Report from UK GEK experiments


Intro

 

I received my GEK in early January 2009 as an unpainted Level IV kit.

 

I spent a couple of days painting it, assembling it and getting it prepared for its first run.  The early stages are documented on my website  www.powercubes.com/listers_12.html

 

Unfortunately,  "Sitebuilder", my web page design tool has crashed on my PC for a couple of days - but I have now transferred to a laptop.  I have now updated my listers_12 site with some photos of the first run and pictures of the gas flare I achieved.

 

Stripdown and Analysis - 30th Jan 2009

 

After 3 runs and 60 litres of fuel, it was time to stripdown the GEK and make some changes.

 

Whilst I was vacuuming out the reactor vessel to recover the vermiculite insulation, my mate Simon was making some 2" tubular spacers to set the nozzles to their correct height.

 

The char residue in the reduction tube consisted of fairly small pieces.  The biggest were 0.5" x 0.25" shreds, but many were 0.25 x 0.125".

 

I emptied out the area under the grate and obtained about 6 litres of char particles and dust. This represents about 10% by volume of the original fuel burned.   I used a sheet of perforated metal (same as filter discs) to riddle the char and separated it into 2 fractions.  There was about 3 litres of char dust, and a similar quantity of granules greater than about 0.125" minimum dimension.  These larger particles were saved for use in the filter can.

 

The filter was a mess - the steel wool soaked with a black mixture of tars, water and other condensates.  I made an additional disc from perf metal, and filled the lower half of the filter can with the saved charcoal granules.  Then I put in the new disc, then some steel wool and finally another disc.

 

I was concerned that I had not previously been able to establish the correct oxidation zone.  These suspicions were confirmed when I tried another run, where I found it impossible to get the fire to move upwards from the reduction tube and into the area around the nozzles.  20 minutes of poor quality gas and I gave up and capped all the air vents.

 

Simon and I revisted the gasifier after a break of some 45 minutes, and a chat about what was going wrong.  I turned on the blower, opened the 1.5" BSP bung in the lid and immediately got flarable gas. I then opened one air vent cap and screwed in the central bung.   The fuel had obviously been drying out over the 45 minute break and quickly flared into action around the one nozzle that was feeding air in. The air velocity at this nozzle would also be quite high - which suggests that for a small engine requirement perhaps 5 air inlets are too much?

 

Mukunda Tube Ideas

 

I am very keen to use this trashy wood chip fuel, and I know I'm going to have to auger it in, so I think I may have to take an alternative tact, and use an extended 6.5" diameter tar fence to concentrate the fire in front of the nozzles, and back fill behind the tar fence with vermiculite.

 

I need to establish the hot oxidisation zone across the whole 6" diameter zone - and constraining it within a tar fence appears to be one way to do this.   By extending the fence upwards, any heat of oxidisation will rise upwards and  pyrolise, torrefy and dry the column of fuel above it.

 

There's a lot of difference between a 6.5" diameter fuel column and a 10" diameter one.  Areas are in the ratio of 10 to 25,   so restricting the fuel to the area immediately above the oxidation zone, will intensify the heat of oxidisation and help pre-pyrolise the fuel.  I intend to do this with either a 6.5" diameter fire extinguisher can or a 6" diameter flue pipe section.

 

This leads on to a new idea about fuel augering.    If  I make a new lid for the reaction vessel, I can have a fire extinguisher can welded in centrally to make a fuel feed tube, and use my auger to drive the fuel into the oxidation zone.  Each turn of the auger will deposit about 2.5 litres of fuel into the oxidation zone.    At 20 litres per hour consumption, that's one turn every 7.5 minutes.  This means that the oxidation fire has 7.5 minutes to dry, torrefy and pyrolise just 2.5 litres of fuel  - thats better than working inefficiently against a  full canister of 16 litres.

 

This is now starting to take on the appearance of an auger feed, Mukunda tube gasifier.  The Mukunda tube above the nozzles would hold about 6 litres of  fuel, or 2.5 litres with an airspace above it. Sensing the temperature above the fuel layer would allow the auger motor control to decide when to add more fuel.

 

If that's what needs to be done to gasify this damp, trashy fuel with low tar output - then it's a route worth following.

 

 

First Trial Runs

 

I've done 3 trial runs with the GEK  Jan 20, Jan 21 and Jan 23.  The first two were about 1 hour duration, the last about 30 minutes.

