In-Situ Emission Factors for Residential Wood Combustion Units

           United States
           Environmental Protection
           Agency
          Office of Air Quality
          Planning and Standards
          Research Triangle Park NC 27711
EPA-450/3-88-013
December 1988
           Air
&EPA
In-Situ Emission Factors
for Residential Wood
Combustion Units

-------
                                           EPA-450/3-88-013
        IN-SITU EMISSION FACTORS FOR

      RESIDENTIAL WOOD COMBUSTION UNITS
         Emissions Standards Division
    U. S. Environmental  Protection Agency
         Office of Air and Radiation
Office of Air  Quality Planning and  Standards
Research Triangle Park,  North Carolina  27711
                December 1988

-------
This report has been reviewed by the Emission Standards Division of the
Office of A1r Quality Planning and Standards, EPA, and approved for
publication.  Mention of trade names or commercial products is not
intended to constitute endorsement or recommendation for use.  Copies of
this report are available through the Library Services Office (MD-35),
U. S. Environmental  Protection Agency, Research Triangle Park, North
Carolina 27711; or,  for a fee, from the National Technical Information
Services, 5285 Port  Royal Road, Springfield,  Virginia 22161.

-------
ABSTRACT

This document recommends participate emission factors (grams per hour, grams per kilogram) for
existing traditional technology woodstoves and catalytic and noncatalytic units which meet (certified) or could
meet (certifiable) the U.S. Environmental Protection Agency's (EPA) 1988 (Phase I) New Source
Performance Standards (NSPS). This information is to assist EPA regional offices and state agencies in
estimating the emission contribution from existing technology woodstoves and the airshed impact of replacing
those woodstoves with units that meet the NSPS.

The woodstove emission factors are based on field emission studies conducted during the 1985-86,
1986-87, and 1987-88 heating seasons in three localities in North America. The emission factors for the
traditional technology woodstoves are based on how those units are presently and in future years are expected
to perform. The emission factors for 1988 NSPS certified/certifiable woodstoves are based on how these units
should perform in future years.

While significant emission rate variability was demonstrated in field studies, there are several cases where
individual certified woodstoves and/or woodstove models (both catalytic and noncatalytic) have shown
relatively low emission rates and variability. These observations indicate that it is possible for new technology
woodstoves to consistently produce relatively low emissions. It is therefore expected as emission control
technology improves for NSPS-certified woodstoves, that a more consistent pattern of relatively low emission
rate performance will be demonstrated in future studies.
111

-------
                                        CONTENTS
                                                                                           Page
Disclaimer	ii
Abstract	iii
Tables	   v

I.    Introduction	   1-1

II.   Conclusions & Recommendations	II-l

in.  Technical Approach	III-l
     a.   Traditional Technology Woodstove Emission Factors	EH-1
     b.   1988 NSPS-Certified/Certifiable Woodstove Emission Factors	IH-IO

IV.  Results & Discussion	IV-1
     a.   Traditional Technology Woodstove Emission Factors	IV-1
     b.   1988 NSPS-Certified/Certifiable Woodstove Emission Factors	IV-6

V.   References	V-l
                                              IV

-------
TABLES
Page
II-l. Recommended Particulate Emission Factors for Woodstoves H-2
III-l. Woodstove Models and Number of Emission Samples Collected During Field Studies
(Traditional Technology Woodstoves) HI-2
III-2. Installation, Burn and Particulate Emission Rate Characteristics - Traditional
Technology Woodstoves HI-4
HI-3. Mean Particulate Emission and Burn Rates by Woodstove Installation - Traditional
Technology Woodstoves III-8
III-4. Mean Particulate Emission and Burn Rates by Woodstove Model - Traditional
Technology Woodstoves ffl-9
III-5. Woodstove Models and Number of Emission Samples Collected During Field Studies
(1988 NSPS Catalytic and Noncatalytic Certified or Certifiable Woodstoves) ID-12
III-6. Installation, Burn and Particulate Emission Rate Characteristics - Catalytic Woodstoves
(1988 NSPS Certified/Certifiable) HI-14
III-7. Installation, Burn and Particulate Emission Rate Characteristics - Noncatalytic
Woodstoves (1988 NSPS Certified/Certifiable) m-18
III-8. Mean Particulate Emission and Burn Rates by Woodstove Installation - Catalytic
Woodstoves (1988 NSPS Certified/Certifiable) 111-20
ni-9. Mean Particulate Emission and Burn Rates by Woodstove Model - Catalytic
Woodstoves (1988 NSPS Certified/Certifiable) m-20
111-10. Mean Particulate Emission and Burn Rates by Woodstove Installation - Noncatalytic
Woodstoves (1988 NSPS Certified/Certifiable) m-21
III-ll. Mean Particulate Emission and Burn Rates by Woodstove Model - Noncatalytic
Woodstoves (1988 NSPS Certified/Certifiable) m-21
IV-1. Data Base Sample Sizes - Traditional Technology Woodstoves IV-2
IV-2. Confidence Intervals - Traditional Technology Woodstoves IV-2
IV-3. Recommended Particulate Emission Factors - Traditional Technology Woodstoves FV-3
IV-4. Relationship Between Area-Specific Average Gram per Kilogram Particulate Emissions
and Burn Rates - Traditional Technology Woodstoves FV-3
IV-5. Burn Rate Frequency Distribution - "Northeast" IV-5
IV-6. Data Base Sample Sizes - Catalytic Woodstoves (1988 NSPS Certified/Certifiable) IV-7
IV-7. Data Base Sample Sizes - Noncatalytic Woodstoves (1988 NSPS Certified/Certifiable) .... IV-7
IV-8. Confidence Intervals - Catalytic Woodstoves (1988 NSPS Certified/Certifiable) IV-8
IV-9. Confidence Intervals-Noncatalytic Woodstoves (1988 NSPS Certified/Certifiable) IV-8
IV-10. Recommended Particulate Emission Factors - Catalytic Woodstoves (1988 NSPS
Certified/Certifiable) IV-9
IV-11. Recommended Particulate Emission Factors - Noncatalytic Woodstoves (1988 NSPS
Certified/Certifiable) IV-9
IV-12. Comparison of 1988 (Phase I) NSPS Emission Limits with Recommended 1988 NSPS
Emission Factors IV-9

-------
                                     I. INTRODUCTION

     The Environmental  Protection Agency (EPA) has promulgated New Source  Performance Standards
(NSPS) for residential  woodstoves.1   The purpose  of this report is to determine  appropriate  participate
emission factors  (gram per hour  [g/hr]  and gram per kilogram  [g/kg]) for existing traditional  technology
woodstoves  and those units which meet  or  could meet NSPS  requirements for catalytic and noncatalytic
stoves.  This  information is  to assist  EPA regional offices  and state agencies in estimating the  emission
contribution from existing technology woodstoves and the airshed impact of replacing those woodstoves with
units that meet the NSPS.

     The NSPS includes emission  limits for new woodstoves based upon testing conducted under laboratory
conditions.  Recent  in-home studies2'3'4-5'6'7 of woodstove participate emission  performance have produced
data which suggest  that woodstoves perform  differently under actual in-the-home operating conditions than
under  laboratory test conditions.   In  addition, the in-home  emission  rate data  collected to  date have
demonstrated  significant particulate  emission  rate (g/hr and g/kg) variability within woodstove technology
categories (traditional and NSPS "certified/certifiable" woodstoves [catalytic and non-catalytic]), and within
specific woodstove models.  The term  "NSPS certified/certifiable" refers  to woodstoves that are  (1) 1988
Oregon certified woodstoves that have been "grandfathered" under  Phase I NSPS  provisions and  (2) 1988
Oregon certified woodstoves  that could  have been  "grandfathered"  under the NSPS1  There  were  no
woodstoves used in any of the  field studies that were directly certified under the EPA 1988 (Phase I) or 1990
(Phase  II) New  Source Performance  Standards (NSPS).   Therefore,  the 1988 NSPS  certified/certifiable
woodstoves used in  the  field studies represent  a "first generation" of woodstove emission control technology
which  appears to be a contributing factor  to the observed in-situ  emission rate variability.   Woodstoves
"grandfathered"  under 1988  NSPS provisions  are  not subject  to  the  more  stringent direct-certification
requirements of the NSPS.

     The observed variability in measured "real world" emission rates for woodstoves has been attributed to
the following factors:

     •    Differences in fueling patterns (e.g., frequency, fuel load density) between woodstove operators;
     •    Differences in woodstove installations and chimney systems;
     •    Differences in burn rates between woodstoves;
     •    Differences in fuel characteristics in each home, or by region;
     •    Differences in how by-pass damper(s) and/or air draft settings are used; and

                                                 1-1

-------
     •     Differences in the integrity of key components of the emission control system  of each woodstove
           over time, e.g., catalysts, by-pass damper gaskets, fueling or ash clean-out door seals, etc.

