Jim,

One way of thinking of this is the "extended tar fence" say 7" diameter with 5 short ejector tubes 1.25" long welded on at the position of the nozzles.

2 birds with 1 stone?

Either way its going to produce a lower pressure near the top of the fuel hopper and help recirculate and burn the tars.

Any thoughts


Ken

Damn you Jim - that's at least 10X simpler than what I had in mind. And I can't see a single thing wrong with your concept. Wow, well done!

yes ken, combing these ejector nozzles with the tar fence is the imagined implementation. i didn't include the tar fence in the above drawing as it would get crowded and confusing in there.

however, i had not realized we could just weld the ejector leg pipe on the tar fence itself, with the nozzle loosely nestling into them. this would certainly ensure a complete seal at the nozzle entry point, and thus no tar leakage around the nozzles. it is also likely a good way to hold everything together and kept oriented, in a fairly simple manner.

there would be a little less accuracy about the ejector detail geometry, which would liklely create variable performance across the ejectors. such may or may not be important. or maybe we weld the fully assembled and welded ejector nozzles to the tar fence, so there is one giant assembly, and nothing can move around. push it all down into the reactor and onto the nozzle riser stubs in one go.

Hello, Jim & Co!

This item was discussed in Finland during the ninties; there were two alternatives:
1. Helping the centre pillar upwards by installing a pyramid made of diam. 10 mm stainless steelbars. = four bars from a centre point about 120 mm above the nozzleplain and sloping out between the nozzle jets to the restriction plane. One realization, and a successful one; 10 km/h more speed. The bars though were not longlived.

2. A proposal like yours, but more modest in height; corresponding 3" height in your GEK.
The reason for this: The downward ring-chanal cools on the outer wall, and if one
takes the gases from too high above the oxidation zone, their "starting temperature"
on the down trip will lead to tar condensation in a place it will no more boil away
from!



The tarfencing also started on my proposal, but has not become "popular" yet. Only 3
cars have it built so far.

Max

One more item: The injector needs to have and retain good suction. The secondary nozzle needs to be narrower (and a bit longer) than in the principal drawing...
Max

Max, Jim, or anyone - could someone give me a pointer to where the "tar fence" is introduced and discussed? I don't know what is trying to be accomplished with it, but would like to...

W.r.t. Jim's proposed ejector geometry, I can think of one shortcoming and one place for improvement. Firstly, the mixing zone isn't nearly long enough. There should be 10-40 (primary jet) diameters of length for the driving gas to fully entrain the pumped gas. Secondly, when the 600C preheated air hits the tar-rich vapour, it's going to immediately start burning (which is great!). This fundamentally alters the combustion process of the entire gasifier. It becomes strictly a tar-burning gasifier, since the first thing that the incoming air sees is an abundance of highly reactive tar vapours. In fact, 100% of the input air is not sufficient to fully consume all the tars (this too is OK).

So once an effective tar ejector is built, the char will never see any oxygen. All it will see is very hot combustion products from the combustion of the updraft gas (N2, CO2, CO, H2, H2O), which is now the sole char gasifying agent.

In order for this to be fully realized, the air and tar vapours need the opportunity to fully react (the three Ts: time, temperature and turbulence), before their very hot combustion products are directed as a jet onto the char. This needs a certain volume, and thought to flowpath. I am thinking that the entire area that is below the present nozzle level, outside of the inverted-V-hearth, and inside of the cylindrical shell, should house the ejector/mixer/combustor-chamber-volume/exit-jet assembly. I am figuring out how this can be done with cut and welded 310 stainless; and would not at all be disappointed to see Jim nicely trump me with something just as good but far more buildable ;-)

Daniel look under Proposed GEK Improvements/Hearth Mods/Tar fence.
Good idea. Execution is going to be tricky. As Max stated inside the hearth apparatus gets killed from the heat and the primary free gasses. Speaking as an operator any tar cooled off area is going to gum and plug up. Think shutting/ shut down cooling off tars condensation. The successful design will be the one that works, keeps itself clean when up to operating temp AND is easy to disassemble for the inevitable "clogged up needs to be cleaned".
Regards SteveU.

"we currently believe pyrolysis in a downdraft to be producing 2x or 3x the amount of tar gas than we can directly combust. (The specific numbers are we still debating in the gasifier mass flow study). "

Where is this debate and mass flow study happening? I want to be in on it, please, if I may!

