As of GEK v2.3, we've upgraded all GEK kits to include a stainless steel tar fence.  We chose the taller design with side penetrations for the nozzles.  This is the design the discussion below decided was the most promising.  Here's the options we considered, as well as the made object in 16awg 304 stainless steel. 

For those who didn't see the original discussion, the purpose of this addition is to prevent passive circulation of tar behind and below the nozzles with inverted V hearth type reduction bells.  With V hearths, the char ash combustion area is porous, and the degree to which the nozzles are off the wall creates opportunities for tar gas to get to the restriction without going in front of the nozzles.

It is likely that J tube style air intakes like the GEK have more potential for this bad convective circulation, as the nozzles end up further off the walls, than direct in through the vessel designs. The tar fence is likely an easy fix for this, while also adding some additional insulation and radiant heat focussing benefits.

Update on 4/4/09

Anecdotal reports to date suggest the tar fence is in fact helping tar handling, but this is coming at the expense of lots of new fuel bridging problems.  Many GEK users are reporting the tall tar fence we chose is creating bridging problems with wood chips.  Thinking through this, as well as running some more fuels lately myself, I do agree and we now need to makes some adjustments.  Here's what seems to be happening.

The tall tar fence creates an exposed sharp edge that will trap fuel on top of it, often with some portion of the fuel chunk cantilievered inward of the fence.  These chunks trapped top of the tall tar fence will not decompose and pass, as they are not yet in the combustion zone.  This create a restriction that is smaller than the 6.5" of the fence itself.   This is not much of a problem for fuels like pellets or shells, but it is certainly a problem for more difficult chip fuels.

Thus we suggest that anyone running chip fuels cut off your tar fence so it is stops at the bottom of the nozzles.  Eliminate the top 3" or so of it, including the holes that the nozzles go through.   In essence, make the shorter tar fence option on the left in the drawing above, the one we didn't choose originally.  With the fence completely below the nozzles, it should not contribute significantly to bridging, as things are in combustion and decomposing, and the angle of repose line from the wall to the constriction is already underway.  The tar fence will likely sit below this line and thus not contribute to increased bridging difficulties. 

This shorter tar fence will still prevent tar from passing behind and below the nozzles.  It will have a little less heat focussing, but this was a minor point of the original.   In the future i believe we're going to go to a two piece tar fence.  A solid base like described above, then a movable top wrapping tube that you can adjust for various nozzle heights, and hold in place with a big hose clamp around the base. 

Making and installing a tar fence

You can make any of the above fence versions using the cut files below.  Roll the cut sheet in to a cylinder with a slip roller.  Or bang them out with a soft sledge around an appropriately sized forming tube.  Full res versions of the cut files are available in .pdf and .dxf at the links below. 

TarFenceOptions6_5cylCutFile.pdf

TarFenceOptions6_5cylCutFile.dxf 

The files assume your nozzle height is 3" above the top of a 6" tall reduction bell.  The tar fence is 6.5" in diameter, and is intended for a 6" nozzle ring.    If you use a different height, you may have to cut off the bottom of the tar fence, or cut it to be taller.  It is likely a good idea to have the tar fence go all the way to the floor of the reactor.  If you use a different diameter, the math is easy to redo.

To install, simply set the tar fence down around the inverted V reduction bell.  You will need to twist the nozzles to the side and wiggle things aronnd to get the nozzles through the tar fence penetrations.  Make sure you nozzle risers remain full pushed down on the riser stubs, and well sealed with clay or some other sealant.

The abover drawings just show the innermost reactor assembly.  This is part of the larger reactor insert and gasifier assembly shown below