W oodstove Durability Testing Protocol


EPA-600 /R-94-193
November 1994

Woodstove Durability Testing Protocol

By:

Roger D. Bighouse, Stockton G. Barnett,
Jaiues E. Houck, and Paul E. Tiegs
OMNI Environmental Services, Inc.
10074 SW Arctic Drive
Beaverton, OR 97005

EPA Contract 68-D0-0120, WAs 1-28,1-41,1-42, and 1-58
(E.H. Pechan and Associates, Inc.)

EPA Project Officer:

Robert C. McCrillis
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711

Prepared for:

U.S. Environmental Protection Agency
Office of Research and Development
401 M Street SW
Washington, DC 20460

-------
Abstract

Woodstove field studies conducted over the course of seven heating seasons have shown ihat new
technology woodstoves designed to have low particulate emissions have frequently shown rapid
degradation in emission control. This degradation has been documented both by measurement of
particulate emission factors with in-home automated emission sampler (AES) and by observable
physical damage to the woodstove components. It appears that most of the damage occurs during
those occasional times when the woodstove is allowed to operate at exceptionally high temperatures.
A method to test the long-term durability of woodstove models in the laboratory in a one to two
week time frame has been developed and has come to be referred to as a stress test.

Two avenues of research have been taken in developing the stress test protocol. First, the
performance of woodstoves while in actual in-home use has been observed over the course of two
heating seasons in three communities. These are: Medford and Klamath Falls, OR, and Glens Falls.
NY. Eight models of stoves in thirteen homes were studied. The field studies permitted records of
woodstove operating temperatures, particulate emission levels, and in some cases, physical
degradation, to be followed in a real world setting. The second line of research used in developing
a stress test was the laboratory "stressing" of various woodstove models under high temperature
operation. This laboratory research has been conducted on six stoves (five models) and, as with the
in-home research, changes in particulate emission rates were measured and physical degradation
documented. Both catalytic and noncatalytic stove models, including EPA Phase 2 certified stoves,
were represented in the tests.

EPA REVIEW NOTICE

This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signily that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

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

-------
Table of Contents

Abstract	ii

List of Tables	iv

List of Figures	v

1.	Introduction	1

2.	Development of a Stress Test	1

2.1.	Preliminary Tests	2

2.2.	Development of Final Protocol	 11

2.3.	Stack Draft	 16

2.4.	Final Stress Test Protocol	 16

3.	Methods	23

3.1 Emission Measurements	23

3 .2 Quality Assurance	28

3.2.1	Calibration Procedures and Frequency	28

3 .2.1 1 Temperature Measuring Device Calibration	29

3.2.1.2	AWES Unit Flow Rate	29

3.2.1.3	Analytical Balance Calibration	29

3.2.1.4	Wood Moisture Meter	29

3.2.2	Quality Assurance Audits	29

3 .2.3 QA/QC Checks of Data Reduction	29

3.2.4 Sample Identification and Custody	30

3 .2.5 Measurement Uncertainty	30

4.	Results	32

4.1.	Blaze King Royal Heir (Stove #1)	32

4.2.	Second Blaze King Royal Heir (Stove #2)	32

4.3.	Country Flame BBF-6	40

4.4.	Regency R3/R9	40

4.5.	Quadrafire 3100	49

4.6.	Earth Stove 1003C	49

5.	Conclusions	58

6.	Acknowledgements	59

7.	References	59

iii

-------
List of Tables

Table Number	Page

2.1-1. Comparison of Field Chimney Temperature Data for One Season to Predicted Laboratory
Data for One Week Using Protocol 1	14

2.1-2.	Comparison of Field Catalyst Temperature Data for One Season to Predicted Laboratory
Data for One Week Using Protocol 1	14

2.2-1.	Summary of Interim Protocols Tested	15

2.2-2.	Rates of Temperature Accumulation for Interim Protocols	15

3-1.	Parameters Measured and Associated Uncertainty	31

4-1.	Observations of Physical Damage Due to Stressing	33

4.1-1.	Particulate Emissions and Bypass Gaps in Blaze King Royal Heir Stoves	36

4.3-1.	Particulate Emissions and Bypass Gaps in Country Flame BBF-6 Stoves	45

4.4-1.	Particulate Emissions and Baffle Warpage in Regency R3/R9 Stoves	48

4.5-1.	Particulate Emissions and Baffle and Secondary Air Tube Warpage in Quadrafire

3100 Stoves	54

4.6-1.	Particulate Emissions and Bypass Gaps in Earth Stove 1003C Stoves	58

iv

-------
List of Figures

Figure Number	Page

2-1.	Fuel sizing templates	 3

2.1-1.	Continuous chimney temperature record - Regency stove in

Home Y20	 5

2.1-2.	Continuous chimney temperature record during typical day -

Regency stove in Home Y20	 5

2.1-3.	Chimney temperature accumulation - Regency stove in Home

Y20	 6

2.1-4.	Continuous catalyst temperature vecord - Blaze King stove in

Home Y02	 7

2.1-5.	Continuous catalyst temperature record during typical day -

Blaze King stove in Home Y02	 7

2.1-6.	Catalyst temperature accumulation - Blaze King stove in Home

Y02	 8

2.1-7.	Continuous catalyst temperature record - Blaze King stove in

Home Y23	 9

2.1-8.	Continuous catalyst temperature record during typical day -

Blaze King stove in Home Y23	9

2.1-9.	Catalyst temperature accumulation - Blaze King stove in Home

Y23	 10

2.1-10.	Continuous chimney temperature - Regency stove during

laboratory fuel density tests	 12

2.1-11.	Predicted chimney temperatures with six 112 kg/m3 (7 lb/ft3)

fuel loads per day - Regency stove	 12

2.1-12.	Predicted chimney temperature accumulation with six 112

kg/m3 (7 lb/ft3) fuel loads per day - Regency stove	 13

2.1-13.	Comparison of laboratory and home chimney temperature
accumulations - Regency stove in Home Y20	 13

2.2-1.	Catalyst temperature accumulation - Blaze King stove interim
protocol 2			 17

2.2-2.	Fire-box temperature accumulation - Blaze King stove interim

protocol 2	 17

2.2-3.	Catalyst temperature accumulation - Blaze King stove interim

protocol 3	 	 18

2.2-4.	Fire-box temperature accumulation - Blaze King stove interim

protocol 3	 18

2.2-5.	Catalyst temperature accumulation - Blaze King stove interim

protocol 4	 19

2.2-6.	Fire-box temperature accumulation - Blaze King stove interim

protocol 4	 19

2.2-7.	Catalyst temperature accumulation - Blaze King stove interim

protocol 5	 20

v

-------
List of Figures (continued)

2.2-8.	Fire-box temperature accumulation - Blaze King stove interim

protocol 5	 20

2.2-9.	Catalyst temperature accumulation in the Blaze King stove by

protocol category	 21

2.2-10.	Fire-box temperature accumulation in the Blaze King stove by

protocol category	 21

2.3-1.	Stack draft - temperature relationship for two stoves used to quantify
proper draft during stress tests	 22

3-1.	Schematic of AWES system	 24

3-2.	AWES data acquisition and control system	 26

4.1-1.	Particulate emissions from Blaze King Royal Heir (stove ti\) before and

after stressing (protocols 2-5)	 34

4.1-2.	Oxidation of top of firebox in Blaze King Royal Heir (stove #1) after

14 days of stressing in the laboratory	 35

4.1-3.	Bypass gap (0.64 cm) in Blaze King Royal Heir (stove 01) after 14 days

of stressing in the laboratory	 35

4.1-4.	Oxidation in firebox in Blaze King Royal Heir (Home Y01) after one

heating season of use	 37

4.1-5.	Bypass gap (1.6 cm) in Blaze King Royal Heir (Home Y02) after one

heating season of use	 37

4.1-6.	Plugging of catalyst in Blaze King Royal Heir (Home Y25)	 38

4.1-7.	Catalyst flaking in Blaze King Royal Heir (Home Y12)	 38

4.2-1.	Particulate emissions from Blaze King Royal Heir (stove ttl)

before and after stressing (protocol 6)	 39

4.2-2.	Comparison of particulate emissions due to bypass leaks and

degraded catalysts in Blaze King Royal Heir stoves	 41

4.3-1.	Particulate emissions from Country Flame BBF-6 stove before

and after stressing	 42

4.3-2.	Flame impingement shield warpage in Country Flame BBF-6

after 25 days of stressing	 43

4.3-3.	Catalyst holder oxidation and warpage in Country Flame BBF-

6 after 25 days of stressing	 43

4.3-4.	Flame impingement shield warpage in Country Flame BFiF-6

(Home Y14) after two heating seasons of use	 44

4.3-5.	Catalyst holder oxidation and warpage in Country Flame BBF-

6 (Home Y14) after two heating seasons of use	 44

4.4-1.	Particulate emissions from Regency R3/R9 stove during

stressing	 46

4.4-2.	Warped baffle removed from regency R3/R9 stove (Home

Y20) after one heating season	 47

4.4-3.	Warped baffle removed from Regency R3/R9 stove after 18

days of stressing	 47

vi

-------
List of Figures (continued)

