Introduction:

For a description of the current test heater setup in our lab, see 2006 - 2007 Heater Testing at Shawville

Because of the non-typical floor level entry into an outside insulated chimney, it is not possible to take stack temperature measurements at the standard height of 8' above floor level. Since there is a damper at this height in the chimney, the first 8 feet of chimney is normally masonry, and is considered part of the heater system rather than the venting system.

For recent tests, stack temperatures have therefore been measured immediately upon leaving the heater. This results in somewhat elevated temperatures, typically around 400F (200C). This represents a small but significant hit on overall efficiency, when calculated by the stack loss method.

We have been intrigued by the extremely low stack temperatures attained in Igor Kuznetsov's "Free Gas Movement" (FGM), or "double bell" system, as demonstrated in recent testing by Alex Chernov.

We decided to do a variation on Kuznetsov's, and add a couple of small in-line "bells" between the test heater exhaust and chimney entry to see how much of a drop there would be in stack temperature, and to see if combustion is affected adversely.

The idea for this came from an observation on the 2007 test heater at Wildacres. Somewhat to everyone's surprise, when the gas path was set directly from the exit of the downdraft channel to the chimney, the three vertical flues still heated up all the way to the top, even though they were dead ends.

Description:

Existing setup. Flue gases exit heater on left, and enter outside stainless chimney on right.

May 3/07: Teardown.

Mockup.

Colin Coveny does a layout to the 4.5" firebrick module.

Floor is covered with sheetmetal. 1x2 steel tubing is used as a spacer and a sheet of cement board is laid on top.

Completed manifold.

(May 4/07) A pivoting damper has been fitted to the bottom of two 16"x16"x24" clay flue "bell" sections

May 31/07

Sept 1/07
16 x 16 flue liners are capped off with ceramic glass offcuts.
Flue system is sealed with clay mortar.

Sept 16/07: New opacity setup.
Arrow points to sensor. The end of the flue is blocked off with ceramic glass.
There is a slot along the left edge, which is the new sliding damper location. The sensor hole doubles
as the hole for the damper handle.

Analytical balance on table will be used to weigh Condar filters. Resolution is 0.1 mg (0.0001 g).
Table will need to be beefed up to provide enough stability for balance. 3" thick aluminum sensor
platform on floor has a 3 point steel pin suspension that rests directly on the concrete slab underneath
the insulated wood floor.

Oct 10/07
View of damper from top of flue liner "bell"

View during burn, damper open. Water condensation on glass.

View of bell during burn. Water condensation dripping down flue liner. Liner outside temp is fairly even from top to bottom.
Flue temperature drops around 20 - 40 degrees, and efficiency goes up about 2%.
Once we get the firing sorted out with the white birch, we will reconfigure the bells with a sheet metal exension.

Comments and Progress Reports:

October 15/07: Testing commenced on Oct 10. 5 runs so far, ironing out various glitches. We are using white birch with 20% moisture, which we will be using in Alaska shortly. Some dirty burns, due to various phenomena with the flammable bark during startup. Useful, since we will need some dirty burns as a reference for establishing what the limits are for clean burning.

Sept 16/07: This week, we are installing a heater where we have added a "bell" heat exchanger en route to the exhaust. Footprint sketch.

We are also shipping one to California where a channelled heat exchanger between the heater and chimney has been reconfigured to the much simpler, and easier to build, FGM design. Old heat exchanger design.

Test Results

Links:

Lopez archive of wood combustion technical information