BTW I agree with you that we don't have enough gasification air to fully burn the tar. Here's my SWAG of it: Wood = 20% fixed carbon plus 80% volatiles. Overall we know we need approx 6.5kg of air to fully combust 1kg of wood. We know that the 0.2kg of fixed carbon needs (32/12/.22*0.2) =2.4kg of air to burn it, therefore the volatiles need 6.5-2.4=4.1kg of air to burn them We are going to run a gasifier at an equivalence ratio of 0.25-0.30, or 1.6-2.0 kg of air per kg of wood; this is just 40-50% of the 4.1kg air we need to fully burn the tars.

Steve, thanks. W.r.t. tars clogging it, I agree that this must be taken care of in the design. I am thinking that if the tar laden gas flows downward, and mixing with the ejecting stream takes place at the very bottom of the tarry gas passage, that there won't be an opportunity for the tars to pool and collect. Also, the ambient temperature ought to rise, as one gets lower in the tarry gas ducts. This ought to keep them freely running.

Also, the ejecting air at 600C ought to be capable of autoigniting the tars. If one arranges the passages so that in the event of a pluggage, the tar plug is exposed to 600C air, it ought to be able to be burned clear.

(Yes, I realize that to a certain extent I am assuming well-behaved co-operative tars here... ;-)

Answ.to Daniel

The disscussion in Finland as far as I know of, was between a few individuals and the enthusiasm was not remarkable. In the early eighties serious experiments were done in the Nederlands at Twente Univerity. Indonesian and Dutch scientists did the opposite to what now is "on the table". A vertical flame burning chamber in the lower central end of the silo, letting out over the reduction region at the perifery of the flame chamber. Part of the flue gas was succed upwards outside the flame chamber inside the silo. These hot flue gases made the pyrolysing on their way up to the upper end of flame chamber. The sucction was provided by a compressed air-driven injector on top of the flamechamber.
Name of the main researchers:

Herri Susanto and Antonie A. C. M. Beenackers

Published by 1996 Esevier Science Ltd.

FUEL Vol.No. 11, pp. 1339-1347, 1996

Printed in Great Brittain


Low tar and medioker gas from this "laboratory baby"...

Max

Uploaded the Susanto paper, for anyone interested:
http://gekgasifier.pbwiki.com/f/moving-bed-gasifier-with-internal-recycle-of-pyrolysis-gas--Susanto-1996.pdf

Max, could you please elaborate on your "mediocre gas" comment? What was wrong with it? (I'm concerned I may have missed something, because nothing about it twigged me as being flawed..)

Daniel RE: Maxes mediocre gas and your air to fuel ratios. I would think if too complete of oxidation is achieved in the gasifier not enough char will remain for reduction process resulting in very stable CO2, H2O and N2 output gasses. Good results if in my space heating wood stove. But bad for fueling my diesel engine. SteveU.

Jim,

Did you consider carbonising of the ejector? There will be some border between gasious tars and burned tars. This carbonising border is always present at a certain height above of the nozzles. In the fuel bin is not much of a problem. In a relative narrow tube it is.

Regards,
DJ

Daniel, and All!

If one does study the ~4 pcs of A4 that the report contains thoroughly, one can see from the heat value diagram, that increased recirculation lessens the tar content AND decreases the heating value!

The authors themselves! “confess” the reason being the increased hot UNINSULATED surfaces in the silo…

Another item is, that the flue gas is barely exceeding 1100 Degrees Celsius!

Blasting charcoal easily reaches 1400 Degrees Celsius!

Does the candle light up?


So far the only way to include a heat value RISE as well as a tar reduction is achieved with BOTH char blasting AND a flame cavity upwards in the centre.

This way all the doubling, outside, cooling, tar condensing surfaces are avoided. The cavity forming materials, on the other hand have to be EXCELLENT!

Regards,

Max

List,

Functioning of the tuyeres as an ejector depends a lot on the ratio primairy air/pyrolysis gas. If the amount of pyrolysis gas is too large, this gas will find its way by the normal way down. We need to consider that the vacuum at the tuyere tip is about the same as in the vertical pyrolysis gas tube. On excess pyrolysis gas volume the ejector principle will not work.

What would normal ratio primairy air/pyrolysis gasvolume be with bone dry wood? Just a wild guess is enough. I understand it depends on the volume of pyrolyse wood and temperature.

Next step: how to dimension an ejector?

Regards,
DJ

Hi John, I've been working on getting an analytical handle on ejector design (and it's taken me far longer than I thought, since I told Jim and Bear 1.5 weeks ago that I was starting work on it!). I made really good progress last night thought, there's a good chance I'll be able to post something today.