4.4-4.	Effect of Baffle warpage on particulate emissions from

Regency R3/R9 stove	 50

4.5-1.	Particulate emissions from Quadrafire 3100 stove before and

after stressing	 51

4.5-2.	Oxidation and minor warpage of secondary air tubes and

baffle in the Quadrafire 3100 after 10 days of stressing	 52

4.5-3.	Oxidation and major warpage of secondary air tubes and

baffle in the Quadrafire 3100 after 20 days of stressing	 52

4.5-4.	Warpage after 20 days of stressing, as seen from top of stove,

in the Quadrafire 3100 stove	 53

4.6-1.	Particulate emissions from Earth Stove 1003C stove before

and after stressing	 55

4.6-2.	Failed catalyst bypass mechanism in the Earth Stove 1003C

after 20 days of stressing	 56

4.6-3.	Oxidation and warpage of catalyst holder in the Earth Stove

1003C after 20 days of stressing	 57

4.6-4.	Warpage of door frame in the Earth Stove 1003C after 20

days of stressing	 57

vii

-------
1. Introduction

Recently, there has been much concern by regulatory agencies and stove manufacturers about
long-term physical degradation of woodstoves and elevated air pollutant emissions due to this
degradation. In the past, the observation of such degradation could only be made in the field
after one or more heating seasons of use, after a particular model has been widely introduced
to the market. Consequently, improvement in the manufacturing and design of woodstoves in
response to degradation has been a slow process.

The development of an accelerated test to simulate in-home woodstove aging and degradation
over a short period of time in the laboratory is reported here. The aging of stoves was achieved
through extreme conditions and is therefore termed the "stress test". The goal of the project
was to develop a protocol by which a woodstove could be operated in the laboratory for a short
period of time (about one week) to simulate one heating season of burning in the field. The
short turnaround time has been deemed necessary to evaluate a stove's long-term performance
and durability so that modifications can be made in stove design and manufacturing during the
time period when a stove model is in development.

2. Development of a Stress Test

Stress testing was done to subject target stoves to a cyclic pattern of high temperature exposures.
To maintain consistency, a protocol was developed to specify all parameters of woodstove
burning. These include fuel type, moisture, size, and configuration; loading density; woodstove
air settings; startup method; length of time doors and bypasses were open; stack height; and a
criteria for refueling.

Throughout the development of a stress testing protocol, many of the above parameters were
held constant, while others were varied to determine the combination of factors which would
lead to the most extreme burning condition. The parameters held constant were:

Fuel

Type

Split lodgepole pine, as free of knots as possible.

Fuel

Moisture

10 to 20% (dry basis).

Fuel

Length

5/6 of the longest firebox dimension.

Fuel

Configuration

Fuel in center of firebox, packed tightly with smaller fuel on

bottom.

Air Settings Air settings on stove set to maximize burn rate and firebox

temperatures.

Kindling Load Maximum of IS minutes in duration.

-------
Length of Bypass
Opening

Stack Draft

Refueling

For catalytic stoves only. Bypass open for additional 7
minutes after stove door is closed.

Minimum of 17.4 pascals (0.07 inches H20) for 90% of
burn cycle.

A temperature threshold was empirically determined for each
stove model by putting fuel wood loads into operating
stoves. The temperature that corresponded to the conditions
when there was first enough space (from the burn-down of
the previous load) to put the full wood load into the stove
was later used as a reloading prompt for each stove model.

Four other parameters were varied throughout protocol development and the effects on
temperature were analyzed. These variable parameters were:

Fuel Size

Loading Density

Length of Door
Openings

Stack Height

Two different fuel sizing regions were used: (I) 70%
"large", 30% "small", and (2) 100% "small" (see Figure
2-1). Wood was considered large if it fit through a 20 cm
(8-inch) diameter hole but not through a 13-cm (5-ir.ch)
diameter hole. Wood was considered small if it fit through
a 13-cm (5-inch) diameter hole but not through a 8-cm
(3-inch) diameter hole.

Loading densities used were 48, 112, and 160 kg/m3 (3, 7,
and 10 lb/cubic feet) of firebox volume.

Stove door was left open between 3 and 45 minutes after
fuel was loaded.

Two stack heights were used: (1) 6.1 m (20 feet) and (2)
8.2 m (27 feet).

2.1. Preliminary Tests

Preliminary tests were done on two stoves to examine the feasibility of achieving high stove and
catalyst temperatures, for a duration of time, similar to those temperatures found in stoves in
homes during a heating season. Of primary importance was the cumulative time that the stove
or catalyst was above a certain temperature. Stove temperatures were monitored by
thermocouples either placed in the chimney 30 cm (1 ft) above the flue collar (Regency stove)
or placed in the metal at the top of the firebox (Blaze King). The catalyst temperatures (Blaze
King) were monitored by a thermocouple placed in the center of the catalyst 2.5 cm (1 inch) in
from the front face.

2

-------
Large

Maximum Diameter between
13 and 20 cm. (5 and 8 inches)

Figure 2-1. Fuel sizing templates.

Small

Maximum Diameter between
8 and 13 cm. (3 and 5 inches)

-------
Temperature data recorded during the Glens Falls, New York study was analyzed. A Pegency
R3/R9 noncatalytic stove (home code Y20) and two Blaze King Royal Heir stoves (home codes
Y02 and Y23) were analyzed for temperature accumulation. These stoves were chosen because
they displayed physical signs of deterioration and/or elevated particulate emissions.

Regency Stove. Home Y20. !n over 500 hours of field testing (four week-long tests), the
stove's chimney temperature reached 427 °C (800 °F) 25 times and 482 °C (900 °F) nine times.
For a 3,000-hour burning season, these would be 52 and 144 times, respectively. The average
duration of a 427 °C (800 °F) event was 0.54 hours, which would yield 77 cumulative hours
when the stack temperature was greatei than 427 °C over a 3,000-hour season. Likewise, the
average duration of a 482 °C event was 0.40 hours, yielding 21 cumulative hours over 482 °C
per season. Figure-. 2. i-1 through 2.1-3 illustrate temperature exposure for the Regency stove
in home Y20.

Blaze King Stove. Home Y02. The hottest-burning stove of the Glens Falls study, the Blaze
King stove in home Y02 experienced catalyst temperatures which exceeded 871 °C (1600 °F)
86 times and 982 °C (1800 °F) 25 times in over 800 hours of field testing. The average 871
°C and 982 °C excursions lasted 0.51 hours and 0.40 hours, respectively. Over a 3,000-hour
burning season, 871 °C would be exceeded 322 times, for a total of 159 cumulative hours and
982 °C would be exceeded 91 times, for a total of 36 cumulative houn>. Figures 2.1-4 through
2.1-6 illustrate catalyst temperature exposure fot the Blaze King stove in home Y02.

Blaze King Stove. Home Y23. The Blaze King stove in hcne Y23 was the second-hottest-
burning Blaze King of the Glens Falls study. In over 750 hours of field testing, catalyst
temperatures exceeded 871 °C 23 times and 982 °C seven times. The average excursions were
0.30 hours and 0.17 hours, respectively. Projected over a 3,000-hour season, catalyst
temperatures in excess of 871 °C would be experienced 89 times, for a total of 27 cumulative
hours, and temperatures of 982 °C would be experienced 27 times, for a total of 4.7 cumulative
hours. Figures 2.1-7 through 2.1-9 illustrate ca'alyst temperature exposure for the Blaze King
stove in home Y23.

Several test burns were made in Regency R3/R9 and Blaze King Royal Heir stoves to ascertain
whether the amount of temperature accumulation recorded in the field for one season could be
duplicated in the laboratory in a reasonably short time.

Three test burn cycles were made following an initial stress testing protocol, protocol 1, which
was based on the in-home data. The parameters used for protocol 1 were:

Fuel Size
Loading Density
Length of Door

Opening
Stack Height

70% large, 30% small

Alternating 48 and 112 kg/m3 (3 lb/ft3 and 7 lb/ft3) loads

5 minutes
6.1 m (20 feet)

-------
Figure 2.1-1.

Time Burned (hours)

Continuous chimney temperature record - Regency stove in Home Y20.

li-
es

a)

u.

(0

I—

CD

a

E
a>
h-

o
ca
+-~

CO

n
O

1.2
1.1
1

0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0



















































A











i













L

\







!









\

\





V







\

1

\n

A

















h

\















J

\













\

Jir















I



320

360

Figure 2

330 ™ 350
Time Burned (hours)

1-2. Continuous chimney temperature record during typical day - Regency stove in Home Y20.