 

The low grade wood chip fuel I am using seems to gasify OK, provided that you shake/rotate the grate at regular intervals and use a poker to stir and agitate the fuel in the integral hopper to prevent bridging.

 

 

 

My wood chips reduce to very small grains of charcoal. Whist these offer a high surface area in the reduction zone, they may be hindering the gas flow - so I have found that I need to keep the blower fan on the full setting.

 

One point regarding these first firings, is the height of the air nozzles has been set at 2.1" above the reduction cone.  This happened as a result of a misunderstanding in the length of spacers needed for the vertical air pipes. Whether this is a major problem is yet to be ascertained.

 

Fallling into the tarpit

 

The GEK is producing flarable gas - but at the expense of tar. After 3 hour run time, there were significant tar deposits in the flexible plastic gas hose. The tar was fairly soft and not the hard toffee-like stuff I had imagined.  This might be as a result of running the blower fan on too high a setting and having too little gas remanence time in the reduction zone.  It looks like I have encountered all of the beginner's mishaps.

 

In the last firing, the fuel hopper was about third full, and the intention was to burn the remaining fuel.  Lighting was not as easy as before, and it is my belief that the fuel around the air nozzles failed to ignite, and the usual oxidation zone lobes failed to become established. Thus all of the reactions were confined within the reduction cone. Using the long thermocouple probe, I confirmed a temperature of almost 1100C at the mouth of the reduction cone and 350C at the air nozzles.  Nevertheless, the GEK produced gas for 30 minutes in this limited mode until the remaining woodchips were consumed and only char remained in the reduction cone. The flare burned with a distinctive blue flame suggesting a high carbon monoxide content - attributable to charcoal gasification, now that water and tars had been consumed.

 

In future trials I will take care to ensure that the woodchips are burning correctly around the air nozzles.  This will increase the gas throughput through the GEK as oxidation and pyrolysis is occurring over a much larger area - approx 6" diameter, rather than being restricted by the 3" diameter aperture of the reduction cone.  Is the failure to get the oxidation lobes around the nozzles as a result of the nozzles being an inch too low?

 

The presence of tars in the output gas confirms that the tar cracking is not occurring properly in the reduction cone.  Jim has suggested that this is probably because the gas flow is too high and the gas is not spending sufficient remanance time in the reduction cone.  It might also be because the nozzles are too low, and not leaving sufficient char bed below the nozzles.  Thirdly, if, as in the last run, the oxidation was clearly occurring in the reduction cone and not around the nozzles - then the depth of the char bed would be severely reduced.  If this is the case, then I will have to ensure correct operation around the air nozzles in subsequent firings. - Clearly there is a lot more to learn for successful operation.

 

Fuel Usage

 

In terms of fuel consumption - In the course of the initial 3 firings, I had flarable gas for almost 3 hours.  I used 60 litres of low grade wood chips fuel, with a bulk density of 180g per litre. The fuel included bark, leaves and twigs but appeared to gasify regardless of being low quality fuel.

 

If we assume the calorific value of the woodchips with 20% MC is about 15MJ per kilo the energy flow through the GEK can be calculated as

 

20 litres or 3.6kg of fuel per hour equates to 54MJ or 15kWh

 

If the gasification process is about 70% efficient, then the thermal content of the gas flared per hour is about 10.5kWh.

 

Performance of Blower?

 

One unknown is the volume of air or gas that the blower can shift per hour. Perhaps Jim might have some figures.  I know that the Lister will draw about 28m3 of air per hour at STP.  If a fuel/air ratio of 1:1.4 is needed then 12m3 of woodgas and 16m3 of air will be needed.

 

With the air vents closed and the fan on full setting I am seeing just 3" of pressure difference on the water manometer.  On low speed, its about 1.25" of water.

 

Vermiculite as Insulation?

 

I insulated the reduction cone gap and the reactor vessel with vermiculite.  There has been a recent comment on the Woodgas group, suggesting that vermiculite and perlite have an R value of about 2, making them less than ideal for insulation, when compared to kaowool. This appears to be contrary to popular knowledge - so the point is certainly worth further investigation.

 

It appears that the thermal conductivity of some ceramics increases with temperature - kaowool starts at about 0.06 W/mK and ends up at 2 W/mK at 1200C. Vermiculite may also show similar behaviour with temperature.

 

http://www.vermiculite.org/properties.htm