While significant emission rate variability has been demonstrated, there are several cases where individual
certified woodstoves and/or  woodstove  models  (both  catalytic and noncatalytic) have shown  relatively low
emission rates and variability. These observations indicate that it is possible for new technology woodstoves to
consistently produce relatively low emissions.  It is therefore expected as emission control technology improves
for NSPS  direct-certified woodstoves,  that a  more consistent  pattern of  relatively  low  emission rate
performance will be demonstrated in future studies.
                                                  1-2

-------
                   H. CONCLUSIONS & RECOMMENDATIONS

     The recommended  participate  emission  factors (g/hr, g/kg) are  based on  in-situ  emission  studies
conducted during  the  1985-86, 1986-87, and 1987-88 heating seasons2-3'4-5'6'7.  Emission factors  have been
developed for the following three catagories of woodstoves:

     •    Traditional Technology (pre-NSPS) woodstoves;
     •    1988 NSPS Certified/Certifiable Catalytic woodstoves; and
     •    1988 NSPS Certified/Certifiable Noncatalytic woodstoves.

     The term  "NSPS Certified/Certifiable"  refers to woodstoves  that  are:   (1)  1988  Oregon Certified
woodstoves that  have been "grandfathered" under Phase I NSPS provisions,  and (2) 1988 Oregon Certified
woodstoves that  could have been "grandfathered" under the NSPS1. Since there were no 1988 or 1990 NSPS
direct-certified woodstoves in any of the in-situ studies, there are no recommended emission factors for these
technology categories.

     The traditional technology woodstove emission  factors  were based on judgements which considered
differences in woodstove  operation, installation, and fueling patterns  in different  geographic  areas.  The
recommended emission factors for this woodstove technology category are based on how these units presently
and should continue to perform.

     The recommended  participate   emission  factors  (g/hr,  g/kg)  for  1988  NSPS  certified/certifiable
woodstoves (catalytic and  noncatalytic) are based on how these units should  perform in future years.  This
approach is based  on observations that:  (1) several of the woodstove units demonstrating elevated emission
rates could be considered  "first generation" technology which is expected  to improve over time, and (2) there
was inferential evidence that  new technology woodstoves properly installed, operated, and maintained could
produce consistently low emissions.

     The recommended emission factors are also based on the assumption that woodstove emission control,
installation, and  maintenance  problems identified  in  field studies have  been or will be addressed by the
woodstove industry and/or regulatory agencies.

     The recommended particulate emission factors (g/hr, g/kg) are presented in Table II-l.  Also included in
this table are several  related statistical measures which are based on the data used to derive  the emission
                                                 II-l

-------
                                          Table 11-1

                   Recommended Participate Emission Factors for Woodstoves

Emission Rate
Standard Deviation
95% Confidence Interval
99% Confidence Interval
Traditional
Technology
Woodstoves
g/hr
21.3
7.4
21.3 ± 2.8
21.3 ± 3.7
g/kg
15.2 *
4.7
15.2 ± 1.8
15.2 ± 2.4
Catalytic Wood-
stoves (1988 NSPS
Certified/Certifiable)
g/hr
6.2
2.5
6.2 ± 2.2
6.2 ± 2.8
g/kg
6.6
2.2
6.6 ± 1.9
6.6 ± 2.5
Noncatalytic Wood-
stoves (1988 NSPS
Certified/Certifiable)
g/hr
9.2
5.5
9.2 ± 3.8
9.2 ± 5.0
g/kg
9.6
5.5
9.6 ± 3.8
9.6 ± 5.0
* For Traditional Technology woodstoves only, the g/kg emission factor may be calculated using the
  following equation:

  y = -9.77 x + 29.2,

  where

    x = Area-specific average burn rate, dry kg/hr; and

    y = Emission rate, g/kg.
                                            H-2

-------
factors.  Since the data bases upon which these emission factors are based are relatively small, the user should
apply the appropriate statistical caveats, e.g., confidence limits, when using the recommended emission factors
on either an absolute or relative comparison basis.

     The equation shown below Table II-l  can be used  to calculate  a gram-per-kilogram  (g/kg) emission
factor  for traditional technology woodstoves based  on the  average burn  rate  (dry kg/hr) in  a specific
geographic area if burn rate information is available.  At this time there are no recommendations regarding
adjusting  either  the   gram-per-hour  or   gram-per-kilogram  emission   factors   for   the   1988   NSPS
certified/certifiable woodstoves due to the limited amount ofin-situ emission and burn rate data available.

     Since the data  bases used to determine the recommended  emission factors are relatively small, it is
recommended  that  additional  in-situ  data  for traditional technology  and   NSPS-certified  woodstoves be
collected to reduce the associated statistical uncertainty. Additional emission  and burn rate data  may provide
the basis upon which area-specific emission factors for NSPS-certified woodstoves could be recommended.
                                                  n-s

-------
                                 . TECHNICAL APPROACH

A.  TRADITIONAL TECHNOLOGY WOODSTOVE EMISSION FACTORS
     There have been four studies2'3-4-6 conducted during the 1985-86 and  1986-87  heating seasons which
evaluated  the in-the-home (in-situ) particulate emission performance of traditional technology or pre-NSPS
woodstoves.  These studies were conducted in three regions of North America:  Northeastern United States as
represented by limited areas of upstate New York  and  Vermont2, the Portland, Oregon metropolitan area3-4,
and the City of Whitehorse, Yukon Territory, Canada.6

     All four studies including two other studies which only evaluated the in-situ emission performance of new
technology woodstoves5'7, used the same type of sampling equipment, the Automated Woodstove Emission
Sampler (AWES).  The operational and design characteristics of the AWES have been described in  detail in
several  reports2'3'4-6'8-9.   This emission sampler  was  designed  specifically  to measure residential wood
combustion (RWC)  particulate emissions under in-situ operating conditions. A recent study determined the
upper-limit absolute difference of emission rates between AWES and EPA reference  sampling Methods 5G
and  5H was 2.0  g/hr4-  However, this study concluded that when considering the sampling accuracies
associated with  Methods  5G,  5H, and  the AWES, there were no  statistically  significant differences in  the
emission rates between the three sampling methods.

     Data used to  produce previous emission factors for woodstoves  were  based primarily on laboratory
tests10.  These tests employed  a wide variety of sampling methods and fueling  techniques which preceded the
adoption of  uniform fueling and sampling method protocols as  developed by the Oregon Department of
Environmental Quality11 and the U.S. Environmental Protection Agency (EPA)1.   It is therefore difficult to
make any direct  comparisons between  the  in-situ data determined by  the AWES system  and  previous
laboratory studies. Since the AWES emission rates  are:  (1) comparable with emission  rates as determined by
EPA reference woodstove particulate sampling methods; and  (2) based on  "real world" operating conditions,
only  AWES-determined particulate emission rate (g/hr and g/kg) data are considered in the determination of
recommended emission factors for traditional technology and NSPS-certified/certifiable woodstoves.

     Presented  in Table in-1  is  a list of woodstove  models and in-situ emission  samples  collected  per
woodstove model for each region.  Since hundreds  of models of traditional technology woodstoves have been
produced over the last several decades,12 the issue of representativeness should be addressed initially.   In
reviewing  the  study designs  for  each of  the  completed  woodstove  field  studies, it appears  only  the
                                               m-i

-------
                  Table 111-1

    Woodstove Models and Number of Emission
      Samples Collected During Field Studies
       (Traditional Technology Woodstoves)
Study Area
"Northeast"2






"Northwest"3-4




"Whrtehorse"6









Stove Model
T-A
T-B
T-C
T-G
T-H
TH

T-B
T-D
T-vJ
T-K

T-B
T-E
T-F
T-L
T-M
T-N
T-0
T-P
T-Q

Emission samples
4
1
3
3
2
1
14
5
5
1
1
12
4
4
4
4
6
6
3
10
20
61
Total Woodstove Models Evaluated     17
Total Emission Samples              87
                    m-2

-------
"Whitehorse" study included woodstoves in direct proportion to the types of traditional technology woodstove
models installed throughout the City of Whitehorse. This was a relatively easy task, since one woodstove
model, T-Q, has been installed in approximately 40 percent of the homes in the Whitehorse area13. For larger
metropolitan areas, or areas with a more diverse population of installed models, the selection of homes based
on the type of woodstove models installed becomes a much more difficult task if resources only allow a limited
number of homes to be included in a particular study. Since the number of woodstove models included in the
"Northeast" and "Northwest" studies is limited, the emission data generated may not be representative of the
full range or average of expected emission rates for these areas.

Table III-2 presents the in-situ emission rate data initially considered in the development of emission
factors for traditional technology woodstoves. Also included in this table is information regarding the type of
chimney system attached to each woodstove and average burn rate during each emission sampling period. A
review of woodstove use and other information associated with each paniculate emission sample led to a
decision not to include eight emission samples from the Whitehorse Efficient Woodheat Demonstration
Study6 (values deleted in parentheses). Seven of the deleted emission samples exceeded the quality assurance
(QA) objective that the ambient oxygen value should be within 20.9±2.0% absolute at the end-of-sampling-
period calibration. Calculated emissions (g/hr, g/kg) have a potentially higher degree of associated uncertainty
when this QA criteria is exceeded. One additional sample was deleted due to the level of uncertainty
associated with the emission rate calculation. This calculation was based on an assumed upper-limit value of
sample collected in the methanol solvent rinse. Since average emission rates reported in both the "Northeast
Cooperative Woodstove Study"2 and the "Northwest" studies3'4 did not include individual emission samples
exceeding these quality assurance objectives in the calculation of mean emission rates, the eight data sets are
deleted from the mean emission rates shown in subsequent tables in this report.