-------
CO

100 200 300 400 500 600
Time Burned (hours)

Figure 2.1-3. Chimney temperature accumulation - Regency stove in Home Y20.

6

-------
(D

~
4—'

(0

l_

0)
Q.

E
a)

n
O

W
>.

(0
«-»
nj

O

100 300 500 700 900
Time Burned (hours)

Figure 2.1-4. Continuous catalyst temperature record - Blaze King stove in Home Y02.

o

D

2

CL o

E -

0) X

w

to

-4-»

0J

o

1.8
1.6
1.4
1.2
1

0.8
0.6
0.4
0.2
0





























,\ h







il







i











A.

I

,\



' i











A



\





i'

'¦I,











i











i





\





\

V









\





\







\ f







< i





V

















































430

440

450

435 Time Burnec? (hours)

460

Figure 2.1-5. Continuous catalyst temperature recoid during typical day - Blaze King stove in Home
Y02.

7

-------
<	v 100 300 500700900

Time Eurned (hours)

Figure 2.1-6. Catalyst temperature accumulation - Blaze King stove in Home Y02.

8

-------
0

0)

k_

D

To

c °
E

CD X
H

V)

_>>

2
«

O

200

400

600

700

800

100 300 ™ 500

Time Burned (hours)

Figure 2.1-7. Continuous catalyst temperature record - Blaze King stove in Home Y23.

u.

o
CD

i—

3

ro

§ ?T
E °

03 X

W
>»

ro

CO

O

£

465

U

J—4

470

475

485

Time Bumec? (hours)

490

495

Figure 2.1-8. Continuous catalyst temperature record during typical day - Blaze King stove in Home
Y23

9

-------
(/)

0 100 200 300 400 500 600 700 800 900
Time Burned (hours)

Figure 2.1-9. Catalyst temperature accumulation - Blaze King stove in Home Y23.

10

-------
These test burns indicated that a 112 kg/m3 (7 lb/ft3) load would produce stack temperatures in
the Regency which exceed 482 °C (Figure 2.1-10). With a four-hour burn cycle (reloading at
approximately 204 °C |400 °F|), six cycles per day or about 42 cycles of 427 °C (800 °F) and
482 °C (900 °F) events per week would be achieved (Figures 2.1-11 and 2.1-12). If the
durations of such events were similar to those in the field, a cumulative total of 22 hours in
excess of 427 °C and 17 hours in excess of 482 CC would be achieved. Table 2.1-1 and Figure
2.1-13 compare the chimney temperature data in the Regency for one season in a home and one
week in the laboratory using protocol 1.

Laboratory tests on the Blaze King indicate that following the initial protocol 1 with 112 kg/m3
(7 lb/ft3) loads, 982 °C (1800 °F) could be exceeded at a frequency of about once every five
hours or a total of 35 times per week. If the 871 °C (1600 °F) and 982 °C (1800 °F)
excursions are similar in duration to those in the stove in home Y02 (0.50 hours and 0.40 hours
respectively) 871 °C would be exceeded for a total of 17 hours and 982 CC would be exceeded
for a total of 14 hours in one week of stress testing. Table 2.1-2 summarizes and compares the
catalyst temperatures data in the Blaze King for one season in a home and one week in the
laboratory using protocol 1.

Upon comparison of home and laboratory data for the Regency stove, very high chimney
temperatures (over 482 °C) can be exceeded in the laboratory about the same number of times
as the stove in home Y20 during a heating season. Episodes of greater than 427 °C, however,
would be far fewer in the laboratory than in the home (42 versus 144). Blaze King home and
laboratory data also show this pattern. The 35 episodes in the laboratory in excess of 982 °C
fall between the 91 and 27 shown for the stoves in homes Y02 and Y23, respectively. However,
the 35 episodes in the laboratory in excess of 871 °C are far fewer than the 322 and 89 seen in
the stoves in homes Y02 and Y23 for one season.

The predicted severity of the protocol 1 stress test, as indicated by the higher temperature
episodes, is less than one heating season experienced by the Regency in home Y20 and more or
less equal to one heating season for the Blaze King in home Y23, but much less than one season
for the Blaze King in home Y02.

2.2. Development of a Final Protocol

Based on the results of the initial protocol 1 tests, more extreme protocols were developed to
simulate the temperature exposure similar to the Blaze King stove in home Y02. The variable
parameters in protocol 1 (fuel size, loading density, length of door opening, and stack height)
were systematically changed and the accumulated times at certain temperatures were analyzed.
Four different interim proUxols were tested (protocols 2 through 5). Each protocol was divided
into at least two different door opening lengths (coded A through I).

A Blaze King Royal Heir was used in this portion of the development stage. Throughout a 14-
day period, the four different protocols were tested. Table 2.2-1 summarizes the different
interim protocols along with the codes for door opening periods for the 14-day tests. Figures

-------
-------
<2

I-

O
JQ
OS

a>
E
i-

3
O

o
<

24
22
20
18
16
14
12
10
8
6
4
2
0

S 80C

° F

90C

Is F

, r1"

r1 r

Time (day)

Figure 2.1-12. Predicted chimney temperature accumulation with six 112 kg/m3 (7 lb/ft3) fuel loads per
day - Regency stove.

>
o
_a
n)

E
I-

TJ
-------
Table 2.1-1. Comparison of Field Chimney Temperature Data
For One Season to Predicted Laboratory Data For One Week Using Protocol 1

Stove

427 °C (800 °F)

482 °C (900 °F)

Events

Duration of
Event

Cumulative
Time

Events

Duration
of Event

Cumul
ative
Time

Home Y20

144

0.54 hours

77 hours

0.40
hours

21
hours

Predicted for
Regency Based
on Lab Tests

0.54 hours

22 hours

0.40
hours

17
hours

Table 2.1-2. Comparison of Field Catalyst Temperature Data
For One Season to Predicted Laboratory Data For One Week Using Protocol 1

Stove

871 °C (1600 °F)

982 °C (!800 °F)

Events

Duration
of Event

Cumulative
Time

Events

Duration
of Event

Cumulative
Time

Home Y02

322

0.50
hours

159 hours

0.40 hours

36 hours

Home Y23

0.30
hours

27 hours

0.17 hours

5 hours

Predicted for
Blaze King
Based on Lab
Tests

0.50
hours

17 hours

0.40 hours

14 hours

-------
Table 2.2-1. Summary of Interim Protocols Tested

Protocol

Door Opening

Load Size

Load Density

Stack
Height

Code

Time Period

1-2.5

70% large,
30% small

112 kg/m3
(7 lb/ft3)

6.1 m
(20 ft)

2.5-5

5-5.4

100% small

160 kg/m3
(10 lb/ft3)

6.1 m

(20 ft)

5.4-6

6-6.8

70% large,
30% small

160 kg/m3
(10 lb/ft3)

6.1 m
(20 ft)

6.8-8

8-8.3

8.3-10

70% large,
30% small

160 kg/m3
(10 lb/ft3)

8.2 m
(27 ft)

10-15

Table 2.2-2. Rates of Temperature Accumulation for Interim Protocols

Door
Opening
Code

Catalyst (hour accumulated/day)

Firebox (hour accumulated/day)

Protocol

982 °C
(1800
°F)

1058 °C
(1900

°F)

1045 °C
(2000
°F)

704 °C
(1300
°F)

760 °C
(1400
°F)

815 °C
(1500)

0.05

0.27

0.07

1.66

0.20

2.92

0.83

2.71

1.25

0.14

4.02

1.11

2.00

0.11

5.22

1.00

2.92

0.28

6.05

0.78

0.14

2.09

0.42

6.05

1.04

5.11

2.33

0.50

6.35

2.08

0.20

3.22

0.74

7.63

2.45

0.05

-------
2.2-1 through 2.2-10 show the temperature accumulation of the catalyst and firebox during test
protocols 2-5. The rates of temperature accumulation are summarized in Table 2.2-2. Except
for the 704 °C (1300 °F) and 760 °C (1400 °F) increments for firebox temperatures, protocol
5 code H had the highest rates of temperature accumulation (corresponding to maximum slopes
of the lines in Figures 2.2-9 and 2.2-10). The highest rates for the 704 °C and 760 °C firebox
temperature accumulation were obtained with protocol 5 code I.

The protocol 5 code H yielded catalyst temperatures above 982 °C for 5.11 hours per day or
35.8 hours per week. This is nearly identical to the 36 hours of catalyst temperatures above 982
°C accumulated in one heating season in the stove in home Y02.

2.3. Stack Draft

During the 14 days of testing with the Blaze King, low stack draft was an issue. Changing from
a 6.1-meter to an 8.2-meter chimney increased draft significantly, but the originally specified
draft—a minimum of 17.4 pascals (0.07 inches H20) for 90% of the burn cycle—was no longer
an accurate description of the levels of draft achieved.

