sche-
matic of the stove. The entire stove is cast in ceramic refractory
material; the exterior then is covered in decorative tile. The firebox
itself is spherical in shape with the combustion air entering near the
bottom of the firebox and to one side. This air is preheated during
its path through the refractory prior to entering the combustion
chamber. On this particular stove there is no air inlet control; the
only means of controlling the combustion rate is a damper in the flue,
as well as the charge rate of wood. This is an interesting feature in
regards to controlling air pollutant' emissions, since this feature
essentially means the operator cannot control the amount of combustion
air; therefore, the stove is unlikely to be operated in an air starved
condition. In order to promote heat transfer, this stove does have an
air space (channel) between the combustion chamber and the outer
ceramic shell. Room air is drawn downward through this channeling via
a blower located at the bottom of the stove and then is distributed to
the room from beneath the stove at floor level. Note that this blower
was not operated during the emission testing.
A single airtight door is provided for charging the fuel to the
combustion chamber. The chamber on this stove has a volume of 3.2
o
cubic feet (.08 m ). Surface area of the stove is estimated at 22.3
square feet. No recommendations are provided by the manufacturer
regarding estimated heat output or area space which this stove is cap-
able of heating.
52
-------
FIGURE 7
Ceramic Stove *
Flue
Gas
Room Air for
Heat Exchange
Combustion
Air
Charge Door
Combustion
Chamber
Recirculated
Room Air
* Not to scale
53
-------
CATALYTIC RETROFIT DEVICE
This retrofit device is designed to add a catalytic combustor
directly to the outlet of any typical wood stove. Figure 8 is a
schematic of the device. The device is inserted into the flue at the
*
point of the stove collar. The primary component of the system is the
precious metal catalyst supported on a ceramic honeycomb. When the
retrofit unit is installed, the catalyst sits directly above the stove
flue outlet and the effluent gases are directed through the catalyst,
over the heat exchanger and to the stack. The heat exchanger consists
of an array of stainless steel tubing situated directly over the
catalyst. A 140 CFM blower forces room air through the tubing to
achieve the heat exchange. A damper is situated to provide a by-pass
of the catalytic combustor. In normal operation this by-pass is
closed, although even in the closed position about a 10% by-pass is
permitted by design. The by-pass damper may be manually activated at
any time; the damper is automatically activated by a thermostatically
controlled solenoid if the operating temperature at the catalyst is
below a certain minimum level. A secondary air control is provided to
permit introduction of secondary combustion air just prior to the
catalyst to assure adequate oxygen for combustion.
During testing, the by-pass damper was open during charging of
the stove (catalyst off position). Once the stove was charged, the
electronic control was set to the catalyst on position; the damper
automatically closed once the catalyst reached a high enough tempera-
ture level to trip the solenoid. The unit was operated with the
catalyst in operation throughout the remainder of the test burn. The
blower serving the heat exchanger also was used during the test burns.
54
-------
FIGURE 8
Catalytic Retrofit Device*
To Flue pipe/Atmosphere
Blower
Secondary
Air Inlet
Damper
Heat Exchanger
Warmed
Room
Air
Catalyst By-pass Damper
TC - Thermocouple
* Not to scale
55
-------
NON-CATALYTIC RETROFIT DEVICE
This device is inserted into the flue immediately following the
stove outlet. The device consists of a large steel wire mesh pad
mounted on a cast iron support. Figure 9 is a schematic of the device.
In principle, the steel mesh traps any particulate in the effluent;
once the particulate is trapped, one of two things may then occur.
The trapped particulate remains on the pad, and thus is prevented from
emission to the atmosphere; the particulate then is removed from the
system when the pad is removed and cleaned. Alternatively, the par-
ticulate which accummulates on the pad during periods of stove operation
at low temperatures, is combusted upon stove operation at elevated
combustion temperatures when the steel pad's temperature is elevated to
a point capable of combusting the entrapped particulate. Only ash will
remain on the pad and this is removed from the system during pad
cleaning. In this scenario, the pad acts first as a collector, then as
a heat sink and reaction area to promote combustion.
This device does not include any additional heat exchangers or
blowers.
56
-------
FIGURE 9
Non-Catalytic Retrofit Device*
Flue
Gas
Stove
Effluent
Inspection/
Clean-out Port
Cast Iron Support
Steel Mesh Pad
* Not to scale
57
-------
RESULTS
The results for each stove are presented separately in tabular and,
in some cases, graphic format. A brief description of the tests is
presented for each stove with special attention given to any problems
encountered or anomalies in the data. Comparison of results among runs
or between stove types, as well as interpretation of the results are
presented in the Discussion of Results section of this report.
The first table in each subsection presents a complete listings of
the emission results, as well as other important test parameters (e.g.,
burn rate). Where applicable, the results also are presented separately
for each test run phase in a second similar table. Two separate tables
present the results of the Bacharach smoke spot tests and the visible
emissions opacity observations. Note that the Bacharach and the creosote
measurements were both taken at multiple sample locations. Bacharach
measurements were taken both at the gaseous emissions sampling port, as
well as at the stack outlet (see Figure 3) and are reported as such.
Creosote measurements were taken both at the sampling (Method 5) and trans-
mi ssometer levels in the stack (see Figure 3). The creosote value reported
in this section is the average value of the two measurements; the creosote
results are presented separately for each sample location in the Discussion
of Results Section (Table 44), as well as in the Appendix (Table B-2). Where
applicable, a single table summarizing the temperatures measured before
and after the catalyst is presented for each run.
58
-------
Separate figures present the variation of the gaseous measurements
(CO, CO > HC) as a function of time for each run and the variation of
both stove surface and combustion chamber temperatures as a function of
time. In both cases, the graphs were prepared using data recorded at
five-minute intervals throughout the duration of the run. Similarly, a
single figure presents the results of the emissions opacity data from
both the visible emission observer and the transmissometer. The Bacharach
smoke spot density results also are presented in the figure. The trans-
missometer data are plotted based on readings taken at ten-minute intervals,
A single average value is plotted for each visible emission observation
period which ranged from 5 to 30 minutes, depending on the test run.
The smoke spot density for each measurement taken at the outlet location
is plotted; generally, a measurement was taken approximately every 30-
minutes.
Finally, the stove operations log for each test run is presented in
tabular form. The operation log summarizes all changes made to stove
operation (e.g., opening and closing of air inlet dampers) throughout
the run and is presented in chronological order. The actual field data
sheet on which stove operations were recorded is presented in Appendix E
along with all other field data forms.
59
-------
AIRTIGHT BOX STOVE
A total of nine test runs were conducted on the airtight box stove.
Three tests (Runs 4, 6, and 8) were conducted with the stove starting
from a totally cold condition. Since the purpose of these three tests
were to determine if emissions during start-up-of the stove would be
significantly increased under cold start conditions, as compared to hot
start conditions, the tests were short runs. Six other test runs were
conducted, two test runs at each of three fuel moisture levels. Runs 1
and 7 were conducted with medium moisture fuel (20%, wet basis); runs
2 and 5 were conducted with dry fuel (12%, wet basis); and runs 3 and 9
were conducted with wet fuel (56%, wet basis).
Several minor anomalies occurred during the first few test runs.
During Run 1, no creosote measurement was taken during phase 2 at the
transmissometer location. In the second run, the first 20 minutes of
transmissometer opacity readings were not valid since the creosote cou-
pon was blocking the light transmission path. Due to the high fuel
moisture content, problems in maintaining combustion were encountered
during Run 8. Consequently, during this run, dry kindling had to be
added during the test in order to maintain combustion. This run was
from a cold start; similar operating problems were not encountered with
the high moisture fuel during the hot start tests (Runs 3 and 9).
The average burn rate for the six normal test runs was 2.7 kg/hr
with a range of 1.9 to 2.9. Average combustion chamber and surface tem-
peratures were 580°F (513 - 650a) and 420°F (336 - 507), respectively.
Range
60
-------
The average test time was 215 (189 to 275) minutes. Excess air ranged
from 161 to 322 percent with an average of 206 percent for the six tests.
The average burn rate for the three cold start tests was 5.3 kg/hr
(4.5 to 17.8). (This compares to an average burn rate of 4.0 kg/hr for
the similar first phase of Runs 1-3.) Average combustion and stove
surface temperatures were 580°F (470 - 675) and 400°F (233 - 575), respec-
tively. The average test time was 80 minutes. Excess air for these
tests averaged 225 percent.
61
-------
TABLE 3
Results — Airtight Box Stove
PARAMETERS
Fuel Charge (Ib, Wet)
Ob. Dry)
Fuel Moisture (S, Wet)
C; Dry)
Sample Time (Minutes)
Fuel Burned (Ib, Wet)
Ob, Dry)
Burn Rate (Ib/hr, Wet)
(Ib/hr, Dry)
Temperatures
Combustion (»F)
Surface (°F)
Stack (»F)
Total Particulate Emissions*
Concentration (g/dscm)
Rate (Ib/hr)
Factor (g/kg wood, Dry)
(g/104 Btu, Net)
Front-half (% of total)
Creosote (mg/m kg)
Hydrocarbons (ppm)
Hydrocarbons (g/kg wood. Dry)
Carbon monoxide (g/kg wood, Dry)
Opacity (observer, %)
Stack Gas Composition
C02 (i)
CO (I)
02 (S)
N2 C.)
Moisture (?)
Excess Air (5)
Stack Gas Flow (dscf/hr)
Efficiency, (net, S)**
1
26.2
20.7
21
26
275
23.7
18.7
5.2
4.1
513
336
265
1.34
0.09
22
17
28.6
969
237
13.8
190
36
4.8
1.0
15.3
50.7
-
228
1060
65
2
27.4
24.1
12
14
250
24.9
21.9
5.9
5.3
543
435
316
3.38
0.28
54
45
26.5
917
294
16.9
189
34
4.4
1.0
13.7
80.9
-
161
1350
60
RUN NUMBER
3 5
34.5
25.0
56
126
181
32.4
14.3
10.7
4.7
654
361
342
1.47
0.16
34
40
33.0
218
127
10.5
160 '
20
3.6
0.6
14,2
81.6
-
181
1750
43
26.0
22.8
12
14
208
23.8
20.9
6.9
6.0
569
450
299
5.00
0.37
•62
50
17.4
592
273
12.1
220
37
6.1
1.5
14,0
78.4
8.3
177
1200
62
7
28.3
22.6
20
25
198
26.2
21.0
7.9
6.4
610
507
372
1.53
0.12
19
14
23.2
568
197
8.8
160
28
7.0
1.1
13.5
78.4
6.2
167
1270
66
9
27.6
12.2
56
126
189
26.2
11.5
8.3
3.7
600
427
361
1.16
0.08
22
23
22.0
240
95
6.6
110
11
4.7
0.5
16.1
78.7
7.2
322
1130
48
•Oregon DEQ Method 7 (EPA Method 5 with back half)
** Since techniques for determining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
•comparison between stoves and test runs under the specific operating
conditions of this program.
62
-------
TABLE 4
Results -- Airtight Box Stove
(Cold Start Tests)
PARAMETERS
Fuel Charge (1b, Wet)
Ub, Dry)
Fuel Moisture (*., Wet)
(X, Dry)
Sample Time (Minutes)
Fuel Burned (lb. Wet)
(lb. Dry)
Burn Rate (Ib/hr, Wet)
(Ib/hr. Dry)
Temperatures
Combustion (»F)
Surface (°F)
Stack (of)
Total Particulate Emissions*
Concentration (g/dscm)
Rate (Ib/hr)
Factor (g/kg wood, Dry)
(g/104 Btu, Net)
Front-half (I of total)
p
Creosote (mg/m kg)
Hydrocarbons (ppm)
Hydrocarbons (g/kg wood, Dry)
Carbon monoxide (g/kg wood. Dry)
Opacity (observer, 5)
Stack Gas Composition
C02 (X)
CO (X)
02 (S)
N2 (X)
Moisture (S)
Excess Air (X)
Stack Gas Flow (dscf/hr)
Efficiency, (net, X)**
4
29.7
26.1
12
14
35
11.8
10.4
20.2
17.8
675
575
544
3.57
0.72
40
36
23.9
216
294
11.9
210
40
7.0
1.6
12.8
78.6
16.2
137
3230
56
RUN NUMBER
6 8
27.3
22.0
19
24
46
12.1
9.8
15.8
12.8
598
388
401
3.75
0.54
42
34
19.7'
291
227
9.2
170
46
7.3
1.3
12.7
78.7
13.8
138
2320
62
38.1
16.9
56
126
155
26.5
11.7
10.3
4.5
470
233
295
0.89
0.11
24
27
25.4
109
115
11.1
190
20
3.1
0.6
17.1
79.3
7.5
407
1950
44
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
** Since techniques for determining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves and test runs under the specific operating
conditions of this program.
63
-------
TABLE 5
Results -- Airtight Box Stove
(By Test Phase)
PARAMETERS
Fuel Charge (Ib, Wet)
Ob, Dry)
Fuel Moisture (:. Wet)
(I, Dry)
Sample Time (Minutes)
Fuel Burned (Ib, Wet)
Ob, Dry)
Burn Rate (Ib/hr, Wet)
(Ib/hr, Dry)
Temperatures
Combustion (eF)
Surface ("F)
Stack (<>F)
Total Particulate Emissions*
Concentration (g/dscm)
Rate (Ib/hr)
Factor (g/kg wood, Dry)
(g/104 Btu, Net)
Front-half (5 of total)
2
Creosote (mg/m kg)
Hydrocarbons (ppm)
Hydrocarbons (g/kg wood. Dry)
Carbon monoxide (g/kg wood. Dry)
Opacity (observer, *)
Stack Gas Composition
CO (5)
o2 (s)
N2 (%)
Moisture (5)
Excess Air (%}
Stack Gas Flow (dscf/hr)
Efficiency, (net, X)
1/1
1/2
RUN NUMBER/PHASE
2/1 2/2 3/1
3/2
68
11.8
9.3
10.4
8.2
538
380
359
0.79
0.11
11
46.9
683
264
12.9
190
36
6.2
1.2
11.9
80.8
7.2
113
2310
26.2
20.7
21
26
207
11.9
9.4
3.4
2.7
496
317
229
1.74
0.09
33
27.4
1051
226
15.2
220
37
4.2
1.0
16.2
78.6
4.9
311
820
42
9.4
8.3
13.4
11.9
552
472
437
7.01
0.82
69
23.9
725
365
12.9
149
47
8.1
1.3
12.6
78.0
17.8
137
1880
27.4
24.1
12
14
208
15.5
13.6
4.5
3.9
464
367
269
2.28
0.18
45
28.6
1038
280
19.8
210
33
3.9
0.9
13.9
81.3
5.7
165
1250
80
18.5
3.2
13.9
6.2
601
356
354
1.55
0.21
35
40.6
124
123
9.8
210
22
3.6
0.8
13.2
82.4
10.9
143
2200
34.5
25.0
27
38
101
13.9
6.1
8.3
3.6
693
364
322
1.37
0.12
33
19.7
344
130
11.1
140 •
19
3.6
0.5
15.0
80.9
7.0
221
1390
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
64
-------
TABLE 6
Bacharach Smoke Spot Results
RUN
1
1356-1831
B
2
1301-1711
3
1234-1535
4
0959-1034
SAMPLE TIME
1413 - 1417*
1445 - 1447
1530 - 1537
1615 - 1620
1645 - 1655
1730 - 1735
1320 - 1825
1837 - 1840
Ave. Temperature
1337 - 1341
1410 - 1412
1441 - 1442
1512 - 1514 '
1545 - 1547
1614 - 1615
1704
Ave. Temperature
1320 - 1324
1345 - 1347
1410 - 1416
1443 - 1444
1510 - 1515
Ave. Temperature
1018 - 1021
Ave. Temperature
SPOT DENSITY
Sample Location
Sampl ing
9
8
8
7
8
2
8
-
253°F
9
9
8
8
8
4
2
281° F
6
8
8
6
7
337° F
8
292 °F
Outlet
9
7
8
6
9
2
6
4
163°F
9
8
8
7
7
2
-
158°F
6
7
7
6
6
223° F
8
180°F
65
-------
TABLE 6 (Cont.)
RUN
5
1242-1610
6
0914-1000
7
1149-1507
8
0905-1140
9
1318-1627
SAMPLE TIME
1320 - 1339
1426 - 1428
1453 - 1455
1531 - 1533 .
Ave. Temperature
0943 - 0945
Ave. Temperature
1254 - 1255
1315 - 1317
1403
1440
Ave. Temperature
0935 - 0940
1012 - 1014
1134 - 1137
Ave. Temperature
1306 - 1308
1456 - 1458
1540 - 1546
Ave. Temperature
SPOT DENSITY
Sample Location
Sampling
8
8
7
5
249° F
8
293" F
8
7
7
-'
393° F
9
5
4
331° F
7
7
7
302 °F
Outlet
7
7
6
4
163° F
8
195° F
8
7
-
4
222° F
8
4
4
247<> F
6
7
6
240 °F
66
-------
TABLE 7
Visible Emissions Observation Log
01 IM
KUN
1
2
3
4
RUN TIME
—~~ •
Clock
• Phase 1
• Phase 2
1356 - 1504
1504 - 1831
1301 - 1343
1343 - 1711
1234 - 1354
1354 - 1535
0959 - 1034
Elasped Min
• Phase 1
• Phase 2
68
207
275
42
208
250
80
101
181
35
VISIBLE EMISSIONS MEASUREMENT
. Time Period
1355 - 1424
1424 - 1454
1515 - 1544
1545 - 1609
1620 - 1649
1650 - 1719
1300 - 1341
1341 - 1406
1410 - 1439
1440 - 1509
1515 - 1544
1545 - 1614
1615 - 1644
1645 - 1714
1230 - 1259
1300 - 1329
1335 - 1404
1405 - 1434
1435 - 1504
1505 - 1534
1013 - 1017
Minutes
Observed
30
30
. 30
30
30
30
41
19
30
30
30
30
30
30
30
30
30
30
30
30
5
Opacity
(Percent)
38
34
38
43
43
25
47
43
42
59
47
27
7
0
28
15
25
20
17
20
40
Opacity
Range
(Percent)
15 - 60
30 - 45
30 - 45
40 - 45
40 - 45
15 - 45
5-60
35 - 45
35 - 45
45 - 60
35 - 60
5-40
0-15
0 - 5
10 - 60
15
20 - 40
15 - 45
15 - 20
15 - 25
30 - 50
Average Opacity
(Percent)
Phases 1,2, total
"=••"-••- -.. •.._ . . _-l _:_ .^- ,„
36
37
T7"
47
00
34~
23
1 q
l :>
21
40
67
-------
TABLE 7 (Cont.)
RUN
5
6
7
8
9
RUN TIME
Clock
• Phase 1
• Phase 2
1242 - 1610
0914 - 1000
1149 - 1507
0905 - 1140
1318 - 1627
Elasped Min.
• Phase 1
• Phase 2
208
46
198
155
189
VISIBLE EMISSIONS MEASUREMENT
Time Period
1315 - 1319
1450 - 1554
1535 - 1539
0940 - 0944
1233 - 1237
1335 - 1339
1430 - 1434
0940 - 0944
1100 - 1104
1347 - 1351
1445 - 1459
1541 - 1553
Minutes
Observed
5
5
5
5
5
5
5
5
5
5
5
5
Opacity
(Percent)
42
41
28
46
28
36
19
25
14
11
14
7
Opacity
Range
(Percent)
35 - 50
30 - 55
20 - 30
40 - 50
20 - 40
25 - 40
15 - 20
20 - 30
10 - 20
10 - 15
10 - 20
5 - 10
Averaqe Opacity
( Percent)
Phases 1,2, total
37
46
28
20
11
68
-------
FIGURE 10
Run 1
Gaseous Component
FIGURE 11
Run 1
Temperature
Co*«lltOII
Stpve Surface
69
-------
FIGURE 12
Run 2
Gaseous Comoonent
I •
" m
S t
100 ISO
TIME, nimtttl
FIGURE 13
Run 2
Temperature
• Stove Surface
50 f=
inn IPS
TIW. Hliuitrs
isn i?s zoo
70
-------
FIGURE 14
Run 3
Gaseous Component
• CO
• HC
« CO,
s s
s §
i i
g 3
I I-
!§•
§ 8 ,
line. MIWICI
FIGURE 15
Run 3
Temperature
100 us
TIN;. Mimitrs
71
-------
FIGURE 16
Run 4
Gaseous Comoonent
CO
HC
TIHC. Himi
•
S 8
§ §
S 8
FIGURE 17
Run 4
Temperature
• Stove Swrftce
*• ClMmtton
SI) 75
inn iz*
Tint. Hlniitr*
150 175
72
-------
-FIGURE 18
Run 5
Gaseous Component
CO
HC
100 ISO
TIME. NliMttl
FIGURE 19
Run 5
Temperature
I *
5
§ *
* Stove Surfac
« Coobufttton
1 §
s s
§ s
Sff
s §
SO 7S
100 12S
TIHF. Hinulc&
ISO 175 700 ??S
73
-------
FIGURE 20
Run 6
Gaseous Comoonent
CO
IIC
100 HO
1IHC. Hliuitts
ii
II
§ 8
55
S 8
FIGURE 21
• Run 6
Temperature
• Stev« Sur
« (Mfcuitto
10(1 IZ5 IW1
TIHC. Himitcs
174 7ITO
74
-------
FIGURE 22
Run 7
Gaseous Component
100 ISO
IIHt. Dilutes
g g
fT. .«.
I 1
II
g s
§ s
FIGURE 23
Run 7
Temperature
7!v ' 111
75 IIX) US ISO
lire. Hiiunei
75
-------
-------
FIGURE 26
Run 9
Gaseous Comoonent
• CO
• HC
• CO,
100 ISO
TIME. miwt€!
FIGURE 27
Run 9
Temoerature
Cevhuttton
Stow Surface
us rou
77
-------
FIGURE 28
Opacity and Smoke Spot Density
vs
Time
(Run 1)
50
40
c
0)
o
0)
o.
o
CO
o
<-»•
CO
Bl
n>
4 -••
3
to
o
3
fD
t/>
to
-------
FIGURE 29
Opacity and Smoke Spot Density
vs
Time
(Run 2)
80 .
(J
O
TO
a.
O
60-
40-
20
0
Start
13:01
Transmissometer
Observer
(Ave. for period)
Smoke Spot
50
100
150
200
Elapsed Time, Minutes
250
End
17:15
9
. 8
. 7
6
5
4
3
2
-1
"-0
oo
3
o
fD
00
•a
o
H-
o
CU
n>
ex
3'
n>
in
_J.
o
79
-------
FIGURE 30
Opacity and Smoke Spot Density
vs
Time
(Run 3)
80-
60
c
O)
(J
i.
n
n>
ex
_j.
n>
3
o'
3
n>
in
-------
50
40
30
0
Ol
Q.
O
10
Q.
O
20
10
, FIGURE 31
Opacity and Smoke Spot Density
vs
Time
(Run 4)
Transmissometer
Observer
(Ave. for period)
Smoke Spot
0
Start
09:59
10
20
30
Elapsed Time, Minutes
81
End
10:34
- 9
-8
o
:*-
o>
in
•o
o
ist
o
o>
ro
4 2:
ro
3 I
CD
10
-------
FIGURE 32
Opacity and Smoke Spot Density
vs
Time
(Run 5)
80
o>
CD
"
o.
_J.
3
(D
3
l/>
o'
3
(D
-------
FIGURE 33
Opacity and Smoke Spot Density
vs
Time
(Run 6)
60
40
03
o
i-
Ol
CL
O
20
0
Start
09:15
25
-h 50
End
10:01
Elapsed Time, Minutes
CO
o
n>
CO
r>
n>
o.
—i.
n>
3
fD
in
Transmlssometer
Observer
(Ave. for period)
Smoke Spot
83
-------
Smoke Spot Scale, dimensionless
cr>
oo
C\J
en
O -3 O
• O C ••
C\J UJ LO
O
Lf)
O
O
Ol
4->
3
O)
T3
O)
Q.
(O
O
LO
o «a
O
in
o
CM
-------
Smoke Spot Scale, dimension!ess
oo
r-H O
-o
V- O
J_
Ol i- . 4)
(T3 in > O
.i- -a < e
ID
CO
UJ
CJ3
i- 3
T3
C
u
(O
Q.
o
o
.10
o
•a <«r
c ••
o
o
oo o
-------
Smoke Spot Scale, dimensionless
CO
r-1 O
s_
QJ
-t-J
QJ
E
O
to t- o
to 1. O Q.
-.- QJ **- 00
QJ
Q.