Since significant intra- and inter-stove model emission rate variabilities are shown in the results from all
three study areas, there appeared to be no justifiable reason to delete other data. There is no reason to believe
the range of emission rates shown in Table III-2 is not to be expected for traditional technology woodstoves.
This variability is demonstrated in the data in Tables III-3 and m-4. Table III-3 summarizes mean emission
rates (g/hr and g/kg) and burn rates (dry kg/hr) by each home (installation) in the three study areas. Table
III-4 summarizes mean emission and burn rates by woodstove model. Inter-home emission rate variability is
most clearly demonstrated by the woodstove model T-Q emission rates. The range of average emission rates
for woodstove model T-Q are 16.8 to 353 g/hr and 7.4 to 23.1 g/kg (Table m-3) with average emission rates of
23.5 g/hr and 16.7 g/kg (Table III-4). The percentage differences in gram-per-hour and gram-per-kilogram
emission rates between the lowest (Home Wll) and highest (Home W01) average values were 110% and
ni-3

-------
                        Table 111-2

Installation, Burn and Paniculate Emission Rate Characteristics
            Traditional Technology Woodstoves
Stove
Model
T-A



T-B







T-C


T-D




T-E


T-F


Study
Area
"Northeast"



"Northeast"
"Northwest"




"Whitehorse"



"Northeast"


"Northwest"




"Whitehorse"


"Whitehorse"


Home
Code
V06



V09
P03




W10



V14


P05




W03


W12


Chimney
Type
II



II
II




V



IV


I




VI


VI


Sample
Period
1
2
5
6
1
1
2
3
4
5
1
2
3
4
1
2
3
1
2
3
4
5
1
2
3
4
1
2
3
4
Burn Rate
dry kg/hr
2.45
1.60
1.52
1.86
1.12
1.68
1.26
1.02
0.94
1.25
1.28
1.15
1.44
1.49
1.67
1.45
0.92
1.37
1.12
0.94
1.01
0.92
1.18
1.04
1.31
1.44
1.76
1.44
1.49
1.82
Emissions
g/hr
2.9
4.7
12.7
17.3
15.4
22.3
9.7
10.8
9.5
16.7
28.6
16.5
22.5
24.6
16.9
23.5
20.3
29.4
28.7
21.4
27.1
20.9
5.7
8.6
(13.3)
10.1
26.7
20.3
(24.7)
29.4
g/kg
1.2
2.9
8.4
9.3
13.7
13.3
7.7
10.5
10.1
13.4
22.4
14.4
15.6
16.5
10.2
16.3
22.0
21.4
25.6
22.9
26.7
22.7
4.8
8.3
(10.2)
7.0
15.1
14.1
(16.6)
16.2
Ref.
2



2
4




6



2


4




6


6


Comments
a
a















b






c



                                                                 (continued)
                           III-4

-------
Table 111-2
(continued)
Stove
Model
T-G


T-H

T-l
•KJ
T-K
T-L



T-M




T-N




T-O


Study
Area
"Northeast"


"Northeast"

"Northeast"
"Northwest"
"Northwest"
"Whitehorse"



"Whitehorse"




"Whitehorse"




"Whitehorse"


Home
Code
N08


N14

N16
O05
O20
W05



W04



W13

W07


W09


W06


Chimney
Type
IV


IV

IV
V
V
VI



VI



V

VI


V


VI


Sample
Period
4
5
6
6
7
1
1
1
1
2
3
4
1
2
3
4
2
3
2
3
4
1
2
3
1
3
4
Burn Rate
dry kg/hr
1.91
2.19
2.00
2.45
1.57
1.55
1.41
1.08
1.67
1.59
1.64
1.73
1.90
1.25
2.01
1.87
1.85
1.74
1.88
2.12
2.57
0.93
0.70
0.88
1.22
1.26
1.13
Emissions
g/hr
32.6
26.6
30.9
34.0
29.0
13.9
23.1
18.8
10.9
14.1
13.4
18.7
40.7
29.2
35.7
26.3
33.0
26.7
32.8
21.5
(35.0)
18.2
17.7
11.3
19.2
18.8
15.7
g/kg
17.1
12.2
15.4
13.9
18.4
9.0
16.4
17.5
6.5
8.9
8.1
10.8
21.4
23.3
17.8
14.1
17.9
15.4
17.5
10.1
(13.6)
19.5
25.4
12.8
16.5
15.3
12.0
Ref.
2


2

2
3
3
6



6



6

6


6


6


Comments
d






e




f



f



c






                                          (continued)
   III-5

-------
Table 111-2
(continued)
Stove
Model
T-P






T-Q












Study
Area
"Whitehorse"






"Whrtehorse"












All Areas
Home
Code
W06


W13




W01


W02




W08

W11



W14


Chimney
Type
VI


V




V







VI

VI



V


Sample
Period
5
6
7
8
9
5
6
7
8
9
1
2
4
1
2
3
4
8
9
2
3
4
1
2
3
4
1
2
3
4
Mean (edited data)
^n-l
Burn Rate
dry kg/hr
1.11
0.77
1.14
0.66
0.61
1.29
1.35
1.19
0.89
0.89
.93
.32
.26
.45
.29
.46
0.90
0.75
0.56
1.65
1.50
1.30
2.37
2.02
2.02
2.60
1.16
1.07
1.04
1.71
1.44
0.45
Emissions
g/hr
(19.1)
(20.9)
16.1
12.6
10.0
33.2
16.4
22.9
17.5
17.8
49.8
30.4
25.8
28.9
17.2
17.3
(11.5)
17.7
13.1
22.5
(31.3)
31.7
10.5
14.1
23.6
18.9
24.7
25.7
(32.6)
28.0
21.0
8.6
g/kg
(17.2)
(27.4)
14.1
18.9
16.5
25.7
12.2
19.3
19.6
20.0
25.8
23.1
20.5
20.0
13.3
11.8
(12.8)
23.6
23.3
15.0
(24.1)
16.7
4.4
7.0
11.7
7.2
21.4
23.9
(31-4)
16.4
15.4
5.9
Ref.
6


6




6


6




6

6



6




Comments
c
c













9


c






c


                                          (continued)
   III-6

-------
Table 111-2
(continued)
Chimney Types
Code
Description
I
II
III
IV
V
VI
Unlined masonry chimney located primarily inside the exterior walls of the house
Tile-lined masonry chimney located primarily inside the exterior walls of the house
Unlined masonry chimney located primarily outside the exterior walls of the house
Tile-lined masonry chimney located primarily outside the exterior walls of the house
Prefabricated metal chimney located primarily inside the exterior walls of the house
Prefabricated metal chimney located primarily outside the exterior walls of the house
Comments:

a. It was observed that the users of this stove maintained a deep ash bed (15 to 25
cm [6 to 9 in.]) when operating this stove. This ash bed condition effectively
reduced the firebox size of the stove from 84 liters (2.95 cubic feet)'to 44 liters
(1.6 cubic feet) and may have been a contributing factor to this relatively low
emission rate.

b. This stove model was an insert mounted into an unlined masonry chimney.

c. Ambient oxygen (O^ value associated with emission sample greater than 20.9 ±
2.0 percent absolute at final calibration. Therefore, calculated emissions (g/hr,
g/kg) have a potentially higher degree of associated uncertainty. Because of the
higher uncertainty associated with this emission sample, it is deleted from the
data base for calculating emission factors.

d. Operators of this stove kept firebox relatively full of fuel and the primary air
inlets at a low setting. This operating practice may have contributed to the
relatively high observed emission rates.

e. While stove T-K had a relatively large firebox of 148 liters (5.2 cubic feet), the
operator of this stove fueled this stove with small loads, i.e. average fuel load
density of 12 kg/m3 (0.73 Ib/ft3).

f. Woodstove model T-M is a "cabinet style" convection heater with an underfire
primary air combustion system. This type of combustion system tends to produce
elevated emissions in wood-fired stoves.

g. Estimated upper-limit emission rate due to loss of methanol solvent rinse. These
emission rates were eliminated from the data base to calculate emission factors
due to the associated uncertainty.
III-7

-------
                          Table 111-3

Mean Particulate Emission and Burn Rates by Wood stove Installation
               Traditional Technology Woodstoves


Study Area
"Northeast"





Avg (unweighted)
Range
"Northwest"



Avg (unweighted)
Range
"Whitehorse"















Avg (unweighted)
Range
ALL AREAS
Avg (unweighted)
Range


Model
T-A
T-B
T-C
T-G
T-H
T-l


T-B
T-D
T-J
T-K


T-B
T-E
T-F
T-L
T-M
T-M
T-N
T-N
T-O
T-P
T-P
T-Q
T-Q
T-Q
T-Q
T-Q






Home
V06
V09
V14
N08
N14
N16


P03
P05
O05
O20


W10
W03
W12
W05
W04
W13
W07
W09
W06
W06
W13
W01
W02
W08
W11
W14




Mean Emission Rates


9/hr (er,^)
9.4 (5.9)
15.4 (0)
20.2 (2.7)
30.0 (2.5)
31.5 (2.5)
13.9 (0)
20.1 (8.9)
9.4-30.0
13.8 (5.6)
25.5 (4.1)
23.1 (0)
18.8 (0)
20.3 (5.1)
13.8-25.5
23.1 (5.0)
8.1 (2.2)
25.5 (4.7)
14.3 (5.0)
33.0 (6.5)
29.9 (4.5)
27.2 (8.0)
15.7 (3.8)
17.9 (1.9)
12.9 (3.1)
21.6 (7.0)
35.3 (12.7)
18.8 (5.9)
27.1 (6.5)
16.8 (5.7)
26.1 (1.7)
22.1 (7.6)
8.1 -33.0