A guideline for stack draft for use with the stress protocol was developed. The relationship of
stack draft versus temperature was determined for two stores with 8.2-meter stacks and .sed to
develop the guideline. Temperature and draft were measured at 30 cm above the flue collar for
both stoves. Figure 2.3 1 shows the temperature relationship for the test stoves, as well as a
linear regression through each set of data. Under this guideline, draft is acceptable if the point
is within 2.5 pascals (0.01 inches H,0) of the regression lines. If the draft is unacceptable,
modifications would need to be made to the stove installation until the stack draft-temperature
relationship is within these specifications.

2.4. Final Stress Test Protocol

The final protocol (protocol 6) developed from the tests can be summarized as follows:

Fuel

Type

Split lodgepole pine, as free of knots as possible.

Fuel

Moisture

8 to 12% (dry basis).

Fuel

Length

5/6 of the longest firebox dimension.

Fuel

Configuration

Fuel in center of firebox, packed tightly with smaller fuel on

bottom.

Air Settings Air settings on stove set to maximize burn rate and firebox

temperatures.

Kindling Load Maximum of 30 minutes in duration.

-------
tn

3
O
SI

E
a>
I-

Q)
>
O

x»
aJ

•a

3
O

o
<

10
9
8
7
6
5
4
3
2
1

ieoo°F
1900-F
2000'F

1 2 3 4 5

Time Stressed (days into test)

Figure 2.2-1. Catalyst temperature accumulation - Blaze King stove interim protocol 2.

3
O
xz

r~
C

O
XI
(0

o
a>

10
9
8
7
6
5
4
3
2
1
0

r-r->1

1300°F

1 2 3 4

Time Stressed (days into test)

1400*F
1500*F

Figure 2.2-2. Fire-box temperature accumulation - Blaze King stove interim protocol 2.

-------
V)

3
O

O
XI
(0
Q3

D
O

E
o
l-

o
n
ca

•o

E
d
o
o
<

1300'F

1400*F
1500'F

5 5^1 5 2 5.3 5'4 6.5 5'6 5.7 5 8 5.9 6

Time Stressed (days into test)

Figure 2.2-4. Fire-box temperature accumulation - Blaze King stove interim protocol 3.

-------

oj
cu

•4—»

ro
a
E

o
o
<

10
9
8
7
6
5
4
3
2
1
0

,-r

ieoo°F

6 6!26,46!66,8 7 7'27!47'67!8 8 8^284

1900°F
2000°F

Time Stressed (days into test)

Figure 2.2-5. Catalyst temperature accumulation - Blaze King stove interim protocol 4.

w
O

0)
I—

_Q
(0

•o
a>

E
u
o
o
<

i—/

i—'

1300°F

1400*F
1500*F

Time Stressed (days into test)

Figure 2.2-6. Fire-box temperature accumulation - Blaze King stove interim protocol 4.

-------
w

o
sz

XJ
(0
o
E
i—

T3
CD
j?3
3

E
3
o
o
<

1800T

1900°F

2000"F

9 ,w 11 "" 13 ,~r 15
Time Stressed (days into test)

Figure 2.2-7. Catalyst temperature accumulation - Blaze King stove interim protocol 5.

u
o

a>
>
o
n
to

h-
o

CD
3

3
O

o
<

bO
70
60
50
40
30
20
10

1300°F

1400*F

1500*F

Time Stressed (days into test)

Figure 2.2-8. Fire-box temperature accumulation - Blaze King stove interim protocol 5.

-------
V)

3
O

a>
J-

a>
>

0
_a
aj
a>

E
I—
tj

o
o
<

40
35
30
25
20
15
10
5
0

2
A

2
B

5
H

1800°F

1900-F

2000'F

6 8 10 12 14
Time Stressed (days into test)

Figure 2.2-9. Catalyst temperature accumulation in the Blaze King stove by protocol category.

3
O
.c

o.
E

I—

O
X3
(0

E
I-

(1)

E
3
o
o
<

1300*F

1400*F

1500'F

6 ' 8 * 10 1 '1213141516
Time Stressed (days into test)

Figure 2.2-10. Fire-box temperature accumulation in the Blaze King stove by protocol category.

-------
0.2 0.4 0.6 0.8 1
Stack Temperature (°F)
(x 103)

~8.2 m Stove A <>8.2 m Stove B

Figure 2.3-1. Stack draft - temperature relationship for two stoves used to quantify proper draft during
stress tests. One inch of water is equivalent to 248 pascals.

-------
Length of Bypass
Opening

For catalytic stoves only. Bypass open for additional 5
minutes after stove door is closed.

Stack Draft

±2.5 pascals (0.01 inch H20) of regression lines on stack
draft-temperature graph.

Refueling Criteria Maximum temperature at which entire fuel load can be

consistently loaded.

Fuel Size

70% large, 30% small.

Loading Density

160 kg/m3 (10 lb/ft3) of firebox volume.

Length of Door
Openings

5, 10, 15 minutes (cyclical).

Stack Height

8.2 meters (27 feet).

3. Methods

3.1. Emission Measurements

Emissions testing of stoves before and after stressing was done in SAIC's Beaverton, Oregon
Laboratory. As with the in-home tests conducted in Glenn Falls, New York1,2 and Klamath
Falls, Oregon34 (used for comparison with the laboratory stress data), Automated Woodstove
Emissions Samplers (AWES) were used to conduct the sampling. The AWES system has been
used in approximately 20 in-home studies of residential wood combustion since 1985 to measure
particulate emissions (for examples, see references 1-14). A number of these projects were co-
sponsored by the EPA. Consequently, the system's comparability with Methods 5H and 5G has
been established and quality assurance plans in the standard EPA format have been developed
for it15,\

Figure 3.1 shows a schematic diagram of the AWES system as it is currently configured. The
AWES unit draws flue gases through a 38 cm (15 in.) long, 1.0 cm (3/s in.) O.D. stainless steel
probe which samples from the center of the flue. The sample then travels through a 1.0 cm (3/„
in.) O.D. Teflon line, a heated EPA Method 5-type glass fiber filter for collection of particulate
matter, followed by a sorbent resin (XAD-2) trap for semi-volatile hydrocarbons. Water vapor
is removed by a silica gel trap. Flue gas oxygen concentrations, which are used to determine
flue gas volume, are measured by an electrochemical cell (Lynn Instruments 6100). The AWES
uses a critical orifice (Millipore #XX500001) to maintain a nominal sampling rate of 1.0 liters
per minute (0.035 cfm). The actual flow rate through each AWES critical orifice is measured
with a bubble flow meter to determine the exact sampling rate. Temperatures are monitored

-------
Thermocouple

(Room
Temperature)

Computer/Data Logger

Computer
ConLog Software

Oj Input
Pump Control

Figure 3-1. Schematic of AWES system.

AWES

Heated Chamber

Thermostat
and Heater

Filt- r

0>2 Sensor

Flow
Rcstriclcr

XAJD
Cartridge

Pump

Silica
Gel

Time
Totalizer

Flow
Controller

Thermocouple

Exhaust Return

Inlet

Thermocouples

Calibration
Gas Inlet
Port

1
f

Residential
Wood burning
Device

-------
using type K ground-isolated, stainless-steel-sheathed thermocouples (Omega KMQSS-000U-L).

The AWES unit returns particle-free exhaust gas to the flue via a 0.6 cm ('/4 in.) Teflon line and
the 38 cm stainless steel probe inserted in the flue downstream of the sampling probe. Some
flue gas exiting the AWES is pumped into a 22-liter Tedlar gas bag (Calibrated Instruments,
Inc.) under positive pressure. The flow to 'he bag is controlled by a solenoid valve connected
to the pump circuit and a temperature sens> to allow gas collection only during combustion.
The solenoid valve is open only when the pump is activated and when the flue temperature is
above 38 °C (100 °F), allowing effluent gas to be pumped into the gas bag. The rate of flow
into the bag is controlled by a needle valve rotameter which is adjusted to acquire sufficient gas
for analysis over the entire test without over-pressurizing the bag.

The data acquisition and control system for the AWES system is shown in Figure 3-2. It
consists of a personal computer (PC) containing a data processing board, a terminal box, and
specialized data acquisition and system control software.

The software can be programmed to control, collect, and store the following software settings
and data:

• Establish starting and ending date and length of sampling
period

• Establish pump cycle length and thermocouple cycle recording
interval

• Record up to eight temperatures, including flue gas and
ambient temperatures, averaged over pre-selected intervals

• Record times and weights of fuel loads

• Record flue gas oxygen measurements, averaged over pre-
selected intervals

• Record data to disk

Instantaneous readings of real-time data are also displayed on the system status screen. Date,
time, temperature for as many as ei^ht thermocouples, and flue gas oxygen content are shown.
The most recent 15 sets of recorded data are also displayed.