In order for the ejector to pump, we need to ensure that there is a low pressure region where the pyrolysis gas meets the air jet. After they combine, mix, and burn, we need to _raise_ their pressure, in order to cause them to flow to the hearth area. Not a lot of pressure rise is needed, but a modest positive amount certainly is, in order to assure positive circulation. This is done by having a diffuser, which will slow the flow down a bit (converting kinetic energy into pressure).

Here's my SWAG on pyrolysis gas volume. Let's say that 1.0kg of bone dry wood is turned into 0.8kg of pyrolysis gas and 0.2kg of fixed carbon (char). Assuming that the temperature of the pyro gas is 150C and its molecular mass is 40g/mol (a bit denser than air at 28g/mol), that would be about 0.9 m^3 of actual gas volume at that temperature, and at atmospheric pressure.

(Now please stand by while I break my comment in two; I've just been told by the wiki software that there is a limit of 2000 characters per comment...)

(...cont'd...)

So how much air are we going to have, which we will use to pump the 0.9m^3 of pyro gas?

To run the gasification process, let's say we are using a 0.25 equivalence ratio. This is pretty much the ideal ratio to use, and it is quite appropriate if you are using for dry fuel and you are running a thermally efficient gasifier (which the GEK is already, and will be even-more-so once it is equipped with internal recirculation). This means that we will be using 0.25*6kg-air/kg-wood = 1.5kg of air per kg of wood that we gasify.

It is this 1.5kg of air that we will use to pump the 0.8kg (0.9 m^3) of pyrolysis gas. This 1.5kg of air is about 1.2m^3 of volume when it enters the gasifier's air inlets. But, when we heat it to 600C, we increase its volume by a factor of 3.2X, so the airflow volume that will be driving the ejector's primary jet is 3.9m^3.

It really ought to be very straightforward to pump 0.9m^3 of pyrolysis gas gas, using a jet of 3.9m^3 of air, and to do so at quite a low cost (in terms of pressure requirements).

Thanks Daniel!

Even if the pyro gas would be a bit warmer (and I guess it will and must be to avoid condensation) and the air a bit colder it should still be functional. What will be important is to have a long slender fuel bin in order not to have stowing of pyrolysis gas just after startup. Excess moist can also increase pyrolysis gas volume. So, again, dry wood or a monorator is important.

Regards,
DJ

As far as I can tell, internal recirculation is thermally superior to (more efficient than) a monorator. Adding a monorator to a gasifier with internal recirculation will result in a net loss of heat.

Daniel you are correct that a drying monitor would result in a net system heat loss. But depending on the locale and fuel stock this is a nessary use of the heat to perform the work of predrying the fuel. Example in the maritime climate of the Pacific North America where I live fresh cut woody fuel will have a 40-50% moisture content by weight. Given a "good " summer of 4-6 weeks drying time I can get this down at best to 7-8%, last year was 15%. Now unless hermatically stored, IMPRACTICAL, it will rehydrate back up to 20-22% within weeks and stay that way until the following summer even when in covered "dry" storage. For me and most Northern Europeans and many others around the world.
Forcing the internal tar recirculation to handle this level of moister will quench the desired oxidization temperatures. Let a monorator drop 20 points of fuel mioster and leave the tar recirculator do its job to direct and blend the air and tars into the oxidization zone. All my own opinions. SteveU.

i don't think a monorator and the ejector nozzle scheme are mutually exclusive. the ejector nozzles are only better directing gasses to where work can be done. they are not creating more heat to do more work (i.e. lead to better tolerance of moisture). their only marginal gain here will come from making sure the tar gas goes through the hottest point in the bed, not rather wander down the middle. this is a gas directing/mixture issue, not a raw heat available issue. in practice this might create a slight greater moisture tolerance, but it is not going to be significant i don't think.

we are not going to see improved moisture tolerance above current until we are predrying the moist fuel and/or heating the dry fuel before pyrolysis and/or running pyrolysis before the combustion zone heat contribution. or in other words, continue to make progress in removing the 5 or so thermal drags on the combustion zone, beyond the current air preheating and improved insulation solutions. (see here for summary: http://gekgasifier.pbwiki.com/Modelling-Gasifier-Energy-Balance) there are add ons in process that will do all these things in situ on the gek gasifier, but in the interim, we need to start with dry fuel.

given the importance of dry fuel, and the difficulty of assessing such by looking at wood, i'm now including a pin/resistive moisture meter with every gek kit. thus we now cover the basics of temp, pressure and fuel moisture with included instruments in the kit. much more is ultimately desireable beyond this instrumentwise, but such is a good starting quiver of tools to acquire relevant info.

jim