21.3 (7.4)
8.1 -33.0

g/kg («rn-i)
5.5 (3.5)
13.7 (0)
16.2 (4.8)
14.9 (2.0)
16.2 (2.3)
9.0 (0)
12.6 (4.4)
5.5-16.2
11.0 (2.4)
23.9 (2.2)
16.4 (0)
17.5 (0)
17.2 (5.3)
11.0-23.9
17.2 (3.6)
7.6 (2.3)
15.1 (1.05)
8.6 (1.8)
19.2 (4.1)
16.6 (1.8)
13.8 (5.2)
19.2 (6.3)
14.6 (2.3)
16.5 (2.4)
19.4 (4.8)
23.1 (2.6)
18.4 (5.5)
15.8 (1.2)
7.4 (2.7)
20.5 (3.8)
15.8 (4.6)
7.4-23.1

15.2 (4.7)
5.5-23.9

Mean Burn Rate
dry kg/hr (a^)
1.80 (0.34)
1.12 (0)
1.35 (0.31)
2.01 (0.11)
2.06 (0.37)
1.55 (0)
1.65 (0.37)
1.12- 2.06
1.23 (0.29)
1.07 (0.18)
1.41 (0)
1.08 (0)
1.19 (0.16)
1.07- 1.41
1.34 (0.16)
1.24 (0.17)
1.67 (0.20)
1.66 (0.06)
1.76 (0.34)
1.80 (0.08)
2.00 (0.17)
0.99 (0.33)
1.19 (0.11)
0.80 (0.29)
1.12 (0.22)
1.50 (0.37)
1.10 (0.41)
1.46 (0.25)
2.25 (0.28)
1.31 (0.35)
1.45 (0.39)
0.80- 2.25

1.46 (0.37)
0.80- 2.25
Emission
Samples
(N)
4
1
3
3
2
1

14
5
5
1
1

12
4
3
3
4
4
2
2
3
3
3
5
3
5
2
4
3

53

79
                             III-8

-------
Table 111-4

Mean Paniculate Emission and Burn Rates by Woodstove Model
Traditional Technology Woodstoves

Stove
Model
T-A
T-B
T-C
T-D
T-E
T-F
T-G
T-H
T-l
T-J
T-K
T-L
T-M
T-N
T-O
T-P
T-Q

Study
Area(s)
NE
NE.NW.WH
NE
NW
WH
WH
NE
NE
NE
NW
NW
WH
WH
WH
WH
WH
WH

Homes(s)
V06
V09, P03, W10
V14
P05
W03
W12
N08
N14
N16
005
020
W05
W04, W13
W07, W09
W06
W06, W13
W01, W02, W08
W11.W14
Average (unweighted)
Range
Mean Emission Rates

g/hr (ff^)
9.4 (5.9)
17.6 (6.6)
20.2 (2.7)
25.5 (4.1)
8.1 (2.2)
25.5 (4.7)
30.0 (2.5)
31.5 (2.5)
13.9 (0)
23.1 (0)
18.8 (0)
14.3 (5.0)
31.9 (5.6)
18.3 (9.6)
17.9 (1.9)
18.3 (7.1)
23.5 (9.2)
20.4 (7.0)
8.1 -31.9

g/kg (
-------
212%, respectively. Since woodstove model T-Q emission rates are based on measurements from five homes
(17 samples), woodstove model-specific average emission rates based on limited observations, i.e. one home,
probably have a high degree of uncertainty. Emission rate variability is also demonstrated by woodstove
models T-B, T-N, and T-P. The ranges of average gram-per-hour emission rates for woodstove models T-B,
T-N, and T-P are 13.8 to 23.1 g/hr, 15.7 to 27.2 g/hr, and 12.9 to 21.6 g/hr, respectively (refer to Table III-3).

From a statistical viewpoint, the sample sizes of the "Northeast" and "Northwest" studies appear to be
too small to adequately quantify the mean emission rates for traditional technology woodstoves in these areas.
As a practical upper limit, the limit (L) around the sample mean should not be greater than 3.0 units at the
99 percent confidence limit for gram-per-hour emission rate data. This value of "L" was selected because of
the estimated 2.0 g/hr absolute upper-limit accuracy of the AWES4. The 95 percent and 99 percent
confidence limits can be expressed by the following equations:

L = 1.96 cr/N*4 (95% Confidence Limit) (Equation 1)
L = 2S7ff/tf* (99% Confidence Limit) (Equation 2)
where: a- = assumed population standard deviation; and
N = sample size.

The minimum number of samples (N) needed can be determined by assuming a representative standard
deviation (
-------
under the direct certification provisions of the NSPS, whereas it is allowed under the Oregon certification
rules. Neither 1988 nor 1990 direct-certified NSPS certified woodstoves (either catalytic or noncatalytic) were
included in any of the woodstove field studies, and therefore recommended emission factors are not developed
for these woodstove categories. Table III-5 lists the catalytic and noncatalytic woodstove models that were
evaluated in field studies and were either certified or certifiable under the 1988 NSPS emission limits. Also
listed is the number of particulate emission samples collected per woodstove model.

Field tests of various models of 1988 NSPS certified/certifiable woodstoves demonstrated significant
emission rate variability both between homes with the same woodstove model and between different
woodstove models. Field tests of several models of woodstoves conducted during the 1985 through 1988
heating seasons have shown that in-situ emission rates of NSPS-certified woodstoves are generally higher (in
the range of 1 to 20 times higher) than laboratory certification levels.

The recommended particulate emission factors (g/hr, g/kg) for 1988 NSPS certified/certifiable
woodstoves (catalytic and noncatalytic) are based on how these units should perform in future years. This
approach to evaluating emission data is based on observations that: (1) several of the woodstove units
demonstrating elevated emission rates could be considered "first generation" technology, and (2) there was
inferential evidence that new technology woodstoves properly installed, operated, and maintained could
consistently produce low emissions. Therefore, the following criteria were used in determining whether or not
individual emission samples should be included in the data bases to determine emission factors for catalytic
and non-catalytic woodstoves:

1. Emission rate data which does not meet quality assurance objectives is deleted.
2. Emission rate data which are representative of a specific combination of catalytic woodstove model
and a catalytic combustor which is no longer available to the consumer as of July 1, 1988 are
deleted.
3. Emission rate data which appear to be anomalous, e.g., outlier, and appear to have resulted from
poor operator practices that are correctable through an improved public education program are
deleted. It is assumed that the woodstove industry, e.g. manufacturers, retailers, chimney
sweeps/installers, etc., and/or regulatory agencies will develop and provide improved public
education programs to address proper woodstove installation, maintenance, and operating
practices.
4. Emission rate data which appears to be anomalous and to have been influenced by an incompatible
chimney system installation is deleted. (Note: There is presently only inferential evidence that
ra-ii
-------
Table 111-5

Woodstove Models and Number of Emission
Samples Collected During Field Studies
(1988 NSPS Catalytic and N on catalytic
Certified or Certifiable Woodstoves)
Study Area
"Northeast"2
"Northwest"4-5
"Whitehorse"6
Technology
Catalytic
Noncatalytic
Catalytic
Noncatalytic
Catalytic
Noncatalytic
Stove Model
C-A
C-D
NC-A
NOB
C-A
C-B
NC-B
NC-C
C-A
C-C
NC-0
Samples
15
1
4
7
5
3
5
4
9
10
10
NSPS Status
Certified
Certified
Certified
Certified
Certified
Certified
Certified
Certifiable
Certified
Certifiable
Certified
Total Woodstove Models Evaluated
Total Emission Samples
Catalytic 4
Noncatalytic 4

Catalytic 43
Noncatalytic 30
m-12
-------
"oversized" or "undersized" chimney systems may influence NSPS-certified woodstoves emission
performance. Additional research is needed to confirm if there is a "cause and effect" relationship
between emission rates and chimney system configuration as installed with NSPS-certified
woodstoves. If future studies confirm that improperly designed chimney systems adversely affect
the emission performance of NSPS-certified woodstoves, regulatory agencies may modify
certification programs to address this issue.)

In selecting this criteria, it is assumed that woodstove emission control, installation, and maintenance
problems identified in field studies have been or will be addressed by the woodstove industry and/or regulatory
agencies.