C OJ QJ
s- xi <:
^ o^—
QJ
co
c
d)
o
ex
00
(II
> ••- =
I- o:
-o
.c
ra
o
ia
CL
o
o
o
CN4
I —
CM
LU l£>
O
LO
to
O)
o T-
o H-
i—i
•o
to
CL
10
o
LO
o
tr>.
o
ro
O
CSJ
-i-> co
l. <-<
o CO
-------
TABLE 8
Operational Log
RUN 1
TIME
1356
1401
1406
1431
1508
1606
1616
1641
FUEL WT.
(lb)
26.2
24.8
21.8
17.5
13.8
9.4
8.7
6.5
COMBUSTION
TEMPERATURE
CF)
483
650
676
518
500
478
465
593
STOVE SURFACE
TEMPERATURE
(-F)
200
270
440
415
320
310
300
350
COMMENTS
Begin run; damper
full open.
Closed damper to 1
turn open.
Closed damper 3/4
turn (1/4 open).
Damper to 1/8 turn
open^
Phase II.
Open to 1/4 turn.
Open to 1/2 turn.
Close to 1/4 turn
open.
1740
4.1
428
300
Open damper to 1/2
turn.
1817
1833
3.2
2.5
437
436
280
295
Open to 3/4 turn.
End of burn.
87
-------
TABLE 8 (Cont.)
RUN 2
COMBUSTION STOVE SURFACE
TIME FUEL WT. TEMPERATURE TEMPERATURE COMMENTS
Ob) CF) CF)
1301 27.4 -- -- Begin run; damper
full open.
1305 25.8 470 400 Damper to 3/4 turn
open.
1311 24.5 503 420 Damper to 1/2 turn
open.
1326 20.7 579 510 Damper to 1/4 turn
open at 1323.
1336 19.2 664 500 Damper to 1/8 turn
open at 1333.
Damper to 1/16 open.
Phase II.
Closed damper.
Changed coupon at
1350.
Open damper to 1/8
turn; closed damper
at 1410.
Closed damper.
Open to 1/16 turn
at 1420.
1516 7.2 547 430 Closed damper at
1515.
88
1342
1343
1351
1356
1406
1411
1421
17.6
16.2 .
15.5
14.8
13.9
12.9
620
623
540
531
572
500
550
540
525
470
450
440
-------
TABLE 8 (Cont.)
RUN 2 (Cont.)
COMBUSTION STOVE SURFACE
TIME FUEL WT. TEMPERATURE TEMPERATURE COMMENTS
Ob) CF) CF)
1556 4.9 574 400 Open door and poke
fire.
1701 2.9 543 370 Open to 1/8 turn.
1715 2.5 541 380 End of burn.
89
-------
TABLE 8 (Cont.)
RUN 3
TIME
1234
1243
1249
1252
1317
1335
1338
1343
1355
FUEL WT.
Ob)
37.2*
32.8
31.7
31.1
25.7
23.0
22.2
20.6
18.5
COMBUSTION
TEMPERATURE
CF)
842
778
588
676
517
523
667
550
519
STOVE SURFACE
TEMPERATURE
CF)
425
470
470
440
315
270
290
350
340
COMMENTS
Begin test; dampers
full open.
Damper to 1/2 turn
open.
Damper to full turn
open.
Close damper to 3/4
open.
Damper open to 1 tun
Damper to full open.
Damper to 1 1/2 turn.
Damper to 1 3/4 turn.
Damper to 2 turns
open.
1356
1400
1408
16.4
998
300
Phase II.
Dampers full open.
Open doors to poke
fire.
1410
16.2
998
290
Close damper to 1
turn open.
Weights include 2.7 Ib tare.
90
-------
TABLE 8 (Cont.)
RUN 3 (Cont.)
TIME FUEL WT.
Ob)
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
COMMENTS
1412
Open damper to 1 1/2
turn open.
1425 13.4
953
320
Close damper to 1
turn open.
1440 10.3
713
400
Close damper to 3/4
turn open.
1530
5.6
526
335
Open damper to full
open.
1532
Close damper to 1 1/2
turn open.
1545
4.8
702
350
End of burn.
91.
-------
TABLE 8 (Cont.)
RUN 4
TIME
0959
FUEL WT.
29.7
COMBUSTION
TEMPERATURE
CF)
-0-
STOVE SURFACE
TEMPERATURE
-0-
COMMENTS
Begin run.
1001
869
220
Doors closed; damper
full open.
1008
810
608
Closed dampers to
1/2 turn open.
1022
19.8
575
625
Closed damper to 1/4
turn open.
1034
17.9
703
625
End of burn.
92
-------
TABLE 8 (Cont.)
RUN 5
TIME FUEL WT.
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
COMMENTS
1242
1243
28.6
447
390
Begin run.
Doors closed; damper
1 turn open.
1252 25.2
721
480
Closed damper to
1/2 turn open.
1307 21.3
563
575
Closed to 1/4 turn
open at 1305.
1353 14.8
518
450
Open damper to 1/2
open.
1358 13.9
570
500
Close damper to 1/4
turn open.
1611
4.8
520
400
End of burn.
93
-------
TABLE 8 (Cont.)
RUN 6
TIME
0915
0916
FUEL WT.
Ob)
27.3
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
COMMENTS
Begin test.
Doors closed; damper
full open.
0920
23.1
796
200
Close damper to 1/2
turn open.
0925
22.0
535
320
Open damper to 1 tur
open.
0930
20.7
659
400
Closed damper to 3/4
open.
0935
19.5
716
440
Closed damper to 1/2
open.
0940
18.7
525
450
Opened damper to
5/8 open.
0955
1001
16.2
15.2
514
569
410
410
Opened damper to 7/8.
End of burn.
94
-------
TABLE 8 (Cont.)
RUN 7
TIME
1149
1154
1159
FUEL WT.
(lb)
28.3
27.0
25.4
COMBUSTION
TEMPERATURE
CF)
648
594
STOVE SURFACE
TEMPERATURE
430
525
COMMENTS
Begin run.
Damper full open.
Close damper to 3/4
open at 1157.
1204
24.7
550
525
Open damper to 7/8
at 1202.
1214
22.4
490
540
Open damper to 1
turn at 1217.
1219
1224
22.0
20.5
428
653
470
460
Damper to full open.
Close damper to 3/4
open at 1225.
1229
1234
1244
1254
19.6
18.4
16.2
14.3
606
735
750
677
550
600
600
580
Damper to 1 turn
open at 1227.
Close damper to 1/2
open at 1235.
Close damper to 3/8
open at 1245.
Close damper to 1/4
at 1259.
1304
12.8
613
580
Open damper to 1/2
open at 1305.
95
-------
TABLE 8 (Cont.)
RUN 7 (Cont.)
TIME
FUEL HT.
Ob)
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
CF)
COMMENTS
1344
7.7
565
500
Open damper to 5/8
open at 1346.
1409
5.1
561
480
Open damper to 3/4
open.
1509
2.1
583
420
End of burn.
96
-------
TABLE 8 (Cont.)"
RUN 8
TIME
FUEL WT.
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
COMMENTS
0905
0910
35.3
32.1
421
120
Begin run.
Doors closed at
0908, damper full
open.
0915
30.5
396
150
Open doors to poke
at 0912.
0920
29.4
264
180
Doors opened at
0919, doors closed
at 0921.
0925
29.7
748
180
Added 1.2 lb of
kindling at 0924.
Door closed at 0925.
0930
28.5
415
200
Added 2.8 Ib of
kindling at 0932,
door closed at 0934.
0955
25.5
639
280
Open doors to poke
fire at 0956.
1005
23.9
395
240
Open doors to poke
fire at 1009.
1020
21.5
289
210
Open doors to remove
coupon rod and poke
the fire at 1015,
closed doors at 1018.
1140
11.6
412
280
End of burn.
97
-------
TABLE 8 (Cont.)
RUN 9
TIME
1318
1323
1333
FUEL WT.
29.8
27.5
23.1
COMBUSTION
TEMPERATURE
CF)
902
1251
STOVE SURFACE
TEMPERATURE
CF)
390
525
COMMENTS
Begin run.
Close doors at 1322.
Close dampers to
1 1/2 turn open.
1338
21.5
783
600
Open dampers to 2
turns open at 1341.
1448
10.4
490
430
Open damper to 2 1/2
turn.
1533
7.5
992
300
Open doors at 1532
to poke the fire.
1627
3.6
407
370
End of burn.
98.
-------
CATALYTIC RETROFIT
No major problems were encountered during Runs 10 and 11. However,
there was a minor problem with grounding of the post catalyst thermo-
couple during both runs. For this reason the thermocouple readings that
were obviously erroneous were deleted when calculating average values.
Combustion temperatures during both runs averaged around 600* F. As
previously mentioned, this combustion temperature is considered near
the lower limit for operation of the catalytic combustor. The average
temperature increase across the catalyst for Runs 10 and 11 was 125 and
290" F, respectively. The manufacturer of this unit reports that
temperature increases across the catalyst of 800*F are achievable with
this unit. Upon reviewing the preliminary test results, the manufacturer
indicated that, in his opinion, it did not appear that proper catalytic
operation had been obtained, and that this likely was due to high excess
air levels causing quenching of the temperature at the catalyst. Excess
air for these runs averaged 310 percent.
The average fuel consumption rate was 2.4 kg/hr. The average stove
surface temperature was 385* F; the average length of the test run was
160 minutes.
99
-------
TABLE 9
Results -- Catalytic Retrofit
PARAMETERS
Fuel Charge (Ib, Wet)
(Ib, Dry)
Fuel Moisture (5, Wet)
(5, Dry)
Sample Time (Minutes)
Fuel Burned (Ib, Wet)
(It), Dry)
Burn Rate (Ib/hr, Wet)
(Ib/hr, Dry)
Temperatures
Combustion (»F)
Surface («F)
Stack («F)
Total Particulate Emissions*
Concentration (g/dscm)
Rate (Ib/hr)
Factor (g/kg wood. Dry)
(g/104 Btu, Net)
Front-half (S of total)
Creosote (mg/m kg)
Hydrocarbons (ppm)
Hydrocarbons (g/kg wood. Dry)
Carbon monoxide (g/kg wood. Dry)
Opacity (observer, %)
Stack Gas Composition
co2 (;)
CO (%)
02 (S)
N2 (".)
Moisture (%}
Excess Air U)
Stack Gas Flow (dscf/hr)
Efficiency, (net, *)**
RUN
10
30.7
25.1
18
22
299
28.1
23.0
5.6
4.6
602
360
290
1.26
0.10
22
16
16.6
223
125
8.0
110
24
5.0
0.5
16.6
77.9
—
381
1320
68
NUMBER
11
27.7
22.5
19
23
223
25.6
20.7
6.9
5.8
599
408
353
1.34
0.09
17
12
19.7
190
137
6.2
90
16
7.4
0.6
14.7
77.8
8.9
240
1130
71
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
** Since techniques for determining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves and test runs under the specific operating
conditions of this program.
100
-------
TABLE 10
Results -- Catalytic Retrofit
(By Test Phase)
PARAMETERS RUN NUMBER
10/1 10/2
Fuel Charge (Ib, Wet) 30.7
(Ib, Dry) 25.1
Fuel Moisture (I, Wet) 18
(5, Dry) 22
Sample Time (Minutes) 59 23°
Fuel Burned (Ib, Wet) 10'4 17'7
(Ib. Dry) 8-5 14'°
Burn Rate (Ib/hr, Wet) 9-° 4-6
'(lb/hr, Dry) 7-4 3'7
Temperatures
Combustion (°F) 587 577
Surface («F) 445 337
Stack (»F) 34S 235
Total Particulate Emissions*
Concentration (g/dscm) 1-48 i-15
Rate (Ib/hr) °-l7 °-08
Factor (g/kg wood. Dry) 24 22
(g/104 Btu, Net)
Front-half (S of total) 16.2 16.8
Creosote (mg/ra2 kg) 81 306
Hydrocarbons (ppm) 80 139
Hydrocarbons (g/kg wood, Dry) *-6 9-3
Carbon monoxide (g/kg wood, Dry) HO HO
Opacity (observer, S) 22 32
Stack Gas Composition
C02 (%) S.7 4.8
CO (5) 0-s °-5
0 (;) 16.8 16.3
N2 (%) 76.9 78.4
Moisture (S) I3-4 6-4
Excess A1r (5) 432 349
Stack Gas Flow (dscf/hr) ' 1890 1150
Efficiency, (net, %)
* Oregon OEQ Method 7 (EPA Method 5 with back-half)
101
-------
TABLE 11
Bacharach Smoke Spot Results
RUN
10
1050-1554
11
1057-1440
•
SAMPLE TIME
1155 - 1200
1257 - 1259
1336 - 1339
1412 - 1415
: 1459 - 1502
Ave. Tempera tur
1133 - 1145
1207 - 1213
1245 - 1250
1338 - 1343
1440 - 1442
Ave. Temperature
SPOT DENSITY
Sample Location
Samp! ing
7
5
7
6
3
175" F
7
6
5
7
295° F
Outlet
7
5
7
6
2
162 "F
7
6
6
5
6
155" F
102
-------
TABLE 12
Visible Emissions Observation Log
RUN
10
11
RUN TIME
Clock
• Phase 1
• Phase 2
1050 - 1159
1204 - 1554
1057 - 1440
•
Elasped Min.
• Phase 1
• Phase 2
69
230
299
223
VISIBLE EMISSIONS MEASUREMENT
Time Period
1055 - 1124
1125 - 1154
1204 - 1213
1055 - 1104
1210 - 1219
1300 - 1309
1400 - 1414
1440 - 1444
Minutes
Observed
30
30
10
10
10
10
15
15
Opacity
(Percent)
22
20
32
9
25
16
16
15
Opacity
Range
(Percent)
10 - 40
15 - 25
25 - 40
0-20
15 - 30
10 - 20
15 - 20
10 - 15
Averaqe Opacity
( Percent)
Phases 1,2, total
21
32
23
16
103
-------
TABLE 13
CATALYST TEMPERATURES
RUN 10
TIME
Clock /
1055
1100
1110
1120
1130
1140
1210
1225
1240
1300
1320
1340
1400
1420
1440
1500
1545
Elapsed
05
10
15
30
40
50
80
95
110
130
150
170
190
210
230
250
295
TEMPERATURE
precatalyst
(combustion)
—
—
—
590
567
553
527
520
522
—
—
—
—
593
498
477
473
Average Value 532
• r-
, " r
post-catalyst
810
984
820
714
723
618
611
627
597
536
565
584
704
624
538
540
558
656
Average AT across catalyst 125"F.
104
-------
RUN 11
TABLE 13 (Cont.)
CATALYST TEMPERATURES
TIME
Clock /
1102
1107
1122
1137
1152
1202
1212
1227
1242
1302
1317
1332
1402
1422
1442
Elapsed
05
10
25
40
55
65
75
90
105
125
140
155
185
. 205
225
TEMPERATURE
precatalyst
(combustion)
—
733
640
517
520
465
—
580
637
608
533
498
—
—
—
Average Value 573
, 'F
post-catalyst
1120
1036
965
921
930
—
825
850
747
894
781
111
830
660
748
863
Average AT across catalyst 290'F.
105
-------
FIGURE 37
Run 10
Gaseous Component
to
HC
100
TIMC, Xtnutcs
FIGURE 38
Run 10
Temoerature
* Stove Surface
I I
8 8
100
lIHf.
106
-------
FIGURE 39
Run 11
Gaseous Component
• en
• HC
* CO,
FIGURE 40
Run 11
Temperature
101)
TIKt.
107
-------
FIGURE 41
Opacity and Smoke Spot Density
vs
Time
^Kun IDJ
— • — Transmissometer
— — Observer
40 .
S 30
o
I.
OJ
0.
-
r^*
•r™
.»
a 20
o
10
(Ave. for period)
• • • Smoke Spot
•
~"
It K A
A /\ /\
/ \ \ / \
1 \ 1 \ / \
/
/
/
'
J
9
- 8
- 7
- 6
5
4
3
2
1
0
0 100 200 300't
Start End
10:50 Elaosed Time. Mlnutp* 15:54
CO
ro
Co
-o
o
CO
o
OJ
"•^
n>
V
CL
3'
n>
3
o'
3
(T>
in
in
108
-------
S-
O)
QJ
O
T3
O
s_
Q.
S_
i.
O) OJ
jQ
O •
O
cu
00
cu
o
Smoke Spot Scale, dimension!ess
co
vo
CM
c
O)
a
CM o
•=r CL
« m „ S'
= 1! > J
C3 o I-
E J J
T3
C
(O
u
re
a.
o
-a •
c
O
o
CM
O
o
to
QJ
cu
•o
(U
(/I
a.
(O
LJJ
cn
O
O
CO
o
CM
s- in
o ie ••
4-> O
C/O ^H
-------
RUN 10
1315
TABLE 14
OPERATION LOG
TIME
1050
1055
1105
mo
1120
1125
1130
1135
1155
1220
1240
1250
1255
1313
FUEL WT.
(lb)
30.7
29.5
26.4
25.3
23.8
23.1
22.7
22.2
20.3
17.8
15.9
14.9
14.5
COMBUSTION
TEMPERATURE
CF)
—
911
733
687
639
637
667
655
608
538
554
559
549
STOVE SURFACE
TEMPERATURE
CF)
500
590
590
480
440
390
360
300
310
295
280
280
COMMENTS
Doors opened, begin i
Closed doors, damper
opened at 1053.
Damper to 1/2 open
at 1105.
Damper to 1/4 open
at 1112.
Secondary air control
off.
Secondary air control
- on.
Secondary air control
off.
Secondary air control
on.
Opened damper to 1/2
open, end of Phase I.
Opened damper to 3/4
open.
Secondary air control
off.
Secondary air control
on.
Dampers to 1 turn opei
Doors open, fire
12.9
534
275
stoked.
Dampers to 1 1/2
turn open.
110
-------
TABLE 14 (Cont.)
RUN 10 Cont.
TIME
'1345
1355
1415
1425
1455
1510
1520
1545
1550
1554
FUEL WT.
db)
9.6
8.2
6.8
6.0
5.1
4.6
3.9
2.8
2.7
2.6
COMBUSTION
TEMPERATURE
CF)
789
792
708
659
488
443
570
490
480
—
STOVE SURFACE
TEMPERATURE
CF)
400
450
390
370
300
280
350
380
360
—
COMMENTS
Dampers to 1 turn
open at 1340.
Closed dampers to 3/4
turn open.
Closed damper to 1/2
turn open at 1400.
Closed damper to 1/4
open.
Open damper to 1/2
turn open at 1455.
Damper to 1 1/2
turn open at 1508.
Damper to 1 turn
open at 1518.
Dampers to 1 1/8
turn open at 1546.
Damper to 1 1/2
turn open at 1551 .
End of Burn.
Note: Stack damper fully open during entire run.
Ill
-------
TABLE 14 (Cont.)
RUN 11
TIME
1057
1100
1102
1107
1143
1157
1202
1212
1222
1227
1232
1247
1257
1402
FUEL WT.
(Ib)
27.7
25.2
25.0
19.5
17.9
17.1
15.8
14.9
14.7
14.2
11.1
9.5
5.3
COMBUSTION
TEMPERATURE
( F)
""• "
1153
776
507
465
465
504
486
553
830
620
627
485
STOVE SURFACE
TEMPERATURE
( F)
_ — ->
500
550
370
330
320
300
300
360
420
540
490
320
COMMENTS
Begin Run.
Doors closed at 1100
dampers opened; cata
by-pass closed.
Dampers to 1 turn
open at 1102.
Dampers to 1/2 turn
open at 1107.
Opened dampers to
3/4 turn open at 114;
Open damper to 1 -turr
open at 1158.
Removed Post catalyst
TC wire and re-twist€
end.
Opened damper to 1 I/
turn open at 1213.
Opened damper to 2
turns open at 1224.
Damper to 1 1/2 turn
open at 1229.
Closed damper to 1 I/
turn at 1234. Change
filters.
Closed damper to 1 tu
open at 1251.
Closed damper to 3/4
turn open at 1301 .
Opened damper to 1
turn at 1403.
112
-------
TABLE 14 (Cont.)
RUN 11 Cont.
TIME
1407
1422
1427
1442
FUEL HT.
4.9'
3.1
2.1
COMBUSTION
TEMPERATURE
CF)
425
1120
562
STOVE SURFACE
TEMPERATURE
CF)
310
480
480
COMMENTS
Opened damper to 1 1/2
turn.
Fire stoked.
Opened damper to 3/4
turn at 1429.
End of Burn at 1444.
113
-------
NON-CATALYTIC RETROFIT
Runs 12 and 13 were conducted with the airtight box stove fitted
with a non-catalytic retrofit device. The first run with this device
was conducted on 8/27/81; prior to this run, the steel mesh in the device
was cleaned and tared. After completing the emission test run on 8/27/81,
stove operation was continued in order to determine if residue would build
up on the mesh and cause operational problems. The stove also was operated
on 8/28/81 and 8/31/81 prior to the second emission test on 9/01/81. The
steel mesh pad was reweighed after the second emission test on 9/01/81.
A total increase of only 2.8 grams over the original tare weight was
measured.
During Run 12, the pump on the 0- meter malfunctioned; therefore,
orsat oxygen data were used for this test. A new pump was added to the
system prior to Run 13. No other anomalies were noted for these test runs.
The average burn rate for the two runs was 2.9 kg/hr. Average
combustion chamber and stove surface temperatures were 565 and 365* F,
respectively. Excess air for the two runs averaged 300 percent. The
average length of the test burn was 260 minutes.
114
-------
TABLE 15
Results -- Non-catalytic Retrofit
RUN N.IMBER
12 13
Fuel Charge (Ib, Wet) 30.8 29.2
Ob, Dry) 24.8 23.8
Fuel Moisture (X, Wet) 19 19
(X, Dry) . 24 23
Sample Time (Minutes) 255 270
Fuel Burned (Ib, Wet) 28.5 27.0
(Ib. Dry) 23.1 21.9
Burn Rate (Ib/hr, Wet) 6.7 6.0
(Ib/hr, Dry) 5.4 4.9
Temperatures
Combustion (°F) ' 574 553
Surface (°F) 352 381
Stack (of} 265 251
Total Particulate Emissions *
Concentration (g/dscm) 1-76 1.84
Rate (Ib/hr) • 0.21 0.17
Factor (g/kg wood, Dry) 38 35
(g/104 Btu. Net) 32 27
Front-half (X of total) 19.6 17.3
Creosote (mg/m2 kg) 273 337
Hydrocarbons (ppro) 152 136
Hydrocarbons (g/kg wood. Dry) -- 11.8 9.3
Carbon monoxide (g/kg wood, Dry) 200 160
Opacity (observer, X) ' 22 23
Stack Gas Composition
C02 (X) ' 3.7 4.1
CO (X) 0.8 0.7
02 (X) 16.4 15.8
N2 (X) ' 79.2 79.4
Moisture (X) 8.5
Excess Air (X) 322 281
Stack Gas How (dscf/hr) 1890 1490
Efficiency, (net, !)** 60 66
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
** Since techniques for datermining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves and test runs under the specific operating
conditions of this program.
115
-------
TABLE 16
Results -- Non-catalytic Retrofit
(By Test Phase)
PARAMETERS RUH NUMBER
13/1 13/2
Fuel Charge (lb. Wet) 29.2
(lb, Dry) 23.8
Fuel Moisture (2. Wet) 19
(X, Dry) 23
Sample Time (Minutes) 50 220
Fuel Burned (lb. Wet) 10.6 16.4
(lb. Dry) 8.6 13.3
Burn Rate (Ib/hr, Wet) 12.7 4.5
(Ib/hr, Dry) 10.3 3.6
Temperatures
Combustion («F) 589 545
Surface (»F) 419 373
Stack (oF) 275 226
Total Particulate Emissions*
Concentration (g/dscm) 3.05 1.35
Rate (Ib/hr) 0.44 0.11
Factor (g/kg wood, Dry) 42 30
(g/104 Btu, Net)
Front-half (S of total) 19.3 15.9
Creosote (mg/m2 kg) 349 329
Hydrocarbons (ppm) 169 129
Hydrocarbons (g/kg wood. Dry) 8.4 10.4
Carbon monoxide (g/kg wood. Dry) ISO 180
Opacity (observer, %) 33 20
Stack Gas Composition
C02 (•„) 6.1 3.7
CO (5) 0.9 0.7
02 (S) 14.8 16.0
N2 (*) • 78.2 79.6
Moisture (%) 13.0 5.6
Excess Air (!) 228 292
Stack Gas Flow (dscf/hr) 2300 1300
Efficiency, (net, 5)
Oregon DEQ Method 7 (EPA Method 5 with back-half)
116
-------
TABLE 17
Bacharach Smoke Spot Results
RUN
12
1020-1435
13
1003-1433
SAMPLE TIME
1125 - 1131
1215 - 1233
1249 - 1254
1336 - 1338
1427 - 1430
Ave. Temperature
1030 - 1035
1211 - 1215
1323 - 1326
1415 - 1418
Ave. Temperature
SPOT DENSITY
Sample Location
Sampl ing
7
7
7
7
5
201° F
8
7
6
3
269° F
Outlet
7
7
6
6
5
140" F
8
6
6
2
134° F
117
-------
TABLE 18
Visible Emission Observation Log
RUN
12
13
RUN TIME
Clock
• Phase 1
• Phase 2
1020 - 1435
1003 - 1053
1053 - 1433
Elasped Min.