Prior to emissions testing, the AWES unit is prepared with a new pre-weighed glass fiber filter
and clean XAD-2 sorbent resin cartridge, stainless steel sampling probe, Teflon sampling line,
and Tedlar gas bag. Arter the sampling period, the stainless steel sampling probe. Teflon
sampling line, filler holder, XAD-2 cartridges and Tedlar gas bags are processed as follows:

1. Filters: The glass fiber filter (102 mm in diameter) is
removed from the AWES filter housings and placed in a petri
dish for desiccation and gravimetric analysis for particulate
catch.

-------
N)

Data Acquisition Board

Amplifier/Multiplexer
and Terminal Boards

To Sample Pump

+¦ To Tedlar Bag
Gas Collection
System

Thermocouple
Ports (8)

Figure 3-2. AWES data acquisition and control system.

-------
2. XAD-2 sorbent resin: The sorbent resin cartridge is extracted
in a Soxhlet extractor with dichloromethane for 24 hours. The
extraction solvent is transferred to a tared glass beaker. The
solvent was evaporated in a low-temperature (ambient) air
dryer, the beaker and residue are desiccated, and the residue
is weighed.

3. AWES hardware: All hardware in the sample stream
(stainless steel probe, Teflon sampling line, stainless steel
filter housing, and all other Teflon and stainless steel fittings)
through the top of the sorbent resin cartridge is rinsed with a
solvent mixture of 50% dichloromethane and 50% methanol.

The solvent wash is placed in a tared glass beaker. The
solvent is evaporated in the ambient air dryer, desiccated, and
weighed.

4. Tedlar gas bag: The gas bag is removed and analyzed for
oxygen content (Infrared Industries model 2200), and carbon
monoxide and carbon dioxide content (both with Infrared
Industries model 732).

The EPA Method 5 procedures for desiccation and the weighing schedules are followed for the
solvent drying/desiccation steps.

The sampling equipment is serviced at the start and end of the sampling period. At installation,
leak checks are performed; the thermocouples, scale unit, and oxygen cell are calibrated; and
the data logger is programmed with the proper sampling interval and start/stop times. At the
end of the sampling period, final calibration and leak-check procedures are performed, and the
AWES, sampling line, filter housing, XAD-2 cartridge, sampling probe, and Tedlar bag are
removed and processed.

The sampling probes are located 30 cm (1 ft) above the flue collar. Up to three thermocouples
are active during the emissions tests. One records ambient room temperature. One
thermocouple is placed in the flue 30 cm above the flue collar. In catalytic stoves, one is placed
in the center of the catalyst at 2.5 cm downstream from its front face.

Each emissions test was completed over a 24-hour period. The control system was programmed
to sample one minute out of every two minutes (1 minute sampling, 1 minute not sampling)
Temperatures and flue gas oxygen percent were recorded every two minutes.

Upon completion of each test, the data file is reviewed immediately to check for proper
equipment operation. The data logger data files, log books, and records maintained by
laboratory staff are reviewed to ensure sample and data file integrity.

-------
The data file is used in conjunction with the AWES particulate and gas samples to calculate the
various emission rates and efficiencies.

Mass of particles emitted per mass of dry fuel burned (g/kg) is the key parameter that has been
used to evaluate woodstove performance before and after stressing and to compare in-home stove
performance with the stressed stoves in the laboratory. It is calculated from the AWES data by
the equation:

MP x SV x DF

emissions =

FR x SD

where:

MP = Total mass of particles captured

SV = Stoichiometric volume of fuel. This is the volume of flue gas
which is produced by complete combustion of 1 kg of dry
fuel. This value is calculated from elemental composition of
the fuel and is typically in the 4500-5000 1/kg range for
cordwood. A small correction is made for carbon monoxide
levels characteristic of the various wood burning devices.

DF = Dilution factor. This is a measure of the dilution of stack
gases by the presence of excess air. The dilution factor is
calculated by the following equation:

DF = 20.9% / (20.9% - average stack oxygen% during burn)

FR = Sampler flow rate. The sampling system is designed to
sample at a nominal rate of 1 l/min, however, each sampler
samples at a slightly different rate and is measured for each
sampler.

SD - Sampling duration. This is the length of time which the
sampler was on during active combustion (defined by the flue
temperature being over 38 °C (100 °F| measured at 30 cm
above the flue collar).

3.2. Quality Assurance

3.2.1. Calibration Procedures and Frequency

This section addresses the calibration procedures for the sampling equipment. Applicable
calibrations were performed in conformance with the EPA publication "Quality Assurance
Handbook for Air Pollution Measurement Systems, Volume III, Stationary Source Specific
Methods".17

-------
3.2.1.1. Temperature Measuring Device Calibration

During sampling, accurate temperature measurements were required. Individual type "k"
thermocouple temperature sensors were calibrated using an ice bath and a boiling water bath.
The CON LOG thermocouple read-out was calibrated every two weeks during the project with
an electronic thermocouple simulator (OMEGA Engineering, Inc., Model CL-300-2100F).

3.2.1.2. A WES Unit Flow Rate

A bubble meter was used as a transfer standard to calibrate individual AWES unit flow rates.
The bubble meter is used to measure the orifice flow rate of each AWES unit. The meter has
a liquid film and a known volume. The AWES unit is first leak-checked and any leaks are
repaired before a calibration is conducted. The bubble meter is connected to inlet of the AWES
unit. The AWES is set up in normal testing configuration with sample line, filter, XAD, and
silica gel in place. The liquid film is interposed in the flow path. While the air flow causes the
film to move from one volume mark on the tube to another, the travel time is measured with a
stopwatch. Travel time and volume traveled by the thin film, along with barometric pressure,
ambient temperature, and vapor pressure of H20 at ambient conditions were then used to
calculate flow rate at standard conditions.

3.2.1.3. Analytical Balance Calibration

The analytical balances were calibrated over the expected range of use with standard weights
(NBS Class S) prior to use each day. Measured values were within ±0.1 milligram.

3.2.1.4. Wood Moisture Meter

The Delmhorst moisture meter was calibrated through adjustment to its internal circuitry. This
calibration is gauged by applying known resistances to the contact pins of the meter. Readings
which deviate from the values associated with the known resistance standards by more than one
percent (absolute) require factory overhaul and/or adjustment. No factory adjustments were
necessary on this project.

3.2.2. Quality Assurance Audits

Systems and performance audits were not conducted for this project. A technical systems audit
(TSA) and a performance evaluation audit (PEA) of the AWES sampling and analytical system
were most recently performed in 1992 by RTI.

3.2.3. QA/QC Checks of Data Reduction

The quality assurance officer for the project performed an independent check of the calculations
with predetermined data before the field tests were conducted. This ensured that calculations
done in the field were correct. Independent checks were also conducted by a staff member

-------
designated by the QA officer to assure that data were being recorded accurately. After the field
tests, the QA officer checked the data input to assure that the raw data were transferred to the
computer accurately.

3.2.4. Sample Identification and Custody

After collecting and recovering samples, the field team members labeled the samples and
completed all chain-of-custody forms.

3.2.5. Measurement Uncertainty

From the uncertainty of the individual measurement parameters shown in Table 3-1, the
uncertainty for the emissions values can be calculated. The standard partial derivative approach
is used, i.e., if...

F = / (xp V' Xn>

then dF = — dx, + — dx, + ... + —dxn
6x, 1 bx^2 6*n '

where: dx„ = uncertainty in individual measurement, and

dF = uncertainty in final value, if absolute values are used; i.e.

Uncertainty Value » 1(uncertainty *.) + [(uncertainty *,) +... + 1-^- ((uncertainty xj
fix, fix, ^ fix.

It should be noted that the final uncertainty values obtained from the above equation represent
the maximum probable propagated uncertainties (i.e., assuming totally dependant variables with
additive uncertainties) rather than the most probable propagated uncertainties (which would
require a root mean square analysis with a covariance term expansion). When variables are
independent of each other, the uncertainties will, to some extent, cancel each other out. Because
the degree of dependence and corresponding covariance terms between many of the variables
are difficult to estimate, a conservative (maximum probable uncertainty) approach has been taken
in this study. The uncertainties for woodstove particulate emission values are typically about
20 percent of the reported value.

