EFFICIENCY STUDY OF A CONTRAFLOW MASONRY WOOD­BURNING HEATER 
 
 
Peter M. Hanley  Christopher Nielson 
Department of Architecture  Department of Architecture 
University of Oregon  University of Oregon 
Eugene, OR  97403  Eugene, OR  97403 
[email protected]  [email protected] 
   
Devin Saez  Hank Warneck 
Department of Architecture  Department of Architecture 
University of Oregon  University of Oregon 
Eugene, OR  97403  Eugene, OR  97403 
[email protected]  [email protected] 
 
Department of Architecture 
University Of Oregon 
Eugene, OR 97403 

 
ABSTRACT create an internal draft, constantly stoking the fire,
producing higher burning temperatures than traditional
This study determines the efficiency of a Finnish contraflow woodstoves. A higher burning temperature increases
masonry wood-burning heater in a residential setting in combustion efficiency by more thoroughly consuming small
Pleasant Hill, Oregon. particulates and combustible gasses. Secondly, after the
combustion process, contraflow masonry stoves increase
Data concerning the thermal properties of the home were efficiency by sending the heat-bearing exhaust gas on a
collected between February 15th and February 23rd of 2009. circuitous route through channels within the masonry
This data was used in conjunction with calculations of heat structure surrounding the combustion chamber before
gain and loss to determine an operating efficiency of 79.5 % exiting out the chimney.
for the contraflow masonry wood-burning heater.

1. INTRODUCTION

Contraflow masonry stoves are thermally massive wood-

burning stoves specifically designed to increase the
efficiency of wood combustion for the purposes of Figure 1.
residential heating. Contraflow masonry wood stoves were Section of a
first produced in Finland and other regions of Northern contraflow
Europe in the 17th and 18th centuries as a response to a masonry stove,
declining supply of wood resources (Tulikivi, 2008). showing the
circuitous route
Contraflow masonry stoves are large wood-burning stoves taken by the
with several specially designed efficiency increasing exhaust gasses
thermal characteristics used for residential heating and within the
baking. They were designed in an attempt to provide masonry
superior efficiency compared to traditional woodstoves or structure.
simple fires. (Modified from
Stein, et al. 2005
Contraflow masonry stoves increase the heating efficiency p. 334)
of wood combustion through the employment of several key
strategies. First, through a controlled air intake, the stoves
Through this process heat is transferred to and stored in the
thermal-mass of the fireplace structure itself. This heat is
then slowly radiated out into the surrounding area over the
course of several hours, or even days (Tulikivi, 2008).

This study will focus on determining the efficiency of a

contraflow masonry stove in a typical, active, single-family
home context.

The Barkman family of Pleasant Hill Oregon allowed the

use of their home and their Tulikivi contraflow masonry
stove for the study. The home has a mostly open floor plan
on the first floor, making it a good candidate for efficient
wood heating as the heat easily spreads from the fireplace
through the rest of the space. The stove faces a kitchen
island and, according to the homeowner, becomes a social Figure 3. First floor plan of the Barkman residence showing
hub around the time of lighting fires. Typically, a single fire the location of the contaflow masonry stove and data loggers.
is lit every evening in the winter. On the coldest of days a
second fire in the morning is required. The fires burn for

To isolate the heat produced by the stove, other internal heat

sources (solar gain, occupant and appliance heat) were
calculated and subtracted from the net heat loss of the
building. The remaining figure represents the net heat loss
and therefore the total heat released into the home by the
stove.

The energy consumed by the stove was calculated by

applying an assumed constant energy content in BTUs per
pound of quarter split Douglas fir to the amount of wood
Figure 2. Photograph of Barkman family contraflow used during the observation period. The homeowners
masonry heater. Pleasant Hill, Oregon. weighed and recorded the amount of wood burned each day
and at what times the burns occurred on a data sheet.

roughly one hour. The ratio between the energy consumed by the stove and the
net heat loss of the home was then calculated. This
2. HYPOTHESIS difference is presumably due to a combination of inefficient
combustion and heat loss up the chimney. The ratio of net
The contraflow masonry stove operates at 90% efficiency. heat loss of the home to energy available from the fuel
represents the efficiency of the stove.
3. METHODOLOGY

To accurately calculate the stove’s efficiency, the available 4. RESULTS

energy from the fuel consumed by the stove was compared
to the net heat loss of the home. Results are broken into four sections: temperature data
collected, heat loss, heat gains and measured fuel.
For the duration of a 7-day observation period, 5 HOBO H8
Pro Series data loggers were placed in the study area as well 4.1 Temperature Data Collected
as outside to measure internal and external temperature.
The average internal and external temperatures were used to Temperature recorded by the data loggers shows a distinct
calculate an average ∆t for this study. Heat loss though the day and night swing as well as small peaks in inside
envelope was calculated using this ∆t, areas derived from temperature after each fire.
architectural drawings of the house, and published R and U
values of pertinent envelope assemblies.
 Figure 4. Inside VS Outside Temperature over time graph with fire burning events.