• Phase 1
• Phase 2
255
50
220
. 270
VISIBLE EMISSIONS MEASUREMENT
Time Period
1020 - 1029
1110 - 1119
1215 - 1224
1310 - 1320
1005 - 1015
1105 - 1115
1215 - 1225
1333 - 1343
Minutes
Observed
10
10
10
10
10
10
10
10
Opacity
(Percent)
25
20
22
20
33
21
19
24
Opacity
Range
(Percent)
10 - 35
20 - 25
20 - 25
20
25 - 45
15 - 25
15 - 25
20 - 25
Average Opacity
( Percent)
Phases 1,2, total
22
33
21
24
118
-------
FIGURE 43
Run 12
Gaseous Comoonent
• HC
« CO,
II
§§
loo isn
tlNC. W flutes
FIGURE 44
Run 12
Temperature
• Stnv* Sortk
* CwHiuttton
119
-------
FIGURE 45
Run 13
Gaseous Component
• to
« nc
I ~
w
a o
o u
s §
II
100 ISO
tint. Ninuiei
FIGURE.46
Run 13
Temperature
Stove Sin-face
§ g
S 8
120
-------
FIGURE 47
Opacity and Smoke Spot Density
vs
Time
(Run 12)
60
Transmissometer
Observer
(Ave. for period)
Smoke Spot
40
30
10
4 2:
3 ^
StSrt
10:20
25
75
"Too T2l> TsTT
Elapsed Time, Minutes
175
200
225
121
-------
FIGURE 48
Opacity and Smoke Sopt Density
vs
Time
(Run 13)
60
30
20
10
Transmissometer
Observer
(Ave. for period)
Smoke Spot
J
2b
50
100
125
150
175
200
225
250
2/1
-------
RUN 12
TIME
FUEL WT.
TABLE 19
Operational Log
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
COMMENTS
1020
1023
30.8
Begin run.
Closed doors, damper
full open.
1025
28.5
1105
460
Close damper to 1
turn open at 1026.
1035
25.8
1325
500
Open damper to 1/2
turn at 1036.
1046
23.5
788
420
Open doors to remove
wood from TC at 1049.
1100
21.2
681
420
Close damper to 1/4
open at 1102.
1105
20.6
643
400
Close damper to 1/8
open at 1107.
1115
19.6
551
350
Open damper to 1/4
open at 1118.
1150
16.5
481
290
Open damper to 1/2
open.
1220
13.8
524
290
Open damper to 3/4
turn open at 1218.
1235
12.3
456
300
Open damper to 1
turn open at 1237.
1240
11.8
466
320
Open damper to 1 1/2
turn open at 1241.
123
-------
TABLE 19 (Cont.)
RUN 12 (Cont.)
TIME
1305
FUEL WT.
Ob).
8.8
COMBUSTION
TEMPERATURE
CF)
622
STOVE SURFACE
TEMPERATURE
CF)
380
COMMENTS
Close damper to
1 1/4 turn open at
1307.
1340
5.6
478
340
Open damper to 2
turns open at 1348.
1410
4.1
453
280
Open doors to poke
fire at 1412.
1415
3.7
592
290
Close damper to 1
turn open at 1416.
1435
2.5
413
360
Close damper to 2
turns open at 1435.
1438
2.3
427
350
End of burn at 1439.
124
-------
TABLE 19 (Cont.)
RUN 13
TIME
1003
1006
FUEL WT.
Ob)
29.2
COMBUSTION
TEMPERATURE
CF)
STOVE SURFACE
TEMPERATURE
CF)
COMMENTS
Begin run.
Doors closed;
damper full open.
1018
26.0
667
420
Close damper to
1 1/2 turn open
at 1020.
1023
25.1
779
470
Close damper to 1/2
turn open at 1025.
1053
18.6
595
400
Close damper to 1/4
turn open at 1054.
1058
1123
17.6
15.1
581
521
400
320
End of Phase I.
Open damper to 1/2
turn open at 1125.
1138
13.5
492
320
Open damper to 3/4
turn open at 1139.
1148
12.6
481
305
Open damper to 1
turn open at 1149.
1213
9.2
682
420
Close damper to 1/2
turn open at 1216.
1218
8.7
717
460
Close damper to 1/4
turn open at 1220.
1248
6.7
504
370
Open damper to 1/2
turn open at 1250.
125
-------
TABLE 19 (Cont.)
RUN 13 (Cont.)
TIME
1308
FUEL WT.
5.9
COMBUSTION
TEMPERATURE
CF)
502
STOVE SURFACE
TEMPERATURE
CF)
350
COMMENTS
Open and poked fire,
damper to 1 turn opei
at 1310.
1333
4.5
487
390
Open damper to
1 1/2 turn at 1338.
1358
3.5
500
400
Open damper to 2
turns at 1359.
1433
2.2
519
360
End of burn at 1434.
126
-------
CATALYTIC BOX STOVE
No major problems were encountered with testing during Runs 14
and 15. However, problems were encountered with erratic thermocouple
readings in the combustion chamber. To partially alleviate this problem,
during Run 15 an additional thermocouple was placed just prior to the
catalyst.
The average burn rate for the two tests was 2.0 kg/hr. Average
combustion chamber and stove surface temperatures were 710 and 400CF,
respectively. Excess air averaged 156 percent for the two runs; the
average length of the test was 180 minutes.
Catalytic combustor operation was marginal during Run 15, as
indicated by temperature change across the catalyst. During Run 15 a
catalyst temperature change of only about 100°F was noted, with an
average combustion chamber temperature of 800°F. For Run 14 with a
significantly lower average combustion chamber temperature of only 600°F,
a temperature rise of about 200°F was noted across the catalyst. During
Run 14 a fairly constant glow was noted on the catalyst. Nonetheless,
Run 14 had significantly higher emissions than Run 15. From these
results it would appear that the higher burn rate of Run 15 (2.2 kg/hr vs
1.6 kg/hr) had a greater impact on emissions, than did catalyst operation.
Note in Figure 52 how at approximately 70 minutes into test Run 15, the
combustion chamber temperature was increased resulting in a significant
increase in C0« and corresponding decrease in CO.
127
-------
TABLE 20
Results -- Catalytic Box Stove
PARAMETERS RUN NUMBER
. 14 15
Fuel Charge (Ib, Wet) 16.8 16.6
(Ib, Dry) 13.8 13.3
Fuel Moisture (J, Wet) 17 20
(t. Dry) 21 25
Sample Time (Minutes) 209 150
Fuel Burned (Ib, Wet) I5-5 15-4
(Ib, Dry) 12.9 12.3
Burn Rate (Ib/hr, Wet) 4-= 6-2
(Ib/hr, Dry) 3.7 4.9
Temperatures
Combustion («F) 609 815
Surface («F) 382 412
Stack («F) 2*4 323
Total Particulate Emissions*
Concentration (g/dscm) 2.15 1.51
Rate Ob/hr) 0.14 0.11
Factor (g/kg wood, Dry) 38 23
(g/104 Btu. Net) 28 16
Front-half (5 of total) 15.6 21
Creosote (mg/m kg) 379 273
Hydrocarbons (ppm) 167 134
Hydrocarbons (g/kg wood. Dry) 10.7 7.3
Carbon monoxide (g/kg wood. Dry) 120 50
Opacity (observer, J) 10 10
Stack Gas Composition
C02 (S) 5.9 6.3
CO (5) 0.6 0.3
02 (S) 12.8 13.5
N2 (%) 80.7 78.9
Moisture (5) « 10.4
Excess Air (X) 140 172
Stack Gas Flow (dscf/hr) 1060 1200
Efficiency, (net, %)** 68 71
* Oregon OEQ Method 7 (EPA Method 5 with back-half)
** Since techniques for determining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves and test runs under the specific operating
conditions of this program.
128
-------
TABLE 21
Results — Catalytic Box Stove
(By Test Phase)
PARAMETERS RUN NUMBER
14/1 14/2
Fuel Charge (lb, Wet) 16.8
(Ib, Dry) 13.8
Fuel Moisture (S, Wet) *7
(5. Dry) 21
Sample Time (Minutes) *2 167
Fuel Burned (lb, Wet) °.7 8.8
(lb. Dry) 5.6 7.3
Burn Rate (Ib/hr, Wet) 9-6 3-2
(Ib/hr, Dry) 8.0 2.6
Temperatures
Combustion (»F) 582 615
Surface (°F) 407 376
Stack (»F) 287 212
Total Particulate Emissions*
Concentration (g/dscm) 2.14 2.15
Rate (Ib/hr) 0.39 0.08
Factor (g/kg wood, Dry) 49 30
(g/104 Btu, Net)
Front-half (S of total) 24.7 12.2
Creosote (mg/m2 kg) 389 371
Hydrocarbons (ppm) 102 183
Hydrocarbons (g/kg wood. Dry) 8.4 7.0
Carbon monoxide (g/kg wood, Dry) 80 110
Opacity (observer, X) 13 9
Stack Gas Composition
C02 (S) 3.9 6.4
CO (5) 0.3 0.7
02 (5) 12.9 12.7
N2 (5) . 82.9 80.2
Moisture (Z) 13.6 8.3
Excess Air (%) 140 140
Stack Gas Flow (dscf/hr) 2940 590
Efficiency, (net, %)
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
129
-------
TABLE 22
Bacharach Smoke Spot Results
RUN
SAMPLE TIME
SPOT DENSITY
Sample Location
Sampling
Outlet
14
1050-1419
15
1035-1305
1136 - 1139
1229 - 1235
1336 - 1339
1412 - 1417
Ave. Temperature
1134 - 1137
1232 - 1238
Ave. Temperature
7
6
6
7
192°F
6
321° F
130
7
6
4
4
109CF
4
145° F
-------
TABLE 23
Visible Emission Observation Log
RUN •
14
15
RUN TIME
Clock
• Phase 1
• Phase 2
1050 - 1232
1232 - 1419
1035 - 1305
Elasped Min.
• Phase 1
• Phase 2
42
'167
109
150
VISIBLE EMISSIONS MEASUREMENT
Time Period
1107 - 1117
1143 - 1153
1255 - 1265
1415 - 1425 ,
1041 - 1051
1203 - 1213
Minutes
Observed
10
10
10
10
10
10
Opacity
( Percent)
13
6
17
<5
14
7
Opacity
Range
(Percent)
10 - 20
5 - 10
10 - 20
<5
5 - 25
5 - 10
Average Opacity
( Percent)
Phases 1 ,2 .total
10
9
10
11
131
-------
RUN 14
TABLE 24
CATALYST TEMPERATURES
TIME
Clock /
1050
1150
1250
Elapsed
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
TEMPERATURE
precatalyst
(combustion)
—
652
535
473
520
870*
738
652
607
598
567
560
640
658 •
584
598
589
613
579
547
, 'F
post-catalyst
_—_
1102
1032
1002
1131
1122
850
725
711
603
817
758
732
756
767
753
714
669
655
681
Thermocouple repositioned.
132
-------
TABLE 24 (Cont.)
CATALYST TEMPERATURES
RUN 14 Cont.
TIME
Clock / Elapsed
200
1419 209
TEMPERATURE
precatalyst
(combustion)
603
602
Average Value 609
, 'F
post-catalyst
623
600
800
Average AT across catalyst 190*F.
133
-------
TABLE 24 (Cont.)
CATALYST TEMPERATURES
RUN 15
TIME
Clock /
1035
1135
1235
1305
Elapsed
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
TEMPERATURE
precatalyst
(combustion)
___
1076
843
401
415
454
558
541
908
. 996*
922
922
0
939
—
Average Value 748
, -F
post-catalyst
— — _.
993
728
734
638
909
935
801
734
789
948
921
915
914
893
847
851
* New thermocouple added and positioned directly below catalyst;
temperatures recorded in test log book, only.
Average AT across catalyst 100'F.
134
-------
FIGURE 49
Run.14
Gaseous Component
CO
NC
100 1511
TIHC. Mlnutf.
—I—
200
! ~
i s
§ § •
FIGURE 50
Run 14
Temperature
1 I •
135
-------
FIGURE 51
Run 15
Gaseous Component
• CO
• HC
* CO,
100
Ttrt.
FIGURE 52
Run 15
Temperature
136
-------
FIGURE 53
Opacity and Smoke Spot Density
vs
Time
(Run 14)
5
o
s_
0)
ex
u
(O
Q.
30
20
10
Transmissometer
Observer
(Ave. for period)
Smoke Spot
0
Start
10:50
25
50
75
100 125
Elapsed Time, Minutes
150
175
200
t
End
14:20
1
0
oo
o
7T
tt>
to
-o
o
01
o
D>
rt>
4 2:
ft)
O
3
n>
-------
FIGURE 54
Opacity and Smoke Spot Density
vs
Time
(Run 15)
50
40
Transmissometer
Observer
(Ave. for period)
Smoke Spot
9
8
OO
o
30
QJ
O
0)
CL
O
n)
o.
o
20
10
4
3
(SI
-o
o
GO
o
cu
3
fD
3
Ol
ft)
to
§.
1
0
8!§!
25
50 75 ' 100
Elapsed Time, Minutes
125
-------
RUN 14
TABLE 25
OPERATION LOG
TIME
1050
1055
1100
1105
1110
1115
1140
1145
FUEL WT.
db)
16.8
15.5
14.5
13.6
12.8
12.1
8.2
7.6
COMBUSTION
TEMPERATURE
CF)
761
652
600
535
526
870
787
STOVE SURFACE
TEMPERATURE
(•F)
290
310
320
500
460
500
525
1150
1213
1216
1217
1231
1248
1335
1340
6.8
5.5
738
557
530
380
2.4
613
310
COMMENTS
Begin Run.
Damper to 2 turns open
at 1057; catalyst glow
noted.
Damper to 3 turns open
at 1104.
Damper -to- 2 turns open
at 1107; catalyst glow
noted.
Changed surface temp.
reading to top of stove.
Damper to 3 turns open
at 1118; glow on 50% of
catalyst only.
Damper to 3/4 turn open
at 1142.
Damper to 3/4 turn open
at 1147.; glow on all of
catalyst.
Damper to 1/4 turn open
at 1150; no catalyst
glow.
No catalyst glow.
Open" doors to poke the
fire at 1218.
Faint catalyst glow.
Medium catalyst glow.
Faint catalyst glow.
No catalyst glow.
Open doors to poke
the fire at 1347.
139
-------
RUN 15
TABLE 25 (Cont.)
TIME
1035
1038
1040
1045
1050
1055
FUEL WT.
16.6
COMBUSTION
TEMPERATURE
(•F)
STOVE SURFACE
TEMPERATURE
14.8
14.1
13.6
13.2
1149
1076
832
843
410
430
450
410
1100 12.8 402
1105 12.3 401
1110 11.9 389
•
1200 5.9 749
1202
1206
1211 4.7 960
1220
1255
1305 1.2 915
400
400
370
390
420
430
COMMENTS
Begin Run.
Closed doors, damper
full open at 1038;
good catalyst glow.
Damper to 1 1/2 turn
open at 1041.
Damper to 1/2 turn
open at 1045.
Damper to 1/4 turn
open at 1047; faint
catalyst glow.
Doors opened at 1055
to poke the fire; no
catalyst glow.
Damper to 1 turn open
at 1059.
Doors opened at 1109
to look; no catalyst
glow.
m
Damper to 2 turns open
at 1110.
Changed M-5 filters.
Moved thermocouple to
directly below catalyst
Faint catalyst glow.
Dampers to 1 turn open
at 1213.
No catalyst glow.
No catalyst glow.
End of Burn at 1306.
140
-------
TABLE 25 (Cont.)
RUN 14 Cont.
TTMC rn-i IIT COMBUSTION STOVE SURFACE
JM FUcL WT. TEMPERATURE TEMPERATURE
U ; (-F) ('F)
1410 No catalyst glow.
1419 1.3 602 290 End of Burn at 1420.
141
-------
CATALYTIC MODIFIED COMBUSTION STOVE
No major problems were encountered with emission testing during
Runs 16 and 17. The only testing anomaly noted is that separate creosote
coupons were not used for each phase of Run 17 as per the test protocol ;
a single coupon at each measurement location was used for the entire run.
Regarding stove operation the burn rate for Run 16 was slightly
higher than called for in the test protocol, and was significantly higher
than Run 17. The burn rates for Runs 16 and 17 were 3.0 and 2.1 kg/hr,
respectively. Average combustion chamber temperatures for the two runs
were 630 and 490* F; this slight change in combustion temperature appears
to have had a significant impact on emission levels; the measured emis-
sions for test Run 16 were approximately one-half the measured value for
Run 17 (14 vs 30 g/kg wood). Excess air for the two runs averaged 422
percent. The average length of the two tests was 150 minutes.
142
-------
TABLE 26
Results — Catalytic Modified Combustion Stove
PARAMETERS RUN NUMSER
16 17
Fuel Charge {lb, Wet) ' 17.6 18.7
(Ib, Dry) 14.2 15.0
Fuel Moisture (I, Wet) 19 20
(%, Dry) 24 25
Sample Time (Minutes) 120 175
Fuel Burned (lb, Wet) 16-4 17-3
(lb, Dry) 13-3 I3-8
Burn Rate (Ib/hr, Wet) m 8-2 5-9
(l.b/nr. Dry) 6-7 *-7
Temperatures
Combustion (»F) 627 491
Surface (»F) 354 259
Stack («F) 375 230
Total. Particulate Emissions*
Concentration (g/dscm) °-57 l-l5
Rate (Ib/hr) 0-09 °-14
Factor (g/kg wood, Dry) I4 30
(g/104 Btu, Net) 10 24
Front-half (I of total) 25 19
Creosote (mg/ra2 kg) 88 317
Hydrocarbons (ppiri) 46 63
Hydrocarbons (g/kg wood. Dry) 3.9 5.8
Carbon monoxide (g/kg wood, Dry) 80 150
Opacity (observer, ") <5 13
Stack Gas Composition
C02 (5) 4.0 3.3
CO (5) 0.3 0.5
02 (S) 16.4 17.5
N2 (S) 79.3 78.7
Moisture (X) 6.5
Excess Air (5) 347 498
Stack Gas Flow (dscf/hr) • 2500 1960
Efficiency, (net, X)** 68 64
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
** Since techniques for determining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves and test runs under the specific operating
conditions of this program.
143
-------
TABLE 27
Results -- Catalytic Modified Combustion Stove
(By Test Phase)
PARAMETERS *
17/1 17/2
Fuel Charge Ob, Wet) 18.7
Ob, Dry) 15.0
Fuel Moisture (X, Wet) 20
(5. Dry) 25
Sample Time (Minutes) 50 125
Fuel Burned (Ib, Wet) 8.2 9.1
(Ib, Dry) 6.5 7.3
Burn Rate (Ib/hr, Wet) 9.8 4.4
Ob/hr, Dry) 7.8 3.5
Temperatures
Combustion (°F) 568 459
Surface (°F) 325 232
Stack (°F) 237 224
Total Participate Emissions*
Concentration (g/dscm) 1.59 0.90
Rate Ob/hr) 0.25 0.10
Factor (g/kg wood. Dry) 32 28
(g/104 Btu. Net)
Front-half (I of total) 19.0 19.3
Creosote (mg/m kg)
Hydrocarbons (ppm) 78.5 55.8
Hydrocarbons (g/kg wood, Dry) 5.6 6.2
Carbon monoxide (g/kg wood. Dry) ISO 150
Opacity (observer, S) 19 10
Stack Gas Composition
C02 (I) 4.3 2.8
CO (5) O-7 °-4
02 W 16.5 17.9
N2 (%) 78.5 78.9
Moisture (X) 10.0 4.6
Excess Air (5) 354 578
Stack Gas Flow (dscf/hr) 2500 1750
Efficiency, (net, %)
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
144
-------
TABLE 28
Bacharach Smoke Spot Results
RUN
16
1015-1215
17
1416-1711
SAMPLE TIME
1029 - 1031
1114 - 1118
1210
Ave. Temperature
1445 _ 1448
1540 - 1544
1634 - 1637
Ave. Temperatur
SPOT DENSITY
Sample Location
Sampl ing
5
5
5
320eF
6
6
6
251°F
Outlet
5
5
-
233°F
6
5
6
157°p
145
-------
TABLE 29
Visible Emissions Observation Log
RUN
16
17
RUN TIME
Clock
• Phase 1
• Phase 2
1015 - 1215
1416 - 1506
1506 - 1711
Elasped Min.
• Phase 1
• Phase 2
120
50
125
175
VISIBLE EMISSIONS MEASUREMENT
Time Period
1019 - 1029
1200 - 1209
1418 - 1427
1540 - 1549
1659 - 1708
Minutes
Observed
10
10
10
10
10
Opacity
(Percent)
<5
<5
19
11
10
Opacity
Range
(Percent)
<5 - 10
0 - <5
15 - 25
10 - 15
10 - 15
Average Opacity
(Percent)
Phases 1,2, total
<5
19
11
13
-------
RUN 16
TABLE 30
CATALYST TEMPERATURES
TIME
Clock / Elapsed
1015 1.5
4.5
10
20
30
40
50
1115 60
70
85
95
105
110
1215 120
Average Value
TEMPERATURE, 'F
precatalyst post-catalyst
Tertiary Combustion
Air Chamber
547
911
644
568
522
447
450
447
421
397
665
408
—
—
535
730
840
686
642
604
582
647
610
599
607
670
669
688
597
655
1046
1121
782
726
632
579
697
646
629
604
598
624
602
559
703
Average AT across, catalyst is 50"F.
147
-------
RUN 17
TABLE 30 (Cont.)
CATALYST TEMPERATURES
TIME
Clock / Elapsed
1416 0
13
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
175
TEMPERATURE, " F
precatalyst post-catalyst
Tertiary Combustion
Air Chamber
470
472
442
451
414
387
—
365
345
—
295
279
261
—
358
—
323
—
Average Value 374
—
583
556
611
550
520
507
502
476
416
396
388
375
461
556
517
457
423
488
804
550
560
530
499
479
457
446
458
396
395
373
358
468
525
489
406
370
476
Average AT across catalyst is O'F.
148
-------
-£_
FIGURE 55
Run 16
Gaseous Component
• co
HC
100
Tine.
I -5
3
S S
s s
FIGURE 56
Run 16
Temperature
149
-------
FIGURE 57
Run 17
Gaseous Component
• to
'«
'to,
ion
TIW. mmit
1 i
FIGURE 58
Run 17
Temperature
150
-------
FIGURE 59
Opacity and Smoke Spot Density
vs
Time
(Run 16)
30
CO
o
r+
CO
ro
3
ro
tn
.j.
O
' o ">
1
151
-------
FIGURE 60
Opacity and Smoke Spot Density
vs
Time
(Run 17)
40
g 30
o
OJ
ex
o
2. 20
0
10
At
14
— * — Transmissometer
Observer
(Ave. for period)
• Smoke Spot
• •
•
X Ax ~ ~
0 25 50 75 100 125 150 175
art End
15 Ha need Ttmo Mi'mii-oc 17. m
' 9
. 8
. 7 t/>
o
7T
n>
• K ^
b -o
0
• 5 2
n>
\»
A Q.
limension'
00
CD
• 2 £
• 1
- 0
-------
RUN 16
TABLE 31
OPERATION LOG
TIME
1015
1020
1026
1035
1214
FUEL WT.
17.6
14.4
COMBUSTION
TEMPERATURE
CF)
740
STOVE SURFACE
TEMPERATURE
(•F)
440
1.2
597
340
COMMENTS
Begin Run, damper full
open; 3rd level air
full open.
Closed damper to 3/8"
open at 1021.
Closed damper to 1/4"
open.
Damper automatically
opened to 1/2".
End of Burn.
153
-------
TABLE 31 (Cont.)
RUN 17
run tiT COMBUSTION STOVE SURFACE
MM TEMPERATURE TEMPERATURE
1 ' CF) f7?)
1416 18.7 Begin Run; damper full
open; 3rd level air
full open.
1421 17.0 569 390 Closed damper to 1/4"
at 1418.