-------
Table 3-1. Parameters Measured and Associated Uncertainty

Measured Parameter

Units

Measurement Uncertainty

Mass of filter

0.0005 g

Mass of rinse

0.0005 g

Mass of XAD-2 extraction

0.0005 g

Mass of blank

0.0005 g

Stoichiometric volume

1/dry kg

10% (relative)

AWES flow rate

l/min

1 % (relative)

Sampling duration

min

1.7% (relative) (one second per
minute)

Stack oxygen

0.1% (absolute)

Gas bag oxygen

0.02% (absolute)

Stack temperature

°F

5 °F (3 °C)

Mass of fuel (wet)

0.1 kg x n (n = # of fuel loads)

Moisture content of fuel

5% (absolute)

Sampling period

hours

1 % (relative)

-------
4. Results

Five stoves were stress-tested using protocol 6. Data for one stove used in the development of
the final stress test protocol (Blaze King Royal Heir #1) are presented and represent the effect
of protocols 2 through 5. Each stove was emissions-tested prior to stressing and once again
afterwards. Some stoves underwent extended stress testing. A summary of the physical
degradation observed is provided in Table 4-1. Particulate emissions and a more detailed
description of physical degradation (both in homes and in the laboratory) are provided by stove
model in sections 4.1 through 4.6.

4.1. Blaze King Royal Heir (Stove #1)

This stove was used to develop protocol 6 and was stressed for 14 days using protocols 2
through 5. Initial particulate emissions were 4.68 g/kg. After the 14-days of cumulative
stressing, the particulate emissions rose 80% to 8.41 g/kg (Figure 4.1-1).

There was physical damage to the stove due to this stressing. After day 4 of stress, some
oxidation and warpage were noticed on the bypass region of the firebox. This oxidation and
warpage continued to worsen throughout the remainder of the test (Figure 4.1-2). After day 14,
the bypass had warped enough to create a 0.64-cm gap on the left front of the bypass (Figure

4.1-3).

The five Blaze King Royal Heir stoves examined in the Glens Falls study (Y01, Y02, Y12, Y23,
and Y25) had emissions, after one season of use, ranging from 5.0 to 16.9 g/kg (Table 4.1-1).
The stove with the lowest emissions (5.0 g/kg) was retested after a second season of use.
Emissions rose to 17.1 g/kg at the end of the second season for that stove. All stoves showed
some oxidation in the firebox and the bypass gap caused by warpage ranged from 0 to 1.6 cm
(Figures 4.1-4 and 4.1-5). Both catalyst plugging and flaking were seen in stoves in homes after
one heating season (Figure 4.1-6 and Figure 4.1-7). In general, similar degradation trends in
physical condition and particulate emissions were seen on the Blaze King stove stressed for 14
days and in the stoves in homes after one season of use.

4.2. Blaze King Royal Heir (Stove #2)

This stove was stressed for 10 days using protocol 6. Particulate emissions from this stove
increased from 4.14 g/kg to 21.02 g/kg—over 400% increase (Figure 4.2-1). This stove showed
physical damage similar to the first Blaze King stove tested, as well as other field stoves.
Midway through the test, accumulation of catalyst temperatures over 982 °C (1800 °F) ceased
(which also occurred in the stove in home Y02 of the Glens Falls study). This is an indicator
of declined catalytic activity.

Particulate emissions from this stove were somewhat higher than those measured in the field
stoves after one season of use (Table 4.1-1).

-------
Table 4-1. Observations of Physical Damage Due to Stressing

Duration of
Stressing

Bl?ze King Ro,al Heir

Blaze King
Royal Heir
#2

Country Flame
BBF-6

Regency R3/R9

Quadrafire 3100

Earth Stove 1003C

4 dayf

Some oxidation &
warpage

7 days

Continued oxidation &
warpage

Catalyst tested; still
fully active

None

10 days

Bypass gap
of 0.64 cm

Baffle plate
oxidized &
moderately warped
(matches Y20,

Y24)

Minor warping &
oxidation of
baffle &
secondary air
tubes

Some warping & oxidation of
catalyst holder

14 days

Bypass gap of 0.64 cm

Test

complete

Catalyst holder oxidized &
slightly warped (similar to
Y14)

20 days2

Test complete

Extensive baffle
warpage

Major warping &
oxidation of
baffle &
secondary air
tubes

Failure of bypass mechanism
(stuck open)

Severe oxidation & warping of
catalyst holder
Warped door frame (not
airtight)

25 days

Catalyst holder oxidized &
warped (identical to Y14
after two seasons)

Test complete

35 days

Still no bypass gap

1. Protocols 2 through 5 were used with the Blaze King Royal Heir #1; all others used protocol 6.

2. Observation after 18 days of stressing for the Regency R3/R9 stove.

-------
25

£0

c
o

CO
CO

UU
CD

¦4—•

_cd

=3
O

-t—»
i—

05
D_

4.68 g/kg

+ 80%
8.41 g/kg

1 3 5 7 9 11 13

Days of Stress

Figure 4.1-1. Paniculate emissions from Blaze King Royal Heir (stove #1) before and after stressing
(protocols 2-5).

-------
Figure 4.1-2. Oxidation of top of firebox in Blaze King Royal Heii (stove tt 1) after 14 days of stressing
in the laboratory. Thermocouple probes used to monitor catalyst and firebox temperatures are visible.

Figure4.1-3. Bypass gap (0 64 cm) in Bla/c King Royal Heir (stove #1) after 14 days of stressing in the
laboratory. Nut, washer, and attachment rod were needed in order to move the bypass damper after warpagc
was visible.

-------
Table 4.1-1. Particulate Emissions and Bypass Gaps in Blaze King Royal Heir Stoves

Field: After One Season

Home Code

End-of-Season Emissions
(g/kg)

Front Bypass Gap
(cm)

Y02

10.8

1.6

Y23

14.2

0.41

Y12

9.0

0.23

Y25

16.9

None

Y01

5.0 (17.1)'

None

Lab: Blaze King #1

New

4.68

None

Day 14

8.41

0.64

Lab: Blaze King #2

New

4.14

None

Day 14

21.02

0.64

1. End of second season.

-------
Figure 4 1 4. Oxidation in firebox in Blaze King Royal Heir (Home YOI) after one heating season of
use

Figure 4 1-5 Bypass gap (1.6 cm) in Blaze King Royal Heir (Home V02) after one heating season of
use.

-------
v i

¦*Vt v
•v«>« »lr#
-------
o>

c/5

c
o

CO
CO

-------
In an effort to determine the contributing sources of the particulate emission elevations, new
catalysts were tested in the two stressed Blaze King stoves, and the two stressed catalysts (as
well as the catalyst removed from the stove in home Y02) were tested in a third, new Blaze King
Royal Heir stove. The results from these tests are shown in Figure 4.2-2 and are compared to
particulate emissions obtained from the two Blaze King stoves used in the stress test prior to
stressing.

From Figure 4.2-2, it can be seen that the bypass leakage of the stressed stove and the
deteriorated catalyst performance each contribute to particulate emission elevation. It should be
noted that the value for the Blaze King stove #1 in the stressed stove/stressed catalyst
configuration appears to be an outlier.

4.3. Country Flame BBF-6

This stove was stressed for a total of 35 days using protocol 6. Emission tests were performed
prior to stressing as well as after 7, 14, 25, and 35 days of stressing. Particulate emissions
values in these tests were 1.47, 3.14, 1.88, 2.24, and 6.54 g/kg, respectively (Figure 4.3-1).

Limited damage to this stove was caused by this test. After day 14, the flame impingement
shield and the catalyst holder became slightly warped and oxidized. This is similar to the
condition of the stove in home Y14 in the Glens Falls study after one season of use. The flame
impingement shield and catalyst holder continued to oxidize and warp throughout the remainder
of the stress period. At day 25, their condition was nearly identical to the condition of the stove
in home Y14 after two seasons of use (Figures 4.3-2 through 4.3-5). After day 35, the bypass
was examined and found to have no gaps or warping. A small and variable amount of widening
of the bypass gap was seen in the stoves in the field (Table 4.3-1). Emissions were also variable
among the stoves in the field, and with the exception of one stove (Home Y13), were generally
low. It was hypothesized that the unusually poor performance of Y13 resulted from the
homeowner continually operating the stove at maximum burn rates.

4.4. Regency R3/R9

The Regency stove was stressed for a total of 18 days with emission tests after 10 and 18 days.
Emission testing prior to stressing was not done. Particulate emissions values from these two
tests were 2.47 g/kg and 5.83 g/kg, respectively (Figure 4.4-1).

Physical damage to ti.e Regency was limited to the baffle plate at the top of the firebox. After
10 days of stress, the baffle was oxidised and moderately warped, a condition very similar to
conditions found in stoves in homes Y20 and Y24 after one season of use (Figure 4.4-2). After
day 18, the baffle warpage was so extensive that it no longer stayed in place (Figure 4.4-3).

After one season of use, the five Regency stoves in the Glens Falls study had particulate
emissions ranging from 3.5 to 16.3 g/kg, the iatter considerably higher than the values obtained
in the lab, even after 18 days (Table 4.4-1).