The above criteria are used to delete all values hi parentheses, ( ), in Tables Ul-6 and III-7 from the data
bases used to determine emission factors for 1988 NSPS-certified/certifiable woodstoves. Tables III-8 and
III-9 show mean emission rates (g/hr and g/kg) and burn rates (dry kg/hr) for each installation and for each
woodstove model for catalytic technology woodstoves (1988 NSPS certified/certifiable). Tables 111-10 and
III-ll show mean emission and burn rates (by installation and by woodstove model) for noncatalytic
woodstoves (1988 NSPS certified/certifiable).
ffl-13
-------
Table 111-6

installation, Burn and Paniculate Emission Rate Characteristics
Catalytic Wood stoves (1988 NSPS Certified/Certifiable)

Stove
Model
C-A

Study
Area
"Northeast"

"Northwest"

"Whftehorse"

Home
Code
V05

V11

N09

N18

C01

P02

W07

W11

Chim-
ney
Type
IV

Catalytic
Combustor
Model
A

Sample
Period
4
5
2
6
7
1
4
6
7
4
5
6
7
1
2
1
2
3
4
5
5
6
8
9
5
6
7
8
9

Burn Rate
dry kg/hr
0.84
0.92
1.02
1.19
1.15
1.23
.31
.23
.37
.37
.57
.19
1.40
0.89
0.84
1.07
0.95
0.87
0.79
0.70
1.72
1.62
1.85
1.48
1.84
1.91
1.64
1.48
1.69
Emissions

9/hr
(9-0)
(31.4)
(6.1)
(6.3)
(7.0)
(15.7)
(21.2)
(17.1)
(29.6)
(20.6)
(41.3)
(31.6)
(29.2)
4.7
5.1
2.7
4.2
4.7
4.3
4.3
(7.4)
(7.5)
(21.0)
(25.4)
(13.1)
(22.0)
(11-3)
(12.0)
(19.4)

g/kg
(10.8)
(34.1)
(6.0)
(5.3)
(6.1)
(12.8)
(16.2)
(13.9)
(21.5)
(15.1)
(26.4)
(26.5)
(20.8)
5.3
6.1
2.5
4.4
5.4
5.4
6.3
(4-3)
(4-6)
(11.4)
(17.2)
(7.1)
(11.5)
(6-9)
(8.1)
(11.5)

Ref.
2

5
6

Comments
a
a
a,b
a,b
a,b
a
a
a
a
a
a
a
a

f
a
a
a
a,c
a
a
a
a
a,c,d
(continued)
111-14
-------
Table 111-6
(continued)
Stove
Model
C-B
C-C

C-D
Study
Area
"Northwest"
"Whrtehorse"

"Northeast"
All Areas
(edited data)
Home
Code
P06
W03

W14

COS
Chim-
ney
Type
V
VI

VII
Catalytic
Combustor
Model
C
D

E
Sample
Period
1
4
5
5
6
7
8
9
5
6
7
8
9
3
Mean
ffn-l
Bum Rate
dry kg/hr
1.83
1.61
1.10
0.99
1.05
1.27
1.11
0.86
1.19
0.86
1.09
1.07
0.86
0.92
0.96
0.14
Emissions
g/hr
(61.9)
(31.2)
(36.8)
8.6
10.5
6.2
9.0
11.9
6.2
10.0
4.3
15.4
8.2
4.3
6.9
3.4
g/kg
(33.8)
(19.4)
(33.0)
8.7
9.9
4.9
8.1
13.9
5.2
11.7
3.9
14.3
9.6
4.7
7.3
3.4
Ref.
4
6

7
Comments
e
e
e

Chimney Types
Code
Description
IV
V
VI
VII
Unllned masonry chimney located primarily inside the exterior walls of the house
Tile-lined masonry chimney located primarily inside the exterior walls of the house
Unlined masonry chimney located primarily outside the exterior walls of the house
Tile-lined masonry chimney located primarily outside the exterior walls of the house
Prefabricated metal chimney located primarily inside the exterior walls of the house
Prefabricated metal chimney located primarily outside the exterior walls of the house
Stainless steel liner (8" diameter) inside masonry chimney primarily located inside the
exterior walls of the house.
(continued)
111-15
-------
Table 111-6
(continued)
Comments:

a. The combination of catalytic wood stove model "C-A" with catalytic combustor model
"A" is no longer available to the consumer as of July 1, 1988. Woodstove model "C-A"
is only available with combustor "B." For this reason, these emission rate values are
deleted from the data base used to develop emission factors.

b. Home V11 had the highest overall mean fuel load (12.7 kg) and the lowest overall mean
loading frequency (0.09 Ib/hr) of all homes, with stove model C-A in the "Northeast"
Study. This type of fuel loading practice may have contributed to the observed
relatively low emission rates.

c. An inspection of the emission control system of the woodstove in this home conducted
in August 1987 revealed: loose bypass damper gasket; weld gaps on the bypass damper
door; bi-metallic coil tension set-point on the thermostat was improperly set; and the
gasket around the catalyst had gaps. Using EPA Methods 28 and 5G, it was
determined the catalyst in this stove had prematurely failed. However, it could not be
determined if the catalytic combustor emission reduction performance failure was due
to inherent material defects in the catalyst and/or due to the stove installation and/or
woodstove manufacturing problems.

d. An inspection of the emission control system in this stove in August 1987 revealed that
the catalytic combustor was not properly seated. This improper seating may have
contributed to the seesaw emission rates values observed. The improper seating of the
catalyst was probably due to improper installation procedures.

e. An inspection of the emission control system of this stove conducted after the
collection of emission samples revealed the following:

1. Approximately 22.9 cm (9 inches) of the glass fiber gasket that seals the catalyst
bypass damper was missing on the bottom edge of the damper. The gasket is
normally fitted into a groove located on the damper door. During fuel loading
events, the bypass damper door is opened, exposing the gasket. It is hypothesized
that the missing gasket may have been caused by fuel, e.g., logs, 2 by 4s, etc.,
contacting the gasket, since the missing gasket material was in the area most
exposed to the type of potential physical abrasion. Since this observation is based
on one stove, it is difficult to determine if the problem is a stove design and/or
operator related issue.

2. A portion of the missing bypass damper gasket had fallen into the hinge of the
bypass door, which did not allow the bypass door to seal properly. A 0.6 cm (V4
inch) gap between the bypass door and its jamb was formed when the missing
gasket acted as a wedge. There was no detectable tactile indication of whether the
bypass door was fully closed or partially open when operating the damper lever.
The lack of tactile "feedback" to determine if the bypass damper door is fully
closed appears to be a design problem.

(continued)
111-16
-------
Table 111-6
(continued)

3. Approximately one-third of the catalyst cells were plugged with ash. The
"upstream" or combustion gas inlet side of the catalyst is the top horizontal side of
the catalyst. This configuration can lead to a buildup of ash on the top surface of
the catalyst versus the more common configuration where the "upstream" side of
the catalyst is either the bottom horizontal or side face of the catalyst. Operator
practices could also have contributed to the buildup of ash if an excessive amount
of paper was used as fuel.

4. The gasket used on the removable ash pan located on the bottom of the woodstove
indicated that the pan was not seating properly. This observation appeared to
indicate that an underfire air condition may have occurred during the study. This
type of combustion gas condition can cause elevated emissions through reduced
combustion efficiency. Since this observation is based on only one stove, it is
difficult to determine if the problem is a stove design and/or operator related issue.

5. It appears the study participant did not regularly inspect the woodstove's gaskets
and mechanical components as recommended in the manufacturer's stove
operating instruction booklet.

6. Operator inexperience: The woodstove was the first integral catalytic woodstove
used by the participants in home P06. The participants in this home claimed that it
took two weeks to learn how to operate the stove. Since there appears to be
evidence that the operators of this stove, given adequate training regarding
operating and maintenance procedures, may have avoided some, if not all, of the
above identified problems, these emission rate values are deleted from the data
base used in the development of catalytic woodstove emission factors.

f. Emission rates are the mean of the emission rates from co-located AWES units.
111-17
-------
Table 111-7

Installation, Burn and Paniculate Emission Rate Characteristics
Noncatalytic Wood stoves (1988 NSPS Certified/Certifiable)
Stove
Model
NC-A

NC-B

NC-C
NC-D

Study
Area
"Northeast"

"Northeast"

"Northwest"

"Northwest"
"Whitehorse"

All Areas
(edited data)
Home
Code
V12
V34

V35
V03

V14

N16

P04

P01
W04

W09
Chimney
Type
II
V

V
II

II
VI

V
Sample
Period
6
5
7
7
5
6
6
7
4
6
7
1
2
3
4
5
1
2
3
4
5
6
7
8
9
5
6
7
8
9
Mean
*n-l
Burn Rate
dry kg/hr
0.67
0.76
0.92
0.90
1.28
1.38
1.07
0.85
0.97
1.10
0.87
1.29
0.99
0.90
0.65
0.70
0.87
1.22
1.07
1.10
1.29
1.08
1.51
1.20
1.11
1.01
0.84
0.91
1.06
0.85
1.02
0.24
Emissions
g/hr
5.2
7.9
5.9
3.6
18.3
2.0
26.3
17.2
10.0
4.3
10.3
6.9
10.0
10.9
6.7
6.9
(19.4)
(17.4)
(13.3)
(24.1)
10.5
9.2
5.3
12.8
10.3
(19.4)
(18.3)
(17.9)
(17.9)
(22.2)
9.5
5.6
g/kg
7.7
10.4
6.4
4.0
14.3
1.4
26.6
20.4
10.3
3.9
11.9
5.3
10.1
12.1
10.3
9.9
(22.2)
(14.3)
(12.5)
(21.8)
8.1
8.5
3.5
10.7
9.3
(19.2)
(21.8)
(18.7)
(16.8)
(26.0)
9.8
5.7
Ref.
2
2