1427 Closed damper to 3/8"
open.
1506 10.5 520 275 End of Phase I.
1514 Open damper to 3/4".
1552 Open damper to 1".
1616 4.7 375 200 Open damper at 1612.
1625 Rearrange wood to
single pile.
1633 Closed damper to 1".
1650 Open damper to full
open.
1711 1.4 423 220 End of Burn.
154
-------
CERAMIC STOVE
Due to the design of this stove, the fuel burn rate during Runs
18 and 19 were significantly higher than the burn rate called for in
the test protocpl; 6.5 and 3.9 kg/hr, respectively. This high burn
rate characterized by a high combustion chamber temperature (1200 -
1500* F), is no doubt a factor in the low emissions rate measured
during these runs. The fact that this stove does operate at such a
high burn rate makes it difficult to compare the emissions results
to the results of the other stoves tested at much lower burn rates.
Excess air for these two tests averaged 80 percent. The average
length of the burn was only 80 minutes, due to the high burn rate.
Several testing anomalies occurred during these runs. First,
no Bacharach smoke spot tests were conducted. Secondly, although
visual emission readings were taken for these runs, the observer was
not Method 9 certified. This is not expected to have any impact on
the results since no visible emissions were noted during the test
runs; consequently, no judgement of opacity was required. Finally,
at the beginning of Phase 2 for Run 19, the thermocouple monitoring
the stack temperature malfunctioned. The obviously erroneous tempera-
ture readings taken prior to fixing the thermocouple were deleted
from the test calculations.
Note the impact on C0/C0? emissions and combustion chamber
temperature in Run 19 when the fire was stoked (045 elapsed time).
155
-------
TABLE 32
Results — Ceramic Stove
PARAMETERS
Fuel Charge (Ib, Wet)
Ob, Dry)
Fuel Moisture (Z, Wet)
("-, Dry)
Sample Time (Minutes)
Fuel Burned (Ib, Wet)
(Ib, Dry)
Burn Rate (Ib/hr, Wet)
(Ib/hr, Dry)
Temperatures
Combustion (8 F)
Surface («F)
Stack (°F)
Total Particulate Emissions*
Concentration (g/dscm)
Rate (Ib/hr)
Factor (g/kg wood, Dry)
(g/104 Btu, Net)
Front-half (' of total)
Creosote (rag/in kg)
Hydrocarbons (ppm)
Hydrocarbons (g/kg wood. Dry)
Carbon monoxide (g/kg wood. Dry)
Opacity (observer, 5)
Stack Gas Composition
C02 (5)
CO (%)
02 U)
N2 (S)
Moisture (I)
Excess Air (I)
Stack Gas Flow (dscf/hr)
Efficiency, (net, 5)**
RUN NUMBER
18 19
19.1
16.0
17.4
21
61
17.8
14.4
17.5
14.2
1519
--
779
0.16
0.02
1
0.7
63.1
56
15
0.4
20
-0-
12.6
0.2
8.2
79.0
15.0
63
1850
68
19.5
15.6
20
25
99
18.0
14.4
10.9
8.7
1195
207
604
0.13
0.01
2
1.5
48.6
27
15
0.6
50
-0-
8.1
0.4
10.7
80.9
--
97
1720
65
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
** Since techniques for determining appliance efficiency vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves and test runs under the specific operating
conditions of this program.
156
-------
TABLE 33
Results -- Ceramic Stove
(By Test Phase)
PARAMETERS RUN NUMB£R
19/1 19/2
Fuel Charge (lb. Wet) 19-5
(lb. Dry) I5-6
Fuel Moisture (5, Wet) 20
(t. Dry) 25
Sample Time (Minutes) 25 74
Fuel Burned (lb, Wet) 8-3 9'7
Ob, Dry) 6-6 7'8
Burn Rate (Ib/hr, Wet) 19-9 7'9
(Ib/hr. Dry) 15-8 5-3
Temperatures
Combustion («F) 126° 1171
Surface («F) 2°° 209
Stack (.F) 634 48S
Total Particulate Emissions*
Concentration (g/dscm) °-36 °-12
Rate (Ib/hr) % 0.05 0.01
Factor (g/kgwood. Dry) i-5 1>6
(g/104 Btu, Net)
Front-half (I of total) 53-5 42'9
Creosote (mg/m2 kg) 51 8
Hydrocarbons (ppm) 26
Hydrocarbons (g/kg wood. Dry) °-9 °-5
Carbon monoxide (s/kg wood. Dry) 70 50
Opacity (observer, 1.) . ~°~ ~°~
Stack Gas Composition
CO, (S) . 10-4 7'3
coo °-6 °'3
0 (.) 10.5 10.8
N2 0 ' 78.5 81.7
Moisture (5) "-1 7-4
Excess Air (I) 96J 97.2
Stack Gas Flow (dscf/hr) 2*0° 139°
Efficiency, (net, S)
* Oregon DEQ Method 7 (EPA Method 5 with back-half)
157
-------
TABLE 34
Bacharach Smoke Spot Results
RUN
18
SAMPLE TIME
No Data
SPOT DENSITY
Sample Location
Sampling
Outlet
19
No Data
158
-------
TABLE 35
Visible Emissions Observation Log
RUN
18
19
RUN TIME
Clock
• Phase 1
• Phase 2
1044 - 1145
1405 - 1544
Elasped Min.
• Phase 1
• Phase 2
61
99
VISIBLE EMISSIONS MEASUREMENT
Time Period
1032 - 1041
1132 - 1141
1142 - 1151
•
Minutes
Observed
10
10
10
Opacity
( Percent)
-0-
-0-
-0-
Opacity
Range
(Percent)
-0-
-0-
-0-
Average Opacity
( Percent)
Phases 1,2, total
-0-
-0-
159
-------
FIGURE 61
Run 18
Gaseous Component
100 ISO
TIKI, (limit**
CO
»c
CO,
FIGURE 62
Run 18
Temperature
160
-------
-f.-
FIGURE 63
Run 19
Gaseous Comoonent
inn
TIMC .
KC
CO,
I §
FIGURE 64
Run 19
Temperature
161
-------
•u
o
Q.
I/I
O
E
Smoke Spot Scale, dimensionless
O1
oo
ro c\j
D;
to
c
a;
O
Q.
OO
oi QJ
^ & E
i»i=
00
•o
c
ro
Q.
O
CO
O
LT>
-------
FIGURE 66
Opacity and Smoke Spot Density
vs
Time
(Run 19)
20
10
Transmissometer
Observer
(Ave. for period)
0 25 50 75 100
Start End
14:05 15:45
163
-------
TABLE 36
OPERATION LOG
RUN 18
TIME
1044
1045
1047
1055
1144
FUEL WT.
19.1
COMBUSTION
TEMPERATURE
(•F)
STOVE SURFACE
TEMPERATURE
(•F)
1.3
1297
COMMENTS
Begin Run; stack
damper full open;
door ajar.
Latch door.
Fully close damper
Stove surface warm
to touch.
End of Burn.
164
-------
TABLE 36 (Cont.)
RUN 19
TIME
1405
1407
1430
1444
1544
FUEL WT.
19.5
COMBUSTION
TEMPERATURE
(•F)
STOVE SURFACE
TEMPERATURE
(•F)
196
1.5
1170
220
COMMENTS
Begin Run.
Close stack damper.
End Phase I.
Rearrange logs.
End of Burn.
165
-------
RESIDENTIAL WOOD COMBUSTION STUDY
TASK 5
EMISSIONS TESTING OF WOOD STOVES
Volume 2 of 4
-------
THIS REPORT CONSISTS OF SEVERAL DIFFERENT PARTS.
THEY ARE LISTED BELOW FOR YOUR CONVENIENCE.
EPA 910/9-82-089a Residential Wood Combustion Study
Task 1 - Ambient Air Quality Impact
Analysis
EPA 910/9-82-089b Task 1 - Appendices
EPA 910/9-82-089c Task 2A - Current & Projected Air Quality
Impacts
EPA 910/9-82-089d Task 2B - Household Information Survey
EPA 910/9-82-089e Task 3 - Wood Fuel Use Projection
EPA 910/9-82-089f Task 4 - Technical Analysis of Wood Stoves
EPA 910/9-82-089g Task 5 - Emissions Testing of Wood Stoves
Volumes 1 & 2
EPA 910/9-82-089h Task 5 - Emissions Testing of Wood Stoves
Volumes 3 & 4 (Appendices)
EPA 910/9-82-089i Task 6 - Control Strategy Analysis
EPA 910/9-82-089J Task 7 - Indoor Air Quality
-------
DISCLAIMER
This report has been reviewed by Region 10, U. S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policies of the U. S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
TABLE OF CONTENTS (Cont.)
page
VOLUME 2
Discussion of Results 166
General 166
Comparison of Particulate Results Among Appliances 167
Stove Efficiency 177
Particulate Emissions - First Phase Tests 179
Effect of Fuel Moisture Content 184
Simplified Test Procedures 229
Reasonable Emission Standard 241
Quality Assurance 264-
Quality Assurance Records 264
Wood Moisture Determination 265
Particulate Sampling 268
Gaseous Measurements 269
Transmissometer 271
Data Reduction 273
References/Bibliography 274
VOLUME 3
Appendix A - Nomenclature and Sample Calculations
Appendix B - Laboratory Data
Appendix C - Quality Assurance Data
Appendix D - Cost Estimates for Simplified Test Procedures
VOLUME 4
Appendix E - Field Data Runs 1-19
-------
LIST OF TABLES
page
Table 1 Emission Summary ' 6
Table 2 Operating Parameters Summary 7
Table 3 Results — Airtight Box Stove 62
Table 4 Results -- Airtight Box Stove (Cold Start Tests) 63
Table 5 Results — Airtight Box Stove (By Test Phase) 64
Table 6 Bacharach Smoke Spot Results -- 65
Airtight Box Stove
Table 7 Visible Emissions Observation Log -- 67
Airtight Box Stove
Table 8 Operation Test Log -- Airtight Box Stove 87
Table 9 Results — Catalytic Retrofit 100
Table 10 Results — Catalytic Retrofit (By Test Phase) 101
Table 11 Bacharach Smoke Spot Results -- 102
Catalytic Retrofit
Table 12 Visible Emissions Observation Log -- 103
Catalytic Retrofit
Table 13 Catalytic Temperatures -- Catalytic Retrofit 104
Table 14 Operation Test Log -- Catalytic Retrofit 110
Table 15 Results — Non-Catalytic Retrofit 115
Table 16 Results — Non-Catalytic Retrofit (By Test Phase) 116
Table 17 Bacharach Smoke Spot Results -- 117
Non-Catalytic Retrofit
Table 18 Visible Emissions Observation Log -- 118
Non-Catalytic Retrofit
Table 19 Operation Test Log -- Non-Catalytic Retrofit 123
Table 20 Results — Catalytic Box Stove 128
Table 21 Results -- Catalytic Box Stove (By Test Phase) 129
Table 22 Bacharach Smoke Spot Results -- 130
Catalytic Box Stove
Table 23 Visible Emissions Observation Log -- 131
Catalytic Box Stove
Table 24 Catalytic Temperatures -- Catalytic Box Stove 132
Table 25 Operation Test Log -- Catalytic Box Stove 139
-------
LIST OF TABLES (Cont.)
page
Table 26 Results -- Catalytic Modified Combustion Stove 143
Table 27 Results — Catalytic Modified Combustion Stove 144
Table 28 Bacharach Smoke Spot Results -- 145
Catalytic Modified Combustion Stove
Table 29 Visible Emissions Observation Log -- 146
Catalytic Modified Combustion Stove
Table 30 Catalytic Temperature -- 147
Catalytic Modified Combustion Stove
Table 31 Operation Test Log -- Catalytic Modified 153
Combustion Stove
Table 32 Results -- Ceramic Stove 156
Table 33 Results -- Ceramic Stove (By Test Phase) 157
Table 34 Bacharach Smoke Spot Results -- Ceramic Stove 158
Table 35 Visible Emissions Observation Log -- Ceramic 159
Table 36 Operation Test Log -- Ceramic Stove 164
Table 37 Comparison of Operating Parameters and Emissions 170
Table 38 Fuel Load and Combustion Rate Summary 176
Table 39 Results: Stove Efficiency 178
Table 40 Summary of Results: Phase 1 Tests 181
Table 41 Results: Cold Start vs Hot Start 183
Table 42 Summary of Results: Fuel Moisture Tests 189
Table 43 Normalized Particulate Results: 191
Fuel Moisture Tests
Table 44 Creosote Results 201
Table 45 Creosote Deposition: Theoretical Percent of 211
Total Emissions
Table 46 Summary of Stack Gas Opacity 225
(Transmissometer and Visual Observer)
Table 47 Summary of Average Emission Factors for 230
Carbon monoxide, Gaseous Hydrocarbons, and
Creosote (Literature Review)
Table 48 Simplified Test Procedures Summarized 234
-------
LIST OF TABLES (Cont.)
page
Table 49 Emissions Standard Range 242
Table 50 Summary of Average Emission Rates 246
(Literature Review)
Table 51 Particulate Emission Data Summary 248
(Literature Review)
Table 52 Gaseous Calibration Gases 270
-------
LIST OF FIGURES
page
Figure 1 Particulate Emissions as a Function of 13
Fuel Moisture
Figure 2 Particulate Sampling Apparatus 33
Figure 3 Sample Location 42
Figure 4 Airtight Box Stove 45
Figure 5 Catalytic Box Stove 48
Figure 6 Catalytic Modified Combustion Stove 51
Figure 7 Ceramic Stove 53
Figure 8 Catalytic Retrofit Device 55
Figure 9 Non-Catalytic Retrofit Device 57
Figure 10 Gaseous Component (CO, CCL, HC) vs Time (Run 1) 69
Figure 11 Temperature (Combustion, Stack Gas) vs Time (Run 1) 69
Figure 12 Gaseous Component vs Time (Run 2) 70
Figure 13 Temperature vs Time (Run 2) 70
Figure 14 Gaseous Component vs Time (Run 3) 71
Figure 15 Temperature vs Time (Run 3) 71
Figure 16 Gaseous Component vs Time (Run 4) 72
Figure 17 Temperature vs Time (Run 4) 72
Figure 18 Gaseous Component vs Time (Run 5) 73
Figure 19 Temperature vs Time (Run 5) 73
Figure 20 Gaseous Component vs Time (Run 6) . 74
Figure 21 Temperature vs Time (Run 6) 74
Figure 22 Gaseous Component vs Time (Run 7) 75
Figure 23 Temperature vs Time (Run 7) 75
Figure 24 Gaseous Component vs Time (Run 8) 76
Figure 25 Temperature vs Time (Run 8) 76
Figure 26 Gaseous Component vs Time (Run 9) 77
Figure 27 Temperature vs Time (Run 9) 77
Figure 28 Opacity and Smoke Spot Density vs Time (Run 1) 78
Figure 29 Opacity and Smoke Spot Density vs Time (Run 2) 79
-------
LIST OF FIGURES (Cont.)
page
Figure 30 Opacity and Smoke Spot Density vs Time (Run 3) 80
Figure 31 Opacity and Smoke Spot Density vs Time (Run 4) 81
Figure 32 Opacity and Smoke Spot Density vs Time (Run 5) 82
Figure 33 Opacity and Smoke Spot Density vs Time (Run 6) 83
Figure 34 Opacity and Smoke Spot Density vs Time (Run 7) 84
Figure 35 Opacity and Smoke Spot Density vs Time (Run 8) 85
Figure 36 Opacity and Smoke Spot Density vs Time (Run 9) 86
Figure 37 Gaseous Components vs Time (Run 10) 106
Figure 38 Temperature vs Time (Run 10) 106
Figure 39 Gaseous Components vs Time (Run 11) 107
Figure 40 Temperature vs Time (Run 11) 107
Figure 41 Opacity and Smoke Spot Density vs Time (Run 10) • 108
Figure 42 Opacity and Smoke Spot Density vs Time (Run 11) 109
Figure 43 Gaseous Components vs Time (Run 12) 119
Figure 44 Temperature vs Time (Run 12) 119
Figure 45 Gaseous Components vs Time (Run 13) 120
Figure 46 Temperature vs Time (Run 13) 120
Figure 47 Opacity and Smoke Spot Density vs Time (Run 12) 121
Figure 48 Opacity and Smoke Spot Density vs Time (Run 13) 122
Figure 49 Gaseous Component vs Time (Run 14) 135
Figure 50 Temperature vs Time (Run 14) 135
Figure 51 Gaseous Component vs Time (Run 15) 136
Figure 52 Temperature vs Time (Run 15) 136
Figure 53 Opacity and Smoke Spot Density vs Time (Run 14) 137
Figure 54 Opacity and Smoke Spot Density vs Time (Run 15) 138
Figure 55 Gaseous Component vs Time (Run 16) 149
Figure 56 Temperature vs Time (Run 16) 149
Figure 57 Gaseous Component vs Time (Run 17) 150
Figure 58 Temperature vs Time (Run 17) 150
-------
LIST OF FIGURES (Cont.)
page
Figure 59 Opacity and Smoke Spot Density vs Time (Run 16) 151
Figure 60 Opacity and Smoke Spot Density vs Time (Run 17) 152
Figure 61 Gaseous Components vs Time (Run 18) 160
Figure 62 Temperature vs Time (Run 18) 160
Figure 63 Gaseous Components vs Time (Run 19) 161
Figure 64 Temperature vs Time (Run 19) 161
Figure 65 Opacity and Smoke Spot Density vs Time (Run 18) 162
Figure 66 Opacity and Smoke Spot Density vs Time (Run 19) 153
Figure 67 Summary of Particulate Emissions Results 168
Figure 68 Total Particulate Emissions vs Burn Rate 172
Figure 69 Total Particulate Emissions vs 173
Fuel Load-Combustion Rate Ratio
Figure 70 Filterable Particulate Emissions vs 174
Fuel Load-Combustion Rate Ratio
Figure 71 Particulate Emissions Results: First Phase Tests 180
Figure 72 Particulate Emissions Results: 185
Fuel Moisture Tests
Figure 73 Particulate Emissions as a Function of 188
Fuel Moisture
Figure 74 Particulate Emissions (Normalized for Burn Rate) 190
as a Function of Fuel Moisture
Figure 75 The Dependency of Appliance Efficiencies on 194
Fuel Moisture Content (From Sheltonlc>°)
Figure 76 Creosote Accumulation as a Function of Moisture 195
Content Using Pinon as Fuel (From SheltonS)
Figure 77 Creosote Accumulation as a Function of Moisture 196
Content Using Oak as Fuel (From Shelton^)
Figure 78 Creosote: Transmissometer vs Sample Location 203
(By Test)
Figure 79 Creosote: Transmissometer vs Sample Location 204
(By Test Phase)
Figure 80 Creosote (Transmissometer Location) vs 206
Total Particulate Emissions
-------
LIST OF FIGURE (Cont.)
page
Figure 81 Creosote (Sample Location) vs Total 207
Particulate Emissions
Figure 82 Creosote (Average) vs Total Particulate Emissions ^08
Figure 83 Creosote (Transmissometer Location) vs 209
Filterable Particulate
Figure 84 Particulate Concentration vs Carbon monoxide 213
Concentration
Figure 85 Particulate Emissions vs Adjusted Carbon 214
monoxide Concentration
Figure 86 Particulate Emissios vs Carbon monoxide 216
Concentrations
Figure 87 Particulate Concentration vs Gaseous 218
Hydrocarbon Concentration
Figure 88 Particulate Emissions vs Adjusted Gaseous 219
Hydrocarbon Concentration
Figure 89 Particulate Emissions vs Gaseous 220
Hydrocarbon Emissions
Figure 90 Particulate Concentration vs Opacity 223
Figure 91 Particulate Emissions vs Opacity 224
Figure 92 Particulate Concentration vs Smoke Spot Density 227
Figure 93 Particulate Emissions vs Smoke Spot Density 228
Figure 94 Wood Moisture Measurement 266
-------
DISCUSSION OF RESULTS
GENERAL
In this section, the emission test results are discussed and
evaluated with respect to the objectives of the study. The results
of the moisture tests, simplified test procedures, and appliance
performance are evaluated.
The tests were conducted at an average burn rate of 2.5 kg/hour.
During the tests the stove surface temperatures averaged in the
350 - 400°F range, the combustion chamber temperature averaged 500 -
600°F, and the stack temperature averaged 250 - 300°F.
The average total emission rate for these tests (excluding fuel
moisture and cold start emissions) was 19 g/kg. Front-half emissions
averaged 21% of the total. Carbon monoxide and hydrocarbon emissions
averaged 115 and 8.7 g/kg, respectively. Creosote deposition averaged
2
330 mg/m kg. Average opacities measured by the visible emission observer
»
and transmtssometer were 21 and 18 percent, respectively.
Numerous'figures correlating -various emission factors are presented
in this section. Several notations are included with the data points
plotted on these figures and have the following meanings, unless other-
wise noted:
L Low moisture fuel test
LC Low moisture fuel test, cold start
H High moisture fuel test
HC High moisture fuel test, cold start
MC Medium moisture, fuel test, cold start
"Star" Not included in correlation
166
-------
COMPARISON OF PARTICULATE RESULTS AMONG APPLIANCES
Figure 67 graphically presents the participate emission test results.
The test runs conducted for fuel moisture evaluation and evaluation of
cold start emissions are not included in the graph. Examination of the
results indicates that, with the exception of the ceramic stove, no
significant reduction in emissions occurred with the improved technology
stoves for the operating conditions under which the stoves were tested.
Although the ceramic stove did have significantly less emissions, the
burn rate for this appliance was significantlly higher. The dependency
of emissions level on burn rate will be discussed later in this section.
The average emissions for the different stoves are as follows:
•
Total, Front-half Burn Rate
g/kg kg/hr
Airtight Box 21 (5.3) 2.4
Catalytic Retrofit 20 (3.5) 2.4
Non-Catalytic Retrofit 37 (6.8) 2.3
Catalytic Box 31 (5.4) 2.0
Catalytic Modified Combustion 22 (4.6) 2.5
Ceramic 1.5 (0.8) 5.2
Note that even with assuming an equivalent heat output (Btu/hr), although
the ceramic stove would consume twice as much fuel as the other appliances
in the same time period, the emissions would be 85% less (8 g/hr vs
approximately 48 g/kg). Of course, it is reasonable to expect that at
double the burn rate, there would be a higher heat output. The problem
is that it is not really possible to determine what percent of emissions
reduction can be attributed to the ceramic stove per se, and what percent
is simply due to the higher burn rate. Nonetheless, it must be noted that
167
-------
OJ
+->
ro
ai
O Z3
•r- O
r- t.
3 H3
O Q-
S_ i—
re IQ
o- 3C
en
OJ
ai 4->
-t-1 re
o s-
(O C O C
+j o a> i-
O S- S- 3
i — u. o ca
E
x
O)
f-.
VJ3
O>
(TJ
a.
O
s_
CM
CM
CM ,
CM
o
CO
CM
en
CM \
tu/6m
6>)/6
008
009
oe
SUOISSIUJT
OOt?
oz
003
01
-------
the high burn rate is a characteristic of this stove; consequently, one
would expect it to operate with this low emission level under normal
operating conditions.
Note that neither the catalytic or non-catalytic retrofit
reduced emissions at the conditions under which tests were run. The
duplicate test runs (for each appliance) were conducted at close to the
same conditions and the emission results compared well for the paired
samples. The non-catalytic retrofit actually yielded higher results
than the airtight box stove; there is no explanation for this result.
The tests with the catalytic add-on device yielded results nearly
identical to the airtight stove alone.
The measured creosote deposition followed the same trend as the
particulate emissions with one major exception. The measured creosote
emissions for the airtight box stove (Runs 1 and 7) were significantly
higher than for the other test runs of similar emission rate. The
creosote results particularly in relation to measured emission rates,
are discussed in more detail in a late'r section of this report.
The second major point which should be noted from the results
relates to the significant variation in emissions between the paired
runs for both the catalytic and catalytic modified combustion stoves.
These data warrant further discussion because they illustrate the
large impact operating conditions can have on appliance emissions.
Table 37 summarizes the operating conditions for the two stoves during
the four test runs. Note that in both cases an approximate 40%
increase in burn rate results in a 40 - 50% decrease in emissions.
169
-------
TABLE 37
Comparison of Operating Parameter and Emissions:
Airtight Box, Catalytic, and
Catalytic Modified Combustion Appliances
STOVE
Airtight
Catalytic
Catalytic Modified
Combustion
RUN
1
7
14
15
17
16
BURN
RATE
1.9
2.9
1.7
2.2
2.1
3.0
TEMPERATURE (AVERAGE ° F)
Combustion
510
610
600
820
• 490
630
Surface
340
510
390
412
260
350
Post Catalyst
(ave) (max)
_
-
800
850
480
700
.