-------
25

21.02

W
CO

LU
O

3
O

¦c

(0
DL

5
0

4.68

BK 1

4.14

BK2

New Stoves
New Catalysts

10.69

9.80

8.78

BK 1 BK 2 Y02

12.42

BK 1

15.77

9.18

BK 1 BK 2

New Stoves Stressed Cat. Stressed Stoves

Stressed Cats. Bypass Leaks New Catalysts

Plugged (Bypass Leakage)

8.41

BK 1 BK 2

Stressed Stoves
Stressed Cats.

Figure 4 2-2. Comparison of particulate emissions due to bypass leaks and degraded catalysts in Blaze King Royal Heir stoves. BK 1
Blaze King BK 2 = Blaze King #2, Y02 is home in Glenns Falls from which a catalyst was removed from stove and used in
laboratory tests with new stove.

-------
CD
XL

05
C

¦4—•

_T0

O
V-¦

CO
Q.

20-

3.14 g/kg

+ 345%
6.54 g/kg

J2.24 g/kg
1 -4^ g/kg ^,.88 g/kg , , L

0 5 10 15 20 25 30 35

Days of Stress

Figure 4.3-1. Particulate emissions from Country Flame BBF-6 stove before and after stressing.

-------
Figure 4 3 2 Flame impingement shield warpage in Country Flame BBF-6 after 25 da>s of stressing

Figure 4 3-3 Catalyst holder oxidation and warpage in Country Flame BBF-6 after 25 days of stressing.

-------
Figure 4.3-4. Flame impingement shield warpage in Country Flame BBF-6 (Home Y14) after two
heating seasons of use.

Figure 4.3-5. Catalyst holder oxidation and warpage in Country Flame BBF-6 (Home Y14) after two
heating seasons of use.

-------
Table 4.3-1. Paniculate Emissions and Bypass Gaps in Country Flame BBF-6 Stoves
Field: After One Season

Home
Code

End-of-Season
Emissions
(g/kg)

Front
Bypass Gap
(cm)

Y14

4.8

0.254

Y13

11.4

0.102

Y10

7.4

0.178

Y07

4.4

0.051

Y19

2.0

0.152

End-of-Season
Emissions
(g/kg)

Front
Bypass Gap
(cm)

7.02

0.508

26.32

7.39

1.56

Laboratory

New

1.47

None

Day 7

3.14

None

Day 14

1.88

None

Day 25

2.24

None

Day 35

6.54

None

-------
25

+136%
5.83 g/kg

2.47 g/kg

Days of Stress

Figure 4.4-1. Particulate emissions from Regency R3/R9 stove during stressing.

-------
Figure 4.4-2. Warped baffle removed from regency R3/R9 stove (Home Y20) alter one heating season.

-------
Table 4.4-1. Particulate Emissions and Baffle Warpage in Regency R3/R9 Stoves

Field: After One Season

Home Code

End-of-Season Emissions
(g/kg)

Baffle Warpage

Y20

3.7

Major

Y24

15.1

None

Y15

3.5

None

Y16

14.4

Minor

Y17

16.3

None

Lab

New

No Data

Day 10

2.47

Minor

Day 18

5.83

Major

-------
To test the effect of baffle warpage on particulate emissions, the stressed Regency stove was
tested with a new baffle as well as with no baffle at all (Figure 4.4-4). Particulate emissions
were 3.80 g/kg with the new baffle and 8.30 g/kg with no baffle. These two values would be
the extreme cases of baffle effects. As expected, the particulate emissions value with the warped
baffle (5.85 g/kg) fell between the two extreme values.

4.5. Quadrafire 3100

The Quadrafire stove underwent 20 days of protocol 6 stress testing with emission tests prior
to stressing as well as after 10 and 20 days. Particulate emissions from these three tests were
8.41 g/kg, 4.71 g/kg, and 11.85 g/kg, respectively (Figure 4.5-1).

After 10 days of stressing, physical damage was limited to minor warping of the front lip of the
baffle as well as some oxidation on its lower surface (Figure 4.5-2). Similar conditions were
seen in the stove in home KF1 from field studies conducted in Klamath Falls, Oregon3-4. After
20 days, the oxidation and warping continued until the front center lip of the baffle was warped
down about two inches. The secondary air tubes also showed some warping (Figures 4.5-3 and
4.5-4). This damage is much greater than seen in the stove in home KF1 after two seasons of
use4.

Particulate emissions after one and two seasons of use from the stove in home KF1 were 2.60
and 5.04 g/kg, respectively (Table 4.5-1). The emission value and physical damage seen after
one to two seasons of use are similar to what was seen in the laboratory after 10 days of
stressing. The laboratory value of 11.85 g/kg after 20 days of stressing is significantly higher
than the field values and the stove used in the laboratory had greater baffle damage than the
stove in the field.

4.6. Earth Stove 1003C

The Earth Stove was stressed for 20 days using protocol 6. Emissions tests were done prior to
stressing as well as after 10 and 20 days. Particulate emissions values for these tests were 4.52,
6.06, and 16.77 g/kg, respectively (Figure 4.6-1).

After ten days of stress, there was limited warping of the catalyst holder, similar to the stove
in home KF2 in the Klamath Falls, Oregon study after one season of use. During the next 10
days of stressing, however, the bypass mechanism failed (Figure 4.6-2). This was due to
excessive warping of both the bypass plate itself and the seat for the bypass. The back edge of
the bypass plate warped downward and the seat warped upward, making . insure of the bypass
impossible. In addition to this, the bypass lever came detached from the bypass plate, rendering
the bypass totally inoperable. Other damage was also noted after 20 days of stressing. This was
oxidation and warpage of the catalyst holder (Figure 4.6-3) and warpage of the door frame
(Figure 4.6-4).

-------
U1

O)
CO

¦o

Warped Baffle

No Baffle

New Baffle

Figure 4.4-4. Effect of baffle warpage on particulate emissions from Regency R3/R9 stove. Emissions tested after 18 days of stressing.

-------
Days of Stress

Figure 4.5-1. Particulate emissions from Quadrafire 3ICO stove before and after stressing.

-------
Figure 4.5 2 Oxidation and minor waipage of secondary air tubes and baffle in the Quadrafire 3100
after 10 days of stressing

Figure 4.5 3 Oxidation and major warpage of secondary air tubes and baffle in the Quadrafire 3100
after 20 days of stressing

-------
Figure 4.5-4. Warpage after 20 days of stressing, as seen from top of stove, in the Quadrafire 3100
stove. Pen illustrates location of depression.

-------
Table 4.5-1. Particulate Emissions and
Baffle and Secondary Air Tube Warpage in Quadrafire 3100 Stoves

Field: After One Season

Home Code

End-of-Season Emissions
(g/kg)

Warpage

KF1

2.6

Slight

Field: After Two Seasons

| KF1

5.04

Moderate 1

Lab

New

8.41

Day 10

4.71

Moderate

Day 20

11.85

Extensive

-------
25

+ 271%
16.77 g/kg

6.06 g/kg

4.52 g/kg

10 14 18

Days of Stress

Figure 4.6-1. Particulate emissions from Earth Stove 1003C stove before and after stressing.

-------
56

-------
F igure 4 6 3 Oxidation and waipuge of catalyst holdei in ihe Barth Stove I003C after 20 days of
stressing

Figure 4,6 4 Warpage ot door frame m the Earth Stove HX)3C after 20 days of stressing.

-------
The particulate emissions values from the stove in home KF2 after one season of use (8.9 g/kg) and
after two seasons of use (6.66 g/kg) are not dramatically higher than the laboratory stove's initial
paniculate emission value (4.52 g/kg) (Table 4.6-1) and particulate emission value after 10 days of

Table 4.6-1. Particulate Emissions and Bypass Gaps in Earth Stove 1003C Stoves
Field: After One Season

Home Code

End-of-Season Emissions
(g/kg)

Bypass Gap
(cm)

KF2

8.9

—

Field: After Two Seasons

| KF2

6.66

0.005 |

Lab

New

4.52

Day 10

6.06

No Data

Day 20

16.77

Large (undetermined)

stress (6.06 g/kg). However, the particulate emission value after 20 days of stressing (16.77 g/kg)
is much higher and is a direct reflection on the failure of the bypass mechanism.

5. Conclusions

Considerable variation in woodstove degradation has been observed in home usage. In some cases,
degradation was observed to be more severe than that produced by the in-laboratory stress test
protocol. In other cases, little in-home degradation was observed even after two heating seasons.
Such variability is not surprising in light of the differences in installation, use habits, and fuel types
seen with woodstoves. The research presented here has shown that the deterioration similar to that
caused by 10 days (240 hours) of stressing a stove following protocol 6 is a reasonable predictor of
the deterioration which may be seen under the more extreme in-home usage conditions after one

-------
heating season. Each of the protocol variables has been quantified so that the protocol can be
standardized and used in a reproducible manner. This protocol can be used as a tool to estimate
particulate emissions from a population of aging stoves. It can also be used by stove manufacturers
during the design stage to ensure that a durable stove with low emissions over a long period of use is
produced.