4.2 Heat Loss Total wall area from design drawings (minus doors and
windows) = 2101 ft2.
4.2.1 Indoor/Outdoor Temperature
U A ∆t = 0.0422 x 2101 x 21.27 = 1885 Btu/hr.
The interior design temperature selected for this analysis
was 65.910F. This temperature was chosen because this was 4.2.3 Doors
the found average indoor temperature. The exterior average
temperature was 44.640F, that gives 65.9 – 44.6 = 21.270F = Total door area from design drawings = 179.9 ft2
∆t. U Value of Door (wood door with light) = 0.39 BTU/0F ft2
hr.
4.2.2 Opaque Above-Ground Walls
U A ∆t = 0.39 x 179.9 x 21.27 = 1492 Btu/hr.
TABLE 1
4.2.4 Windows
Component R-Value oF ft2 h /
Btu Total window area from design drawings = 515 ft2
Exterior Moving Air Film .17 U Value of window (double glaze, ½”) = .51 BTU/0F ft2 hr.
(winter)
Wood Shingles .87 U A ∆t = 0.51 x 400 x 21.27 = 5587 Btu/hr.
Vapor Permeable Felt .06
1/2” Plywood Sheathing .62 4.2.5 Roof
5 ½” Fiberglass Batt R-21 21
1/2” Gypsum Board .32 Insulated Ceiling area: 1707 ft2
Interior Still Air Film .68 U Value of roof system with R-30 batt from Stein, et al. =
Total 23.72 oF ft2 h / .034 BTU/0F ft2 hr.
Btu
U = 1/R = 0.0422 U A ∆t = 0.034 x 1707 x 21.27 = 1235 Btu/hr.
(Source: Stein, et al. 2005 pp. 1549-1567)
4.2.6 Floor
 1day
East: 163Btu / ft 2 / day×76 ft 2 × = 516 Btu/hr
2 24 hr
Total floor area from design drawings = 1707 ft
U Value of floor system from Stein, et. al. = .074 BTU/0F ft2 West: 163Btu / ft 2 / day×207 ft 2 × 1day = 890 Btu/hr
24 hr
hr. Total: = 5308 Btu/hr

U A ∆t = 0.074 x 1707 x 21.27 = 2687 Btu/hr. 4.3.2 Internal Heat Sources

4.2.7 Infiltration 230 Btu/hr per occupant for 4 occupants was used to
determine total heat gain from occupants as 920 Btu/hr
This construction falls into the medium category (Stein et al. (Stein, et al, 2005 p 1611).
2005 p.1602) Since we are working with a winter outdoor
design temperature of 44.60F, the table gives us a design Heat gain from equipment was estimated as 1400 Btu/hr
infiltration rate (ACH) of .73 cfm. (Stein, et al, 2005 p 1611).

Volume of interior space from design drawings = 20243 ft3. Total Heat Gain = 7628 Btu/hr

(ACH )(volume, ft 3 ) .73x 27670

V= = 60
=246.29 cfm.
60 min/ hr 4.4 Measured Fuel
Heat loss is calculated as The study was conducted over 7 days starting at 3:30 pm on
February 15, 2009 until 5:30 pm on February 23, 2009.
q = cfm x 1.1 x ∆t = 336.7 x 1.1 x 21.27 Over that period of time, 239 pounds of douglas fir were
= 5762 Btu/hr. loaded into the masonry heater and burned to heat the house.
4.2.8 Latent Heat Loss TABLE 2
The relative humidity in the house is not intentionally Date Time of Burn Amount of Fuel
controlled (which allows latent heat loss to be “ignored.”)

