Bell experiments with a Contraflow heater

January 19, 2006
This was the original test setup at the Shawville Lopez Lab.

Space considerations dictated an exterior insulated 8" chimney.
This shows the connection via a 3' horizontal run of bare 8x12 clay liner, incorporating a sliding shut-off damper.
The horizontal run was used for opacity measurements.

May 3, 2007
The following year, we modified the exit as shown in order to try some experiments with a
bell section between the heater and the chimney.
There is an air space underneath, to protect the wood floor, and allow measurements.

Two 12"x12"x24" flue liners were modified with pivoting dampers as shown.

Two small "bells" in place. The yellow arrow shows the location of the sliding shut off damper.
The idea was to start with small bells, and get progressively larger, to study the relative effects.

The bells were capped off with ceramic glass.
Water condensation is visible on the underside of the glass, during a burn. The glass eventually got black from soot.
Both flue liners got quite warm along their entire height, due to the buoyancy effect of hotter gases.
The effect on stack temperature was measurable, but was small - roughly 5 degrees F.

January 17, 2009
The flue liner bells were removed.

A firebrick shiner bell section was created by adding height to the connecting piece.

The height was extended by four courses (18") and capped off with ceramic glass.

February 20, 2010
After the results of the July 2009 Rocket-Bell experiments, a diverter was added to the bell.
13.5" high, with a bypass space at the edges.

Four tests were run with the bell in this configuration. HK-L17, HK-L18, HK-L19, HK-L20

March 3, 2010.
With the diverter in place, 5 courses (22.5") in height were added to the bell.
The numbered X's are surface temperature measuring locations. #4 is centered on the right face, close to the top edge.

March 5, 2010.
Test HK-L21
Graph shows the surface temperatures on the bell (F). Bottom scale is time (minutes).
See previous photo for locations.
Note that #2 is cooler than #1 and #3

Summary of 7 tests.
The first test was with no diverter.
The last 2 runs are with the height extension.
The extension seems to add about 1.5% efficiency due to 10 - 15 degrees (F) lower average stack temperature.
The last run had a lower burn rate, due to a cooler firebox, damp wood, smaller load, resulting in higher oxygen, which dropped efficiency.
So, the extension seems more beneficial for a large load than a small load.

• Standardized height for measuring stack temperature is 8'-0"
• The exterior insulated chimney setup necessitates measuring stack temperature at floor level.
• Calibration curve, above, was generated from a similar heater with an interior masonry chimney, in the same building.
• Calibration curve is used to convert a floor reading to an 8'-0" reading.
• For test HK-L21, it results in a calculated efficiency increase of 4.6%, if the heater had 8' of interior masonry chimney below a shut-off damper.
• The efficiency increase from the bell is roughly 3 - 5%, if we assume the original height and the extension to contribute equally.
• This yields approximately 7 - 9% efficiency increase, for a large load, compared to the heater venting directly into an insulated chimney at floor level.
• Theoretical maximum efficiency for a non-condening system is approx. 84%, calculated using the higher heating value of wood (North American method).
• This translates to about 96% theoretical maximum, European method (lower heating value).
• This heater configuration, with a masonry chimney, has a European efficiency of 88% for burn HK-L21.
• About 3% of the loss would be from unburned carbon monoxide. The balance would be split between the need to maintain a chimney temperature to provide draft, and the need to heat about 150% excess air to allow good combustion.
• Efficiency comparisons can only be made in an "apples-to-apples" fashion, and are specific to the testing method used. The tests above used the Colorado method. It requires the damper to be left wide open until the coals are burned (95%) oxygen recovery. The air control also has to be left wide open. Only one poke is allowed during the burn. Kindling is with a coal bed on the bottom.
  

This page was updated on March 7, 2010
This page was created on March 5, 2010