-
1130
1100
800
1120
Catalyst
AT
-
200
100
-0-
50
EMISSIONS
Part.
g/kg
22
19
38
23
30
14
CO
gAg
190
160
120
50
150
80
170
-------
Particularly for the catalytic modified combustion appliance it is
clearly seen that the slight change in operating conditions as measured
by the combustion and surface temperatures results in this significant
decrease in particulate emissions. This same trend occurred for carbon
monoxide emissions (see Table 37). Note that for Run 17, the stove was
clearly operating at a temperature (490° F) insufficient for proper
catalyst operation (manufacturer indicates minimum temperature of 600°F
required at point measured). The importance of operating conditions
cannot be overstressed.
Figure 68 graphically presents a correlation between burn rate and
particulate emissions. The correlation does not include the data
obtained during the cold start or fuel moisture tests, although the'se
data points are indicated in the figure. The correlation indicates a
strong exponential increase in emissions for a decreasing burn rate.
This type relationship is similar to that previously reported in the
literatureld, although this correlation indicates a greater impact than
«
previously reported. The relationship determined from 13 test runs
conducted for the Oregon DEQ also is plotted on this figure, for
comparison.
Figures 69 and 70 graphically present the relationship between
burn rate and emissions in a different manner. Figure 69 is for total
particulate emissions and Figure 70 for front-half emissions only.
An additional factor, the mass of fuel charged, is taken into account
in these correlations. This type correlation is extremely useful
because it essentially allows one to indirectly normalize -the burn rate
171
-------
FIGURE 68
Total Participate Emissions
vs
Burn Rate
100
80 .
.51
01
c
o
"5 50
EE
LjJ
O)
• I
03
| 4°-
S_
0.
're
o
20.
i
I o 04 • Low (L) and Hig
y = 156e ' moisture tested
1 r = -.93 '^Cold Start Test
i
1
t
1
I
\
\
\
t • L
\
\
•
**»A ^
>•-..
H^ ***"*•••
H" \® ""••-...
*. Oregon DEQ
•^ Correlation
@ Sr
4 6
Burn Rate, kg/hr
10
172
-------
62
-------
O-
4-
10.
FIGURE 70
Front Half Particulate
vs
Fuel Load/Combustion Rate
to
d
o
in
to
15
Low (L), High (H)
Moisture Test
Cold Start Tests
(Not included in
correlation)
o
s..
5.
1
Fuel Load Per Average Combustion Rate, kg/ Btu/hr X 10
174
-------
for stove size. For example, it is not necessarily reasonable to
3 3
compare a stove with a 6 ft volume and a stove with a 3 ft volume at
the same burn rate (2.0 kg/hr); since the larger stove holds twice as
much fuel, it is reasonable to expect this stove to be rated for a higher
heat output rate "and therefore, to be operated at a higher burn rate;
4.0 kg/hr, for example. Given a combustion rate of 4.0 kg/hr for the
larger stove and assuming twice the fuel load for twice the volume,
then the "fuel load per combustion rate" factor (M/Q) for the two
stoves would be equivalent. This assumes that some constant factor for
fuel load to stove volume (M/V) also is used. Table 38 presents
a summary of the fuel load combustion data.
The relationship presented here and in Figure 68 (emissions
increasing exponentially with burn rate) are somewhat contradictory.
If one assumes a constant fuel load then the correlation in Figures 69
and 70 would indicate a linear increase in emissions with burn rate.
The correlation factor for the exponential relationship is significantly
better (.93 vs .70). Nonetheless, it seems logical that when character-
izing stove operation, both the fuel load and burn rate must be taken
into account. Consequently, the fuel/load combustion rate correlation
is really of more practial use. This parameter used in conjunction
with a constant fuel mass/stove volume ratio could be used to help
establish desired stove operating parameters (i.e., burn rate) for tests
conducted to compare emissions results.
The correlation obtained during Oregon DEQ's tests of thirteen
stoves is shown in Figure 69 for comparison. The DEQ tests were conducted
in the same basic fuel load/combustion rate range (1.5 - 3.5 kg/Btu/hr x 10" ).
For this test series, emission levels slightly higher than those obtained
by DEQ were measured at the corresponding fuel load/combustion rate.
175
-------
TABLE 38
Fuel Load and Combustion Rate Summary
RUN
1
7
10
11
12
13
14
15
16
17
18
19
2
3
5
9
4
6
8
FUEL LOAD
(Wet)
M
kg
11.9
12.9
14.0
12.6
14.0
13.3
7.6
7.5
8.0
8.5
8.7
8.9
12.5
15.7
11.8
12.5
13.5
12.4
17.3
HEAT
RELEASE RATIO
Q
10^ Btu/hr
3.8
5.8
4.2
5.2
4.8
4.4
3.4
4.4
6.0
4.2
12.7
7.8
4.8
4.2
5.4
3.4
16.1
11.5
4.0
"UEL COMBUSTIOf
RATIO
M/Q
3.1
2.2
3.3
2.5
2.9
3.0
2.3
1.7
1.3
2.0
0.7
1.1
2.6
3.8
2.2
3.7
0.8
1.1
4.4
STOVE
VOLUME
V
m3
0.17
0.17
0.17
0..17
0.17
0.17
0.07
0.07
0.08
0.08
0.09
0.09
FUEL VOLUME
RATIO
M/V
kg/mj
70
76
82
74
82
76
108
107
100
106
97
99
74
92
69
74
79
73
102
FUEL LOAD
(Dry)
M1
kg
9.4
10.3
11.4
10.2
11.3
10.8
6.3
6.0
6.5
6.8
7.3
7.1
11.0
6.9
10.4
5.5
11.9
10.0
7.7
FUEL COMBUSTION
RATE RATIO(Dry)
M'/Q
2.5
1.8
2.7
2.0
2.4
2.5
1.9
1.4
1.1
1.6
0.6
0.9
2.3
1.6
1.9
1.6
0.7
0.9
1.9
FUEL VOLUME
(Dry)
M'/V
55
61
67
60
66
64
90
86
81
85
81
79
-------
STOVE EFFICIENCY
Stove efficiencies (net useable heat output) did not vary sig-
*
nificantly from test run to test run with the exception of the high
fuel moisture tests. Table 39 summarizes the results. The average
stove efficiency was 66 percent. For the high moisture fuel tests the
efficiency dropped significantly to 45 percent, primarily due to the
additional heat loss from the latent heat content of the water.
The ceramic stove efficiency was determined to be 67 percent, even
though stack gas temperatures for this stove were particularly high
(650CF), resultin in excessive stack losses. However, the high combus-
tion efficiency of this stove compensated for the large stack losses;
a heat exchange system on the stack of this unit could increase the net
efficiency.
Since techniques for determining appliance efficiencies vary
tremendously, it is recommended that the results reported here be used
solely for comparing results among the stoves tested in this program.
177
-------
TABLE 39
Results:
Stove Efficiency
STOVE
Airtight Box
Airtight Box
(Low Moisture Fuel)
Airtight Box
(High Moisture Fuel)
Airtight Box
(Non-Catalytic Retrofit)
Airtight Box
(Catalytic Retrofit)
Catalytic Box
Catalytic Modified
Combustion
Ceramic
RUN
1
7
AVE
2
5
AVE
3
9
AVE
10
11
AVE
12
13
AVE
14
15
AVE
16
17
AVE
18
19
AVE
NET EFFICIENCY (%Jf
65
66
65
60
62
61
43
48
45
68
71
69
60
66
63
68
71
69
68
64
66
68
65
66
* Since techniques for determining appliance efficiencies vary tremendously,
the reported efficiency results should be used only as relative values for
comparison between stoves tested under the specific operating conditions
of this program.
178
-------
PARTICULATE EMISSIONS — FIRST PHASE TESTS
As previously mentioned for some of the test runs, a separate
particulate sample was taken during the first part fphase) of the
test burn. A total of 11 first phase samples were taken with 3 of
these samples being taken from a test burn started from a cold stove
condition (i.e., no hot coal bed). Figure 71 graphically presents the
first phase test results; the results are summarized in Table 40.
One reason for taking the first phase samples was to determine if the
emissions from this phase of the test burn were significantly higher
than during the rest of the burn. From Table 40 (last column) it is
seen that in terms of g/kg, total particulate emissions for Phase 1
were generally the same or higher than emissions from Phase 2 (Run 1
is the exception). Similarly, the measured concentration for Phase 1
was more often higher than the Phase 2 concentration. However, the
data can be examined in a different manner; columns 2 and 4 list the
percent of the total fuel consumed and total mass emitted during the
first phase of the test. Examination of these data indicate that in
general, the percentages compare favorably. In other words, the percent
of the total emissions during Phase 1 with the percent of fuel consumed
is consistent. For example, in Run 10, 37 percent of the fuel was
consumed during the first phase and 39 percent of the total mass emitted
was emitted during the first phase. Nonetheless, on an absolute basis
a significant portion of the total mass emitted was emitted during the
first phase of the test (usually first hour of the test run). On the
average, 46 percent of the total mass emitted during the entire run was
emitted during the first phase of the test (23% of the total run time).
179
-------
FIGURE 71
Participate Emissions:
:::X;x Total Participate
f!fl Front Half Participate
Creosote
en
CXJ
en E
en E
° 0
00
s
° 00
CO ^>
s
// -P
5 / 5'
'*&//
**•'•//
&y
i
\
ru:». ruac.^ .v-^o. x x Burn Kate f Kg/ nr
* Missing Value
3.6
5.8 4.7
8 I g ::•:?
*"* •* *."•
x :'• x|
8 >:< ?|:
: ''•• •'• :::'
• >j: x
v>
*.*.*
•:•:•
2.,0 3,4 |j:
: x::-i :|:': p:-x
*.y
^•*»
*•%
•:•:
•v
•:.::
j;:j 3.5
:•:• c-x
x:: Q::::-
.'.*. >r-^'.'.'
'.•',•: +->.•:•:•
•'.•'.• i/>',-:-'.
:x: ^:'-:'-'
:•:•' -Q-:-:-
;:•:•' e>:::
•:•: o'.-:-:
:•:• <->•:•:•
"• o v.!:- • ox. u:x ^.:.:: -;:x .v... QJ .x
covX °°W "x: M!W QJ •••:•: :xX "- '•:•:
• ^^ ^:x: +JX: o¥:: ^::xx ::::XJ M- x:
•: £:!:•:! ^: -=:S ij::-:- oH^ ^ ^ ::::
.* Ul*.*.* *-**.'.• >-"-,-.- • t . » •"'"•v\. •••.*XS <—\ *.*
.' *•"'*• .»— '.*.* »r- • •' QJ'.'.' "r~ •"**• X *.**Sl \ '•".
T~ *«*«*• * ',* • V * t "•'•' f\*' *,' * * •'•'•VN *.* *\. JL" • *.
. i *.*• . ^-i •,*.*v"\ 4-^ • «" ^^".',' "^^ «*«*• X^ *. »^v » ^~^ ".'.
tr^'-V E
fO1!'!* fO
"Ja:'::: OJ
L JL.
3 46 8 10 13 14 17 19
Cold Start
180
-------
TABLE 40
Summary of Resul ts:
Phase 1 Tests
RUN
1
2
3
10
13
14
17
19
AVE
FUEL
CONSUMED
kg, Dry
P-la 5T
9.3 50
8.3 38
8.2 57
8.5 37
8.6 39
5.6 43
6.5 47
6.6 46
45
MASS
EMITTED
9
P-l %
47 25
260 48
131 59
93 39
164 48
125 56
95 50
4.5 44
46
CREOSOTE
mg
P-l %
56 35
53 30
9 33
6 13
27 41
19 45
__
4 84
PARTI CULATE
CONCENTRATION
g/dscm
P-l P-2C
0.8 1.7
7.0 2.3
1.6 1.4
1.5 1.2
3.1 1.4
2.1 2.2
1.6 0.9
0.4 0.1
MASS
EMISSIONS
gAg
P-l P-2
11 33
69 45
35 33
24 22
42 30
49 30
32 28
1.5 1.6
Phase 1
Phase 1 v total x 100
Phase 2
' 181
-------
Comparison of the cold start Phase 1 tests to the corresponding
hot start tests (a total of three pairs of tests at three different fuel
moisture contents) does not yield any discernable trend. Only three
tests were conducted under each start condition and each of the three
tests were conducted using different fuel moistures. In two of the
three test pairs, the hot start generated higher particulate emissions.
Carbon monoxide and hydrocarbon emissions were nearly the same. These
results are summarized in Table 41.
182
-------
TABLE 41
Results :
Cold Start vs Hot Start
RUN
la
4
2a
6
3a
8
FUEL
MOISTURE
Low
Low
Med
Med
High
High
START
PROCEDURE
Hot
Cold
Hot
Cold
Hot
Cold
BURN
RATE
kg/hr
3.7
8.1
5.4
5.8
2.8
2.0
PARTICULATE EMISSIONS
gAg
11
40
69
42
35
24
d/dscm
0.79
3.57
7.01
3.75
1.55
0.89
CARBON
MONOXIDE
gAg
190
210
150
170
210
190
HYDROCARBON
gAg
12.9
11.9
12.9
9.2
9.8
11.1
First test phase only.
183
-------
EFFECT OF FUEL MOISTURE CONTENT
A series of tests on the airtight box stove were tested with the
specific purpose of determining the effect of fuel moisture content on
emissions. A total of six tests were conducted; two tests at three
different moisture levels — "dry, medium, and-high" moisture. The
protocol established that the moisture ranges of interest were: 15-20,
25-30, and greater than 40 percent, dry basis. Figure 72 graphically
presents the particulate results obtained during the six emission tests.
The objective was to operate the stove in a similar manner for all tests
so that the only variable would be fuel moisture content. However,
although an attempt was made to operate the appliance at a uniform burn
rate for all tests, this attempt was not totally successful; burn rates
varied from 1.9 to 2.9 kg/hr. The variation in fuel moisture likely
contributed significantly to the inability to operate at a single standard
burn rate. Nonetheless, the variation in burn rate does not seem to have
a significant impact on interpretation of the results. One other sig-
nigifcant point regarding stove operation needs to be mentioned. In order
to try and maintain a "low to medium-low" burn rate as established by the
test protocol the stove was operated in a "dampered down" condition during
the low moisture tests. With the low moisture wood, a very rapid com-
bustion rate condition was immediately established; in order to control the
burn rate, the air inlets were nearly totally closed. However, it is
believed that this represents typical operation of an appliance by the
homeowner, since this is the action which would normally be taken to
control the heat output rate.
184
-------
O) U
ID 4-)
r— S-
CSJ
CD
(/) I/)
O> 4->
Oi 01
O)
O
E -r-
(5 0
S
Q,
*-> 1 —
IO O)
U
s_
tT3
a.
i- f—
Q. a: -M re
o ce
F— -4-> WJ
re c o c
+•» o a> s-
o s- s_ 3
I— u_ o cfl
E
LT)
03
CTl
-o
OJ
vffigm*
"T^THtrifT™**—**™*
c
cr
6>) m/6ui
O
63|/6
0001
OS
008
Ofr
009
0£
OOt?
02
002
01
-------
Figure 73 presents the emission results from the six tests graphed
as a function of fuel moisture; Table 42 summarizes the data. Emission
results are presented in terms of both g/kg and g/10 Btu. The low
moisture fuel resulted in significantly higher emission levels than the
medium moisture fuel. The high moisture fuel tests also resulted in
higher emissions, but the increase in emissions was not nearly as great
as with the dry fuel. Note that both dry moisture runs were conducted
at higher burn rates than the low moisture tests (see Table 42 ); one
would normally expect lower emissions at a higher burn rate. In Figure 68 ,
presented in the previous section, it is apparent that the low moisture
test emissions are well above the level expected at the respective burn
rates of the two tests. (For Run 2 at a burn rate of 2.4 kg/hr an emission
rate of 20 g/kg is predicted as compared to the measured 54 g/kg.
Similarly for Run 5 an emission rate of 18 g/kg is predicted as compared to
the actual rate of 62 g/kg.) The data also were presented in terms of
4
g/10 Btu in order to account for any difference in net efficiencies for
the tests at different fuel moisture levels. The efficiency of the high
moisture fuel tests was significantly lower than for the other test
conditions; this results in a slight upward shift of the emissions, relative
to the medium moisture fuel tests.
As previously mentioned, it was originally intended to conduct all
tests at a single burn rate. Previously, in this report, it has been
shown that emissions are dependent upon burn rate. To determine if the
differences in burn rate among these runs might have a significant impact
on interpretation of the results, the data were normalized to a single
186
-------
burn rate of 2.4 kg/hr. The data were normalized based on the relation-
ship determined in this study.
1C, -.84x
y = 156 e
where y equals emissions, g/kg and x equals burn rate, kg/hr.
The calculation used for determining the normalized emission rate is
as follows:
Eni = ^i X E.
yi
Where:
En. = Normalized emission rate for run i
y? . = Calculated emission rate for burn rate of 2.4 kg/hr
y. = Calculated emission rate for actual burn rate of run i
E. = Actual measured emission rate for run i
Table 43 summarizes the normalized emission rate data. Figure 74
presents the normalized emission rates as a function of fuel moisture
content. Normalization of the data did not signficantly change the
results. The emission rates for the high moisture fuel tests did decrease
slightly in relation to the other fuel levels, since these tests were
conducted at slightly lower burn rates than the normalized rate of 2.4 g/kg.
However, basic conclusions are not changed.
Table 42 also presents the results for creosote, carbon monoxide,
and hydrocarbons. A similar trend in emissions was not obtained for
these three pollutants. In all these cases, the high moisture tests had
the lowest emission rates (g/kg).
187
-------
70
FIGURE 73
Particulate Emissions As A Function
of
Fuel Moisture
60-
=3
-!->
CO
50
CD
_ii
Ol
co
CO
40
30
i.
n
Q_
20-
® g/kg
• g/kg, average
a g/104 Btu
• g/104 Btu, average
10-
(Wet)
(Dry)
10
11
20 30 40
25 43 67
Fuel Moisture, %
50
100
60
150
-------
TABLE 42
Summary of Results:
Fuel Moisture Tests
RUN
2
5
AVE
1
7
AVE
3
9
AVE
4
6
8
FUEL MOISTURE (%)
Wet
12
12
12
21
20
20 -
56
56
56
12
19
56
Dry
14
14
14
26
25
25
126
126
126
14
24
126
BURN RATE
kg/hr
2.4
2.7
2.5
1.9
2.9
2.4
2.1
1.7
1.9
8.1
5.8
2.0
PARTI CULATE
g/kq
Total
54
62
58
22
19
20
34
22
28
40
42
24
Frt. Half
14
11
22
6.3
4.4
5.3
11
5
8
10
8
6
CREOSOTE
mg/m^ kg
917
592
754
969
568
769
218
240
229
216
291
109
CO
gAg
189
220
205
190
160
175
160
110
135
210
170
190
HC
g/kg
16.9
12.1
14.5
13.8
8.8
11.3
10.5
6.6
8.6
11.9
9.2
11.1
189
-------
Normalized* Particulate Emissions g/kg; g/10 Btu
ro
o
to
o
tn
o
en
o
co
o
o
m
I-1 O
ro ro
in o
c
n>
o .p.
o -
rt
c
-s
n>
o tn
O O
tn en
o O
D
IQ Id IQ
o o
-j
cu
to
CD
<-»•
C
CU
<
ft)
-s
CU
Id
(D
O L2!
-* O
ro 3
cu
M
n>
Q.
rl-
O
cr
-s
3
ro
"O
CU
-S
CU
(-*•
O)
3
cu
«= -1' CD
2 =3 N cr
o o ro ^3
-j. rt a. rn
to -•.
r+ o co -^j
c 3 c j^
-5 -S
n> o ^
cu
rf
fD
CO
l/l
O
3
CD
-------
TABLE 43
Normalized* Participate Results:
Fuel Moisture Tests
RUN
2
5
AVE
1
7
AVE
3
9
AVE
FUEL MOISTURE («)
Wet
12
12
12
21
20
20
56
56
56
Dry
14
14
14
26
25
25
126
126
126
BURN RATE
kg/hr
2.4
2.7
2.5
1.9
2.9
2,4
2.1
1.7
1.9
TOTAL PARTICULATE EMISSIONS
g/kg
54
62
58
22
19
20
34
22
28
g/lO* Btu
45
50
42
17
14
15
40
23
31 .
g/kg*
54
80
67
15
29
22
26
12
19
g/104 Btu*
45
64
55
11
21
16
31
18
25
* Normalized to Burn Rate of 2.4 kg/hour; see text for explanation of procedure.
191
-------
Comparison of Results to Information in the Literature
Review of the literature indicates three other studies (two by
She!ton ' and one by Barnett ) have information pertinent to the
effect of fuel moisture content on emissions.
She!ton conducted two separate studies. In one study he evaluated
the impact on heat output efficiency as a function of fuel moisture.
Q
In the second study , he measured creosote accummulation as a function
of fuel moisture content at several heat output rates. In both cases
the tests were conducted with the objective of maintaining a constant
heat output rate with the different fuels; this is an important
consideration. Shelton's results generally indicate that higher emis-
sions are expected with dryer fuel; the impact of wet fuel is less clear.
In the study of heat output, higher emissions would be expected, but in
the other study, lower creosote emissions were actually measured. These
results are briefly discussed below.
Figure 75, summarizes the results She!ton obtained when measuring
efficiencies as a function of fuel moisture content. Note that in this
study emissions were not measured actually. However, if one assumes
emissions are inversely correlated with combustion efficiency (a reasonable
assumption), then these results indicate higher emissions at both low and
high fuel moisture levels. Optimum performance (least emissions) would
be expected either in the 25-35%, or 15-25% moisture range (dry basis)
depending upon whether one expected emissions to be better correlated with
combustion efficiency or overall efficiency. Just because the overall
energy efficiency is greatest at 20 percent does not necessarily mean one
192
-------
should expect the lowest emissions at this point. The peak combustion
efficiency was measured at 25-30 percent moisture; one might expect that
the lowest emissions (g/kg) would occur at this point. Calculating emis-
sions in terms of useable heat output (g/10 Btu), would likely shift
the point of lowest emissions towards the point of greatest overall
efficiency; however, how much of a shift to expect is not clear, and
would be pure speculation, since the quantitative relationship between
combustion efficiency, heat transfer efficiency, and mass of emissions
is not apparent.
The results obtained during this study actually correlate quite well
with She!ton's work. The minimum emission rate actually measured during
this study occurred at 25 percent moisture (dry) although the function
(graph) estimated from the three points measured would indicate minimum
emissions occurring somewhere between 25-40 percent moisture. According
to Figure 75 (Shelton's) one would expect minimum emissions at 25-30%
moisture.
o
In Shelton's second study , emissions were actually measured by
determining creosote accumulation over a period of days with the appliances
operating at constant heat output for the test period. Six units were
simultaneously operated, two units each with low, medium, and high moisture
wood, 5, 25, and 50 (Oak), 33 (Pine), percent, dry basis for dry, medium
and high moisture, respectively. Figures 76 and 77 summarize the results
obtained by She!ton for pine and oak fuel respectively. With the
exception of the tests conducted to simulate a fireplace (doors open and
air inlets wide open) creosote decreased with increasing fuel moisture.
193
-------
FIGURE 75
MOISTURE CONTEMT (% - DRY WOOD BASIS)
0 10 2O 3O MO 50
90 -
COMBUSTION EFFICIENCY
MEAT TRANSFER
EFFICIENCY
10 ' 20 30
MOISTURE CONTENT (% -MOIST WOOCr BASIS)
The dependence of efficiencies on fuel moisture content in an
airtight stove. The air inlet setting was varied to maintain an
average power output of about 17,000 Btu per hour for all moisture
contents. The fuel load volume was approximately constant.
_. , lc,6
Source: She!ton
194
-------
FIGURE 76
MOISTURE CONTENT (PERCENT, DRY WOOD BASIS)
O 1O 2.O 3D ^O 5O
LL)
U-
IL
o
CD
o
a:
O
I
o
o
$
UJ
o
LOWANDEXXIAL
POWER OUTPUT (SERIES 7)
SERIES 5)
MEDIUM AND EQUAL
POWER OUTPUT
HIGH AND EQUAL
POWER OUTPUT%(SERIES
AIR INLETS
WIDE OPEN (SERIES
DOORS OPEN -X(SERIES 6}
o 10 20 30
MOISTURE CONTENT (PERCENT, MOIST WOOD BASIS)
Creosote accumulation as a function of moisture content, using pinon as fuel
Source: She!ton
195
-------
FIGURE 77
MOISTURE CONTENT (PERCENT, DRV WOOD BASIS)
o
LU
UL
O
20k
a
10
or
ui
O.
