The GEK Gasifier was set up in the 20kW configuration and run off of softwood chips. 

There are a few differences between the 10kW and the 20kW GEK all of which facilitate the higher gas flows for the reactor and flare. The reactor has a wider opening for the bottom plate to accomodate the 8.5'' bottom opening for the reduction bell. The reduction bell itself has a 4'' restriction and the air nozzles are situated in a larger diameter ring than the 10kW. Other hardware differences include a larger burner, larger air lines, and larger nozzle for the ejector venturi for the gas drive on the flare. I'm using an ejector venturi for these tests because of its simplicity, but for the palleted gasifier genset system, we will be using a blower/fan for the gas drive for our more complete system. 

The feedstock used in this run was softwood chips by our disk chipper. The moisture content was around 19% by dry weight which is a little higher than the drier walnut shells used in last weeks run. Due to the differences in void space, energy density, and moisture content, the temperatures are slightly cooler than observed in last weeks run. 

After the reactor was started and brought up to temperature, the pressure across the reactor was set at flow set points and changes over about 10-20 minutes were observed, such as the temperature profile and the packing of the reduction bell due to settling of char particles and their decrease in size. 

The set points are recorded as the pressure across the reactor in inches of water:

pressure:     minute:

2''               29

4''               56

5''               73

6''               94

8''               115

Right before each set point, the grate was shaken and the hopper was filled. Because of these activities the temperature fluctuates as well as a drop in pressure of the system.

Below is the temperature data log through out the run. The temperatures of two locations were taken, one at the bottom of the reduction bell and one at the top of the reduction bell. The bottom of the reduction bell temperature was taken one inch in the inside of the 8.5'' opening at the bottom of the bell. The top of the reduction bell temperature was taken about a 1/8th of an inch inside the 4'' restriction of the bell. 

The temperature rises subtly with respect to increased flow but the bottom of the reduction bell temperature seems to be slightly higher than the top temperature. These are only seen to reverse during filling and shaking the grate. The temperatures are seen reversed with respect what we have typically seen on previous runs. 

The data log of the flow of the wood gas is a bit noisy, but seems to fluctuate between 35-55 m3/hr during the run. In this run the 20kW GEK was driven by the ejector venturi which tops out at a pressure drop of 8''of water over the reactor. The maximum flow rate of the engine/gasifier system is not seen here yet, but we are planning to run this at the end of the week to see our stall flows of the complete system.

The combustion pressure and the reactor pressure are compared here over the course of the run. As the flow rate increases the pressure drop across the reduction bell increases because of turbulent flows and also possibly because of the increased reaction rate, smaller particulates are formed faster and need to be purged from the system. The saw blade type fluctuation in the pressures are seen here, but this is not due to the shaking of the grate. The grate was only shaken 6 times through out the run to observe stand alone interactions within the reactor. I am not quite sure what causes this, although we have seen this before in our runs, I believe. Does anyone have any ideas?

The sawtooth pattern is likely from the large air consumption used by the ejector driving the gas flow. The change in shop air pressure and compressor turning on/off at ~3 minute intervals would cause this. Note the upward curved rising and downward curved falling signature from tank filling/emptying. - BK

The pressure ratio between the combustion pressure and the reactor pressure seem to fluctuate around 0.3 which is similar to what we have seen in our previous runs with the 10kW. This is thought to be a good number in the intention of sizing the nozzles properly, though we do not know for sure.

We use the pressure ratio of the combustion and reactor pressure to determine when to shake the grate. When the pressure surpasses 0.4 for example, the logic programmed on the PCU will shake the grate to purge the reduction bell of smaller particles thus maintaining proper gas flows through out the reduction bell.

Questions/Comments

email: jay@allpowerlabs.org