2
2

4
6

6
Comments

a
a

b
b
b
b

c,d
c
c
c
c

(continued)
111-18
-------
Table 111-7
(continued)
Chimney Types
Code
Description
I
il
III
IV
V
VI
Unlined masonry chimney located primarily inside the exterior walls of the house
Tile-lined masonry chimney located primarily inside the exterior walls of the house
Unlined masonry chimney located primarily outside the exterior walls of the house
Tile-lined masonry chimney located primarily outside the exterior walls of the house
Prefabricated metal chimney located primarily inside the exterior walls of the house
Prefabricated metal chimney located primarily outside the exterior walls of the house
Comments:

a. According to Reference 2, there are no conclusive factors to explain the elevated
emission rates observed for stove model "NC-B" in home V14.

b. The flue pipe of this stove entered a tile-lined masonry chimney that measured 17.8
cm by 27.9 cm by 3.7 meters high (7 inches by 11 inches by 12 feet high). Since the
manufacturer of this stove recommended using a 15.24 cm (6 inch) diameter
chimney, the attached chimney system may have been "oversized" for this
woodstove model. An inspection of the stove between the first and second sampling
periods revealed no obvious problems with the stove's emission control system.
While it is possible operator practices could have contributed to the elevated
emission rate performance, it is hypothesized that the oversized chimney system is a
significant contributing factor to this stove's emission performance. For this reason,
these emission values are deleted from the data base used to develop emission
factors for noncatalytic woodstoves (1988 NSPS-certified/certifiable).

c. It is hypothesized that the configuration of the chimney system used in home W09
may have been a contributing factor to the observed emission rates. The relatively
short length, 3.4 m (11.2 feet), of the chimney coupled with the mismatching of a
15.24 cm (6 inch) diameter flue pipe with 20 cm (8 inch) packed pipe may have
resulted in relatively poor draft conditions and thereby influencing the combustion
efficiency of the stove. The manufacturer recommends a uniform run of 15 cm (6
inch) diameter pipe. An inspection of the emission control system of this stove
revealed no significant problems. For the above reasons, these emission rate values
are deleted from the data base used to develop emission factors for noncatalytic
woodstoves (1988 NSPS-certified/certifiable).

d. Ambient oxygen (0^ value associated with emission sample greater than 20.9 ± 2.0
percent absolute at final calibration. Therefore, calculated emissions (g/hr, g/kg)
have a potentially higher degree of associated uncertainty. Because of the higher
uncertainty associated with this emission sample, it is deleted from the data base for
calculating emission factors for noncatalytic woodstoves (1988 NSPS-
certified/certifiable).
111-19
-------
Table 111-8

Mean Particulate Emission and Bum Rates by Woodstove Installation
Catalytic Woodstoves
(1988 NSPS Certified/Certifiable)
Study Area
"Northeast"
Avg (unweighted)
Range
"Northwest"
Avg (unweighted)
Range
"Whitehorse"
Avg (unweighted)
Range
ALL AREAS
Avg (unweighted)
Range
Model
C-A
C-D

C-A

C-B
C-B

Home
C01
cos

P02

W03
W14

Mean Emission Rates
g/hr (
-------
Table 111-10
Mean Particulate Emission and Burn Rates by Woodstove Installation
Noncatalytic Woodstoves
(1988 NSPS Certified/Certifiable)
Study Area
"Northeast"
Avg (unweighted)
Range
"Northwest"
Avg (unweighted)
Range
"Whitehorse"
Avg (unweighted)
Range
ALL AREAS
Avg (unweighted)
Range
Model
NC-A
NC-A
NC-A
NC-B
NC-B
NC-B

NC-B

NC-D

Home
V12
V34
V35
V03
V14
N16

P04

W04

Mean Emission Rates
g/hr (*„.!)
5.2 (0)
6.9 (1.0)
3.6 (0)
10.2 (8.2)
21.8 (4.6)
8.2 (2.8)
10.1 (7.5)
3.6-21.8
8.3 (2.0)
8.3 (2.0)
9.6 (2.7)
9.6 (2.7)
9.2 (5.5)
3.6-21.8
g/kg (*„.!>
7.7 (0)
8.4 (2.0)
4.0 (0)
7.9 (6.5)
22.5 (2.1)
8.7 (3.5)
10.2 (7.0)
4.0-22.5
9.5 (2.5)
9.5 (2.5)
8.0 (2.7)
8.0 (2.7)
9.6 (5.5)
4.0-22.5
Mean Burn Rate
dry kg/hr (v^)
0.67 (0)
0.84 (0.08)
0.90 (0)
1.33 (0.05)
0.96 (0.11)
0.98 (0.09)
0.94 (0.22)
0.67- 1.33
0.91 (0.26)
0.91 (0.26)
1.06 (0.14)
1.06 (0.14)
0.96 (0.19)
0.67- 1.33
Emission
Samples
(N)
1
2
1
2
2
3
11
5
5
5
5
21
Table 111-11
Mean Particulate Emission and Burn Rates by Woodstove Model
Noncatalytic Woodstoves
(1988 NSPS Certified/Certifiable)
Stove
Model
NC-A
NC-B
NC-D
Study
Area(s)
NE
NE.NW
WH
Homes(s)
V12, V34, V35
V03, V14, P04
W04
Average (unweighted)
Range
Mean Emission Rates
9/nr Ovj)
5.6 (1.8)
10.8 (6.8)
9.6 (2.7)
8.7 (2.7)
5.6-10.8
g/kg K-i)
7.1 (2.7)
11.3 (6.9)
8.0 (2.7)
8.8 (2.2)
7.1-11.3
Mean Bum Rate
dry kg/hr (r^J
0.81 (0.12)
1.00 (0.23)
1.23 (0.17)
1.01 (0.21)
0.81 - 1.23
Emission
Samples
(N)
4
12
5
21
ra-2i
-------
IV. RESULTS & DISCUSSION

A. TRADITIONAL TECHNOLOGY WOODSTOVE EMISSION FACTORS
Table IV-1 lists the number of emission samples used in the development of emission factors, woodstove
installations, and woodstove models per study area. Confidence limits based on the data in Tables III-2, m-3,
III-4, and IV-1 are shown in Table IV-2. While there are relatively small differences between the calculated
emission factors, the confidence interval is influenced by the data base size (N) and standard deviation (tr).
Since (1) the data in Table III-3 showed significant emission rate variability between homes with the identical
woodstove models and (2) there are potentially thousands of different woodstove model/chimney system
installation combinations, it is recommended that the data in Table ffl-3 be used as the basis for calculating
traditional technology woodstove emission factors and related statistics. Presented in Table IV-3 are the
recommended emission factors (g/hr, g/kg) and related statistics based on data in Table ffl-3.

The average area-specific gram-per-kilogram emission factors and burn rates in Table III-3 show a
correlation coefficient (r) of -0.96 (Table IV-4). This inverse relationship supports laboratory data which
indicates that emission rates decrease with increasing burn rate for most traditional technology woodstoves.
Higher burn rates will generally result in lower emissions due to increased firebox temperatures and more
complete combustion of fuel and its combustion gases. While this relationship is based on a limited data base,
it may be of assistance to users in calculating a gram-per-kilogram emission factor based on the "average"
burn rate conditions in a specific geographic area. The range of burn rates upon which this relationship was
developed is 0.80 to 2.25 dry kg/hr. Based on the data in Table ffl-3, there appears to be no statistically
significant correlation between the area-specific gram per hour emission factors and burn rates (r = 0.05).
There is no significant correlation between woodstove model-specific emission and burn rates (r = 0.40,
r = -0.38, respectively) based on the data in Table ffl-4.

As noted in the comments in Table ffl-2, there may have been factors, e.g. fueling practices and chimney
system configuration, that may have contributed to relatively low or elevated emission rates. However, the
range of average emission rates (g/hr, g/kg) did not vary significantly between all three study areas
(Table ffl-3). In each area there were traditional technology woodstoves that produced emission rates under
10 g/hr. Therefore, it did not appear to be justifiable to remove low emission rates from the data base just
because there were below the expected emission rates for traditional technology woodstoves.

An unresolved question still remains as to the "true" population gram-per-hour (or gram-per-kilogram)
mean emission rates in each of the study areas. The variability of emission rates between homes with the same
woodstove model is significant. While the amount of data showing inter-home emission rate variability with
IV-1
-------
Table IV-1

Data Base Sample Sizes
Traditional Technology Wood stoves

Study Area
"Northeast"
"Northwest"
"Whitehorse"
All Areas
Emission
Samples
14
12
53
79
Woodstove
Installations
6
4
16
26
Stove
Models
6
4
9
17*
One of the stove models (T-B) was installed in all three
study areas.
Table IV-2

Confidence Intervals
Traditional Technology Woodstoves
Emission
Factor
g/hr
g/kg
Confidence
Limit
95%
99%
95%
99%
Data Base Used for Calculations
"Emission Samples"1
21.0 ± 1.9
21.0 ± 2.5
15.4 ± 1.3
15.4 ± 1.7
"Woodstove Installations"2
21.3 ± 2.8
21.3 ± 3.7
15.2 ± 1.8
15.2 ± 2.4
"Stove Model"3
20.4 ± 3.2
20.4 ± 4.4
14.6 ± 2.2
14.6 ± 2.9
1. Calculations based on individual emission sample emission rates (Table III-2).
N = 79, ffgfla-9.9, (rg/kg=5.9.