2.0
:z
o
|
I
!—
t>
U.I
O.H
0.25-
I
i
i
O.I I-
!__„„-
O
10
20
30
LOW FIRE
MEDIUM FIRE
WISH FIRE
to
20
(SERIES 3)
(SERIES 2)
(SERIES i)
30
50
MOISTURE CONTENT (PERCENT,MOIST WOOD BASIS)
Creosote accumulation as a function cf moisture content, using oak as fuel
o
Source: She!ton •
196
-------
Shelton's results confirm the results of this study, that dry wood will
cause increased emissions if a constant power output is maintained.
The lower creosote deposition obtained with wet wood contradicts the
results obtained during this study which indicate slightly higher
emissions than with medium moisture wood would be obtained. However,
during this study, the burn rate for the medium and low moisture fuel
were maintained near constant (2.4 and 1.9kg/hr, respectively) with the
wet fuel tests actually being conducted at a slightly lower rate (increased
emissions expected). On the other hand, in order to maintain a constant
heat output, Shelton probably operated the low moisture fuel tests at
a slightly higher burn rate (burn rate data not provided). A higher burn
rate may help to explain the lower emissions for the wet fuel tests.
This brings up an interesting point -- operation of the appliance is
actually the major controlling factor; the impact of fuel moisture on
emissions is largely due to the effect fuel moisture has on stove operation,
A point worth noting is the impact firing rate has on creosote
formation, according to Shelton's work; this impact can be compared to
the impact from fuel moisture. The greatest dependency of creosote
deposition on combustion rate was noted for the oak fuel (Figure 77).
Whereas the greatest variation in creosote due to fuel moisture (high
fire series 1, low to high moisture) was 1.6 g/kg (2.0-0.4), or a factor
of 4 times, the impact on creosote emissions from a change in high fire
(series 1) to low fire (series 3) was 14 g/kg (17-3) or a factor of 6 times
for dry fue; similarly for wet fuel the change in creosote deposition
due to burn rate was 20 g/kg (20-0.2) or a factor of 50 times. Similar
although not as dramatic results were obtained with pine. The maximum
range in creosote emissions
197
-------
due to fuel moisture occurred for the high burn rate (series 4b) 5 g/kg
whereas the variation due to burn rate between high fire (series 4b) and
low fire (series 7) was 4 g/kg for dry wood and 10 g/kg for wet fuel.
A final 'point should be made regarding Shelton's results. A direct
correlation between emissions emitted to the atmosphere and creosote
deposition is not certain. Changes in excess dilution air, stack flue
gas temperature, and stack gas volume might affect creosote deposition
without affecting actual emissions to the atmosphere or vise-versa.
Consequently, the discrepancy between Shelton's results for creosote
and this studies results for measured particulate emissions may in
fact not be totally inconsistent. Finally, the limited amount of
emission data must be considered in drawing firm conclusions.
Barnett conducted two tests at different moisture levels in
*
conjunction with a series of emissions tests. One test was conducted
at a fuel moisture content of 24%, the other test at 3%. Fuel moisture
for the standard test runs was 30-40%. Barnett shows a decrease in
emissions with the very dry fuel which is contradictory to the results
so far discussed. During Barnett's study, the dry fuel (24%) moisture
indicated a slight increase in emissions. However, close examination
of the data indicates that a significantly higher burn rate was used
for the dry fuel tests. The burn rates for the 35, 24, and 3 percent
tests were 1.1, 1.2, and 2.2 kg/hr, respectively. This would likely
explain the lower emissions obtained with the 3 percent moisture fuel.
In conclusion, the optimum fuel moisture range appears to be 25-35%,
dry basis. Dryer fuel is expected to cause increased emissions. The
198
-------
emissions expected for very wet fuel are less clear; this study
o
indicates slightly greater emissions whereas Shelton's work indicates
less emissions (creosote deposition). However, the fact that more fuel
has to be burned when wet fuel is used (due to the heat loss from
vaporization of water) warrants using properly seasoned wood, even if
no emission reduction is obtained.
199
-------
SIMPLIFIED TEST PROCEDURES
Five basic emission parameters other than participate emissions
as measured by EPA Method 5 and Oregon DEQ Method 7 also were measured
during this test program. The primary purpose for measuring these
parameters was to determine if they could be utilized as a simplified
procedure in lieu of Method 5 testing to evaluate particulate emissions
from wood burning appliances. The parameters measured were: creosote
deposition, carbon monoxide concentration, gaseous hydrocarbon concentra-
tion, stack gas opacity, and smoke spot density. With the exception of
smoke spot density, all parameters correlated to some degree with measured
particulate emissions. The results for each of the five test parameters
are separately discussed in this section.
Creosote Deposition
Table 44 summarizes the creosote deposition results. Creosote was
measured at two locations within the stack; results from both locations
are reported in terms of mass of creosote per unit area per kill gram
2
fuel consumed (mg/m kg).
In general, the correlation between the measurements taken at the
two locations was not particularly good. Figures 78 and 79 graphically
present the correlation between the two test locations graphed by test
run and test phase, respectively. In Figure 78, the correlation has a
slope of nearly 1 (.92) indicating that one would not expect a particular
bias from either measurement; however, the correlation coefficient is
only 0.62 indicating a great deal of scatter and imprecision in the
results. Even a poorer correlation is obtained when evaluating the
results for each test phase. This simply indicates that combining
200
-------
TABLE 44
Creosote Results
RUN
1-1
1-2
1
2-1
2-2
2
3-1
3-2
3
4
5
6
7
8
9
10-1
10-2
10
11
12
13-1
13-2
13
CREOSOTE, mg/n/ kg
Transmissometer
Location
479
-
-
893
458
618
159
578
337
167
508
- 292
748
190
331
25
191
130
193
221
289
174
219
Sample
Location
886
1051
969
557
1619
1217
89
110
98
264
676
290
388
27
148
136
421
316
187
324
409
484
455
Average
683
1051
969
725
1038
917
124
344
218
216
592
291
568
109
240
81
306
223
190
273
349
329
337
PARTICULATE, g/kg
Total
11
33
22
69
45
54
35
33
34
40
62
42
19
24
22
24
22
22
17
38
42
30
35
Front
Half
5.1
9.0
6.3
16.0
13.0
14.0
14.0
6.5
11.0
10.0
11.0
8.3
4.4
6.1
4.8
3.9
3.7
3.7
3.3
7.4
8.1
4.8
6.1
201
-------
TABLE 44 (Cont.
RUN
14-1
14-2
14
15
16
17-1
17-2
17
18
19-1
19-2
19
CREOSOTE, mg/m2 kg
Transmissometer
Location
300
236
264
209
108
_
-
308
16
8
11
9
Sample
Location
478
505
493
337
67
_
-
325
95
93
5
45
Average
389
371
379
273
88
_
-
317
56
*
51
8
27
PARTICULATE, g/kg
Total
49
30
38
23
14
30
28
30
1
1.5
1.6
2
Front
Half
12.0
3.7
5.9
4.8
3.5
6.1
5.4
5.7
0.6
0.80
0.70
1.0
202
-------
Creosote: Method 5 Sample Location, mg/m kg
ro *» oi co O |SJ
o o o o o o
o o o o o o
-i
CO
o
in
o ro
r+ O
fD O
-H
CU
^3
3 -P*
_.. o
c/i o
in
3
3 ' |
*-~ CT
ro
-s
i —
o cr>
o o
fa O
r*
— i.
o
3
**
3
3 00
ro o
o
to
* ^
*
\
* \ *
N m T
A a> » ^
W * v 3
KL .^ Ln
* IB *W —3
v =
« 9 l/l
i» \ ui
W -/v. o
\ -—TO
* • » CD <-f
\ ^ ™ "
* • . i (T)
\«— ( i y
n> < o
% t/1 t/l Ul
v rt 0
\ CO rt
\ • g| "
\ ^m
%N> °
* • 0
\
. . rt
\ 0
* ^3
\%
\
%
-s*< \
II II ^
f
00
CTv U3
co ro
X
~n
V— 4
c:
m
— i
CO
-------
1400
1200
1000
FIGURE 79
Creosote:
Transmissometer vs Sample Location
(By Test Phase)
@
1600
800
X y = 0.73x + 146
r = 0.46
600 • /
400
X
200
200 400 600 800 1000
Creosote: Transmissometer Location, mg/m^ ^
-------
(averaging) the test phase results to obtain a single result for each
run, resulted in increasing the precision, as would be expected.
The major question these data pose is whether the lack of precision is
due to the measurement technique used (experimental error) or in fact,
is inherent in the method, in that the creosote deposition varies randomly
from point-to-point in the stack and from run-to-run.
Nonetheless, despite the poor precision obtained, a useful correla-
tion does exist between the creosote deposition and the measured particu-
late emission rates. Figures 80 through 83 present correlations for creosote
deposition and particulate emissions for four different cases: total
particulate emissions versus creosote measured at the transmissometer
location, the sample location, and the arithmetic average of the two
measurements. In addition, the transmissometer location creosote measure-
ment is correlated with front-half particulate. Note that in each of the
figures, two correlation lines are presented, and one or two data points
are highlighted by a star. These one or two data points appeared to be
inconsistent with the rest of the data. Correlations are presented both
including these data points and deleting these data points. Deletion
of these data points has a dramatic impact on the correlation, resulting
in a change in the correlation coefficient from 0.55 to 0.85. All correla-
tions are based on creosote measurements by test run (as opposed to by
test phase) unless otherwise noted.
The correlation of the transmissometer location with the measured
particulate emissions was better than the creosote at the sample location.
This could be for any number of reasons including random error, more
turbulent flow conditions, at the sample location, less uniform temperature,
205
-------
90?
-------
o
to
O
a, cr
-------
g
-------
FIGURE 83
Creosote
• • _
•. •u,.,^:.r.:....
209
-------
etc.. Regardless, the expected results one would obtain from the three
correlations for total particulate are actually very similar especially
at the lower creosote levels:
Creosote Expected Emissions, g/kg
(mg/m2 kg) Transmissometer Sample Average
100 17 ' 17 18
300 34 30 31
500 52 43 43
700 70 56 56
As a matter of interest, a theoretical calculation was conducted
to determine how the measured creosote collected on the test coupons
would translate to deposition on a surface area equivalent to a 15-foot
stack. These results are reported in Table 45. The total grams of
creosote that would be deposited on the surface area of a 15-foot, 8-inch
diameter stack assuming uniform deposition on the entire stack surface
2
equal to the measured deposition (mg/m ) was calculated. See Appendix A
#31 for calculations. This theoretical deposition (grams) was then
compared to the total particulate and condensible particulate (back-half
of sample train) emitted during the test run to determine the theoretical
percent deposition (reported in last two columns of Table 45, respectively).
The creosote deposition ranged from 1.3 to 13 percent of the total emissions,
with an average of 4.5 and a median of 3.1 percent. The creosote deposi-
tion ranged from 1.8 to 32 percent of the condensible particulate emissions
with an average of 6.9 and a median of 3.7 percent.
-------
TABLE 45
Creosote Deposition
Theoretical Percent of Total Emissions
RUN
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
PARTI CULATE MASS EMITTED, g
Total
187
530
219
191
583
188
180
129
114
226
152
405
347
221
125
82
185
9.2
7.5
Condensible3
Fraction
134
400
147
145
482
151
138
96
89
188
122
326
287
187
99
62
150
3.4
3.9
CREOSOTE DEPOSITION
mqD
m^ kg
969
917
218
216
592
291
568
109
240
223
190
273
337
379
273
88
317
56
27
gc
24.7
27.4
4.2
3.1
16.9
3.9
16.2
1.7
3.8
7.0
5.3
8.6
10.1
6.7
4.6
1.6
6.0
1.1
0.66
DEPOSITION^ %
Total
13
5.1
1.9
1.6
2.9
2.1
9.0
1.3
3.3
3.1
3.5
2.1
2.9
3.0
3.7
2.0
3.2
12
9
Condensi'bl e
18
6.8
2.9
2.1
3.5
2.6
11.8
1.8
4.2
3.7
4.4
2.6
3.5
3.6
4.6
2.6
4.0
32
17
Sampling Train Back-half catch
Creosote measured during test
c Calculated total deposition on stack
15 ft length by 8 inch diameter
Creosote deposition •=• Particulate Mass x 100
211
-------
Carbon Monoxide
Carbon monoxide was continuously monitored during each test run;
data were recorded at 5 minute intervals. An average carbon monoxide
(CO) concentration was calculated for each test run. These results were
compared to the measured particulate emissions to determine if a relation-
ship existed. Figure 84 presents total particulate concentration (g/dscm)
as a function of carbon monoxide concentration, %. A linear correlation
with a coefficient of 0.87 is obtained. However, this has limited value
since the correlation for particulate is in terms of concentration; in
order to obtain an emission rate, (g/hr, g/kg, or g/Btu) one needs to
know the stack gas volumetric flow rate. The volumetric flow rate can
be calcuated based on a carbon balance if carbon dioxide and the mass
of fuel consumed also is measured. This calculation would assume all
carbon is accounted for in the CO and CO,, which is not totally accurate.
A more useful correlation would be for particulate emissions, g/kg,
directly as a function of CO concentration. However, it is obvious that
the amount of excess air will affect the CO concentration measurement,
but not the measured particulate emission rate, g/kg. -Consequently, a
correlation would be expected only if the CO concentration is first corrected
for excess air. Figure 85 presents total particulate emissions as a function
of adjusted CO concentration. In this case, the correction applied is
based on the ultimate CO,, concentration (C02 concentration expected for
complete combustion of wood at zero percent excess air) and the actual
measured C0? concentration. This factor essentially corrects the CO
concentration to a zero percent excess air basis. Other correction
functions also could be used such as calculated percent excess air (this
212
-------
-------
FIGURE 85
Participate Emissions
vs
Adjusted* Carbon Monoxide Concentration
60
• L
50
en
(C
3
O
(O
OL-
40-
30
20.
10.
y = 9.Ox + 2,
r = 0.8G
• LC
HC
X % C02 ultimate
where C02ult = IS.8
234
Carbon Monoxide*, percent
214
-------
factor requires oxygen also be measured). As indicated in Figure 85
using the ultimate C02 correction, a reasonable correlation is obtained
between total particulate emissions and CO concentration.
Figure 86 presents total particulate emissions (g/kg) as a function
of CO emissions (g/kg). This is presented as a matter of interest; this
correlation also requires that volumetric flow be measured in order to
calculate the emission rates.
215
-------
60
FIGURE 86
Particulate Emissions vs Carbon Monoxide
en
cn
o>
ro
fO
4->
o
50
40
30
20
10
40
80 120 160
Carbon Monoxide, g/kg
' .12
200 240
216
-------
Gaseous Hydrocarbons
Gaseous hydrocarbons were continuously measured using a non-
dispersive infrared analyzer. The emissions were measured as hexane
(i.e., the monitor was calibrated with Hexane). Data from the monitor
were recorded at 5-minute intervals and an average hydrocarbon concen-
tration was then calculated for each test run.
In Figure 87 total particulate concentration is presented as a
function of hydrocarbon concentration. A linear correlation with a
coefficient of 0.86 is obtained. As previously mentioned regarding CO,
the concentration correlation is of limited value since volumetric flow
rate must be determined in order to obtain results in terms of emission
rates. Figure 88 presents the correlation obtained when the hydrocarbon
concentration is adjusted to account for dilution by excess air. The
correction factor used is the ratio of the ultimate CCL to the measured
CO-. Figure 89 presents total particulate emissions (g/kg) as a function
of hydrocarbon emissions (g/kg); again use of this correlation requires
that volumetric flow rate be measured.
Although reasonable correlations are obtained using the gaseous*
hydrocarbon emissions, these correlations may be of limited use due to
the unknown sensitivity of instrument response to different hydrocarbon
compounds. Both Non-Dispersive Infrared (NDIR) and Flame lonization
Detectors (FID) (but especially the NDIR), have a varying response to '
different organic species. Consequently, a significant variation in
response might be obtained from stove to stove or test to test depending
on test conditions and the organic species emitted. The degree of
variation expected is unknown; this may not be a major problem. Nonetheless,
217
-------
5.0-
FIGURE 87
Particulate Concentration
vs
Gaseous Hydrocarbon Concentration
(By Run)
4.0-
« MC
3.0-
y = .012x + .091
r = 0.86
2.0-
1 J
o
1.0-
50
100
150
200
250
300
350
Hydrocarbon (as Hexane). com
-------
FIGURE 88
60
Participate Emissions
vs
Adjusted Gaseous Hydrocarbon Concentration1
HC
•L
04x + 5.7
0.80
CO, Ultimate
where C02 ^ - 19.8
200
400 600 800 1000
Hydrocarbon (as Hexane), ppm
1200
1400
219
-------
ro
r\3
o
60
50
40
-------
the measurement of hydrocarbons certainly should be useful for monitoring
and measuring instantaneous changes in appliance emissions. Although
some variation in system response might be expected, the data indicate
that a-relationship does exist between HC concentration and participate;
consequently, the HC analyzer would serve well as a real time monitor
for emissions. In any event, in order to minimize this phenomenon use
of an FID is recommended when measurement of gaseous emissions is desired.
221
-------
Opacity
Opacity was measured by two methods-- visual observation and
transmissometer. Visual observations were taken for discrete periods
ranging from five to thirty minutes during each test run. Trans-
missometer measurements were continuously recorded on a strip chart;
in addition, readings were recorded at 5-minute intervals. Figure 90
presents both the average (time weighted) observer opacity and average
transmissometer opacity as functions of total particulate concentration
(g/dscm). As previously mentioned, a correlation based on particulate
concentration is of limited use unless volumetric flow rate also is
measured. However, there is no simple correction factor for opacity.
for dilution due to excess air. Consequently, measurement of opacity
is probably best correlated to concentration. Figure 91 presents -
opacity (observer and transmissometer) as a function of total par-
ticulate emissions, g/kg. A linear correlation is apparent. However,
a significant spread in the data exists, especially in the 10-30 percent
opacity range.
Table 46 presents a summary comparing observer and transmissometer
data by run. For most runs the measured opacities for the two methods
compared well. Visual observations were not made continually throughout
the run; consequently, the reported value represents the average value
for the periods observed only.
222
-------
FIGURE 90
Total Particulate Concentration
vs
Opacity
50-
40-
30-
OJ
o
O)
a.
2. 20-
o
10-
t>
JBt.
y
r
8.1x + 6.1.
0.79 ^-^
y
r
7.9x + 4.2
0.81
Observer
Transmissometer
0.5
I
1.0
i I i I
1.5 2.0 2.5 3.0
Total Particulate, g/dscm
i
3.5
i
4.0
4.5
223
-------
-------
TABLE 46
Summary of Stack Gas Opacity
^Transmissometer and Visual Observer)
Average Opacity, percent
Run • a
Visual Observer Transmissometer
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
36
34
20
40
37
46
28
20
11
24
16
22
23
10
10
<5
13
-0-
-0-
26
50
15
28
39
29
28
12
5
20
13
21
18
14
10
5
6
7
3
Observations not made for all periods of the test run
225
-------
Smoke Spot Density
Smoke Spot Density measurements were taken at approximately one-half
hour intervals during the test runs. An average smoke spot density was
calculated for each test run (complete results are reported in the Results
Section). Unfortunately, due to a sampling problem, no smoke spots were
taken for the ceramic stove (the lowest emission rate).
Figures 92 and 93 present smoke spot density as a function of total
particulate concentration and emission rate, respectively. No real correla-
tion is noted for smoke spot density and measured particulate. The majority
of the data are clustered around an average smoke spot density of 5 to 7.
Basically, it appears the method simply is not sensitive enough to be highly
useful. Examination of the test data for individual runs does however,
generally show a declining smoke spot value as the test burn progresses.
One would expect that concentration is decreasing as the run progresses;
consequently, the smoke spot may be useful as an indicator of changes in
emissions during monitoring of a particular test burn.
226
-------
FIGURE 92
Total Particulate Emissions
vs
Smoke Spot Density
(By Test Phase)
ro
r-o
c
o
c
CL)
E
X
01
-C
o
(d
s_
ta
jd
u
en
-------
Average Barharach Index, dimension!ess
oo
4
m
x
a>
a.
c
3
O
O
(-(•
Oi
00 -o
3 n>
•cr£. -s
CD ^^ {•$••
*^T ^ ^^"
r>
-H oo c
n> -a —'
to o < m
rt- rt- in n-
£75
-a o
3- fl>
0> 3
1/5 00
n> -i.
O
3
n
-------
Evaluation of Simplified Test Procedures and Discussion of Results
Relative to Information In the Literature
Table 47 summarizes average emissions factors for carbon monoxide,
gaseous hydrocarbon, and creosote found in the available literature;
participate emission values also are provided. The primary purpose for
providing this table is simply to determine if emission rates for the
various pollutants fall into the same range. If a correlation exists
between particulate and carbon monoxide, for example, then for par-
ticulate emissions measured by different researchers within a given
range, then carbon monoxide emission rates also should be similar.
Of course, "should" is important here, because many variables including
stove operation, and measurement methods (not to mention quality assur-
ance) enter into these data collected by the different investigators.
Nonetheless, a listing of average ranges seemed useful. For carbon
monoxide and gaseous hydrocarbons, average emissions do fall into fairly
narrow ranges 45-200 g/kg CO and 7-32 g/kg THC. For creosote a much
2
broader range, 17-11,000 mg/m kg, is found. (Particulate emissions
were not measured by any other method for the 11,000 value.) Particulate
emissions (filterable) were in the 2-15 g/kg range.
No attempt was made to develop correlation coefficients (such as
those prepared for this study) with these data obtained from the lit-
erature, or to incorporate any of these data into the correlations pre-
pared in this report.
Some general information relevant to this discussion of simplified
test procedures was found in the literature. Cooke, et al indicates
229
-------
TABLE 47
Summary of Average Emission Factors
for
Carbon Monoxide, Gaseous Hydrocarbon, and Creosote
(Literature Review)
Investigator
DGA/OMNI
Oregon DEQ
Knightlj (Oak)
Knight J
( Fir Brands)
Butcherlk (1980)
Monsanto
11
PEDCo
Battell e4
Hubble12
California Air.
Resource Board
CCRL4
Rudl ing
8
Shelton
Total
Particulate
Matter
g/kg
27
32
_-
__
_ _
10
28
_ _
__
22
__
__
--
Fi i terabl e
Particulate
Matter
g/kg
6
(0.6-14)
15°
(0.8-45)
2
(0.2-3.5)
4
(0.3-8.3)
4
(1.6-6.4)
4
--
•• «-
(0.5-22)
13
..
2.61
(0.1-10)
--
Carbon
Monoxide
9/kgA
115
(20-220)
191E
(42-363)
92
(26-185)
196
(67-310)
100
(63-158)
180
(91-370)
204
100
(33-400)
(55-196)
45
( 4-147)
136
(87-184)
90 J
( 2-300)
--
Gaseous
Hydrocarbon
9/kgB
9c
(0.4-17)
« _
28F
(8.5-78)
29F
(11.3-56)
__
—
21F
(2-112)
7H
(.25-16.5)
32 H
(20-43)
_ _
—
Creosote
mg/m - kg
330
(27-969)
17
(5.7-36.6)
__
__
--
(380-3800?
__
6000
(170-11,000)
230
-------
Footnotes for Table 47
A Method of measurement either orsat or continuous NDIR.
B Method of measurement either Flame lonization Detection (FID) or
Nondispersive infrared analyzer (NDIR).
C NDIR
D Includes data from furnaces.
E Calculated by this author from data in original test reports.
F FID
G Values calculated from reported results as g/kg and collection
area of .70 m (3 sections at .57 m length xllx .13 m diameter)
H GC/FID
I Glass thimble filter.
J Values estimated from graph.
231
-------
that carbon monoxide and total hydrocarbon emissions "vary" together
(no statistical correlation was done); participate emissions were not
le
measured during this study. Nubble's data indicates that both carbon
monoxide and creosote increase with increasing participate matter, and
similarly Knight's data indicates that carbon monoxide and total gaseous
hydrocarbons follow the trend of the measured particulate emissions.
Although neither of these investigators reported a statistical comparison
of measurement techniques, Knight does go so far as to recommend an
appliance performance rating based on a combined CO/THC/Efficiency
measurement.
The most interesting data found in the literature regards creosote
measurements. The most significant result was Shelton's data indicating
a 75% reduction in creosote deposition from use of a barametric damper.
