Here's a proposal for how we might open up the constriction size on small scale gasifiers, so as to not have so much bridging difficulty, (while at the same time not recreating the large tar problem of open core designs).  The requirement for small constriction sizes with the standard Imbert design is the main reason we end up with so many fuel bridging problems.  Tar conversion is better with the Imbert, but at the expense of mechanical flow.  Open cores have much better mechanical flow, but at the expense of tar. 

Mukunda et al improved the tar conversion performance of open core designs by adding a tar reburning second stage air inlet near at the base of the reactor.  Tom Reed, CPC and others made progress solving the same problem with the multi point variable air injection designs.  Both of these strategies create a flaming pyrolysis partial gasifier on top of a tar reburning charcoal gasifier.  While this improves the performance of the open core tremendously, there are still tar problems in practice, or at least great fuel sensitivities lest the tar problem reappear.

One of the most significant reasons for the continued fuel sensitivity follows from the lack of closed top.  Working with an open top, one cannot generate much blast pressure at the nozzles, thus it is difficult to get good bed penetration.   Good bed penetration requires high velocity out the nozzles, with is only possible with a reasonable pressure differential across the nozzles.  Without good bed penetration, there will again be tar.   I believe this might be the reason why the CPC unit had such difficulty with the Mt Shasta wood chip project, but seems to be working great on walnut shells in Sacramento.  Walnut shells are very easy to get good bed penetration, given their open shape.  Walnut shells are the most forgiving fuel I have run so far.

Unfortunately, the world of waste biomass is not often in the form of halved walnut shells.  We need a design that can handle common wood chips.  So here's another swing at solving the same problem, while hopefully not creating a new one this time.   For now let's call it the "Simplified Straight Reduction Tube Tar Recycling Hearth".

See here for full res .pdf file:   SimplifiedStTubeTarRecyclingHearth.pdf

The proposed architecture is about halfway between a throated gasifier and an open core tar reburning gasifier.  The nozzle ring is a little narrower than you would have in an Imbert, but the minimum constriction in the combustion/reduction passage is much wider.  The minimum diameter the fuel has to pass through is about twice what we expect with an imbert configured for the same flow rate.

This architecture is an elaboration of the internal tar recycling ejector venturi nozzle idea I put up a few weeks ago.  See here: http://gekgasifier.pbwiki.com/Internal-Tar-Recirculating-Ejector-Nozzles.  I've now combined these nozzles with a straight reduction tube that comes all the way up to the bottom of the nozzles. 

Given the main combustion plane were trying to maintain is at or near the nozzles, there's something to be said for putting our constriction right at the nozzles.  Such would in principle reduce tar control problems on lower gas flows when the combustion heart lobe no longer reaches all the way down to the constriction.  With the "constriction" at the nozzles, you combine your most focussed combustion area (the ring of nozzles), with the area you can most easily control flow (i.e. the smallest fuel/gas passage space)

The problem of course is whether having the "constriction" at the nozzles will provide adequate space for expansion of the hot combustion gasses.  It certainly would not if the constriction were done at Imbert ratios.   But if the "constriction" is about the size you'd want your reduction bell to taper out to anyway, and there is no immediate combustion cup taper inwards that the gas is having to accelerate through, I think we might find some interesting and workable balance.  I think we might discover there's an acceptable balance of flow with just a straight tube, between hot combustion gasses flowing through the larger between char spaces in the combustion zone, and cooling gasses passing through progressively smaller interstitial spaces as char sizes reduce.  While this combustion reduction geometry is likely not flow optimized, it might be good enough that the improvments in mechanical flow make it a reasonable compromise to work around.

The tar solution in this proposed architecture follows from the ejector venturi based internal tar recycling nozzles.  Tar gasses are not left to passively wander the bed, but are rather actively scavenged into and through the nozzles.  Tar gas is purpose directed to the hottest points in the reactor (i.e. in front of the nozzles), and thus has much less of a chance to bypass without being cracked.

Of likely equal or greater importance, is the change in pyrolysis dynamics we can expect from the internal tar recirculation.  This recirculation will create an upwards convection flow in the center of the reactor, to the top of the inlets to the tar downcomers.   This upward draft drawn is drawn by the ejector venturis, and will bring heat upwards to the venturi tube entrance, and thus stretch pyrolysis upward inside the reactor.  This stretching of the pyrolysis zone will extend pyrolysis times, as well as reduce pyroysis temps.  The results should be better completion of pyrolysis, and at lower pyrolysis temps, which will produce less refractory tars. 

A significant problem of the standard Imbert is the pyrolysis zone is very shallow, hot and short residence time-- thus the produced tars are refractory ones, and much harder to break than low temp tars.  Also, larger fuel chunks can fail to complete pyrolysis quickly enough, and end up still in active pyrolysis after passing into the combustion and reduction zones.   Given these problems, the pyrolysis zone extension and pyrolysis temp reduction we will get from this internal tar recycling system, might prove rather significant. 

The resulting process architecture will be somewhat of a slow updraft pyrolysis unit, sitting on top of a tar reburning downdraft gasifier.   This architecture is similar to the Beijer / Susanto internal tar recyling gasifier-- a unit which apparently had impressive tar control over a large gas flow range.   My proposed design is in some ways the Susanto design turned inside out.  The ombustion comes from the outside of the bed inward, instead of from an inner cup outward, as in Susanto.  In my design fuel remains in a single center open passage, not split into the annular passage surrounding the combustion cup (which most likely created flow difficulties with the Susanto unit).

The proposed design keeps a closed top, thus good bed penetration at the nozzles is still possible.   Again, you need a sealed vessel to generate adequate pressure differential at the nozzles to generate adequate air velocity to break through the fixed bed.  About the best an open top can do is leak air out the holes in the walls (otherwise known as nozzles) and hope this distributes inward through the bed on the longish trip downward.   This loss of bed penetration is why the tar conversion performance of an Imbert plummets when you open the top to refill fuel.

Closed tops are also needed to generate adequate draft to pull air through an air preheating / syngas cooling system like used on the default GEK downdraft.  While I've found the drag from these SS lines to be below measurement on the manometer at typical flow rates, it is still more drag than an open top.  It is much easier for the air to come down through the bed than wind down the SS tubes. 

Open tops and good heat transfer architectures do not go well together.  For all these reasons and more, I find the benefits of a closed top to outweigh their liabilities.  Yes, they are much more difficult with wet fuels, but we have other ways to solve that.  Too much mositure is an easier problem to solve than too much tar.

So maybe we call the proposed architecture a Closed Top, Tar Recirculating, Semi-Open Core?   (or CTTRSOC for clarity and brevity?)

Note: This concept and drawing are distributed under the Creative Commons Attribution-Noncommercial-Share Alike License.  Many rights are given, but some rights are reserved.  Talk to Jim Mason for the specifics.