2. Calculations based on average "woodstove installation" emission rates (Table III-3).
N = 26, ff^ = 7.4, tr^g = 4.7.
3. Calculations based on average "woodstove model" specific emission rates (Table III-4).

N=17>
IV-2
-------
Table IV-3

Recommended Paniculate Emission Factors
Traditional Technology Woodstoves

Emission Rate
Standard Deviation
Range
95% Confidence Interval
99% Confidence Interval
g/hr
21.3
7.4
8.1 - 33.0
21.3 ± 2.8
21.3 ± 3.7
g/kg
15.2
4.7
5.5 - 23.9
15.2 ± 1.8
15.2 ± 2.4
Table IV-4

Relationship Between Area-Specific Average Gram per
Kilogram Particulate Emissions and Burn Rates
Traditional Technology Woodstoves

Location
"Northeast"
"Northwest"
"Whitehorse"
Average
Emission Rate
(g/kg)
12.6
17.2
15.8
Average
Burn Rate
(dry kg/hr)
1.65
1.19
1.45
r = -0.96 r2 = 0.92

Regression Equation: y = -9.77 x + 29.2,

where

x=Area-specific average
burn rate, dry kg/hr;
and

y=Emission rate, g/kg.
rv-3
-------
the same woodstove model is limited (only 5 woodstove models), the data do indicate a high degree of
uncertainty of determining average model-specific emission rates based on data bases which include only one
or two homes. Since all the data bases in the "Northeast" and "Northwest" studies are small and based on
data sets of one specific woodstove model in only one home, it is highly probable that the calculated area-
specific mean emission rates in Table III-3 are not representative of the "true" population. It may be more of
a coincidence than any real relationship that the absolute difference in mean emission rates between all three
study areas is 2.0 g/hr.

There is some inferential evidence that the area-specific mean emission rates (g/hr, g/kg) for the
"Northeast" and "Northwest" studies may be higher than indicated by the data in Table ni-3. The average
area-specific burn rate for the "Northeast" study (1.65 kg/hr) is based on fourteen (N = 14) emission samples
listed in Table m-2. A limited analysis of the frequency distribution of traditional technology woodstove bum
rates from homes V09, V14, and N16 during the 1985-86 heating season indicated a mean burn rate of 1.39 dry
kg/hr (refer to Table IV-5). This analysis includes all the time the woodstoves were operational (including
non-AWES sampling periods) as determined by electronic data logging equipment. Burn rates were
determined on twenty-four-hour time periods (5AM-5AM) between December 20, 1985 and April 7, 1986
(N = 166). Since it appears that at least area-specific gram-per-kilogram emission rates are inversely
proportional to burn rate (equation in Table IV-4), the "true Northeast" gram-per-kilogram emission rate may
be approximately 15 to 16 g/kg rather than the calculated 12.6 g/kg (Table IQ-3).

The area-specific emission rates for the "Northwest" study are based on emission rates for woodstove
models T-B, T-D, T-J, and T-K. There is some inferential evidence that for at least woodstove model T-K
typical "Northwest" emission rates for this woodstove model may be higher than indicated in Table III-3. A
study performed for the Oregon Department of Environmental Quality (ODEQ)14 evaluating a catalytic add-
on device determined mean before-catalyst emission rates of 29.2 g/hr and 28.5 g/kg for woodstove model T-K.
These rates were based on four emission samples collected in two homes (two samples per home). Assuming
approximately a 10 percent increase b before-catalyst emission rates (based on analysis of data in the
"Whitehorse" study6) due to the "dampering" effect of the add-on devices, the corrected flue collar emission
rates would be approximately 26 g/hr and 24 g/kg at a burn rate of 1.1 dry kg/hr.

Since the recommended emission rates are based on a relatively small data base, the user of the
recommended emission factors should recognize the limitations of the data and apply the appropriate caveats,
e.g., use confidence limits, when using the recommended emission rates on either an absolute or relative
comparison basis.
IV-4
-------
Table IV-5
Burn Rate Frequency Distribution - "Northeast"

OMNI Environmental Services, Inc.
»*»»**********»#*»»•******•»»*******
Daily (5 am to 5 am) Burn Rates Calculated as TC#1>10O
Files:
VO9.HRW:
V14.HRW:
NO8.HRW:
N16.HRW:
N16.HRW:
O1-21-86
01-24-86
01-17-86
12-20-85
01-29-86
(2100)
(150O)
(1500)
(160O)
(18OO)
to 03-03-86
to O3-31-86
to O4-02-86
to 12-23-85
to O4-07-86
(1400). Number o-f points: 38
(11OO). Number o-f points: 43
(14OO). Number o-f points: 38
(O6OO). Number of points: 2
(090O) . Number o-f points: 45
Total # Data Points: 166
Burn Re
Interval 1
; o.ooo -
> O. 1OO -
> O.20O -
> O.300 -
> 0.40O -
> O.50O -
> 0.6OO -
> 0.700 -
> O.BOO -
> O.9OO -
> l.OOO -
> 1.1OO -
> 1.2OO -
> 1.300 -
> 1.400 -
> 1.500 -
> 1.600 -
> 1.7OO -
> 1.8OO -
> 1.9OO -
> 2.OOO -
> 2.100 -
> 2.200 -
> 2.300 -
> 2.400 -
> 2.500 -
> 2.600 -
> 2.700 -
> 2.800 -
> 2.9OO -
> 3.OOO -
> 3. 1OO -
> 3.200 -
> 3.300 -
> 3.4OO -
> 3.5OO -
> 3.6OO -
> 3.7OO -
> 3. BOO -
> 3.9OO -
> 4.OOO -
> 4. 1OO -
> 4.2OO -
> 4.300 -
> 4.4OO -
> 4.5OO -
> 4.6OO -
> 4.7OO -
> 4. BOO -
> 4.900 -
»te
:kg/hr)
0. 10O
0.2OO
0.3OO
0.400
O.5OO
O.6OO
0.7OO
O.8OO
O.9OO
l.OOO
1. 10O
1.20O
1.3OO
1.400
1.500
1.6OO
1.7OO
1.8OO
1.9OO
2.OOO
2.100
2.2OO
2.3OO
2.4OO
2.500
2.600
2.700
2.8OO
2.9OO
3.OOO
3.1OO
3.200
3.300
3.400
3.500
3.6OO
3.700
3.800
3.9OO
4.OOO
4.1OO
4.200
4.300
4.4OO
4.5OO
4.6OO
4.7OO
4. BOO
4.9OO
5.000
n of
Points
0
O
0
2
2
5
3
4
8
7
15
14
11
14
14
12
11
9
2
3
1
5
3
3
1
0
1
1
O
1
1
O
1
1
O
O
O
0
0
0
0
O
0
0
0
O
0
0
0
0
Frequency
O.O
O.O
0.0
1.3
1.3
3.2
1.9
2.6
5.2
4.5
9.7
9.O
7. 1
9.0
9.0
7.7
7.1
5.8
1.3
1.9
O.6
3.2
1.9
1.9
O.6
O.O
O.6
0.6
O.O
O.6
O.6
O.O
O.6
O.6
0.0
O.O
0.0
0.0
0.0
O.O
O.O
O.O
0.0
0.0
O.O
0.0
0.0
0.0
O.O
O.O
Cumulative
Frequency (X)
O.O
O.O
O.O
1.3
2.6
5.8
7.7
1O.3
15.5
20.0
29.7
38.7
45.8
54.8
63.9
71.6
78.7
84.5
85.8
87.7
88.4
91.6
93.5
95.5
96.1
96.1
96.8
97.4
97.4
98.1
98.7
98.7
99.4
1OO.O
100.0
100.0
100.0
100.0
100. O
10O.O
1OO.O
10O.O
100.0
100.0
1OO.O
1OO.O
1OO.O
1OO.O
10O.O
1OO.O
Days when stove
was not in use
11
Mean Burn Rate :
Standard Deviation:
1.397
O.55O
IV-5
-------
B. 1988 NSPS-CERTDFIED/CERTIFIABLE WOODSTOVE EMISSION FACTORS
As of July 21, 1988, there were 72 catalytic and 43 noncatalytic woodstove models that had been
grandfathered or directly certified under the 1988 NSPS.15 Only 4 woodstoves out of a total of 115 were
directly certified. Therefore, the NSPS certified/certifiable woodstove models used as the basis for
determining emission factors represent only 4 percent of the approved catalytic units and 7 percent of the
approved noncatalytic units. In addition, none of the woodstoves evaluated in the field studies were directly
certified under NSPS requirements. Tables IV-6 and IV-7 present the number of acceptable emission samples
collected, woodstove installations, and stove models used per study for catalytic and noncatalytic technologies.
Confidence intervals for calculated mean emission rates are shown in Tables IV-8 and IV-9 for catalytic and
non-catalytic technologies, respectively.

Because of the limited data available, and the demonstrated degree of variability in emission rates
between homes (installations) with the same stove model, recommended emission factors are based on the
data in Tables III-8 (catalytic technology) and HI-10 (noncatalytic technology). Tables IV-10 and IV-11
present the recommended emission factors and related statistics for 1988 certified/certifiable catalytic and
noncatalytic woodstove technologies. A comparison of the recommended gram-per-hour emission factors for
1988 NSPS certified/certifiable catalytic and noncatalytic woodstoves with the 1988 (Phase 1) NSPS emission
limits is shown in Table IV-12.