6. Acknowledgments

The authors wish to thank several people for their contributions to this research: Craig Chase
- consultant; Dave Collier, Steve Crane, and John Kowalczyk - Oregon Department of Environmental
Quality; John Crouch and Gary Satterfield - Hearth Products Association; Pat Fox - Bonneville Power
Administration; A. C. S. Hayden - Canadian Combustion and Carbonization Research Laboratory; Phil
Lusk and Russell O'Connell - Coalition of Northeastern Governors Policy Research Center; Steve
Morgan - Citizens Conservation Corp.; Jeff Peterson - New York State Energy Research and
Development Authority; and Alex Sifford - Oregon Department of Energy.

In addition to the Work Assignments under EPA contract 68-D0-0120, identified on the title page of
this report, OMNI Environmental's effort was also funded under: EPA Cooperative Agreement
CR815271-01-0 (subcontract to New York State Energy Research and Development Authority), EPA
Contract 68-02-4464, WA 1-007 and 1-C30 (subcontract to Pacific Environmental Services, Inc.), EPA
Purchase Order 3D0954NTSA, and EPA Contract 68-D0-0141, Task 93-194 (subcontract to Acurex
Environmental Corp.).

The authors would also like to acknowledge the sponsorship of the research by:

• Department of Energy - Pacific Northwest and Alaska Region Biomass Energy Program
(administered by the Bonneville Power Administration)

• Coalition of Northeastern Governors Policy Research Center

• Canadian Combustion and Carbonization Research Laboratory/Canada Centre for Mineral and
Energy Technology/Energy, Mines and Resources

• Hearth Products Association (formerly Wood Heating Alliance)

• New York State Energy Research and Development Authority

• Oregon Department of Energy

• Oregon Department of Environmental Quality

7. References

1. Bamett, S. G., 1990, Field Performance of Advanced Technology Woodstoves in Glens Falls.
NY. 1988-1989. Volumes 1 and II. EPA-600/7-90-019a and -019b (NTIS PB91-125641 and
-125658).

2. Barnett, S.G. and J. Fesperman, 1990, Field Performance of Advanced Technology Woodstoves
in Their Second Season of Use in Glen Falls, New York, 1990, OMNI Environmental Services,
Inc. report to Canada Centre for Minerals and Energy Technology, Energy, Mines, and
Resources.

-------
3. OMNI Environmental Services, Inc., 1990, In-Home Evaluation of Emission Characteristics
of EPA-Certified, High-Technology, Non-Catalytic Woodstoves in Klamath Falls, Oregon,
report to EMRC, 145Q.23440-9-9230.

4. Barnett, S. G. and R. D. Bighouse, 1992, In-Home Demonstration of the Reduction of
Woodstove Emissions from the Use of Densified Logs, OMNI Environmental Services, Inc.
report to BPA, DOE/BP-35836-1.

5. OMNI Environmental Services, Inc., 1987, An In Situ Performance Evaluation of Two
Woodstove Catalytic Retrofit Devices, report to Oregon Department of Environmental
Quality.

6. Burnet, P. G., 1987, The Northeast Cooperative Woodstove Study: Volumes I and II, EPA-
600A7-87-026a and -026b (NTIS PB88-140769 and -140777).

7. OMNI Environmental Services, Inc., 1987, Whitehorse Efficient Woodheat Demonstration,
report to the City of Whitehorse and EMRC.

8. Simons, C. A.; P. D. Christiansen; L. C. Pritchett; and G. A. Beyerman; 1988, Woodstove
Emissions Sampling Methods Comparability Analysis and In Situ Evaluation of New
Technology Woodstove, OMNI Environmental Services, Inc. report to BPA Pacific
Northwest and Alaska Regional Biomass Energy Program, DE-AC79-85BP18508.

9. OMNI Environmental Services, Inc., 1989, Performance Evaluation of the Best Existing
Stove Technology (B.E.S.T.) Hybrid Woodstove and Catalytic Retrofit Device, report to
Oregon Department of Environmental Quality.

10. Barnett, S. G. and R. P. Roholt, 1990, In-Home Performance of Certified Pellet Stoves in
Medford and Klamath Falls, Oregon, OMNI Environmental Services, Inc. report to BPA,
DOE/BP-04143-1.

11. Barnett, S. G.; R. P. Roholt; and J. E. Houck, 1990, (1) Field Performance of Best Existing
Stove Technology (B.E.S.T.) Hybrid Woodstoves in Their Second Year of Use, and (2)
Evaluation of a Modified Evacuated Cylinder Particulate Emissions Sampler (MESC), OMNI
Environmental Services, Inc. report to ODEQ.

12. Elements Unlimited, 1990, Woodstove Field Performance in Klamath Falls, Oregon, report
to the Wood Heating Alliance.

13. Barnett, S. G., 1991, In-Home Evaluation of Emissions from Masonry Fireplaces and
Heaters, OMNI Environmental Services, Inc. report to Western States Clay Product
Association.

-------
14. Barnett, S. G. and P. G. Fields, 1991, In-Home Performance of Exempt Pellet Stoves in
Medford, Oregon, OMNI Environmental Services, Inc. report to BPA, DOE/BP-04143-2.

15. Houck, J. E., 1986, Quality Assurance Plan, Performance Monitoring of Catalyst Stoves,
Add-ons, and High Efficiency Stoves: Field Testing for Fuel Savings, Creosote Build-up, and
Emissions, CONEG Policy Research Center, Inc.

16. Houck, J. E., 1987, Interim Quality Assurance Plan for Northwest Woodstove Study, U.S.
EPA Contract No. 68-02-3996.

17. Von Lehmden, D. J.; W. G. DeWeese; and C. Nelson, 1979, Quality Assurance Handbook
for Air Pollution Measurement Systems, VoL III. Stationary Source Specific Methods, EPA-
600/4-77-027b (NTIS PB80-112303).

-------
TECHNICAL REPORT DATA

/Please read InOructions on the reverse before eompleri, || | |||| || |||||| 1II11 III 11II1 III

1 . REFORT NO. 2

EPA-600/R-94-193

3. f in mini mm mil hi i¦ iii in

, PB95-136164

4. TITLE AND SUBTITLE

Woodstove Durability Testing Protocol

5. REPORT DATE

November 1994

6. PERFORMING ORGANIZATION CODE

7. AUTHOR(S)

Roger D. Bighouse, Stockton G. Barnett, James E.
Houck, and Paul E. Tiegs

8. PERFORMING ORGANIZATION REPORT NO.

9. PERFORMING ORGANIZATION NAME AND ADDRESS

OMNI Environmental Services, Inc.
10074 SW Arctic Drive
Beaverton, Oregon 97005

10. PROGRAM ELEMENT NO.

11. CONTRACT/GRANT NO.

68-DO-1020, Tasks 1-28,-41,
-42, and -58 (E. H. Pechan)

12. SPONSORING AGENCY NAME AND ADDRESS

EPA, Office of Research and Development

Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711

13. TYPE OF REPORT AND PERIOD COVERED

Task Final; 6/89-2/83

14. SPONSORING AGENCY CODE

EPA/600/13

is. supplementary notes AEERL project officer is Robert C. McCrillis, Mail Drop 61. 919/
541-2733.

i«. abstract rj-^e rep0rt discusses the development of an accelerated laboratory test to
simulate in-home woodstove aging and degradation. Known as a stress test, the pro-
tocol determines the long-term durability of woodstove models in a 1- to 2-week time
frame. Two avenues of research have been taken in developing the stress test. First,
the performance of woodstoves in actual in-home use has been observed during two
heating seasons in three communities: Medford and Klamath Falls, OR, and Glens
Falls, NY. Eight models of stoves in 13 homes were studied. The field studies per-
mitted records of woodstove operating temperatures, particulate emission levels,
and (in some cases) physical degradation tc be followed in a real-world setting. The
second line of research was the laboratory ^stressing* of various woodstove models
under high terr perature operation. This laboratory research has been conducted on
six stoves (five models) and, as with the in-home research, changes in particulate
emission rates were measured and physical degradation documented. Both catalytic
and noncatalytic stove models, including EPA Phase 2 certified stoves, were repre-
sented in the tests.

17. KEY WORDS AND DOCUMENT ANALYSIS

a. DESCRIPTORS

b. 1DENTIFIERS/OPEN ENDED TERMS

c. cos ATI Field/Croup

Pollution Degradation

Stoves Particles

Wood

Stresses

Tests

Aging

Pollution Control
Stationary Sources
Woodstoves
Particulate

13 B 14G

13 A
11L
20L

14 B
13 H

18. DISTRIBUTION STATEMENT

Release to Public

19. SECURITY CLASS /This Report)

Unclassified

21. NO. OF PAGES

20. SECURITY CLASS (This page)

Unclassified

22. PRICE

EPA Form 2220-1 (»-73)

-------