The purpose of this test is to run the 20kW GEK and the 20kW PowerPallet to be able to characterize the unit in varied run scenarios and feedstocks so that we can continue to improve the system and system components to provide reliable support and equipment. 

Last Friday, we did a short hour an a half run of the GEK scaled up for the 20kW skid. The main feature is the reduction bell dimensions which have a 8.5'' opening at the bottom and a 4'' diameter restriction in the bell and longer diameter for the circle of air nozzles in the reactor. This will support higher gas flows needed for the 20 kW engine.  To motivate the gas, we used an ejector venturi with an 1/8th'' nozzle diameter and a larger burner (dia. 6'') and a 1'' air pre mix valve to accommodate the higher volume of flared gas.

For this run, we used dry walnut shells and started out the reduction bell with hard wood mesquite charcoal around 1-2'' in diameter.

After the GEK gasifier was started, the pressure across the reactor was ramped up and held for about five minutes. In between each pressure ramp, the grate was shaken, and the drying bucket was opened to fill with more raw biomass. During this refill and shaking activity, the pressure of the system decreased and the temperature increases slightly with the influx of extra oxygen from the opening. 

Within 50-60 minutes into the run, there was condensate observed in the pressure line, and it was purged immediately. This does offer some doubt to the pressure readings in this range, even though they seem reasonable, I would suggest a second run for verification. 

Above is the temperature of the top and bottom of the reduction bell. We started the reactor using about a half a cup of kerosene. The reactor was up to temperature after about 10 minutes from lighting. The temperature is relatively constant with a few oscillations around filling and shaking the grate. 

Typically we see that the bottom of the reduction bell is cooler than the top of the reduction bell in most of the runs, however this shows that even though both of the locations follow a temperature very close to one another, they are reversed throughout the run. 

The zero points around 15 minutes is due to the plugging and unplugging of the third thermocouple from the board. A third thermocouple was set up through the top of the reactor to study the vertical profile of the reactor, however it was at this time of the run that I noticed it had failed. Unplugging to verify this failed thermocouple was what caused T1 to drop signal. These data points could have been taken out for simplicity. 

Above is the combustion pressure in comparison with the reactor pressure. A greater difference is observed at higher flows which is not terribly surprising. 

Bear has been using the ratio of the combustion pressure and the reactor pressure to determine when to shake the grate to purge fine particles from the bell which clog the flow of gasses through the reactor. The pressure ratio here starts at 1 then climbs linearly down to a pressure ratio of .2. The grate was shaken every 5 minutes through out this run. Disregard the lowest points in this graph for they are at long refill points in the run. A possible explanation for this direct linear fall in pressure ratio could be most related to the larger chunks of hard wood charcoal used in the fresh 20kW GEK. It does take time for the particles to reach a homogeneous state from first runs, and we have seen that first runs are typically more different than the previous runs with charcoal left over from a previous run. This may very well have to do with the char particles settling to their most stable run state. I would like to try this run again while using the charcoal left over from this current run, I believe we will get better results. 

This above graph is of the gas flows through out the run. We are reaching upwards to 67m3/hr just using the ejector venturi with a 1/8th'' air jet nozzle and a 1'' air pre mix valve. We topped out at about 8 inches of water on the reactor pressure due to both the air compressors ability to keep up, and the closing of the O2 valve so as not to compete with the pull on the system (even though at this point the flare was running far too rich). Even though 8''of water was the max for the system it is not recommended with the current set up, although 6''of water was reasonable. 

-Jay