Since the barametric damper can reasonably be expected to have little
effect on the emissions actually generated during combustion and emitted
to the atmosphere (She!ton did not measure emissions by any means other
than by creosote deposition), this implies that creosote deposition can
not be expected to correlate well with particulate emissions. The effect
of reduced creosote emissions during Shelton's study is likely due to
the effect of dilution air and increased stack gas velocity.
Nonetheless, Barnett reports an "excellent correlation" between
creosote deposition and particulate emissions measured for.five appliances
(several runs each). (A correlation coefficient of 0.95 based on average
values measured from five appliances was calculated by this author.) A
fairly reasonable correlation also was obtained during this study. However,
232
-------
in each case, appliances were operated under very similar conditions
for the test runs. Thus, one might conclude that for similar test
conditions (e.g., excess air, stack gas temperature) creosote deposition
will be a good indicator of emission levels; however, if operating
conditions change singificantly, this relationship is likely to fail.
Table 48 summarizes the advantages and disadvantages of the
simplified test procedures evaluated during this study. An estimate of
costs for each method also is provided. Carbon monoxide and/or hydro-
carbon measurements appear to be the best potential candidates for a
simplified test method. A correlation coefficient with total particulate
emissions of 0.8 for each of these procedures was determined for the
data from this study. This correlation was for carbon monoxide concentration
(adjusted for excess air) compared to total emissions, g/kg. This is
of particular interest because a correlation based on concentration means
stack gas volumetric flow rate does not need to be measured; this is a
definite advantage for a simplified procedure. Another advantage of this
technique is that continuous measurements can be taken and the results
obtained instantaneously. This makes the technique especially useful
for "screening" stove performance at various operating conditions. Another
advantage of the method is that, depending upon the quality of the data
required consistent with the purpose of the testing, CO can be measured
by several techniques with a wide range of expense. For example, for
the crudest screening tests, an orsat could be used to take grab samples;
or a portable NDIR unit could be used to make continuous measurements
with a reasonable accuracy; finally, for the most accurate measurements,
233
-------
TABLE 48
Simplified Test Procedures Summarized
Method
Total Particulate
Filterable Particulate
High Volume
•
Carbon monoxide
Correlation to
Total Particulate
1
20-50% of total
particulate mass
basis. Correla-
tion coefficient
not determined
--
0.8
Advantages
•Accurate measurement of emissions
including condensibles .
• Integrated sample over entire
burn cycle.
• Accurate measurement of emissions.
• Integrates sample over entire
burn cycle.
•Slightly less expensive than
total particulate method.
• Short sample time enables
measurement of discreet periods
within burn.
• Provides Instantaneous and
continuous output, excellent
for monitoring burn cycle.
• Inexpensive to use once capital
investment incurred.
•Suitable for screening method,
using cheaper, less accurate
instrumentation
• Suitable for field monitoring.
Disadvantages
• Expensive
• Long sample time required (1-hr
minimum); not ideal for measuring
discreet periods within burn.
• Experience with method required.
• Does not measure condensible
particulates .
• Expensive.
• Long sample time required (1-hr
minimum); not ideal for measuring
discreet periods within burn.
• Experience with method required.
• Expensive
• Collection efficiency for con-
densible organics unknown.
• Multiple samples required to
obtain measurements for entire
burn cycle.
. No direct measurement of
particulate.
Estimated Cost^
Capital
Investment
$20,500c>d
20,500c>d
16,500
13.5006
$600-2600f>h
Per Run
$2,100
1 ,900
1 ,900
560
2BU
Per Series
(9 tests)
$15,000
12,500
12.bOO
4,200
234
-------
TABLE 48 (Cont.)
Method
Total Hydrocarbon
Creosote
Opacity (Transmlssometer
or visible emission
inspector)
Smoke spot density
Correlation to
Total Participate3
0.8
0.8
(0.5)
0.8
0.6
Advantages
• Same as Carbon monoxide
• Uncomplicated to use.
• Inexpensive (low level of effort).
•No significant capital Investment.
• Capable of measuring discreet
periods within burn cycle.
• Inexpensive to use.
• Little or no capital investment.
•Suitable for field monitoring
(visible emission observer).
• Inexpensive to use.
•Low capital Investment.
* Easy to use .
•Very short measurement time;
may be used to monitor discreet
periods within burn cycle.
•Convenient for field monitoring.
Disadvantages
• Same as Carbon monoxide
• Potential for variable response
to different HC species.
• Not direct measurement of
particulate emitted to atmos-
phere.
• Results likely dependent upon
numerous variables such as
stack temperature, and excess
air.
• Does not directly measure
particulate.
• Results highly dependent upon
excess air levels.
• Results dependent upon excess
air levels.
• Large number of measurements
required over entire burn cycle.
Estimated Cost1'
Capital
Investment
*23.500e
$4500-670(Ph
1.000
0 - 1 ,000h
75.00*'
Per Run
$560
280
250
280
280
Per Series
(9 tests)
$4,200
--
2.800
2.800
2.800
235
-------
Footnotes for Table 48
Correlation coefficient determined from this study.
See Appendix D for basis of cost estimates.
C Includes laboratory quality, CO, CO?, and (L monitors for accurate
and continuous determination of stack gas composition for determining
stack gas volumetric flow by stoichiometry. Subtract $12,500 for
monitors and add $3,000 for orsat and low velocity flow measurement
instrumentation equipment (net change - $9,500) if orsat/velocity
methods to be used.
Add $5000-10,000 for flame ionization detector if total hydrocarbons
analysis is desired.
p
Delete $1,000 if platform balance not to be used to monitor fuel
consumption rate (necessary for emissions as g/kg basis) and
results to be reported as concentration adjusted for excess air.
If used as a screening test with orsat or less accurate instrumentation
to determine emissions (CO concentration adjusted for excess air).
^ If used as a screening test with orsat or less accurate instrumentation
(e.g., NDIR instead of heated FID) to determine emissions (hydro-
carbon concentration adjusted for excess air).
Add $1,000 if platform balance to be used to monitor fuel consump-
tion rate.
236
-------
a laboratory quality continuous NDIR properly calibrated would be used.
Capital investment for a quality laboratory instrument (CO/CC^) is
estimated at about $8,500. A quality FID (heated system) would cost
$10,000 (under $5,000 for unheated system).
Creosote deposition has some advantages as a simplified method. •
For one thing it requires almost no capital investment or significant
level of effort to conduct the tests. Furthermore, the method
is simple to use; no instrument calibrations or use of gaseous sampling
systems is required. An advantage of the method is that it can be used
over an extended time period of stove operation (if this is desired to
meet the purpose of the evaluation) without any additional effort from
the tester. In other words, a sample can be taken for.1-hour, 1-day,
or 1-week without any difference in level of effort (excepting stove
operation). However, the results of this study indicate a problem with
precision (for the coupon technique used) between duplicate samples.
Furthermore, She!ton's work casts some doubt regarding the effect
variable stove operating conditions has on results.
Opacity correlates well with particulate concentration; however,
this really is of little use unless volumetric flow measurements also
.are taken. The main problem with this technique is the impact excess
air has on the opacity. The method has potential as a screening technique
with a high probability of type II error (false acceptance), because of
the dilution air effect. In other words, if a unit fails an opacity test,
one could be reasonably sure that the appliance was not meeting a par-
ticular emission level; however, if the unit met a given opacity level,
237
-------
a high uncertainty would exist that the unit actually was meeting
the corresponding particulate emission level. Furthermore, as the
emissions decrease towards an opacity of zero, it would become increas-
ingly more difficult to measure any further decrease in opacity even with
a transmissometer (e.g., below 5 percent opacity it would be very difficult
to measure emission changes).
The evaluation of smoke spot density measurements during this study
indicated no reasonable correlation with total particulate. No additional
data on this method's applicability to small wood burning appliances was
found in the literature.
Measurement of filterable particulate emissions (Method 5 front-half)
will result in a slightly reduced level of effort; however, the reduced
level of effort seems negligible compared to the total level of effort
required by either method. Consequently, it would seem logical to proceed
with a back-half analysis whenever a Method 5 is conducted. For this
reason, no correlation coefficient between total and filterable particulate
was calculated for the test data obtained during this study or for data
in the literature. However, filterable particulate ranged from 20-60
percent of the total emissions measured for the data reviewed; this is
a fairly broad range.
Measurement of filterable emissions by the high volume technique
was not evaluated during this study due to budgetary constraints. However,
it has been listed in Table 48 as a simplified method because of the
interest in this technique. Basically, when one considers that the level
of effort required for conducting an emissions evaluation (stove operation,
238
-------
volumetric flow rate measurements, fuel combustion measurements) this
method does not result in any significant reduction in level of effort
from that required to use EPA Method 5. The major advantage of this
method is that a quick sample can be taken and thus discrete emission
samples can be taken at various times in the burn cycle. Thus this
approach is particularly useful in evaluating changes in emissions
during the burn cycle. Of course, depending on the purpose of the
evaluation, this can, in fact, be a liability; i.e., when an integrated
sample over an entire and long burn period is of interest, it may
actually be an inconvenience to have to take multiple samples.
Barnett7 has conducted numerous tests with his high volume
technique to show that his system provides repeatable results if the
measurements are taken at a stack gas temperature of less than 250 -
275°F (i.e., at temperatures up to 275°F the amount of material condensing
and collecting on the filter is consistent; at sample temperature above
275°F, some material does not sufficiently cool prior to reaching the
filter and is not collected resulting in a measured emission level less
than that measured when the sample is taken at 275°F). However, no
tests have been conducted to determine how these measured emission rates
compare with either the filterable or total particulate emission rates
measured by the Method 5 sampling system. A comparative test of this
nature would be informative.
In summary, the carbon monoxide or hydrocarbon methods appear to
offer the greatest advantages and flexibility as simplified procedures.
Additional work to verify the correlation with total particulates reported
239
-------
in this study and to determine if the relationship will hold over a variable
set of conditions is warranted. Although the correlation of creosote with
total participate emissions is more shakey, additional work in this area
to improve the precision of the method and to identify its boundaries would
be beneficial since this method is easy to use, requires no knowledge
of instrumentation or emissions sampling, and is inexpensive. Finally, a
comparative study of the emissions measured with Method 5 and a high volume
technique would be extremely useful since such a study would give a basis of
comparison for the emissions data already collected by the different methods,
as well as define the potential of the high volume method as a "simplified"
technique for obtaining emissions data.
240
-------
REASONABLE EMISSION STANDARDS
One of the objectives of this test program was to evaluate improved
technology stoves in order to provide data for defining the level of a
reasonable emissions standards. Of interest was an achievable emission
level when operating appliances at a moderate to low burn rate (less
than 2.5 kg/hr).
Of the three improved technology stoves and two retrofit devices
tested, only one stove resulted in emissions significantly below the norm.
This appliance (ceramic stove) also was operated at a higher burn rate
than the desired rate, which undoubtedly is expected to result in lower
emissions. Nonetheless, this emission rate can serve as an indication
of an emission rate which may be achievable for appliances operating at
lower fuel combustion rates. None of the other stoves tested in this
program came close to meeting the emission rate of the ceramic stove.
In fact, only one of two tests for the catalyst-modified combustion stove
even resulted in a measurably lower emission rate than the airtight
box stove.
Table 49 summarizes the levels obtained for the various emission
parameters measured for Run 19 with the ceramic appliance and Run 16
with the catalytic-modified combustion appliance. Run 19 was the higher
emitting test of the two test runs for the ceramic stove, but also had
a burn rate which was considerably lower than Run 18. The two runs for
these two stoves can serve to illustrate the range for a reasonable
emission standard.
It is very important to mention that any meaningful emission standard
241
-------
E
TABLE 49
Emission Standard Range
Total participate, g/kg
Front-hal f participate, g/kg
2
Creosote, mg/m kg
Carbon monoxide, g/kg
Hydrocarbon, g/kg
Adjusted ppm
Opacity, average percent
Run 16
Catalytic
Modified
Combustion
14
3.5
90
80
4
230
Run 19
Ceramic
Appliance
2
1
27
50
1
40
0
r
r
Parameters during these tests
b
Burn rate , kg/hr
Fuel load to combustion ,
rate ratioC, kg/Btu/hr x;10
*
Fuel load to stove 3
volume ratio0> kg/m
3.0
1.3
100
3.9
1.1
99
[
L
Adjusted for dilution air
Dry basis
Wet basis
242
-------
The results of this emission test program were disappointing in
terms of providing emission data which justifies an emission standard
below the normal level of a wel1-operated typical stove (15-20 g/kg
total participate). Nonetheless, the emissions obtained from the ceramic
stove point towards a reduced emission level which should be achievable.
Discussion of Results Relative to Information in the Literature
Numerous test programs have been conducted on various residential
wood combustion appliances during the past several years. The purpose
of these various test programs has ranged from evaluating emission
characteristics from innovative and or different stove designs (e.g.,
Oregon DEQ , Barnett a) to determining an -emission factor from typical
home-operated units (e.g., Sandborn ). Since it has been shown that
appliance emission factors can be affected by a number of variables,
results obtained from different studies must be evaluated and compared
with extreme caution, if at all. Factors which may affect emission rates
include, but are not limited to: emission measurement method, stove
operating technique (fuel combustion rate, fuel charge size, whether or
not start-up and burn down are included in test), fuel characteristics
(type, size, moisture content), and type of appliance.
Nonetheless, in attempting to define what constitutes a reasonable
emission level, it is certainly necessary to discuss the results obtained
by other investigators and reported in the literature. The purpose of
this review is 1) to provide a basis by which to compare the results
obtained during this study, 2) to identify data obtained by other investigators
243
-------
is heavily dependent upon stove operating conditions. Consequently,
established emission rates are somewhat meaningless unless stove operating
conditions also are established. The results in Table 49 should be
reviewed keeping this in mind. This point is exemplified by the results
obtained with the catalyst-modified combustion stove. As already noted,
significantly higher results were obtained during the second test run
for the catalyst-modified combustion stove. This can reasonably be
explained by the operating parameters during the test run. In an
attempt to reduce the burn rate (2.1 kg/hr vs 3.0 kg/hr) stove combustion
chamber temperatures were below the level necessary for proper catalyst
operation. A similar result was obtained during testing of the catalytic
box stove; a change in burn rate from 1.7 kg/hr to 2.2 kg/hr resulted in
a reduction of emissions from 38 to 23 kg/hr.
The number of tests on the improved technology appliances in this
program is very limited and in the authors opinion is not adequate to
clearly characterize the emission levels which might be obtained from the
»
appliances. In order to accurately access the emission characteristics
of these stoves, a complete series of multiple emission tests' at different
fuel combustion rates would be required. Whether or not any of these
appliances warrant further testing is another question. It is the author's
opinion that the ceramic stove should be operated under conditions to
determine the actual instantaneous heat output of the unit. Such testing
would enable one to determine if the stove does in fact have adequate
heat storage properties permitting it to produce an even heat output while
being operated in a batch type mode.
244
-------
which might be particularly useful in helping to define a reasonable
emission level, and 3} to identify data which would suggest particular
appliances that potentially have low emissions and warrant further
investigation. Consequently, although the majority of the RWC par-
ticulate emission data known to be available is tabulated herein, the
discussion of the data is not intended to be all encompassing (e.g.,
no attempt is made to discuss such factors as affect on emissions of
fuel size, fuel type, etc.). Readers are urged to refer to the original
publications prepared by the investigators for further information and
insi ght.
Table 50 summarizes the emission rates obtained by various investiga-
tors and is useful in providing a general idea of typical emission levels.
As average emission levels, even for each investigator, the data may
inherently encompass a wide range of stove types, operating parameters,
etc. (such as in this study). A cursory examination of these data indicates
that an emission rate of greater than 20 g/kg total particulate and 5-10
g/kg filterable particulate is typical.
Table 51 presents the same data in a format which provides more detail.
The data are presented first by stove type (design category), with each
investigators data listed separately. Ranges also are presented along
with the average values. The particulate emission rates in this table are
1) total particulate -- this generally includes both the filter catch and
analyses of "back-half condensibles" by EPA Method 5. In some cases (e.g.,
Oregon DEQ and this study) an extra unheated filter was included as part
of the back-half to catch any material leaving the condenser train.
245
-------
TABLE 50
Summary of Average Emission Rates
Investigator
Del Green/OMNI
Del Green/OMNI
(Excluding fuel
moisture tests)
Oregon DEQld
' u lh
Sanborn
Barnett13'7
Knight1^
(Oak)
( Fir Brands)
Butcher10 (1979)
1 k
Butcher (1980)
Monsanto
Furnaces :
Oregon DEQld
. , lh
Sanborn
Burn Rate
kg/hra
2.5
2.5
3.1
2.4
1.5
5
7
2.1
2.4
7.2
9.3
5.7
M/Q
kg/104 Btu-hr
2.3
2.0
4.9
1.5
2.3
0.8
0.8'
2.4
1.4
Particulate Carbon
Total Hi Vol Filterable Monoxide
g/kga g/kga g/kga g/kga
27 -- 6 115
19 ._ • 4 115
32 -- 15
27 -- 18
7
2 92
4 196
9
3.5 -- 100
10 ' -- 4 180
9 -- 2.6
12 -- 5.5
Comments
g
b,c
b,d
e
e
f
b,e
246
-------
Footnotes for Table 50
Dry basis.
° M/Q estimated values,9000 Btu/lb, dry basis fuel, 15 ib fuel charge
assumed for Butcher's work.
c Investigator reported values on wet basis fuel; factor of 0.8 (20%
moisture fuel) used to adjust values.
Average value (unweighted for number of runs) for 14 different
appliances; some tests conducted with modifications to appliances;
investigator reported results on wet basis, factor of .70 (30%
moisture fuel) used to adjust values.
e Tests operated at three distinct burn rages: low, medium, and high;
load size adjusted for each burn rate (e.g., less wood loaded for
low burn rate) presumably resulting in fairly constant M/Q (actual
load sizes not provided with data).
Includes results from tests with both seasoned and green fuels.
^ All tests conducted from cold start(i.e., no hot ash bed and a cold stove)
247
-------
TABLE 51
Particulate Emission Data Summary
(Literature Review)
Stove Type
Investigator Burn Rate
Emissions, g/kg'
Comments
.
Ceramic
Catalytic
Catalytic/Modified
Combustion
Airtight Box
Modified Box
DGA/OMNI
DGA/OMNI
Oregon DEQ
7
Barnett
7
Barnett
Corning
DGA/OMNI
„ . Ih
Sanborn
la
Barnett
10
Butcher
Butcherlk
Barnett13
Kg/nr*
5.2
(3.9-6.4)
2.0
(1.7-2.2)
2.2
(1.8-2.6)
1.4
1.4
(0.8-2.0)
NA
2.5
(2.1-3.0)
3.2
(2.6-3.8)
1.7
(0.8-2.4)
2.1
(0.9-4.0)
2.4 .
(1.7-3.4)
1.6
(1.1-1.9)
type
DF
DF
DF
RO/SM
RO/SM '
NA
DF
MHW
RO/SM
0/P
0/H
RO/SM
net no a
5/BH
5/BH
5/BH
HV
HV
C
5/BH
5/BH
HV
HV
HV
HV
or ourn
2
2
2
1
3
NA
2
2
6
26
10
9
S Total d
1.5
(1-2)
31
(23-38)
25
(21-30)
85-95%
22
(14-30)
14
(13-15)
Filterable6 Hi VoT1"
0.8
(0.6-1.0)
5
(4.8-5.9)
6
(4.7-6.6)
7.0
1.0
5
(3.5-5.7)
6
(4.5-7.0)
7.2
(2.6-16.9)
9.3
(1.3-24)
3.5
(1.6-6.4)
9.1
(3.5-14.5)
p
g
h
1 1
i
i
248
-------
TABLE 51 (Cont.)
Stove Type
Investigator Burn Rate
Wood,
Emissions, g/ka
Comments
Step DGA/OMNI
DGA/OMNI
Id
Oregon DEQ
Barnettla
Sanborn
Monsanto
11
PEDCo
Modified Step Barnettla
Baffled Barnettla
Ih
Sanborn
le
Hubble
11
Monsanto
kg/hr°
2.3
(1.7-2.9)
2.4
(1.9-2.9)
2.6
(1.5-3.6)
1.7
2.7
(2.2-3.6)
7.4
(6.6-7.8)
6.2
1.6
(1.2-2.0)
1.5
(0.9-2.0)
2.1
(1.4-2.6)
NA
(.8-7.7)
7
(6.0-8.4)
Type"
DF
DF
DF
MHW
MHW
0/P
P
RO/SM
RO/SM
MHW
0
0/P
Metnod1-
5/BH
5/BH
5/BH
HV
5/BH
5/BH
5/BH
HV
HV
5/BH
CI
5/BH
of burns
6
2
4
1
3
4
1
4
3
6
NA
4
Total d
36
(19-62)
21
(19-22)
43
(23-74)
38
(24-53)
10
(4.1-19)
*
28.3
26
(13-44)
10
(6-19)
Filterable6 Hi Vol f
9
(5-16)
5
(4-6)
25
(13-43)
7.2
21
(14.0-28.1)
3
(1.8-6.3)
21
6.4
(3.0-12.1)
7 7
(5.7-11.4)
18.9
(8.3-22.4)
NA
(0.5-22)
4 .
(2.5-7)
j
P
i
k
•j
h
1
k
249
-------
TABLE 51 (Cont.)
Stove Type
Investigator Burn Rate
Emissions, g/kg
Comments
p-y/iir ijpc IIKLIIUU UT uurnb 7\ 5 — -r
Total0 Filterable6 Hi VolT
• — - --- --- . •
Cross Draft
Down Draft
Convective
Modified Convective
Gasifier
Unidentified/Misc.
Barnettla
Barnettla
Barnettla
Sanborn
Barnett 3
Oregon DEQ
Knight1 j
Knight1"-"
1.4
(1.0-2.2)
2.0
1.3
(1.0-2.3)
1.6
1.3
(0.7-2.0)
4.5
(4.5-4.5)
5
(2-9)
7
(2-13)
RO/SM
RO/SM
RO/SM
MHW
RO/SM
OF
0
FB
HV
HV
HV
5/BH
HV
5/BH
5
5
6 4.5
(1.5-6.7)
1 5.2
26 14.4
(4.1-36)
1 26 16
5 10.0
(2.9-17.6)
2 26 8
(20-31) (5.1-10.9)
15 1.5
(0.23-3.5)
9 4.4
(0.3-8.4)
i
i
i
h
i
p
m'
250
-------
TABLE 51 (Cont.)
Stove Type
RETROFITS:
Non-catalytic
Catalytic
FURNACES:
Investigator
DGA/OMNI
Oregon DEQ
Sheltdn2'3
DGA/OMNI
Shelton2'3
Oregon DEQ
c u In
Sanborn
Burn Rate
kcj/hra
2.3
(2.2-2.4)
4.2
NA
2.4
(2.1-2.6)
NA
9.3
(4.7-16.4)
5.6
(3.6-8.2)
Wood
Typeb
DF
DF
HW/SW
DF
HW/SW
DF
DF
Test
Method0
5/BH
5/BH
--
5/BH
--
5/BH
5/BH
Number
of burns
2
1
NA
2
NA
4
3
tm
Totald
37
(35-38)
10
(23-41)
20
(17-22)
(45-51)
9
(3-19)
12
(8-18)
issions, g/Kg
Filterable6 Hi VolT
7
(6.1-7.4)
2.6
3.5
(3.3-3.7)
2.6
(0.8-4.2)
5.5
(4.1-6.9)
LommeriLb
P
n
n
P
h
251
-------
-------
2) Filterable participate -- this generally constitutes the front-half
heated filter catch for EPA Method 5 and 3) High volume particulate --
this constitutes filterable particulate collected on the unheated filter
of a "high-volume" (high air sampling rate relative to Method 5) sampling
unit. Typically with this method, ambient dilution air is added to the
sample prior to its contact with the filter. The proponents of this
method generally believe that the emissions collected by this technique
is nearly equivalent to the total particulate emissions measured by
Method 5 and/or that this technique of measurement more adequately repre-
sents the "particulate" emissions which would be measured in the ambient
atmosphere (e.g., by an ambient monitor). Because of the uncertainty as
to whether high volume data is more nearly equivalent to total or filter-
able particulate (as defined above), these data have been listed separately
by this author. The conservative approach is to assume high volume emis-
sion data underestimates total particulate emissions and more closely
approximates the results which would be expected from the filterable
fraction; this is the approach taken here when discussing and evaluating
the test data.
It has already been shown in this study and by other investigators ' '
that emissions vary inversely with burn rate; consequently, higher emissions
are expected at lower'burn rates. In this study, a burn rate of 2.5 kg/hr
was assumed to be typical of a moderate to low burn rate which might typically
be used in the Pacific Northwest. Barnett a> has done extensive studies
in New York State and has generally determined that burn rates were in
the range of 0.8 - 2.8 kg/hr, with a burn rate of 1.6 kg/hr being "typical".