The differences between the 1988 NSPS (Phase I) emission limits and recommended emission factors can
be attributed to differences between the standardized fueling protocol used for certified woodstoves (EPA
Method 28)l and the variability associated with in-situ woodstove fueling, operating, installation, and
maintenance conditions. An additional contributing factor could be that the woodstoves used as the basis for
determining emission factors basically represent "first generation" emission control technologies, which were
certified under the less stringent (as compared to EPA NSPS direct-certification provisions) Oregon rules. It
is expected as emission control technology improves for NSPS-certified woodstoves that a more consistent
pattern of relatively low emission rate performance will be demonstrated.

Due to the limited number of emission samples used to derive the recommended participate emission
factors, no adjustment factors based on burn rates are proposed. As in the case with the recommended
emission factors for traditional technology woodstoves, the user should recognize the limitations of the data
and apply the appropriate caveats, e.g., use confidence limits, when using the recommended emission factors
on either an absolute or relative comparison basis.
FV-6
-------
Table IV-6

Data Base Sample Sizes
Catalytic Woodstoves
(1988 NSPS Certified/Certifiable)

Study Area
"Northeast"
"Northwest"
"Whitehorse"
All Areas
Emission
Samples
3
5
10
18
Woodstove
Installations
2
1
2
5
Stove
Models
2
1
1
3*
* One of the stove models (C-A) was installed in two
study areas.
Table IV-7

Data Base Sample Sizes
Noncatalytic Woodstoves
(1988 NSPS Certified/Certifiable)

Study Area
"Northeast"
"Northwest"
"Whitehorse"
All Areas
Emission
Samples
11
5
5
21
Woodstove
Installations
6
1
1
8
Stove
Models
2
1
1
3*
One of the stove models (NC-B) was installed in two
study areas.
IV-7
-------
Table IV-8

Confidence Intervals
Catalytic Wood stoves
(1988 NSPS Certified/Certifiable)
Emission
Factor
g/hr
g/kg
Confidence
Limit
95%
99%
95%
99%
Data Base Used for Calculations
"Emission Samples"1
6.9 ± 1.6
6.9 ± 2.1
7.3 ± 1.6
7.3 ± 2.1
"Woodstove Installations"2
6.2 ± 2.2
6.2 ± 2.8
6.6 ± 1.9
6.6 ± 2.5
"Stove Model"3
5.9 ± 3.0
5.9 ± 3.7
6.3 ± 2.7
6.3 ± 3.6
1. Calculations based on individual emission sample emission rates (Table Ill-e).
N = 18, g/kg
3. Calculations based on average "woodstove model" specific emission rates (Table III-9).
Table IV-9

Confidence Intervals
Noncatalytic Wood stoves
(1988 NSPS Certified/Certifiable)
Emission
Factor
g/hr
g/kg
Confidence
Limit
95%
99%
95%
99%
Data Base Used for Calculations
"Emission Samples"1
9.5 ± 2.4
9.5 ± 3.1
9.8 ± 2.5
9.8 ± 3.2
"Woodstove Installations"2
9.2 ± 3.8
9.2 ± 5.0
9.6 ± 3.8
9.6 ± 5.0
"Stove Model"3
8.7 ± 3.1
8.7 ± 4.0
8.8 ± 2.5
8.8 ± 3.3
1. Calculations based on individual emission sample emission rates (Table 111-7).
N=21,
2. Calculations based on average "woodstove installation" emission rates (Table 111-10).
3. Calculations based on average "woodstove model" specific emission rates (Table
N=3»
1-11).
rv-8
-------
Table IV-10

Recommended Paniculate Emission Factors
Catalytic Woodstoves
(1988 NSPS Certified/Certifiable)

Emission Rate
Standard Deviation
Range
95% Confidence Interval
99% Confidence Interval
g/hr
6.2
2.5
4.0 - 9.2
6.2 ± 2.2
6.2 ± 2.8
g/kg
6.6
2.2
4.7-9.1
6.6 ± 1.9
6.6 ± 2.5
Table IV-11

Recommended Paniculate Emission Factors
Noncatalytic Woodstoves
(1988 NSPS Certified/Certifiable)

Emission Rate
Standard Deviation
Range
95% Confidence Interval
99% Confidence Interval
g/hr
9.2
5.5
3.6-21.8
9.2 ± 3.8
9.2 ± 5.0
g/kg
9.6
5.5
4.0 - 22.5
9.6 ± 3.8
9.6 ± 5.0
Table IV-12
Comparison of 1988 (Phase I) NSPS Emission Limits
with Recommended 1988 NSPS Emission Factors
Technology
Catalytic
Noncatalytic
Phase I
Emission Limits (g/hr)
5.5
8.5
Recommended 1988 NSPS
Emission Factors (g/hr)
6.2
9.2
IV-9
-------
V. REFERENCES

1. United States Environmental Protection Agency. Standard of Performance for New Stationary Sources;
New Residential Wood Heaters. Federal Register, Volume 53, Number 38, Section 40 CFR, Part 60.
February 26,1988.

2. OMNI Environmental Services, Inc. "Performance Monitoring of Catalyst Stoves, Add-Ons, and High-
Efficiency Stoves. Field Testing for Fuel Savings, Creosote Build-Up and Emissions", Volume 1.
Prepared for Coalition of Northeastern Governors, New York State Energy Research and Development
Authority, and the U.S. Environmental Protection Agency. October 1987.

3. Burnet, P.B. and Simons, CA. "Identification of Factors Which Affect Combustion Efficiency and
Environmental Impacts from Woodstoves", Task D, U.S. Department of Energy's Pacific Northwest and
Alaska Regional Biomass Energy Program (as administered by the Bonneville Power Administration).
Contract No. DE-AC79-85BP18508. July 1988.

4. Simons, CA.; Christiansen, P.D.; Houck, I.E.; and Pritchett, L.C. "Woodstove Sampling Methods
Comparability Analysis and In-Situ Evaluation of New Technology Woodstoves - Task G." U.S.
Department of Energy's Pacific Northwest and Alaska Regional Biomass Energy Program (as
administered by the Bonneville Power Administration). Contract No. DE-AC79-85BP18508. June 1988.

5. Simons, C.A. "Final Report - Particulate Emission Test, Emission Control System Inspection and Leak
Check - Blaze King 'King' in Home P02." Prepared for the Oregon Department of Environmental
Quality and the Environmental Protection Agency. April 1988.

6. Simons, C.A.; Christiansen, P.D.; Pritchett, L.C.; and Beyerman, G_A. "Whitehorse Efficient Woodheat
Demonstration." Prepared for the City of Whitehorse and Energy, Mines and Resources Canada.
September 1987.

7. "Catalyst for Success," Volume I, Issue 1, Corning Glass Works Company, March 1988.

8. Edmisten, N.; Simons, CA.; and Tiegs, P.E. "Major Studies and Preliminary Findings on Woodstove
Performance in 91 Homes in Oregon, New York, and Vermont." APCA Paper 86-74.4, Annual Meeting,
Air Pollution Control Association, Minneapolis, Minnesota. June 1984.

9. Houck, J.E.; Simons, CA.; and Burnet, P.G. "A System to Obtain Tune-Integrated Woodstove Emission
Samples." Proceedings of the 1986 EPA/APCA Symposium on Measurement of Toxic Air Pollutants.
Research Triangle Park, North Carolina. April 1986.

10. United States Environmental Protection Agency. "Emission Factor Documentation for AP-42: Section
1.10, Residential Wood Stoves." EPA-340/4-82-003. May 1983.

11. Oregon Department of Environmental Quality. Administrative Rules for Certification of Woodstoves,
OAR-340-21-100 through 340-21-190. June 8,1984.

12. Satterfield, Gary, Wood Heating Alliance. Personal communication. July 25,1988.

13. "Statistical Record—Whitehorse Woodheating Appliances." Building Inspector Branch, City of
Whitehorse. 1986.

V-l
-------
14. Christiansen, P.D.; Simons, C.A.; and Pritchett, L.C. "An In-Situ Performance Evaluation of Two
Woodstove Catalytic Retrofit Devices," Oregon Department of Environmental Quality. May 1987.

15. U.S. Environmental Protection Agency. "Residential Wood Heaters Certified by the U.S. Environmental
Protection Agency." Stationary Source Compliance Division. July 21,1988.
V-2
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverie before completing)
REPORT NO.
IPA-450/3-88-013
RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
In-Situ Emission Factors for Residential Wood
Combustion Units
5. REPORT DATE
December 1988
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
I. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
2. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
5. SUPPLEMENTARY NOTES
6. ABSTRACT
This report presents particulate emission factors (grams per hour,
grams per kilogram) for existing traditional technology woodstoves and
catalytic and noncatalytic units which meett or could meet the Environmental
Protection Agency's 1988 new source performance standards.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Air Pollution
Pollution Control
Woodstoves
Residential Wood Combustion
Wood Heaters
Catalytic Stoves
Noncatalytic stoves
Particulate Matter
Standards of Performance
Air Pollution
13b
18. DISTRIBUTION STATEMENT

Unlimited
19. SECURITY CLASS I Tins Report/
Unclassified
21. NO. OF PAGES
43
20. SECURITY CLASS (TlliS page)
Unclassified
22. PRICE
EPA farm 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
-------