4.2.9 Total Heat Loss 2.15.09 3:25 pm – 54 lbs

5:53pm
The sum of above mentioned heat losses is17400 Btu/hr.
The occupants heat 76.8% of their home with this system 2.16.09 8:40 pm – 27 lbs
(the rest of the rooms are closed off) which gives: 9:35pm
2.17.09 6:17 pm – 8:40 35lbs
17400 Btu/hr X .768 = 14608 Btu/hr pm
2.18.09 NO BURN NO BURN
Total Calculated Heat Loss = 14608 Btu/hr
2.19.09 7:50 am – 10:30 56 lbs
am
4.3 Heat Gains
2.20.09 NO BURN NO BURN
4.3.1 Solar Heat Gain 2.21.09 4:20 pm – 7:00 43 lbs
pm
The Solar heat gain was calculated using the Window Heat
Gain Calculator (Gronbeck, 2005). Clearness factor for 2.22.09 11:30 am – 24 lbs
Eugene was determined to be 42% (Kusterer, 2009). The 1:15pm
outside surface was variable, so the default reflectance of
0.2 was used. The SHGC for double glazed clear wood
windows is 0.58. Window areas were determined from
design drawings. Douglas fir has 18.1 MBtu/cord and weighs 2900 lbs/cord
(Sweep’s Library, 2009).
Solar Heat Gain By Window Orientation:
2 2 1day 18.1MBtu / cord
South: 398Btu / ft / day × 226 ft × 24 hr = 3748 Btu/hr 2900lbs / cord
= 6241 Btu/lb.
2 2 1day
North: 37Btu / ft / day × 82 ft × 24 hr = 126 Btu/hr
6241 Btu/hr X 239 = 1,492,000 Btu
 perhaps in the kitchen – where it provides a natural focus for
239 lbs of wood therefore gives 1,492,000 Btu total for the talk and dreams and thought” (‘Pattern Language’, #181).
7 days, two hours. The data obtained from the data loggers suggests that
heating a home with fire provides a feeling of thermal
1492000Btu comfort at lower temperatures, as the temperatures in the
170hrs
= 8776 Btu/hr.
space fell below American Society of Heating Refrigeration
and Air-conditioning Engineers (A.S.H.R.A.E) standard 55-
4.5 Total Heat Losses and Gains
2004 (Stein et. al. 2005 p.88) on a few different occasions.
Conversations with the homeowner revealed that the stove
Net heat loss is calculated by subtracting the total internal was not lit on two of the seven days because they considered
heat gains due to solar, occupant and equipment sources
the space “warm”. The data loggers indicate that on these
from the heat loss of the building: days the outdoor temperature was higher than the other
days, but interior temperatures went well below the
14608 Btu/hr – 7628 Btu/hr = 6980 Btu/hr temperatures experienced on the days when the stove was
used. This suggests that heating with such a stove allows
Potential Heat from Masonry Heater = 8776 Btu/hr
occupants to customize how often and to what temperature
they heat their space on a more intimate level than a
Stove efficiency = (6980 Btu/hr) / (8776 Btu/hr) x 100 = traditional system and, since the owners are in such close
79.5% efficiency contact with the method of heating, they realize its
significance and importance and are able to live with and
5. ANALYSIS appreciate the variations in temperature.
The masonry heater in conjunction with other sources of
heat gain supplied sufficient heat to overcome the home’s 7. ACKNOWLEDGEMENTS
heat loss. It kept the home at an average temperature of
65.90F despite relatively cold temperatures outside. Heat The authors of this study would like to thank the Barkman
loss through the thermal envelope accounted for 14608 family for allowing us to study their home. Also, thanks to
Btu/hr. The single largest source of heat loss was
Rob Thallon for setting up the initial meeting the family.
predictably through the windows, which accounted for 38%
of the total heat loss. Heat gains from sources other than the 8. REFERENCES
masonry heater produced 7628 Btu/hr. By subtracting the
heat gains (7628 Btu/hr) from the total heat loss through the Gronbeck, Christopher. "Window Heat Gain." Sustainable
envelope (14608 Btu/hr), the net heat loss of the building is by Design. 2005. 11 Mar 2009
determined (6980 Btu/hr). This value represents the heat <http://www.susdesign.com/windowheatgain/index.php>.
produced by the masonry heater. Comparing this number to
the potential available energy from the wood (8776 Btu/hr)
Kusterer, John M. "NASA Surface meteorology and Solar
provides the efficiency of the stove. The stove was found to Energy:." Atmospheric Science Data Center. March 2009.
be 79.5% efficient. 11 Mar 2009 <NASA Surface meteorology and Solar
Energy:>.
6. CONCLUSIONS
Stein et. al. Mechanical and Electrical
The initial hypothesis that the masonry heater would operate Equipment for Buildings, 10th
at 90% efficiency proved incorrect. The heater in fact Edition. New York, NY: Wiley, 2005.
operated at 79.5% efficiency. This is most likely due to the
process of burning wood, which can only operate at a "Sweep's Library: Firewood Comparison Charts."
limited efficiency level. Chimney Sweep Online. 2009. 11 Mar 2009
<http://chimneysweeponline.com/howood.htm>.
Further study of the stove’s efficiency would benefit from
testing in a closed environment. The variables of internal Tulikivi Product Brochure: 2007 - 2008. Juuka, Finland:
heat gain from occupants and appliances and the passive Tulikivi, 2008.
solar heat gain were estimated in this study but could be
eliminated entirely in a closed laboratory experiment.

In ‘A Pattern Language’, Christopher Alexander highlights

the importance of fire in human existence: “There is no
substitute for fire…build the fire in a common space –