253
-------
Emission tests conducted at burn rates much greater than 3.0 kg/hr are
orobaoly of little value in establishing meaningful emissions data for RWC
*
appliances. Note that use of the burn rate (kg/hr) parameter is actually
a simp!ication; since large appliances are intended for higher heat out-
puts and will require higher burn rates. Consequently, a "high" burn
rate is appropriate for large units intended to provide high heat output.
However, the point is that most homes do not have a high heat demand
from the appliance. Therefore, the appliances would be operated at low
burn rates. The factor M/Q (kg wood charged/10 Btu/hr) is valuable in
establishing a norm in this respect, since larger appliances will hold
a larger charge and are intended to be operated at a higher burn rate.
Nonetheless, for this discussion, it is assumed that evaluation of emis-
sions at burn rates less than 3.0 kg/hr is desirable.
The results obtained in this study have already been discussed in
detail; with the exception of the ceramic stove, no significant emission
reductions from the norm were obtained from the improved technology
appliances tested. The Oregon Department of Environmental Quality was
involved with emissions testing conducted by two consulting firms (OMNI
Environmental,and Seton, Johnson and Odell) with the purpose of evaluating
improved technology appliances. The results of these two test programs
also yielded little in the way of a measurable emissions reduction from
these devices. The appliances tested included a catalytic box stove,
gasifier stove, non-catalytic control retrofit, and two furnaces, in
addition to a standard step stove. An average total particulate emission
rate of 32 g/kg was obtained during this test program. Only the furnace
254.
-------
and the non-catalytic add-on device yielded emission reductions. The
furnaces had an average emission rate of 9 g/kg with one furnace
operating at 3.6 g/kg. However, both furnaces were operated at a high
burn rate. Although the burn rate was high, the fuel load to combustion
rate ratio (M/Q kg fueled/104 Btu/hr) was not particularly low (2.4); in
other words, these units were built to provide a large heat output.
The significance of these results are that like the ceramic stove, a low
emission rate (less than 5 g/kg) is achievable when efficient combustion
N
is attained.
The one test run for DEQ with the non-catalytic add-on device (same
device as tested in this program) did indicate a reduction of 50% in
emissions from the airtight stove operated at similar conditions (10 vs
20 g/kg). About 80% of the reduction was achieved in the filterable portion
and 30% in the condensible which would be expected, since this device
basically consists of a filtering medium. As previously indicated, during
this study, no emission reduction was noted for two tests conducted with
p o
this device. She!ton recently conducted tests with this device (and
two other retrofit devices) to determine if these devices were useful
in reducing creosote deposition in the flue pipe. She!ton found that this
device reduced creosote emissions by 21% to 45% (21% when liquid creosote
collecting in the flue pipe tee was considered). Shelton's test were
conducted by operating eight stoves simultaneously under identical
conditions with one pair of stoves fitted with each of the three devices
to be tested and one pair of stoves with no device to serve as the control.
The stoves were operated for a period of ten days under different operating
255
-------
modes (low to high burn) and with different fuels. The creosote
deposition at the end of the test period was gravimetrically determined.
The other two retrofit devices tested by Shelton were a catalytic retrofit
(same unit as tested in this emission test program) and a barometric
damper. Reductions in creosote deposition of up to 45-50% were obtained
by Shelton during his study with the catalytic unit. The best performance
(i.e., greatest reduction) was obtained under low fire conditions. Under
low fire conditions a temperature rise of 450 to 615°F across the catalyst
was typical; at high fire a rise of 350 to 450°F was noted. This perfor-
mance is significantly better than the results obtained during this emission
study -- 125 to 300°F temperature rise across the catalyst with no measur-
able reduction in emissions when compared to the two runs conducted without
the device under similar conditions (burn rate and fuel moisture content).
These results reinforce the belief that proper catalytic action is critically
dependent upon several factors including flue gas temperature, oxygen
availability, residence time at the catalyst, and the proper mixing of gases
and combustion air. Nonetheless, Shelton's results indicate reduction of
emission levels by catalytic action can be achieved under some operating
conditions. To what degree the conditions necessary for emissions reduction
by catalytic action is dependent upon appliance design, appliance operating
conditions and/or operator actions is unclear at this point. Proper
catalytic action is no doubt a function of all three of the above; the
limiting factors have yet to be clearly defined. The third device tested
by Shelton, the barometric damper, resulted in the most significant
creosote reduction — 75%. This is a significant finding since this
256
-------
indicates creosote deposition is not necessarily a good measure of par-
ticulate emissions to the atmosphere. There is no reason to expect such
a significant reduction in emissions to the atmosphere from use of a
barometric damper (some reduction might be obtained due to more steady
burn conditions). Therefore, the conclusion is that creosote deposition
was significantly reduced due to such factors as dilution of the stack gas
effluent and increased stack gas velocity (i.e., decreased residence time);
however, emissions to the atmosphere were probably not significantly
reduced. This raises some question as to the correlation between creosote
reduction and actual reduction of emissions to the atmosphere achieved
by the other two devices. It is quite possible that due to other factors
(e.g., stack gas dilution from secondary air required for the catalytic
unit) the creosote reductions measured for the other two devices are, in
fact higher than the actual emissions reductions achieved.
Barnett also tested two catalytic appliances. One appliance yielded
no significant emission reduction while the second appliance significantly
reduced emissions. Barnett has tested a number of appliances (with and
without modifications such as secondary air, added fire brick, added
baffles, and thermostatic control) in order to evaluate performance of
the different appliance design types. Appliance types evaluted include
thin wall, box, step, cross draft, down draft, and catalytic; all of
Barnett1s tests were conducted at a low to medium-low burn rate (0.8 -
2.2kg/hr) with the majority of the tests conducted at about 1.6 kg/hr;
this is the combustion rate at which Barnett achieves a stove surface
temperature of approximately 300CF. Barnett does use a high volume
257
-------
sampling technique so that the numerical emission values he has obtained
can not directly be related to the data obtained by the total EPA Method
5 train (this author's opinion). However, taking the conservative approach,
Barnett's results can be considered equivalent to the filterable portion
of the Method 5 sample. This is relatively unimportant since the trend
in Barnett's results are of interest here. For the majority of Barnett's
tests, an average emission rate of 7 g/kgawas measured. Emissions from
the thin walled stove (without modifications) at 14 g/kg were considerably
higher than this mean value. No significant reduction in emissions was
noted for the catalytic box stove (same appliance as that tested by Oregon
DEQ ). As with most devices tested by Barnett and other investigators,
the catalytic box stove exhibited a significant increase in emissions with
a corresponding decrease in burn rate. However, the second catalytic
appliance recently tested by Barnett (catalytic appliance of Barnett's own
design) exhibited a significant (85%) reduction in measured emissions
(1 vs 7 g/kg). Furthermore, this emission rate did not vary tremendously
over the range of burn rates tested (approximately 0.8 - 2.2 g/kg). This
emission rate of 1 g/kg measured by this high volume method would (taking
the conservative approach) be equivalent to 2-5 g/kg total particulate by
EPA Method 5 with back-half analysis.
Sanborn conducted a series of tests on appliances actually installed
in homes with the purpose of estimating an emission factor from wood stoves.
Note that Barnett's results were reported on a wet basis; the results
and burn rates have been adjusted to dry basis using a factor of 0.7
(30% moisture fuel).
258
-------
The average emission rate measured by Sanborn, 27 g/kg total participate,
is very similar to the average rates measured by the Oregon DEQ, 32 g/kg,
and measured during this study, 27 g/kg (19 g/kg without fuel moisture
tests). Sanborn's data provides little valuable information, however,
in regards to evaluating improved technology appliances. It is worth
mentioning that two furnaces were tested and an average emission rate of
12 g/kg was obtained with the units operating at about 5.6 kg/hr
combustion rate (2.4 kg/hr overall combustion rate for Sanborn's study).
These results compare well with the Oregon DEQ's results for furnaces
9 g/kg at a burn rate of 9.3 kg/hr (one furnace was measured at 14 g/kg
and the other at 4 g/kg during the DEQ tests).
The emission tests summarized in Tables 50 and 51 by the other inves-
tigators (Knight, Monsanto, PEDCo, and Butcher) are not discussed in further
detail, because they do not generally include results useful for evaluating
emissions from improved technology appliances and/or because the tests
were conducted at higher burn rates (5-7 kg/hr range).
Several relatively new appliances which have not been emission
tested offer the potential to reduce emissions; several of these units are
briefly mentioned here. These happen to be appliances of which the author
is familiar and there is likely to be other improved technology units currently
available on the market with which the author is not familiar. Lack of
mention of any particular appliance is no direct reflection on the worth-
iness of a particular appliance, as a potential candidate for low emissions.
Cooke and Allen ' characterized the effect on emissions of various
combustion modification techniques. This evaluation was conducted by
259
-------
measuring changes in carbon monoxide and total gaseous hydrocarbon emis-
sions, as well as various combustion parameters (e.g., temperatures).
One of the modifications identified as resulting in reductions was
secondary combustion; however, they noted that it was extremely difficult
to attain and maintain the conditions (temperature and air) needed to
sustain secondary combustion. Nonetheless, several appliances have
recently entered the market which purport to utilize the principle of
secondary combustion to increase efficiency (and consequently, reduce
emissions). Three of these units are unique in that they utilize baffle
plates with air jets to direct the gases emitted to a secondary chamber
for combustion before exiting the appliance. One of these units, the
Brugger Ugly Duckling (from New Zealand) was considered for testing in
this study, but due to financial limitations, could not be included.
A second unit manufactured in New Zealand, the Kent Stove, also uses this
secondary combustion approach. Limited test data is available on this
unit (provided by distributor) which indicates a filterable emission rate
in the 1-5 g/kg range with the unit operating in the 2-3 kg/hr combustion
rate range. Actual test conditions and methods used were not clear at the
time of preparation of this report.
The new Jtftul Model 201 also uses this basic technique. Emissions
tests have not been conducted on this unit; however, Jtftul reports
efficiencies of 75-78% for the unit while operating at burn rates in the
0.8 - 2.0 kg/hr range. A large number of catalytic appliances have
entered the market; a recent issue of Wood'n Energy Journal lists 23
different catalytic units that are commercially available; the emissions
reduction potential of these units is unknown.
260
-------
In summary, the results of this study indicated only one unit tested --
the ceramic stove -- resulted in a significant reduction. However, due
to the high burn rate of this unit, the meaningful ness of the low emission
rate obtained (2 g/kg) is somewhat obscured. Further testing of this
unit's heat output characteristics is warranted. However, the low emission
rate obtained from this unit, as well as from one furnace tested by the
Oregon DEQ suggests that an emission level of 5 g/kg (1 g/kg filterable) is
achievable but not necessarily at the low burn rates typically used in
small home appliances. Test data from the literature reviewed indicates
one catalytic unit has been able to achieve an 85" emission reduction
from the emission levels obtained by typical appliances when operated
at low burn rates; this percentage reduction results in an emission factor
of 1-5 g/kg total particulate (depending on how one interprets the results
from the high volume emissions testing method).
Based on the data obtained during this study and on the literature
reviewed, an emission level of from 15-20 g/kg total particulate (5-10
g/kg filterable) is reasonable to expect from well operated and designed
commercial units currently available. To suggest an emission level of
5 g/kg total particulate (1 g/kg filterable) for small home appliances
operated at low burn rates is "technology forcing" at this point in time;
limited data are available to indicate that one or two units meet this
level, and that larger units operated at high burn rates meet this level.
Although no significant emissions reduction was noted for the retrofit
2 3
devices (catalytic and non-catalytic) during this study, Shelton's '
results indicate a reduction'of from 20-50% (based on creosote deposition).
261
-------
This emission reduction would not result in an emission rate near the
level expected to be achievable from improved technology stoves as dis-
cussed above (5 g/kg total particulate), but would likely result in average
emissions in the 15-20 g/kg range from existing units. Nonetheless, in
any event, any emission reduction of 20-50% for existing units is sig-
nificant.
The potential effect of emission test methods and stove operation
on emissions already has been discussed. Consequently, it should be
understood that in defining any emissions level certain parameters
regarding these two variables must be established. For this discussion
the author has basically assumed the appliance is operated at a burn.rate
of less than 3 kg/hr and that emissions are measured by EPA Method 5 with
back-half analysis. For any emissions measurement program to determine
and/or compare emissions from various appliances (whether for certification
or research purposes), stove operating procedures and test methods should
be carefully defined (particularly stove operating procedures) if the
results are to be meaningful. Development of such a standard test protocol
is well outside the scope of this project.
One final note regarding emissions levels is in order. The entire
discussion has revolved around total or filterable "particulate matter"
and innovative technology to reduce these emissions. No mention has been
made of Polycylic Organic Matter (POM) emissions. Since these emissions
are known to be hazardous, attention also should be given to determining
the level of POM emissions from appliances. Unfortunately, methods for
determining POM emission levels tend to be expensive. Nonetheless,
262
-------
certainly on a >-esearch basis, more information regarding POM emis-
sion levels should oe obtained.
263
-------
QUALITY ASSURANCE
The quality assurance program for Task 5 focused on instrument
calibration, adherence to standard test procedures, and calculations
review. Key aspects of the quality assurance program followed are
discussed below. Supporting data for instrument calibrations, etc.
are presented in Appendix C.
QUALITY ASSURANCE RECORDS
As required by the contractual agreement a quality assurance log
book was maintained by the Task Manager for recording general information
such as records of meetings, telephone conversations, etc..
In addition, two separate log books were maintained at the test site.
The Laboratory Book was maintained for the purpose of recording the wood
moisture measurement information. The Test Log Record was maintained
as a general log book for recording any information pertinent to operation
of the appliances and to performance of the emissions tests, both during
pretest trial burns and during the emissions tests. These log books
(three) are to be submitted to EPA as part of the file for this project.
264
-------
WOOD MOISTURE DETERMINATION
Wood moisture was measured by a Delmhorst moisture meter. Prior
to beginning the emission testing, the results obtained by this technique
for six different pieces of fuel were compared to the moisture content
obtained by measuring weight loss of the wood when heated to 105 C. All
tests were performed on logs split into halves. A total of 22 readings were
taken for each log at 11 different positions. At each position a reading
was made at 1/2" and 1" depths. Four positions on the split face were
tested in an evenly spaced traverse, beginning and ending 1" from the ends;
the same was done for the bark side of the long. Additional readings
were taken at each end and at a point on the split face midway between the
•
ends and 1/2" from the edge. Immediately after the moisture readings were
made, 3 one-inch thick sections were cut. One section was cut from the
center and one each cut from a point 1 1/2" from each end respectively.
Each section was sealed in a plastic bag and sent to a laboratory for
analysis by the ASTM Oven Drying Method. Figure 94 indicates the locations
*
of the moisture meter measurements (A - K) and the cross-section cuts
(A, B, C, and D). Three logs were tested on 6/30/81 ancl then three more
logs were tested on 7/07/81 to confirm results. Based on the results of
these tests, a standard measurement procedure for the emission tests was
decided upon.
The best correlation of the moisture meter with the results from the
Laboratory Oven Method was an average of readings A-D and G-J at a 1/2"
depth. For each log the values were within ±2%, except for one sample
which was ±3%. The average for all six logs, readings A-D and G-J was
265
-------
FIGURE 94
Wood Moisture Measurement
1"
A
t—
B
_j
K
,i"
C
u
D
_^,
i"
I'J—
H I
266
-------
25.5% moisture, while the average oven method for all six logs was 25.8"
moi sture.
The 1/2" depth readings give a better correlation since they detect
the high moisture content just beneath the bark. The end readings (E & F)
were consistently lower than the average. The edge reading (K) is
unnecessary. Measurements at these eight points was used for the duration
of the tests. The results for the fuel moisture quality assurance tests
are reported in Appendix C. Additional quality assurance checks were
conducted on 7/27/81, 8/07/81, 8/31/81, and 9/15/81.
Moisture meter readings for the high moisture fuel did not correlate
at all with the laboratory values obtained. This is not surprising since
the moisture meter is not expected to g'ive accurate readings (biased low)
for a moisture content greater than 30%. However, for the last set of QA
samples taken, 8/31-9/15/81, the laboratory procedure yielded higher
moisture values than the meter even when the meter measurement was less
than 30%. For the last three samples, the meter readings were 23, 25,
and 19 percent, while the laboratory values were 40, 45, and 28 percent,
respectively. On the last two samples, only a single center cross-section
was taken; this might account for part of the discrepancy. One possible
explanation is that these samples may have been taken from the second lot
of wood which was not totally cured; a portion of the laboratory weight
loss measured as moisture might have actually been volatile organics.
Nonetheless, with the exception of the high moisture values, the meter
measurements were used as the fuel moisture content.
267
-------
PARTICULATE SAMPLING
Participate sample collection was conducted according to Reference
Method 5 procedures with the exception that single point sampling was
conducted. Prior to testing, the gas and orifice meters were calibrated
according to standard procedures (calibration data in Appendix C). The
pitot tubes also were calibrated. As per Method 5, reagent and filter
blanks were taken and analyzed along with the samples.
No irregularities were encountered, with the exception of the
velocity measurements, which already have been discussed.
268
-------
GASEOUS MEASUREMENTS
Carbon monoxide, carbon dioxide, oxygen and hydrocarbon cylinder
gases were procured and utilized for calibrating all gaseous measurement
instruments. One gas cylinder for each compound (except HC) was obtained
from the EPA quality assurance laboratory in order to provide an inde-
pendent audit. The NDIR and Oxygen meter were calibrated prior to and
immediately following each days testing. A three point CO and COp cali-
bration was conducted; two point calibrations for oxygen and hydrocarbons
were conducted. The calibration gases available and the daily calibrations
followed are summarized in Table 52.
Orsat calibrations generally were conducted at only a single point.
Problems were encountered with CO and orsat measurements during the test
period. The CO adsorbing solution was changed during the test program
(7/23/81); on 9/03/81 orsat instruments were switched due to problems
with CO adsorption.
269
-------
TABLE 52
Calibration Gases
Carbon Monoxide Daily Calibration
(percent)
0 (.ambient)
0.260*
0.501 /
7.25 /
Carbon Dioxide
(percent)
3.57* /
7.47 /
13.0
Oxygen
(percent)
3.70*
20.9 (ambient) /
Hydrocarbon
0 (ambient) /
710 /
EPA certified cylinder gas
270
-------
TRANSMISSOMETER
The transmissometer system was constructed according to the
specifications for the Smoke Generator Transmissometer System of EPA
Reference Method 9 -- Visual Determination of Opacity of Emissions from
Stationary Sources. Construction details are provided in the Appendix.
Prior to testing the instrument, linearity was checked using a
three point calibration with neutral density filters. The filters used
were 195., 50.1, and 79.4 percent transmission. These filters were used
to check instrument calibration prior to each test. Calibration was not
checked after each test due to the fact that residual (low level) emissions
generally continued after the testing ended. A series of certified neutral
density filters were obtained from the Quality Assurance Division of EPA
and were used to audit the calibration. This audit indicated the instrument
was properly calibrated. The results of the initial and final calibration,
as well as the audit calibration are reported in Appendix C. The daily
calibrations are not tabulated, but are noted on the permanent strip
0
chart record.
Although the instrument indicated proper lineraity throughout the
testing, some problems with upscale zero drift were encountered. This
was indicated by the instrument not returning to baseline after the end
of the test. However, it must be noted that the transmissometer is more
sensitive than the visual observer and consequently, a visual observation
of zero might be expected to indicate at least a 5% instrument reading.
Consequently, it is probably not accurate to assume the final instrument
reading is totally attributable to drift. Nonetheless, it appears more
271
-------
drift (5C0) than should be tolerated (1-2%) was occurring. Significant
drift (+10'-) was noted for Run 2 and possibly Run 7. This normally
would be caused by a change in output of the power source, or in align-
ment problems caused by expansion and contraction of the stack due to
temperature extremes. On 8/27/81 prior to Run 12, the power source was
changed. This seems to have significantly improved the drift problem.
At the end of Run 12, the instrument and observer compare favorably
(5-10%); the next morning without recalibration, the instrument calibrated
accurately after having operated overnight (strip chart in Appendix C).
272
-------
DATA REDUCTION
All calculations were conducted on a programmable calculator. Each
program was independently checked (i.e., equations checked and at least
one run compared to value calculated without program). A random audit
of all calculations also was conducted.
273
-------
REFERENCES/BIBLIOGRAPHY
1. Residential Solid Fuels: Environmental Impacts and Solutions;
Oregon Graduate Center, 1982, Cooper, John.
a. Effects of Wood Stove Design and Operation on Condensible
Particulate Emissions; Barnett, et al, pp. 227-266.
b. The Effect of Catalytic Combustion on Creosote Reduction,
Combustion Efficiency and Pollution Abatement for Residential
Wood Heaters; Zimar, et al , pp. 924-940.
c. Thermal Performance Testing of Residential Solid Fuel Heaters;
Shelton, J., pp. 1119-1159.
d. Particulate Emission from New Low Emission Wood Stove Designs
Measured by EPA Method 5; Kowalczyk, John, et al , pp. 54-78.
e. Experimental Measurements of Emissions from Residential Wood-
Burning Stoves; Hubble, B.R., et al , pp. 79-139.
f. Characterization of Emissions from Residential Wood Combustion
Sources; Cooke, W.M., et al, pp. 139-163.
g. A Characterization of Emissions from Wood-Burning Fireplaces;
Muhlbaier, J.L., pp. 164-187.
h. Particulate Emissions from Residential Wood Combustion in
Vermont; Sanborn, C.R., Blanchet, M.A., pp. 188-198.
i. Wood Combustion Emissions at Elevated Altitudes; Peters, J.A.,
et al, pp. 199-209.
j. Measurement of Wood Heater Thermal and Emissions Performance;
Harper, J.P., and Knight, C.V., pp. 210-226.
k. Particulate Emission Factors for Small Wood and Coal Stoves;
Butcher, S.S. and Ellenkecker, M.J., pp. 289-303.
1. Chemical and Biological Characterization of Emissions from Combustion
of Wood and Wood-Chips in Small Furnaces and Stoves; Rudling,
Lars, et al, pp. 34-53.
2. "Testing Creosote - Removing Devices", Shelton, J. and Barczys, C.;
Mother Earth News, Volume 73, January/February 1982, pp. 113-116.
274
-------
3. "Testing Creosote - Removing Devices: The Results", Shelton, J.,
ana Lewis, C.; Mother Earth News, Volume 74, March/April 1982,
op. 130-123.
*
4. Wood Combustion: State-of-Knowledge Survey of Environmental, Health
and Safety Aspects; Muller Associates, prepared for the Office of
Environmental Programs, U.S. Department of Energy, October 1981.
5. "Jtftul 201: At the Leading Edge Again"; Flagher, G., Wood & Energy
Sol id Fuel Journal , February 1982.
6. Wood Stove Testing Methods and Some Preliminary Experimental Results;
Shelton, J., ASHRAE Transaction, Volume 48, Part 1, 1978.
7. The Effects of Stove Design and Control Mode on Condensible Par-
ticulate Emissions, Flue Pipe Creosote Accummulation and the
Efficiency of Wood Stoves in Homes; Barnett, S., Proceedings:
Wood Heating Alliance 1982 International Trade Show and Seminar,
Louisville, Kentucky, pp. 152-177.
8. The Effects of Fuel Moisture Content, Species, and Power Output on
Creosote Formation; Shelton, J. , Shelton Energy Research, 1981.
9. Catalytically Assisted Combustion in Residential Wood-Fueled Heating
Appliances; Shelton, J., Wood Heating Alliance 1981 International
Trade Show, New Orleans, Louisiana.
10. A Study of Wood Stove Particulate Emissions; Butcher, S.S. and
Sorenson, E.M., Journal of the Air Pollution Control Association,
Volume 29, 1979.
11. Source Assessment: Residential Combustion of Wood; DeAngelis, D.C.,
et al, prepared for U.S. Environmental Protection Agency, EPA
600/2-80-042b, March 1980.
12. Results of Laboratory Tests on Wood-Stove Emissions and Efficiency;
Hubble, B.R. and Harkness, J.B., Wood Heating Alliance 1981 Trade
Show, New Orleans, Louisiana.
13. Control of Emission from Residential Wood Burning by Combustion
Modification; Allen, J.M. and Cooke, M.W., prepared for U.S.
Environmental Protection Agency, EPA 600/7-81-091, May 1981.
14. "The Wood 'N Energy Catalytic Appliance Directory"; Wood 'N Energy
Journal. February 1982, pp. 43-47.
275
------- |